Use F11 on your keyboard to switch between "normal view" and "full screen" (without browser toolbars)

The update status of all pages of this website is now listed here.

There has been a number of Hellschreiber types and models produced by the Siemens-Halske company, other than the Feld-Hell, Presse Hell models, Hell 72 GL, and Hell-80 discussed elsewhere on this website. Below is an attempt to provide a description (with varying levels of detail) of those that apply the spindle-principle.

If you have any additional information or documentation that would be appropriate for this particular page, please contact me.

Hellschreibers that were actually built by the Hell company are discussed on this page.

Note that all major teleprinter manufacturers (e.g., Siemens, Lorenz, Creed, Morkrum-Kleinschmidt/Teletype) produced teleprinters that print on paper tape, but that are not Hellschreibers. However, some did develop and manufacture Hellschreibers. They are discussed on my "Other Manufacturers" page.

[T empf 39d "L"]       [T empf 40 "F"]       [T empf  44 ]      [Teletypewriter speed & distortion tester]

[Blattschreiber]         [FuG 120 "Bernhardine"]

The T empf 39d (a.k.a. model "L") dates from the early-mid 1950s. It is a printer-only model, used to print out messages sent with a Siemens-Hell-Schreiber T typ 72 "GL" or T typ 73 "AGL". Hence, it operates in the start-stop (synchronous) mode, with either 1000 Hz or 3000 Hz tone pulses (selectable). These machines were used by the German railroad (UK: railway) system in the Neanderthal region (i.e., near Düsseldorf) until ca. 1995 (ref. 18). The machine looks similar to models 40a "F" and 44c/e - but the "L" has an indicator light on the front, with a square lens.

An early model "T empf 39", with rounded housing
 

A later model "T empf 39", with re-styled housing

CHARACTERISTICS (source: p. 26 in ref. 8):

	Width of printed text: 2.5 mm
	Width of printed text with space: 3 mm
	Width of paper tape: 9.5 mm (3/8 ").
	Printing speed: 6.1 characters/sec (300 Bd)
	Paper consumption (max speed, continuous printing): 1.1 m/min
	Apparent input impedance: 600 ohm
	Minimum input signal level: - 3.5 Neper (-30 dB) at 600 ohm
	Power input: 220 Vac (40-60 Hz)
	Power dissipation: ca. 50 VA
	Fuses: 500 mA and 250 mA (amplifier)
	Motor dissipation: 25 W
	Motor speed (regulated): 2100 rpm
	Tone filter: 1000/3000 Hz (selectable)
	Amplifier tube (valve): EF80
	Keying tube: EL83
	Size: 32x26x20 cm (WxDxH, ≈12¾ x 10¼ x 8")
	Weight: ca. 11 kg (≈24.3 lbs

The start-stop synchronization accommodates up to 2% speed difference between sender and receiver/printer. The motor is a brushless DC type. Its speed regulator is based on a 125 Hz tuning fork. A strobe disk allows visual checking of the speed. The unit has an output (marked "Mh" = "Mithören") for a 4000 ohm headset. The unit responds to a remote control start command: a pulse of at least 180 msec (1 sec typ.) of 25 Hz, 1000Hz, 3000 Hz, or DC. The remote stop command is a 4 sec (min, 6 sec typ.) pulse.

	Ref. 6: "Siemens-Hell-Schreiber "L" - T empf 39 d", Siemens Fernschreib Technik, Siemens & Halske AG, Wernerwerk für Telegrafen- und Signaltechnik, Beschreibung St Bs 1223/11 [incl. schematic], July 1957, 32 pp. (courtesy Heinz Blumberg, DC4GL)
	Ref. 18: "Wann wurde eigentlich der letzte Hellschreiber außer Betrieb genommen??" ["When was the last Hellschreiber taken out of service??"], 2008 thread in the "Historische Bahn" forum of "Drehscheibe" ("Turntable) website of railroad enthousiasts.

As the type designator "T empf 40" suggests, this Hellschreiber is an "Empfänger", i.e., printer-only. It is for asynchronous operation ("quasi-synchronous") like the Feld Hell (though at twice the speed). Hence, the printer spindle has a 2-turn winding, and two parallel lines of text are printed onto the 15 mm wide paper tape. Actually, the "F" can be thought of as the replacement for the "Presse Hell" (T empf 14; 7x14 pixel font) - fully compatible, and with a built-in tone detector and keying device; it can simply be connected to the loudspeaker output of any radio receiver. This model was used with Siemens-Hell-Geber "S" (T Send 62a) and  Siemens-Hell-Geberanlage "Z". The two-stage amplifier uses an EF80 and an EL83 tube.

The machine looks similar to models 39d "L" and 44c/e - but the "F" is the only model with a toggle switch in the lower right-hand area of the front.

The Siemens-Hell-Schreiber "T empf 40 a" - "F"

(source: ref. X)

Modular design of  the "T empf 40 a" - "F"

(source: fig 8 in ref. X)

Print-out of  a "T empf 40 a"
(courtesy R. Lampe)

Ref. X: "Siemens-Hell-Schreiber "F" - T empf 40 a", Beschreibung St Bs 1221/3, Siemens & Halske AG, Wernerwerk für Telegrafen- und Signaltechnik, February 1960, 14 pp., SuW 6165 R 260.0,2  (courtesy Ralf Lampe)

CHARACTERISTICS:

	Transmission rate: 5 or 5½ characters per second (selectable)
	Telegraphy speed: 245 Bd or 269.5 Bd (5 vs 5½ cps)
	Modulation: On-Off-Keyed single-tone of 1000 Hz
	Size of printed characters: 2.5 x 4.5 mm (WxH).
	Paper tape width: 15 mm (like the Feld-Hell and Hell-80)
	Printer helix: 2-turn (like the Feld-Hell and Hell-80)
	Ink roller: same as Feld-Hell, 72 "GL", and Hell-80
	Motor: single-phase induction motor with electronic regulator; the speed control knob under the lid of the unit, next to the roll of paper tape.
	Nominal motor speed: 1900 or 2100 rpm (5 vs 5½ cps; selectable)
	Input bandpass filter: center frequency 1000 Hz. Filter bandwidth is 700 Hz.
	Line impedance: 600 ohm (standard POTS telephone land-line) or 4000 ohm (radio)
	Minimum input signal: 10 mV into 600 ohm
	Paper tape speed: 1.05 or 1.15 m per minute (5 vs 5½ cps)
	Power: 220 Vac, 40-60 Hz, 45 VA
	Remote control: tone pulse of 1 sec (paper transport "on") or 6 sec ("off")
	Accessory: "Bandaufwickler" (winder for the printed tape), switched on/off via the remote control.
	Dimensions: 32x26x20 cm (WxDxH)
	Weight: 12.6 kg (≈28 lbs)

The T empf 44 is basically a T empf 39 receiver/printer, expanded with a character drum for a limited character set (0-9 and A-G). These units were used in telephone exchanges, for the purpose of recording fault codes. There is no keyboard or punch tape reader. Instead, there is a dedicated control input for each selectable character. Upon receipt of a fault signal, the associated character is printed locally. Compared to model 44c, model 44e has an additional contact, that allows the selected character to be transmitted to a T empf 39 in the local telephone exchange, or elsewhere. These models date back to the 1960s. The housing of model 44c has rounded corners and edges, whereas model 44e has right-angles .

CHARACTERISTICS:

	Characters set: 0-9 and A-G
	Transmission speed: 6.1 characters/sec - Model 44e only
	Telegraphy speed: 300 Baud - Model 44e only
	Audio frequency of sent pixel pulses: 1000/3000 Hz (selectable; to be confirmed) - Model 44e only
	Power: 60 Vdc (possibly also 48 Vdc (4x 12 V), to be confirmed. Telephone exchanges normally has a central 48 Vdc battery that is trickle-charged from the main Vac power. Different countries uses different DC voltages, but they are normally 36, 48, or 60 Vdc.)
	Dimensions: 31½x24x19½ cm (≈12x9½x7½")
	Weight: 9.3 kg (≈20.6 lbs)

Ref. Y:  “New International Telephone Switching Center in Düsseldorf” in “NTZ-Communications Journal“ (NTZ-CJ), No. 3, 1967, p. 146, 147  [T empf 44c]

The Siemens-Hell-Schreiber "T empf 44c" stacked on top of a T empf 44e"
(photo courtesy Heinz, DC4GL)

The Siemens-Hell-Schreiber "T empf 44c"

The inside of a Siemens-Hell-Schreiber "T empf 44c"
(photo courtesy Heinz, DC4GL)

"T empf 44c" installed in a German international telephone exchange (1967)

(source: ref. Y above; for complete photo, click on image or here)

"T empf 44c" installed in an APrE 50 test rack

(APrE = automatische Prüfeinrichtung = automatic test set)

(photo: J. Hartl)

"T empf 44" installed in an RPE test rack of the telephone exchange in Wien-Fünfhaus (Austria)

(RPE = Routineprüfeinrichtung = automatic test set)

(in private collection of N. Dulosy in Laa a/d Thaya, Austria)

As we all know, Hellschreiber printers (other than start-stop models) are really simple and "dumb": the printer blindly prints all received pulses, without any interpretation. Hellschreiber characters are pixel (bit) sequences with a fixed-length. Each character of the alphabet is mapped to a particular sequence, but is not encoded. This is not all that different from teletypewriters ("telex"). Here, each character of the alphabet is typically encoded as a sequence of a fixed number of bits, preceded by a start-pulse and followed by a stop-pulse. Again, all characters have the same length.

In case of the Baudot code (e.g., the CCITT-2 or ITA-2 alphabet), there are five data-bits, one start-bit, and stop pulse with the duration of 1-2 bits (typ. 1.5 bits). That is, 7.5 bits in total. A character that is sent with a telegraphy speed of 50 baud has a duration of 1000 / 50 = 20 msec per bit, and a total duration of 7.5 x 20 = 150 msec.

In the end, both Hellschreiber and teletypewriter characters are fixed-length pulse sequences. Hence, a Hellschreiber printer can print the pulse sequence of any teletypwriter - as long as the printer solenoid is fast enough. For standard 7-column Hellschreiber font, the solenoid can print pulses of 8.16, 4.08, 3.7, or 3.3 msec (at standard 2½, 5, 5½, 6 and 6.1 chars/sec respectively). printing individual 20 msec teleprinter pulses is no problem whatsoever.

A single column of the Hell-font is printed every 57.1, 28.6, 23.8 or 23.4 msec. As stated above, a 50 Bd character (with a 1.5 bit stop pulse) has a duration of 150 msec. If the spindle of a Hell-printer is slowed down enough, each printed column will represent exactly one complete teleprinter character. The printed columns will be vertical, if both Hell-printer and teletypewriter are running at exactly the same speed. Note that that printer should print a single line, not two identical parallel lines. Hence, the spindle only has one thread that makes one winding.

If the speed of the Hell-printer is calibrated, it can be used to verify the speed of a teletypewriter - like a simple electro-mechanical signal analyzer! Such Hellschreibers are called "Drehzahlfehlerschreiber", i.e., an "rpm-error printer". Note that speeds can also be checked with tuning-fork and stroboscopic disk methods.

Printed tape segments from a Hellschreiber "rpm-error printer"

(source: figure 538 in ref. A)

As illustrated in the figure above, the angle between the printed columns and the edge of the paper tape is directly related to the rpm-error. The angle is measured with a transparent template, overlaid on top of the paper tape. The practical rpm-error range that can be measured with this method is +/- 5%. For a printer with a synchronous AC-motor, the measurement accuracy depends on the accuracy of the public power grid frequency (50/60 Hz), assumed to be within +/- 0.3%. The allowed error is +/- 0.75% per CCITT standards, though typically +/- 0.5% is applied.

Siemens-Halske Hellschreiber "rpm-error printer"

(original unedited photo: courtesy B. Rothe)

Equipment label of the Siemens-Halske Hellschreiber "rpm-error printer"

Power supply and other components of the S-H "rpm-error printer"

(original unedited photo: courtesy B. Rothe)

Synchronous motor and gear train of the S-H "rpm-error printer"

(original unedited photo: courtesy B. Rothe)

A Hellschreiber-printer can also be used to check the characteristics of telex networks. Such a special printer is called a "Verzerrungsschreiber", i.e., a "distortion recorder". Here, the spindle speed is the same as the baud rate: e.g., 50 rps (3000 rpm) for 50 Bd. Each printed column now has the duration of a single bit, rather than a complete character. For this to work, the printer must also be start-stop. Upon receipt of a start pulse, the printer spindle makes a fixed number of revolutions. The concept is illustrated in the figure below.

Concept of a "distortion recorder", based on a Hell-printer

(source: figure 528 in ref. B)

This particular printer is not based on an inked spindle and a solenoid-driven printer blade. Here, a stationary blade is used, as well as special paper tape. The white paper is coated blue on top. The arrival of a tone pulse is transformed to high-voltage pulses. The voltage is applied between the blade and the spindle. The high-voltage pulses cause arcing that obliterates the blue coating of the paper. Thus, each pulse causes a white pixel to appear. Note that the pixel is associated with the leading edge of the received pulse; the size of the pixel is not related to the duration of the pulse. The paper transport speed is very low: 1.8 m/hr, compared to 28 m/hr in a Feld-Hell machine. When the printer is operated over a longer period of time, a distribution of pixels is printed, that should be (tightly) centered around the middle of the paper tape. A 175 Hz tuning fork is used to calibrate the speed of the printer motor. 

 

	Ref. A: "Geschwindigkeitsmessungen", pp. 781-784 in "Fernmeldetechnik", Band 9 of “Lehrbücher der Feinwerktechnik“, Fritz Schiweck, 4th ed., 1962, C. F. Winter'sche Verlagsbuchhandlung, 894 pp.,
	Ref. B: "Fernschreib-Verzerrungsschreiber (Bezugverzerrungsschreiber)", pp. 768-770 in "Fernmeldetechnik", Band 9 of “Lehrbücher der Feinwerktechnik“, Fritz Schiweck, 4th ed., 1962, C. F. Winter'sche Verlagsbuchhandlung, 894 pp.

During the first half of 1948, the Hell company developed a "Blattschreiber" based on the Hellschreiber principle; series-production by Siemens-Halske began in 1949 (ref. 9). I don't know what its Siemens-Halske model name/number or type designator is.

One known user of Blattschreiber was the Norsk Telegrambyro (NTB, Norwegian Telegram Bureau). They used the "Presse Hell" system" since the German occupation during WW2, and later switched over to Hell Blattschreiber (cf. p. 114 in ref. 11).

If you have any additional information or documentation on this model, please contact me!

The Siemens-Hell-Blattschreiber

(source: figure 8 in ref. 8; also figure 56a in ref. 13)

As the name suggests, this is a Hell page-printer instead of a Hell tape printer. Tape printers have the inconvenience that the printed tape has to be cut up and glued onto telegram forms or other sheets, or be transcribed manually or with a typewriter. The competition (regular teleprinters) had already adopted the page-printer approach decades earlier. Also, the teleprinters print single lines of text, instead of two identical lines, as typical of the pre-1950s Hell tape printers.

In principle, one can conceive a page-printer that uses the Hell column-printing concept: the printer-spindle is mounted on a carriage, and the carriage is moved across the width of the paper sheet. The sheet of paper is nothing more than series of paper tape segments. This is what Rudolf Hell actually proposed in his patent 825277 (filed in 1948).

The Hell-Blattschreiber as proposed in patent 82577 - item 6 is the printer helix

(source: figure 1 & 2 in the patent)

The main components of the proposed printer are:

	a carriage (item 5) with printer helix (item 6) and felt ink roller (item 10). The thread of helix only makes one turn, as only a single line of text is to be printed - not two identical parallel lines.
	a T-shape printer bar ("Magnetleiste", item 4 in Figure 2). The bar is as long as the width of the paper sheet, and as wide as the height of the characters that are to be printed. It is actuated by "one or more" printer-magnets (item 9). A paper-guiding surface ("Führungsschiene", items 2 & 3) in front of and behind the T-bar ensures that the paper is flat and taught, when and where it is printed.
	a motor that drives the printer helix via a drive shaft (item 7); the coupling gear-wheel (item 19) turns with the drive shaft, but can slide freely along the length of that shaft. The same motor also drives a long spindle shaft (jack screw, item 16); the carriage has a small lever arm (item 7) that rests in the groove of this drive shaft.
	an electromagnet (item 23) that can decouple the carriage lever arm from its drive shaft. When decoupled, the carriage is immediately pulled back to its starting position at the left-hand edge of the paper, by a spring-loaded pull cord (item 25, 26).
	a chain form or roll of paper (item 1). The paper is either transported continuously, or a line-feed is performed at the end of each printed line of text (i.e., a combined carriage-return/line-feed, CR/LF). Rudolf Hell proposed to somehow execute the carriage return during a space between characters or words (claim 3 in patent 825277), or use a special carriage-return signal ("Zeilenwechselzeichen" pulse with the duration of a single character, claims 5 & 6 in patent 825277). The same pulse could also be used to enable the printer.

As stated above, this is the originally proposed embodiment of the Hell-Blattschreiber. It was never turned into a working commercial product. In a follow-on patent (848970) from 1951, Hell recognizes difficulties with the mechanical construction, due to the relatively large dimensions of the printer-head, combined with the very short time allowed for carriage-return. Solution of this problem requires a special control character for commanding a carriage-return/line-feed. Also, in his 1961 patent nr. 1157258 he actually states (column 3, lines 46-59) that all efforts to make a page-printer with a helix-carriage have so far been without success: "... sind alle Versuche, das Schreibsystem des Streifenschreibers in der geschilderten Weise für einen Blattschreiber zu verwenden, bisher gescheitert." Primary difficulties (besides designing a sufficiently fast printing relay) are:

	the inertia of the carriage (esp. noticeable during carriage-returns),
	feeding the advancing paper sheet correctly between the spindle and printer-hammer while the carriage moves across the paper (effectively, the paper simultaneously moves in two perpendicular directions with respect to the printer-head).

So, back to the drawing board and back to basics. Recall that the Hell tape-printers only use a helix, because it is the simplest mechanical means to obtain a small striking-surface that scans across a paper tape. As the helix turns continuously, the scanning movement is repeated continuously. At the same time, the paper tape continuously moves by the spindle. It is this basic concept of continuous scanning action and moving the paper relative to the striking surface (or vice versa), that is essential - and not necessarily the implementation with a helix.

This is exactly what Hell did! He actually already foresaw a Blattschreiber 20+ years earlier, in his 1933 patent 668821. An essential part of this approach is the platen: a large rubber-covered steel cylinder, as used in standard typewriters and in some types of conventional teleprinters. The rubber of the platen cushions the striking force of the printing hammer, type wheel, or typewriter type - as the case may be. This provides a cleaner print and reduces wear. The platen has a fairly large diameter to provide a local striking surface that is relatively flat.

But there is a critical difference: in typewriters and teleprinters, the paper sheet is wrapped around the platen and in constant contact with it. That is, the platen is part of the paper transport mechanism. In the Hell Blattschreiber, this is not the case! Recall that in Hell tape printers, the paper moves freely between the printer hammer and only touches the rapidly spinning printer helix when the hammer is actuated. The paper is slowly advanced with motorized pinch rollers. The exact same concept is applied in the Hell Blattschreiber. 

Blattschreiber as proposed in a Hell patent from 1933

(source: figure 4 in patent 668821)

In the 1933 patent, a standard Hell printer head (magnet + armature + hammer) is moved across the paper sheet. The hammer is oriented tangentially to the platen, aligned with the direction of the paper. See item 7 in the figure above. An inked ribbon or carbon tape is fed between the hammer and the paper.

In the "real" Blattschreiber, we do not want to print two identical lines of text. Not a problem, just make the height of the hammer blade less than the distance between the ribs! Now the hammer can only strike one rib at a time.

At this point, we still need to solve the mechanical problems with the carriage-return. Hell's patent from 1948 (848970) provides the solution. The continuous scanning action across the paper is obtained by mounting the hammer blade ("Schreibfähnchen", "Schreibschneide") on a continuous belt or chain. The belt moves in a race-track pattern. The length of the straight sections of this pattern is slightly smaller than the width of the paper sheet. The pattern is placed across the paper sheet, parallel to both the sheet and the ribbed platen. The speed of the belt is the same as the paper transport speed of Hell tape printers.

Blattschreiber as proposed in the Hell patent from 1948

(source: figure 1 in patent 848970)

If we use a belt with only a single hammer, then the hammer will still have to do a "carriage return" at the end of a scan, and instantaneously fly back from the right-hand edge of the paper to the left-hand edge. This is impossible. On the belt, the blade actually has to move all the way around in the pattern, to end up at the far left of the paper again. No problem: just add a second hammer to the belt and place it opposite the first hammer. When the first hammer is at the end of its scan, at the far right of the paper, the second hammer is ready to start the next scan at the far left of the paper. Well.... no! The overall length of the race-track pattern is always larger than the two straight segments of the pattern combined. In other words: the belt is longer than twice the printable width of the paper sheet. But the distance between successive blades must be equal to that printable width. This problem is solved by using three blades. This is what has been done in the Blattschreiber (see items 7, 8, 9 in the figure above). Of course, we could use four blades (ref. 1949 Hell patent 832444 and 1950 US patent 2656240), but then the belt becomes unnecessarily long (the distance between blades is fixed), and more difficult to fit inside a printer. Note that the belt speed is independent of the number of blades.

Blattschreiber with four hammer blades

(source: figure 4 in Hell's US patent 2656240; similar to figure 1 in 1949 Hell patent 832444);

The printer blade has to be tapped against the inked ribbon and the paper, in the rhythm of the pixels that are to be printed. We could mount three solenoids on the belt, one behind each blade. But the wires to the solenoids would get twisted and destroyed within several turns of the belt, unless a cumbersome and unreliable slip-ring were used. It is much easier to mount two solenoids in a fixed position, one at each end of the scan pattern. Via simple levers, they actuate a hammer bar. The bar is installed along the belt, behind the scan-path of the blades. Voilà, the essence of the Hell Blattschreiber printer!

Simplified mechanism of the page-printer with 3 hammers

There are two options for advancing the paper sheet:

	advance in one step, at the end of a printed line of text. The step size is the height of the font (including blank pixel-rows for line spacing).
	slowly advance continuously. The speed is such that the above step height is achieved at the end of the text line.

In the first case, the printed line of text will be straight across the paper, perpendicular to the side of the paper sheet. Fine. However, this approach requires an additional actuator, such as a stepper motor. This violates the "single motor" approach... 

The slanted installation of the printer mechanism is clearly visible

(source: figure 9 in ref. 8)

The slanted installation of the printer mechanism is clearly visible

(source: figure 56b in ref. 13)

Right hand view of the opened Siemens-Hell-Blattschreiber

(original photo: courtesy Jan Smeets, ON4ASZ)

Left hand view of the opened Siemens-Hell-Blattschreiber

(original photo: courtesy Jan Smeets)

Bottom view of the opened Siemens-Hell-Blattschreiber

(original photo: courtesy Jan Smeets)

PHASE CORRECTION

As stated above, the Blattschreiber printing mechanism is in principle compatible with the standard Hell-system (at 5 chars/sec). However, only a single line of text is printed, not two identical parallel lines. To be able to use the Blattschreiber with existing standard Lochstreifensender (punch tape reader + Hell-format sender), the Blattschreiber must somehow synchronize itself to the received pulse stream. If not, the text line will be slanted an become illegible, as with a hell tape printer and a single-turn spindle:

Downward slant for receiver that is slower than the transmitter
 

Contrary to some publications, the Blattschreiber does not use start-stop synchronization at character level. However, it does use the pulses of received characters to detect a phase difference between the printer and the sender. Recall that the top and bottom rows of the Hell-font are blank. The Blattschreiber "knows" the momentary scan position of the rib that is in front of the printer hammer. If this position corresponds to a pixel in the top or bottom row of a character, then no pixel-pulse should be received. If, however, a pixel-pulse is received during this time, then there is a phase difference between the sender and the Blattschreiber. The phase difference is eliminated by slightly increasing or decreasing the printer's motor speed.

Two cam wheels are used to detect if the rib in front of the hammer is at the bottom or top pixel-position of the current column scan. This is indicated by notches N₃ and N₄. The cams are on a shaft in the platen drive-train. The shaft makes one revolution per column scan (5 char/sec and 7 columns/char —> 28.57 msec/rev :: 2100 rpm). Each notch closes a contact for about half the duration of a single pixel (4.08 msec). If a pulse is detected while N₃ is closed, then the local motor is ahead of the sender's motor. Conversely, if a pulse is detected while N₄ is closed, then the local motor is lagging the sender's motor. In other words, notches N₃ and N₄ are used to determine the direction of the required phase correction.

Cam wheels indicate the rib position 
 

There is a second pair of cam wheels on the same shaft, with notches N₁ and N₂. The associated contacts close for 10 msec (≈2½ pixel durations) per revolution. N₁ closes just after N₄ opens, whereas N₂ opens just before N₃ closes. When N₁ is closed, a capacitor (C₃ in the diagram below) is positively charged during pixel pulses that arrive during that closure. Conversely, that capacitor is negatively charged during pixel pulses that arrive during closure of N₂. The resulting capacitor voltage reflects the (vertical) symmetry of the received characters. When the voltage is sufficiently positive or negative ( = significant asymmetry), it energizes a bi-directional relay (G in the diagram below). This relay enables the "direction" contacts of the notches N₃ and N₄. The printer pulses that arrive during closure of N₃ and N₄ now energize a second bi-directional relay (F). Its contacts temporarily alter the speed set-point of the motor by plus 1.4% or 1.4% (via relays L ("langsam" = slow) and S ("schnell" = fast)). The (filtered) control pulses are also shown on a lead/lag indicating needle instrument in the front of the Blattschreiber.

Cam wheels indicate the rib position 

(source: fig. 293 in ref. 7b)
 

So much for synchronization at character level. A bit complicated, as it does not involve sync-pulses or sync-characters sent by the Hell sender. It does meet the stated objective of being able to use existing Hell senders - without modification.

The Blattschreiber printer is dimensioned to print lines of 69 characters. Without additional measures, the 69^th character of a line would typically be printed partially at the end of one line, and partially at the beginning of the next line. One way around this, is to reduce the distance between the printer hammers by the width of one character (ref. lines 23-30 in Hell patent 848970). The 69^th character will then always be printed in full at the beginning of the next line, even if it is only printed partially at the end of the preceding line. Two hammers are actuated simultaneously. Note that this trick is also applied by all software-implemented Hellschreibers - many decades after the Blattschreiber!

Clearly, partial characters and/or repeated characters do not look professional on a printed sheet, and are disturbing. This cannot be fixed without some form of synchronization at text-line level ("Zeilensynchronisierung"). To this end, a special "start/stop" character is used. It is simply a long pulse, with the duration of an entire Hell character. The leading edge of the start/stop pulse is used to start the motor. In the Blattschreiber, the trailing edge is used to engage the electromagnetic clutch that couples the belt with the printer hammers (and the platen), to the drive shaft. In the sender, the trailing edge is used to start transmission of a new text line. Upon receipt of the stop-pulse, the clutch is disengaged when a printer hammer is exactly at the far left edge of the next text line.

Complete printer of the Siemens-Hell-Blattschreiber

(source: Figure 291, p. 369 in ref. 7b; also figure 2 in ref. 8)

This line-sync approach only works if the sender asserts the start/stop pulses, and sends strings of no more than 69 characters. To be able to do this with existing Hell punch-tape senders ("Lochstreifensender"), a special control unit ("Steuergerät") is required. It interacts with the Lochstreifensender, in order to ensure the correct maximum string length, and to insert the start/stop pulses.

"Steuergerät" for the Blattschreiber

(source: figure 58 in ref. 13)

Print-out without line synchronization

(source: figure 6 in ref. 8)

Print-out with line synchronization - all text lines are left-justified

(source: figure 7 in ref. 8)

The Blattschreiber is designed for 5 characters per second, i.e., a telegraphy speed of 50 WPM. This implies almost 14 sec for a full line of text (69 characters). Assuming a Hell font with 7 columns per character, there are 69 x 7 = 483 pixels per printed matrix line. In his Swiss patent 378940 of 1964, Hell proposes to increase the printing speed to that of fax machines: about 3000 pixels per sec. A Blattschreiber row-line comprises nearly 500 pixels (69 characters per line, seven row-pixels per character). Each character font uses five printable pixel rows (not counting the blank two line-spacing rows). I.e., 5/6 sec per line instead of 13.8 sec: almost 12 times as fast (i.e., 60 cps). He proposed to do this with a fully electronic control unit (with magnetic core buffer memory), for row-scanning entire lines of text.

Dr. Hell conceived a number of Blattschreiber printer implementations, with various technologies.

His patent 1157258 describes a two-step method, in which text is first printed column-by-column on an endless carrier tape with a conventional inked printer helix. The only difference is that text is printed in mirror image. The printed tape passes from left to right in front of the paper sheet. When an entire line of text is in front of the paper, a hammer bar (item 12 in the figure below) taps against the tape. This causes the entire line of text to be transferred onto the paper. The tape is wiped clean after printing.

A line-printing Blattschreiber

(source: figure 1 in Hell's patent 1157258)

In an similar two-step method, text is first electrostatically "printed" column-by-column on an endless carrier tape, and then transferred to paper, where the carbon particles are heat-fused. The tape is wiped clean after printing. This technology is commonly used in mimeographs ("xerographic" copying machines).

Line-printer, using xerographic technology

(source: Hell's patent 1178459)

Hell's US patent 2819941 proposes a single-thread multi-turn printer helix that is installed across the paper. The helix is inked with a long ink roller, or inked ribbon is used. An endless belt or chain is installed behind the paper. One or more hammer blades are attached to it. A hammer bar extends across the paper, and can actuate the blade that sweeps behind the paper. The hammer blade and the thread of the helix sweep across the paper at the same speed. As with the "real" Blattschreiber, this design prints a single line of row-pixels across the paper.

Many-turn spindle across the paper

(source: figure 1 in Hell's US patent 2819941)

Hell's 1950 US patent 2656240 proposes a Blattschreiber with a column-printing mechanism. In this concept, there is a single printer hammer that scans across the paper and then makes a spring-assisted carriage-return. This is not essential, and the endless belt with three hammers would of course also work. The equivalent of a spinning helix is placed behind the paper. It consists of a stack of five thin bars (items 5-9 in the inset Fig. 3 below). This corresponds to five printable pixels per character column. The bars are installed across the paper. The bars are actuated in sequence, by two turning shaft that are placed at opposite ends of the bars (items 28 and 29). The shafts are are equipped with five cams (items 25). As the shafts turn in unison, first the lowest bar is pushed forward against the paper, then retracted while the bar immediately above it is extended, etc. Clearly this creates a continuous column-scanning action. The blade of the printer hammer has the width of the pixels that are to be printed. The height of the blade is that of a full character. When the hammer is actuated, it taps an inked ribbon against the paper, and against the extended bar.

Blattschreiber with column-printing printer head

(source: figure 1-3 in Hell's US patent 2656240)

REFERENCES

	Ref. 7: "Die Schreibenden Drucktelegraphen" [47 MB], Fritz Schiweck, section 9.5 (pp. 358-378) of "Fernschreibtechnik", Band 9 of “Lehrbücher der Feinwerktechnik“, 4th ed., 1962, C. F. Winter, 894 pp. Ref. 7a: Section 9.5.2.2 (pp. 360-363) Ref. 7b: "Siemens-Hell-Blattschreiber", Section 9.5.2.6 (pp. 369-370) Ref. 7c: pp. 388, 389 in "Faksimiletechnik", §10 (pp. 381-392) of "Fernschreibtechnik", Fritz Schiweck, Band 9 of "Lehrbücher der Feinwerktechnik", 4th ed., 1962, C. F. Winter, 894 pp.
	Ref. 8: "Der Blattschreiber für das Siemens-Hell Verfahren", R. Zimmermann, Entwicklungsberichte der Siemens & Halske Aktiengesellschaft, Jg. 15, 25 May 1952, pp. 9-12 (also appeared in VDE-Fachberichte, Vol. 15, 1951)
	Ref. 9: p. 10 in "Die Firmenneugründung der Firma Dr.Ing. Rudolf Hell GmbH in Kiel 1947 - Interview zwischen Christian Sütel und Christian Onnasch, gehalten am 1. September 2001" [post-war re-start of the Hell company in Kiel], 12 pp.
	Ref. 10: "Fernkopierer", Adalbert Kukan, in "Kultur & Technik", Nr. 2, 1988, pp. 112-115
	Ref. 11: pp. 40, 42, 48, 49, 51, 54, 56, 58, 64, 65, 76, 82, 84, 85-90, 94, 95, 98, 99, 102, 103, 110, 113-116, 118, 19, 122, 123, 128-130 in "News agencies. Their structure and operation.", UNESCO, 1953, 208 pp.; reprinted by Greenwood Press, 1969, 207 pp., ISBN: 0837125014
	Ref. 12:"Anwendung von Faksimile-Hellschreibern für den Boden-Bord-Verkehr" [Hell fax, Zetfax], Rudolf Hell, pp. 67-75 in "Flugsicherungs-Verfahren und -Technik: Grundsätzliches über die Technik für Flugsicherung und Betrachtung der wichtigsten technischen Hilfsmittel außer Radar“, Teil IIIB in "Flugnavigation und Flugsicherung“, Band 7 of "Bücherei der Funkortung“, proceedings of the "Internationale Jahrestagung des Ausschuß für Funkortung“, Berlin; Verkehrs- und Wirtschafts-Verlag, 1958, 107 pp.
	Ref. 13: p. 67-68 in "Siemens-Hell-Geräte", pp. 61-77 in "Telegrafentechnik", Band 6, Teil 6 of "Der Dienst bei der Deutschen Bundespost - Leitfaden für die Ausbildung", Fritz Schiweck (ed.), R. v. Decker's Verlag, G. Schenck, 1960, 970 pp.

REFERENCES

Below is a listing of patents from Rudolf Hell (or his employees) related to the Hell-Blattschreiber.

Patent office abbreviations:

	BD = Bundesrepublik Deutschland, Deutsches Patentamt
	DDR = Deutsche democratische Republik, Amt für Erfindungs- und Patentwesen
	CH = Swiss Patent Office (Schweizerische Eidgenossenschaft - Eidgenossisches Amt für geistiges Eigentum)
	US = United States Patent Office

Patent source: DEPATISnet, the on-line public database of the Deutsches Patent- und Markenamt (DPMA, German Patent and Trademark Office).

The above Hell Blattschreiber model is not to be confused with Blattschreiber fax machines that Hell developed years later ("BS" models such as BS109, BS110, BS112, BS116, BS133, BS134, BS1035, WF103, WF104; ref. 7b, 7c, 10), and regular (non-Hell) Siemens-Halske page-teleprinters.

Hellfax-Blattschreiber model BS109 and BS116- these are not "Hellschreibers"
(source: figure 306 and 307 in ref. 7b, 7c above)

Here are some of Dr. Hell's patents related to Hell-fax printers/plotters:

The Luftwaffe also used Hellschreiber technology for aerial navigation, in particular to assist fighter aircraft with the intercept of enemy bombers. This system comprised the Bernhard beacon transmitter ground station with a large rotating antenna, and its airborne counterpart: the FuG 120 UKW-Richtstrahl-Drehfunkfeuer Empfänger "Bernhardine" Hell-printer. It printed the bearing information from the station on a paper strip. The system went into operational service mid-1944.

Basic characteristics:

	Frequency: 30 - 33 MHz
	Maximum range: 120 km (75 mi), at a flight altitude of 500 m (≈1600 ft). Ref. 27. 400 km (250 mi), at a flight altitude of 5000 m (≈16000 ft)
	Accuracy: initially +/- 1°, later improved to +/- 0.5°

The original concept (ref. 1936 patent 767354 in the table below) comprises a primary and a secondary beacon signal. The primary signal is a constant carrier that is transmitted via a rotating directional antenna with a narrow beam ("Richtstrahl-Drehfunkfeuer"). The beam pattern of the primary antenna has two lobes that are slightly offset (20º in the figure below), such that there is a deep, steep null between them. The secondary signal marks passage of the directional beam through the reference direction (typically magnetic north). This signal is transmitted via an antenna with an omni-directional pattern. In a subsequent update to this concept, the omni-directional reference signal is replaced with a second rotating directional signal. Its beam pattern is at maximum where the primary beam has its null between the two lobes.

Antenna radiation patterns of the antennas for the primary signal (curves D, a, b) and the secondary signal (curves K, C)
(Left: figure 2 in patent 767354, right: figure 1 in patent 767523)

The associated receiver continuously records (prints) the signal strength level of the primary rotating beam. This is done with a simple Hellschreiber printer. The antenna beam pattern makes it easy to accurately detect the passage of the beam at the azimuth of the receiver, as shown in the diagram below.

Signal strength received from the rotating constant-carrier directional beam vs. time
(annotated figure 4 in patent 767354)

The printer consists of an inked spindle that is placed across, and slightly above, a moving paper tape. Below the paper tape is an electromagnet with a hammer. When the magnet is energized, the hammer pushes the paper tape against the continuously turning spindle. This causes a line segment to be printed across the paper tape. The length of the printed line depends on the amount of time that the electromagnet is energized. See the "How it works" page. The signal strength is determined by rectifying the received tone. This signal level is then converted to the width of an energization pulse that is fed to the printer's electromagnet. The result is a bar graph of that signal strength.

Signal strength bars, as printed with the Hellschreiber
(source: ref. 15)

As stated above, the printer spindle turns continuously. Only a single bar graph is printed. So, the spindle has a single thread (1-start) that makes a single turn. This implies that the leading edge of the energization pulse is aligned with the spindle position (the starting point of the winding, to be precise). I have not yet been able to figure out how this was done, as only a constant tone was transmitted (no start-pulses).

The received signal strength depends on the distance from the ground station. Subsequent patent 767526 foresees an automatic gain control in the receiver, using yet a third signal transmitted by the ground-signal.

In the original concept, a second printer spindle - with its own electromagnet - records the reference marker signal on a second paper tape (standard 15 mm wide Hellschreiber printer tape). The bearing from the ground station is determined from the time ( = distance on the print-outs) that elapses between the reference "tick", and the null of the rotating beacon signal. Accuracy requires constant rotational speed of the rotating beam, and accurate timing at the receiver. A first improvement to this concept was to print both the signal strength and reference pulse on a single (but wider) paper strip. This still requires accurate timing at the receiver (paper transport speed). One major disadvantage remains: the operator must analyze and interpret the printed information from the two printed traces. Obviously this is not practical when being bounced around in a dimly lit and maneuvering aircraft.

To avoid a paper tape mess in the cockpit, the paper tape is unwound from one spool, and wound onto second spool (like in a reel-to-reel tape recorder). To minimize paper consumption (and the hassle of having to replace rolls of paper tape in flight), the paper transport motor was only enabled when the strength of the secondary signal exceeded a certain minimum value. I.e., a "squelch" function. This way, the paper moved during less than 10 sec per minute (3-5 sec per antenna revolution, equivalent to 36-60 deg change in antenna azimuth).

There were also ideas (patent 767513) to use an endless 3-layer printing tape (a sandwich of wax-covered dark paper, thin silk tape, and thin celluloid tape). The layers are joined just before entering the printer window. Where the printer-head applies pressure to the tape, the pressure points become visible. Erasing was done by separating the three layers after passing the printing window. This method was actually implemented and deployed; "Berhardine 1m mit Wachsschreiber", Telefunken 1936/37, ref. 14. It proved too complicated and unreliable, and was taken out of service. This technology actually dates back to ca. 1923, when it was patented (but not invented) by R.A. Watkins of Illinois/USA (later Watkins-Strathmore Co, until acquired by Western Publishing in 1957). The original implementation consisted of a piece of black waxed cardboard and a translucent grey plastic sheet. One would write or draw on it with a plastic stylus, not unlike today's Personal Digital Assistants (PDAs). This quick-erase slate was called "Magic Slate", and became popular all over the world. In the USA, one can easily patent pre-existing devices and methods of foreigners, so the Bernhardine Wachsschreiber was (re-)patented there in 2003 (US Patent 6,578,615) by the Hewlett-Packard (HP) company.

"Magic Slate" note pad or doodle pad
 

The Hellschreiber-printer is basically a remote dot-matrix printer. It does not decode or interpret the signal that it prints. Without any modification, it is fully capable of printing signal strength bars, text, and low resolution images. This versatility was used for a major improvement. So far, the secondary antenna is used for transmitting the reference marker signal. Now it is used to continuously transmit the momentary value of the azimuth of the directional antenna, as a sequence of binary pixels in Hell-format. The printed pixel stream is directly readable. This approach is already mentioned in the 1936 patent (767354): transmission of azimuth data via the principles of fax or TV ("Bildfunk oder Fernsehprinzip").

In Hellschreiber-format, the graphical information that is to be transmitted, is decomposed into a number of consecutive pixel columns. Each column is transmitted pixel-by-pixel, scanned bottom-to-top. A 360-degree compass card comprises a band of tick marks and numbers. This band is decomposed into columns of pixels as shown below. Consecutive columns are transmitted tail-to-head.

A compass card decomposed into pixels
(source: patent 767524)

Patent 767524 refers to the following pixel format (see figure above):

	columns of 24 pixels: the figure above appears to show a column height of 12 pixels. However, the characters 8, M, 1, and 9 show "half pixels", so effectively there are 24 pixels per column. note: the standard Hellschreiber "2-pixel rule" applies. I.e., within a column, each black or white element is at least 2 pixels high. The same applies at the transition of the top of one column, to the bottom of the next column.
	3 columns per degree: 3 x 360 = 1080 pixel columns total 1080 x 24 = 25920 black & white pixels total (12960 with "2-pixel rule").
	degree tick-marks. 1 pixel wide: 10-degree tick marks: 20 pixels high 5-degree tick marks: 10 pixels high 1-degree tick marks: 6 pixels high
	azimuth numbers: a matrix of 5x10 pixels (WxH)
	ground station identifier letter (e.g., "M" in the figure above): 5x10 pixels

The antenna of the Bernhard system made one full revolution every 30 sec. Assuming the above Hell-format (an assumption that I cannot confirm), this implies 25920 / 30 = 864 pixels per sec. Hence, a pixel duration of 1000 / 864 = 1.15 msec per pixel. Applying the "2-pixel rule" implies a minimum pulse duration of 2 x 1.15 = 2.3 msec. This is about 3.5 times shorter than the minimum Feld-Hell pulse duration (8.16 msec). The equivalent telegraphy speed is 430 Bd. I have not yet been able to confirm whether the radio transmitter, receivers, and printers of the Bernhard system could actually handle this speed (bandwidth). Note that this is different from standard Hellschreibers of that time: they operated at 112.5 or 225 Bd.

The pixel sequence of the entire compass card is captured as an optically encoded track at the circumference of a disk ("Kennzeichenscheibe"). The disk is mechanically synchronized to the shaft of the rotating antenna system. The pixel track passed between a light source and a photocell. The output of the photocell was used to key the Telefunken transmitter. The signal was input to the control grid of a negatively-biased keying tube. The constant-tone modulation was connected to an other grid of the tube. Patent 767529 discusses shaping the photocell signal by amplification and clipping, to obtain square keying pulses.

The disk was made of glass (p. 62 in ref. 42). I have no information on the size of the disk, nor if the engraved pixels were dots or lines. Patent 767524 suggests lines. Patent 767528 includes the cross-section diagram (shown further below) of what a disk attachment might look like. Let's suppose that the disk measured about 30 cm (12") in diameter - i.e., the size of a old 78 rpm vinyl LP phonograph record. The circumference would be about 94 cm, and the pixels a mere 70 microns (0.07 mm) long! This would suggest a photo-chemical process, rather than engraving. With a very large disk, say 1 meter diameter (≈40"), an individual pixel would still measure less than 0.25 mm (for the font assumed above). This would be feasible with engraving (as stated on p. 124 of ref. 43 and in ref. 42), but that would have been an extremely (and unnecessarily) laborious process, given the number of pixels.

Optical encoding disk (only part of the pixel track shown)
(source: patent 767524)

Patent 767524 also considers other implementations of the disk: a pixel track implemented as interconnected metal patches that are scanned with a slip contact (similar to the character drum of the Feld-Hellschreiber), or an optical pixel track with holes through the disk.

The Bernhard navigation system comprised a chain of ground stations (see further below). So the signals had to be identified. This was done by adding a station-identifier letters between the 10-degree numbers. The associated pixels could be implemented in several ways:

	on the same disk as the compass card data, integrated with the azimuth data track.
	on the same disk as the compass card data, but as a separate track.
	as a dedicated track on a separate disk, mechanically aligned and synchronized with the compass card disk. This has the advantage that the same compass card disk could be replicated for all ground stations, and the station identifier be put on a "personalized" disk.

In case of a separate track (on the same disk or on a separate disk), the output of the associated photocell is simply combined (logical "OR") with the output of the photocell of the azimuth track. Patent 767528 suggest that the third option was used: a stack of two optical disks. One disk with the azimuth tick marks and degree numbers ("Gradeinteilungen" & "Gradzahlen"), and a second disk for the station identifier letter ("Funkfeuer kennzeichnender Buchstabe").

Two stacked optical disks (d, e), light sources (f, g; only 2 of 4 shown) and photocells (h, k) 

(source: Figure 4 of patent 767528)

Patent 767528 states that the limiting factors for the upper limit of the antenna's rotational speed, are the printing speed of the Hellschreiber and the required pixel resolution of the printed information. Given the large size and weight of the antenna system, there are obviously also mechanical considerations for the upper speed limit. The patent proposes to resolve this, by quadrupling the number of antenna beams, spaced at 90º intervals. Each optical disk would simply have four "light source plus photocell" pairs (two pairs shown in the diagram above), that could be adjusted to account for angular offsets between the beam centerlines.

On the printed paper strip, the sharp dip in the bar graph now points directly at the aircraft's current azimuth (bearing from the station), as printed on the trace right below it. See the figure below. This is insensitive to variations in the rotating speed of the antenna and the paper transport speed of the printer. Also, the print-out does not need to be interpreted immediately. The achievable accuracy of the system is primarily limited by the pointing accuracy of the twin-lobe beam, and the visual alignment of the printed Hellschreiber traces. Later on, accuracy was improved to 0.5 deg, by adding 0.5 deg tick marks on the azimuth scale.

Signal strength bar graph, azimuth data, and station identifier as printed with the Hellschreiber.

The strip indicates that the receiver is at a bearing of 239 degrees from ground station "R".
(source: ref. 15)

Re-created signal strength bar graph, azimuth data, with station identifier "M"

I have used FontStruct™ to capture the above signal-level track and compass-rose segment as Feld-Hell character of a "TrueType" font that can be used in regular Windows^® and MacOS^® programs (e.g., Word^®, PowerPoint^®). Click here for the two-character font (capital letters A and B) for the signal level. Click here for the 18-character font (capital A-R) for the 360º compass-rose.

The format of the azimuth track in this print-out appears to be similar to that of patent 767524. The azimuth track of this print-out shows an interesting effect: the track is not straight, but has a slight wave to it. See the annotated figure below. This may be caused by variations in the speed of the printer spindle. The bar-graph track does not have this wave, even though its printer spindle is on the same shaft as the spindle of the azimuth track. This would be explained by the bottom of the signal-strength bars somehow being synchronized with the spindle position (start of the winding on the spindle). However, a start-stop mechanism with a start-pulse (such as used in the 1950s Hell-72 machines), would have required a mechanical clutch in the printer, made sending the azimuth data more complicated, and made the entire much more sensitive to interference and other noise. The HS 120 Bernhardine printer definitely does not work with a start-stop system. Ref. 48b suggest that the Psch120 Bernhardine printer did work per the start-stop principle, but that reference already incorrectly states that the printer had a single printer magnet (i.e., a single printed track).

Print-out with wavy azimuth track
(source: ref. 15)

The motor in the Bernhardine Hell-printer is not synchronized to the optical disk in the ground station. Speed differences result in slanting of the printed track (upward slanting if the printer is faster than the transmitter). Even if both sides are running at exactly the same speed, phase differences result in the printed track being shifted up or down. The slant, or shift, causes part of the track to end up beyond a vertical limit (upper or lower) of the track. Due to the nature of the printer spindle, it is actually printed at the opposite limit of that track. See the blue oval in the figure above.

In other Hellschreiber printers, the classical solution to this problem is a printer spindle with a thread that has two windings: two identical text lines are printed, one above the other; hence, at least one text line is always fully legible. This is combined with manual adjustment of the printer's motor speed. It is unclear how this problem was solved in the Bernhardine Hellschreibers... The printers (e.g., Psch 120) only appear to have a manual gain adjustment knob, and a knob for brightness of the "dial" light ("hell/dunkel").

Patent 767512 refers to the option of transmitting an extra heavy tick mark in the azimuth trace. This could be used for tactical purposes, e.g., to indicate a target azimuth (the radial from the ground station) that is to be intercepted. This could be simply implemented with a notch (cam) on the same shaft as that of the optical disk. The position of the notch could be adjusted to the desired target azimuth. The notch would actuate a switch that is simply connected in parallel with the output of the photocell of the optical disk.

Target azimuth bar superimposed on azimuth data
(source: hand-drawn figure 3 in patent 767512)

The above target-azimuth bar is the simplest form of a (unidirectional) data-link for sending tactical data to an aircraft, for guidance or command purposes ("Kommando-Übertragung"). A more sophisticated data link was implemented by transmitting text messages in Hell-format on the azimuth data channel. Note that these messages were transmitted instead of the azimuth data, not in addition to, or in parallel with the azimuth data. There was no additional, third printer track for these messages. An other advantage of this messaging system, is the fact that it did not occupy the already busy frequencies for voice communication. This same concept is making a comeback in modern aviation: Controller Pilot Data Link Communications (CPDLC). This is for up-linking routine air traffic control instructions and clearances to aircraft. However, contrary to the Bernhard system, the pilot can now respond to messages, request clearances and information, and declare an emergency. Apparently, late 1943 / early 1944the Lorenz company also experimented with a Hellschreiber-based command data-link system, referred to as "Sägezahn" ("sawtooth", ref. 55).

These were short messages, consisting of simple coded groups of letters and numbers. The message could consist of information about an enemy bomber formation, e.g.: altitude (in 100s of meters) of the lead-aircraft of the formation, a two-digit identifier of the "Jägergitter" air defense "box" in which the lead-aircraft was located (e.g., "QR" for the box around the town of Mainz), the two-digit course (not heading; in 10s of degrees), size of the group (estimated number of aircraft). The message always started with a "+" as delimiter. The Bernhard/Bernhardine system was the first and only ground-to-air data-link system of the second World War that had freely formattable messages!

(source: ref . 32)

The figure above suggests that the printed message characters had a height of about half the azimuth track. However, this (hand-drawn) figure already incorrectly suggests that the data-link messages were printed as a separate, third track...

In some literature, allied bomber formations are generically referred to as "bomber streams". However, the term only refers to a specific form of sequencing (night) bombers, used by the RAF from the end of May 1942 until the end of the war. The purpose of this tactic was to create a string  of bombers (with designated altitude bands and time slots), that would pass through the narrow German (night) air defense system via a minimum number of "boxes". This defense system, established in 1940 by then-colonel Josef Kammhuber, comprised a chain of rectangular airspace zones ("boxes"). The chain eventually reached from Denmark to the north of France, and was referred to by the British as the "Kammhuber Line". The zones had search and tracking radars (first Freya and Würzburg systems, later also Würzburg Riese) and groups of search lights (some radar controlled). Funneling all bombers through one or a few boxes, quickly overloaded the defense capability of the boxes (two night fighters per box, an estimated 6 intercepts per hour).

Generating the required Hell-signals was relatively easy: use a standard "Handlocher" (keyboard-based tape puncher) to create a section of punched tape with the message in regular 5-bit Baudot character coding, and feed the punch tape to a Hell "Lochstreifensender" perforator. The latter reads the punch tape and converts each character to the associated Hell-pixel stream, with the correct speed. When the message had to be sent, the output of the photocell of the optical azimuth disk was simply disconnected, and the output of the punch-tape transmitter used instead. The message was either repeated multiple times on the same punch tape, or the punch tape was restarted several times.

The Hell-format of patent 767524 (see further above), suggests printed columns with a height of 24 pixels. Note that the printer itself is fully unaware of the number of pixels used per column: the printer spindle turns at some predetermined speed, and the printer solenoid is actuated as long as a signal is received. That's all! Hence, the printer can be used for any number of pixels-per-column that the (de-)actuation speed of the printer's electromagnet can accommodate. The only requirement is that the transmitter and printer be set up for the same column duration (printer spindle speed). So, in principle, any commercial off-the-shelf (COTS) Hell-sender could be used to send the data-link text messages, provided its speed was adapted to that of the Bernhardine printer (or conversely). Alternatively, a Hell-sender with a customized character set could be used. With a COTS Hell-sender, the printed characters would almost be as high as the entire azimuth track. I have not seen any pictures of original Bernhardine print-outs with a data-link message.

The "Bernhard" UKW-Richtstrahl-Drehfunkfeuer (VHF directional beacon), is the ground station part of the Bernhard system. It comprises a rotating antenna system with two antenna arrays. Each antenna array has its own transmitter. The original trial systems were developed by Telefunken ca. 1935-1938. Note that Telefunken was one of the pioneers of radio navigation beacons: 1907-1918 they had already developed and operated the Telefunken-Kompass-Sender, primarily for long-distance navigation by Zeppelin dirigibles (ref. 10, 14, 44, 45, 54).

The original Bernhard system operated at a frequency around 300 MHz (λ = 1 m) and the transmitters had an output power of 20 W. Tests were carried out by Telefunken at several locations (ref. 39, 50):

	in 1935 at Groß-Ziethen (Berlin-Schönefeld) with 1x 20 W.
	in 1936 at Rechlin with 2x 20 W. In 1935, the airfield of Rechlin (about 100 km / 60 mi north-northwest of Berlin) had become the Erprobungsstelle der Deutschen Luftwaffe (E-Stelle, official test site).
	in 1940 at Mietgendorf (this is in the same area as the Bernhard Be 0 station at Trebbin, see further below).
	for the Kriegsmarine, at the naval station at Schillig, just northwest of Wilhelmshafen - a major Reichskriegshafen.

Several versions of this system were evaluated: ground-test only, a system with two receivers without Hellschreiber, a dual system with four receivers and a 2-channel Hellschreiber, and a simplex system with two receivers and a 2-channel Hellschreiber

Development was abandoned until ca. mid-1940, when it was re-introduced - this time as a replacement for another radio navigation system: "Knickebein" (i.e., "crooked leg"), a nickname it received due to the slight V-angle of its antenna array when viewed from above. Knickebein was successfully jammed and spoofed ("diversion" of the apparent beacon beams) by the British. The Bernard system was converted to the same operating frequencies as Knickebein: 30-33.3 MHz  (λ = 10 m). Hence, no new or additional receivers were required in the aircraft. Also, ground station transmitters could output well over 200 W in that frequency range, which was not the case at 300 MHz.

Rotating antenna system of the prototype "Bernhard ", operating at 300 MHz (λ = 1 m)

(source: p. 83 in in "Die deutschen Funkführungsverfahren bis 1945", by Fritz Trenkle

Telefunken also developed the "Bernhard" ground stations that finally entered service. They are generically referred to as a "Funk Sende-Anlage", i.e., a complete radio transmitter installation or system - including the antennas. This is abbreviated as "FuSAn". The initial FuSAn developed for the Bernhard system was FuSAn 724. Its two transmitters each had an output power of 500 W. Later on, a 2 x 5000 W version was developed: FuSAn 725. It is unclear whether the 725 ever entered into active service. Note that most literature uses "FuSAn 724/725", (incorrectly) suggesting that this was a single FuSAn type. These same FuSAn were also evaluated and tested by the Lorenz company during the period 1941-1943, for their "Erika", "Mond", and "Erich" navigation systems. Over a dozen different fixed and mobile FuSAn types were developed for various radar and radio-navigation systems of the Luftwaffe, primarily by Telefunken, Lorenz, and the DVL (Deutsche Versuchsanstalt für Luftfahrt), ref. 31.

I have no photos or details regarding the FuSAn 724/725 transmitter equipment. Bernhard comprised two identical transmitters: one for the twin-beam primary signal (two closely spaced lobes, with a sharp null between them), and one for the secondary beam antenna for the azimuth data in Hell-format. The latter points in the null-direction of the primary antenna. The beams were swept by rotating the antenna system about a vertical axis. Patent 767532 suggests a rotating antenna array that comprises multiple ½λ-dipoles, with ½λ spacing between them. An operating frequency of 30-33 MHz is equivalent to a wavelength λ ≈ 10 m and ½λ = 5 m. Transmitted signals were monitored at a nearby receiving station. The antennas were built by Hein, Lehmann und Co. of Berlin. This was Telefunken's standard builder of large antenna installations (incl. those of the gigantic 1 megawatt Goliath transmitter station).

Arrangement of the two Bernhard antenna arrays and their radiation patterns

In summary, the Bernhard ground station has the following characteristics:

	Frequency: 30 - 33 MHz
	Transmitter power: 2 × 500 W (FuSAn 724) or 2 x 5000 W (FuSAn 725)
	Transmitter antenna system dimensions: ≈28 x 35 m (HxW, 92x115 ft)
	Antenna system track diameter: 22.6 m (74 ft)
	Antenna system weight: 120 t (265000 lbs; some literature states the weight as 102 t or ca. 256000 lbs, or 100 t (ref. 42)) - I have tried to do a "sanity check" on this number, as it appears to be rather high. With many assumptions, I arrived at at least 50-60 tonnes (see this spreadsheet). However, if extreme precision is required (ref. 42 suggests 1 mm (!), an extremely stiff construction would have been necessary. This would imply a much heavier construction...
	Antenna rotational speed: 12 degrees per second (2 revolutions per minute). This means that the small locomotives that turned this enormous antenna installation, moved at a respectable speed of 8.3 km per hour (5.2 mph)!

Based on the local situation (access to the local public power grid), the installation often included a high-voltage bunker ("Hochspannungsbunker"). Ref. 27. It contained one or more transformers, to connect to the multi-phase regional Hochspannugsnetz (high voltage public power grid, 110-220 kV in Germany), or to the local Mittelspannungsnetz (6-60 kV, typically 10 or 20 kV). It is unclear if there were diesel-generators as a back up.

Patent 767525 explains how two stationary transmitters could be connected to two rotating antenna systems. The coupler comprises two sets of stator and rotor disks. All disks are installed coaxially: the rotor plates fixed to a rigid shaft, the stator plates to the housing of the coupler. The transmitters are wired to the edge of the respective stator plate.

The "Bernhard" station at Hundborg/Thisted (northwest coast of Denmark)
(source: www.gyges.dk, used with permission)

The photo above clearly shows the two stacked antenna sub-systems. Each sub-system is a collinear array: the vertical dipoles are parallel, and distributed along a common horizontal axis. To generate the two primary beams, two sub-arrays are used. The combined array consists of 2 x 4 = 8 vertical dipoles. With ½λ spacing (i.e., ≈ 5 m), they would span roughly (8-1) x 5 = 35 m. This is consistent with the diameter of the Bernhard antenna system. However, in the photo above, it appears that the actual dipoles have an overall length of about 10 m. I.e., λ, rather than the standard ½λ... 

The dipoles have a reflector rod placed behind them, to obtain the desired radiation pattern. The reflectors have about the same length as the dipoles. Dipoles and reflectors are probably spaced at a distance of 0.15-0.2 λ (i.e., ≈ 1.5 - 2 m), as in a regular 2-element beam antenna. It is unclear from the available photo material, whether all radiator dipoles were actually driven (via coax), or only one per beam - as was done in other navigation beacons (e.g., ref. 45b) - and all others were carefully tuned reflectors. Also, not all installations have a reflector for all dipoles. This appears to have evolved over time. E.g., the first installation (at Trebbin), where the system was perfected, only had reflectors for the outboard dipoles.

The upper antenna array is used to transmit the azimuth data in Hellschreiber format. This array comprises 3 vertical dipoles. A large reflector screen behind the array acted as a director and to to suppress undesired side lobes of the radiation pattern. Not surprisingly, the Bernhard ground station was easily mistaken for a radar installation.

"Bernhard" station Be01 at Trebbin (near Berlin)
(source: ref. 17)

Upper framework of the"Bernhard" station Be09 at Bredstedt, ca. August 1945

(source: Australian War Memorial photo SUK14635, public domain)

"Bernhard" station Be09 at Bredstedt, ca. August 1945

(source: Australian War Memorial photos SUK14632 and 14634, public domain)
 

In the right-hand photo above, and in the  photo below, note the protective "hood" that covers the tracks. It moves around with the rotating antenna system. This hood was not installed at all locations.

The antenna system was installed on four double bogies (US: trucks) on a circular rail track. The bogies were powered electrically.

Two double-bogies of the "Bernhard" station at Hundborg/Thisted

(source: www.gyges.dk, used with permission)

The motors of the electro-locomotives were built by Ziehl-Abegg Elektrizitätsgesellschaft m.b.H., of Berlin-Weißensee (ref. 27). Their power was a mere 10 kW (13.6 metric horsepower, 13.4 US HP), but with a large down-gearing ratio. Apparently four of such motors was enough to generate the required torque and speed... It is unclear how the electric motors were powered. It is unlikely that the required power was supplied via large slip-rings, from power supplies in a nearby bunker. Possibly small diesel generators and fuel tanks were located in the large doghouse, right above each of the locomotives, but there are no chimneys or exhausts visible in the available photos...

Ziehl-Abegg is a accompany specialized in electric motors, founded in 1910 by Emil Ziehl and his Swedish partner Abegg (who dropped out of the partnership the same year); prior to 1910, Ziehl had developed electric motors and tested generators at AEG, and developed gyro-compasses at Berliner Maschinenbau AG (BEMAG, frmr. Eisengießerei und Maschinen-Fabrik von L. Schwartzkopff), manufacturer of locomotives powered by steam, compressed air, and electricity. These days, Ziehl-Abegg AG is a manufacturer of electric motors for ventilation and air-conditioning systems.

Early 1900s advertising stamp of Ziehl-Abegg (4x5 cm) and listing in the 1943 Berlin phonebook

Side view of a Bernhard station, with the central structure (Be12?)

(source: brdy.org)

In the photo above, note the small diameter of the support between the locomotive bogie and the antenna platform. It supported 1/4 of the 120 tons weight of the antenna system. Electrical power cables and/or  speed regulator wiring may have passed through it.

A damaged Bernhard station, 6 July 1944 (Be7?)

("Captured Nazi radar equipment somewhere in France", source: www.footnote.com, no copyright (US Gov't))

The photo above clearly shows that the bridge with the wooden barrack is suspended from the center of the antenna bridge. The photo also shows a small structure between the barracks and ground level - right at the very center of rotation of the antenna. It is a small round building (brick, or brick and concrete), with a flat concrete roof.

The purpose of remain unclear and controversial.

	the wooden barrack across the base of the antennas system,
	the small round building below it,
	the "closets" above each of the four locomotives.

No conclusive documentary evidence has emerged so far...

The barrack may have contained the two transmitters and associated antenna feed lines, and splitters - if any. If the transmitters were located in a nearby equipment bunker ("Gerätebunker"), a rotary RF-coupler (as described further above) would have been required below the antenna system, i.e., in the small round building.  The barrack may also have housed the speed control system for the electric locomotives. The photo below shows an electrical control panel of the Bernhard system; it does not appear to be located in a bunker (e.g., based on the ceiling trim). There are no indications (e.g., exhaust chimneys) that the barracks contained diesel power generators for the transmitters. Obviously, the (expensive) transmitters would not have been protected against bombing or strafing attacks if they were located in the barrack instead of a nearby bunker.

The round building below the antenna system has an outside diameter of about 3.5 m (≈11-12 ft). This is too small for housing an auxiliary antenna drive system. Also, the required torque would simply shear off the drive shaft through the ceiling. In the round building of the Bernhard station Be 10 at Thisted, there are still electrical (or RF) cables that go up along the ceiling braces, around and to (or from) the hole in the ceiling. The two 5 kW transmitters required a significant amount of electrical power, say 20 kW total (depending on modulation type and efficiency); transferring this power to transmitters located on the bridge of the antenna system would have required heavy-duty slip rings with large brushes, and heavy-gauge cables to the slip-rings. At some Bernhard sites (e.g., Hornstein, near Vienna/Austria), there appear to be remnants of cable trenches between the foundation of a bunker and the concrete base of the antenna system. It is unclear whether they contained power cables and/or RF cables.

Round building at center of Be 4 Le-Bois-Julien (left) and Be 14 at La Teste-de-Buche (right)

(photo Be 4: © T. Oliviers; Be 14 photo source: www.atlantikwall.superforum.fr, thread in the "La détection navale et aérienne" forum)

Round building at center of Be 12 Nevid (Plzen)

(photo: © Jacek Durych, used with permission)

The building has a hole through the center of its roof. The top-side flange is embedded into the roof. It looks like this flange may have had ball bearings inserted into it. There is a large square flange on top of the roof, with a hollow shaft that passes the through-ceiling flange. This flange must have been attached to the bridge of the rotating antenna system. The bottom of the hollow shaft terminates with a small round flange. It may have been connected to the required optical disk(s) with Hellschreiber encoding, as described further above. However, in that case, the transmitter keying-signals must have been passed from the encoder disk assembly to the transmitters - again via slip rings. If the disk assembly were located with the transmitter in the barrack, then the shaft of the disks had to be stationary. This is inconsistent with a rotating shaft entering the top of the building.

Top side of the hole through the roof of the round building at Be 12 (Plzen), and the square flange at Be 6 (Marlemont)

(photo on left: © Jacek Durych, used with permission; source photo on right: www.atlantikwall.superforum.fr, thread in the "La détection navale et aérienne" forum))

On the inside of the building, there are four heavy steel reinforcement braces against the ceiling. The braces start at the hole through the roof, and go down the inside wall of the building, to the floor and possibly into the foundations. This is a very solid construction. Either because something heavy was suspended from the ceiling, or whatever was suspended needed to be completely immobilized, or to protect equipment from attacks.

Cruciform braces on the ceiling of Bernhard station Be 6 (Marlemont)

(source: www.atlantikwall.superforum.fr, thread in the "La détection navale et aérienne" forum)

Each of the four braces has three pairs of mounting bolts, that protrude downward. The inner pair of bolts appear to secure the "through hull" ring on top of the roof. A small L-bracket was mounted on the middle pairs of these bolts. This is still visible at the Bernhard station Be 10 at Thisted (Denmark).

The airborne counterpart of the Bernhard ground station is referred to as the "UKW-Richtstrahl-Drehfunkfeuer-Empfangszusatz mit Kommandoübertragung" (accessory for VHF directional beacon reception, with command data-link). This system, referred to as FuG 120 "Bernhardine", comprises (ref. 57):

	the Hellschreiber printer,
	an amplifier unit ("Schreibverstärker" SV 120),
	a filter unit ("Schreibgabelschaltung" SG 120); it uses two filters to separate the continuous tone and the Hellschreiber tone pulses.
	a switch unit ("Umschaltgerät" UG 120), and
	a power supply unit ("Umformer" U 120).

Several versions of the FuG 120 were developed: 120, 120a, 120b, 120k, and 120s. The difference is basically in the Hellschreiber printer. I do not have the particulars on these types. FuG 120 had a much more compact "Peilschreiber" printer than the FuG120. FuG 120b had a printer with an endless, self-erasing wax tape or disk ("Wachsschreiber") that is described further above. It proved inadequate in operation, and was retrofitted back to FuG120a. A further development was FuG120k. This system dispensed with the filter unit. Hence, the the tones of the two Bernhard signals could no longer be separated, and fed to separate Hellschreiber printer tracks. The printer therefore only had a single track, and the azimuth and signal-strength bars were superimposed onto a single printer track. As a result, the achievable accuracy was reduced from 0.5º or 1º, to ≈4º (cf. p. 87 in ref. 14, p. 125 in ref. 43).

Additional references are to Hellschreiber "HS 120" (cf. p. 99 in ref. 50). This is Hellschreiber with a single wide paper tape, or two parallel paper tapes (standard 15 mm width). It is powered by a 24 volt AC synchronous motor, running at 3000 rpm. The paper tapes are re-wound onto a second set of reels inside the printer. The upper part of the front panel is hinged, to facilitate inserting the paper tapes across the front of the printer.

Left: rack with FuBl 2; Right: rack with FuG 120 "Bernhardine" with HS 120 at top

(source: p. 99 in ref. 50; this HS 120 uses a single wide paper tape)

Front panel of an HS120 Hellschreiber with two parallel paper tapes

(controls for audio gain and lighting brightness)
 

The photo above shows that the lower printer-track actually prints two text lines when regular Feld-Hell characters are sent to it (2.5 characters per second). This means that the printer spindle has two windings. This appears to be inconsistent with published depictions of the printed tapes. Possibly, this particular printer is a prototype. Note that no matter at what speed the characters are sent to this printer, a 2-turn spindle will always print at least two identical lines. The 7x14 Hell font has 2+2=4 white pixels above and below the 10 character pixels (total 14). Reducing the character transmit speed will indeed increase the height of the printed characters. But the height of the printable space is fixed: it is equal to the pitch of the thread of the spindle. The Hell character speed can be reduced by up to ≈30% (4/14) before the top of both printed lines is cut off:

Printing characters transmitted at normal, 1/2, and 1/4 speed

As discussed further above, this is not a start-stop printer. Hence it is unclear how the bar-graph was printed on the upper printer track for the signal strength display. It is not all that complicated to convert the strength of the received signals to an equivalent pulse duration ( = bar height). However, it is unclear which mechanism was used to ensure that the bar graph lines started precisely at the bottom of the printed "text" line. Even if two identical bar-graph tracks were printed, this would not be easy to read...

Top-view of an HS120 Hellschreiber

(left to right: stacked electromagnets, stacked paper rolls, motor, shaft for spools for printed paper)
 

Rear-view of an HS120 Hellschreiber
 

Right-hand side of an HS120 Hellschreiber

( two stacked felt ink rollers on the left)

This printer measures 35.7x15x14 cm (WxHxD, ≈14x6x5½"); that is, not the 60x30x20 cm that is usually stated for this model. It is still relatively large. This means it was either a prototype, used in Bernhard ground monitoring station, or used in larger aircraft (two or more seats, e.g., in front of the radio/radar operator). The associated electronics were housed in two additional units.

The stacked electromagnets, with printer hammers against the paper tapes

Here is a 25 sec video clip of the printer in action:

(click here to start your own player)

The following marking is engraved on the back side of the plastic backlighting panel of the HS120: "A. BAJANZ, Berlin -- DRGM.-DRP.angem." The Bajanz company manufactured acrylic glass parts (common trade names are plexiglas and perspex), such as canopies for military planes. The abbreviations DRGM, DRP, and angem. respectively stand for "Deutsches Reichsgebrauchsmuster" (model - or patent, in the proper sense of the latter word; this implies a registration without patent protection), "Deutsches Reichspatentamt" (patent office), and "applied for" ("pending").  

Marking of the manufacturer of the plexiglas front panel

There is a white "BAL 378" Waffenamt acceptance stamp on top of the chassis. This BAL number has appeared on products of “Dr. Ing. H. Kimmel, Fernmeldetechnisches Entwicklungslaboratorium“ ("development lab for telecom equipment", though they also made measurement equipment) in Munich. Their 3-letter manufacturer's code was "bes". Kimmel also made the "NF Phasenuhr" (audio frequency phase-indicator, with 360º scale) that was part of the on-board equipment of the 1943 Lorenz VHF rotating-phase navigation system "Erich" (very similar to the VOR system developed in parallel in the USA). Like "Bernhard/Bernhardine", "Erich" also used the EBl 3 receivers.

"BAL 378" acceptance stamp on top of the chassis

The "Peilschreiber" (bearing printer) was built by Siemens Luftfahrtgerätewerk (LGW) Hakenfelde GmbH, in the borough Berlin-Spandau, just across the Havel river next to Siemensstadt. LGW was a 1940 spin-off of Siemens Apparate und Maschinen GmbH (SAM). Its manufacturer code was hdc. LGW made a wide variety of aircraft equipment, such as gyros, temperature gauges, radio altimeters, course indicators, switches, indicator lights, course guidance equipment, and autopilots.

Peilschreiber (bearing printer) "Psch 120" of the FuG 120a - paper tape printer
(source: figure on page 11 in ref. 15)

Peilschreiber (bearing printer) "Psch 120" of the FuG 120a - paper tape printer
(source: figure 38 in ref. 48b)

Patent 767536 proposes to print the signal strength bar graph and the azimuth information on a paper disk, instead of paper tape. This is illustrated in the figure below. Note that the threads of the printer spindles (c & d) are oriented incorrectly with respect to the printed traces.

Paper disk Hellschreiber printer with two spindles (items c & d)

(source: Reichspatent Nr. 767536, see patent table below)

Peilschreiber "Psch 120a" of FuG 120 b  - a paper disk printer
(Figure on page 11 in ref. 15; the large round "Hell-Dunkel" knob is for the brightness of the "dial" light)

An estimated 2500 Bernhardine units were installed in various aircraft types (ref. 14, primarily night fighter versions), such as the Messerschmitt Me262 A-1a (one of my all-time favorite aircraft), Junkers Ju 88G (ref. 52), Arado Ar234 "Blitz", and Dornier Do-335 A-6 "Pfeil". It may also have been installed in the Dornier P.254 (Do 435). FuG 120k appears to also have been intended for the advanced Gotha GO-P60 (Gothaer Waggonfabrik). The FuG 120 system actually was not used on the Me-262 day-time fighter. It was, however, considered for the Me 262 A-1a/U2 all-weather fighter, but was cancelled due to lack of availability of the equipment in sufficient numbers. It was removed from the aircraft equipment list early January of 1945 (see handwritten note in ref. 33). A drawing exists (ref. 34) that shows installation of the FuG 120a Peilschreiber suspended from the canopy between the pilot and radar operator. While purported to be for the Me 262 B-2, the drawing appears to be of the interim night-fighter Me 262 B-1a/U1. Several captured Me 262 B-1a/U1's showed something suspended from the canopy, though so far, no photos are available to positively identify the equipment. An RAF report (ref. 35) on the Me 262 B-2a, also describes the FuG 120a as equipment intended for that type.
 

Standard radio-navigation receiver sytems were the Fu Bl 1 ("Funk-Blind-Lande-Gerät, ref. 26, 40, 53) and the Fu Bl 2. The latter was basically an Fu Bl1 with extended frequency range, to make it work with the Knickebein, Bernhard, and Hermine systems. The Fu Bl 2 system included both a Lorenz EBl 2 and an EBl 3 receiver. The EBl 1 "UKW-Blind-Einflugzeichen-Anlage" covered the 30-31.5 + 33.3 MHz frequency range. The EBl 2 (a "Markierungsfunkfeuerempfänger", marker-beacon receiver) had two two fixed frequencies (33 and 38 MHz), whereas the EBl 3 "UKW-Blindflug-Leitstrahl-Anlage" ("VHF guidance beam system") covered the 30-33.3 MHz with 34 channels (100 kHz channel spacing). These receivers where normally used in conjunction with landing guidance systems (instrument landing system for "blind" flying). This meant that these receivers where available for other functions during all flight phases other than approach and landing. The tone-modulated directional signal and the Hellschreiber azimuth signal of the Bernhard station were close enough that they could be received by a single (relatively wideband) receiver. The two tones were separated by an audio filter unit (part of the FuG120) and amplified before being passed to the printer unit. There were two versions of the EBl 3 receiver: EBl 3H ("Handbedienung" = manually controlled tuning) and EBl 3F  ("Fernbedienung" = remote tuning).

EBl 1 Ex receiver

(source: figure 7 in ref. 20; also ref. 48)

The only aircraft actually equipped with the EBl 3 were (ref. 21, p. 106 in ref. 14): the bomber version of the Ar-234, the nigh-fighter version of the Ju-88, Do-217 (ref. 41, 56), Do-335, and the bad-weather fighter ("Schlechtwetterjäger") version of the Me-109, the Me-262 (ref. 25), Ju-388 (ref. 27), FW-190, and Ta-152, and the reconnaissance version of the Me-109, Me-262, Ar-234 (planned only, per ref. 13), and Ta-152. The latter is an early 1945 high-altitude interceptor-fighter derivative of the ubiquitous Focke-Wulf FW-190. The designator "Ta" refers to the renowned FW chief-designer, Kurt Tank.

EBl 3H receiver (manual tuning)

(original unedited photo: courtesy Erich Werner)

Landeempfänger EBl 3 in a Bordfunkanlage FuG 10 P (lower equipment row, right of center)

(source: Figure 4.4 in ref. 49)

The EBl 3 superhet receiver was developed by Lorenz AG out of the FuG16 (ref. 51), as a replacement for the EBl 1. The latter had insufficient sensitivity to be used for long range navigation (i.e., to targets in England) in combination with navigation beacons such as the Knickebein system. The EBl 3 was manufactured at AEG Sachsenwerk in Dresden-Niedersedlitz. Initially, the unit had a light-metal ("Elektron") die-cast construction, built by Mahle. Towards the end of the war, the chassis housing and front face were made of simple sheet metal. The receiver comprised seven tubes of type RV12P2000 (schematic: ref. 36), and had an IF frequency of 6 MHz. After the war, the small company Curt Höhne Radiomechanik (radio sales & repairs) of Dresden-Radebeul, somehow "acquired" the EBl 3 inventory of the Sachsenwerk factory from the US Allied authorities, and converted parts of them to car radios (ref. 37). Conversions were also published for conversion to the 2m (145 MHz) amateur radio band (e.g., ref. 38, 1952).

The night-fighter version of the Me-262

("source: www.footnote.com, no copyright (US Gov't))

Dornier DO-335 "Pfeil" with push-pull propeller arrangement
 

Junkers Ju-88

Arado Ar-234 "Blitz"
 

With a single Bernhard ground station, only relative bearing to/from that particular station can be determined. I.e., neither distance (range) nor position. Position determination is done by combining the bearing of at least two ground stations with know location. I.e., by means of conventional triangulation. Note that the bearing from a ground station to the aircraft has nothing to do with the aircraft's heading (the way the nose is pointing).

Triangulation with two ground stations
(D = beacon ("Drehfeuer"), α = azimuth = bearing from station = angle measured clockwise from north)
 

Patent 767537 proposes to facilitate triangulation by simultaneously printing the signals (strength + azimuth) of two ground station on double-wide paper strip. This does require an double printer, as well as a second receiver.

Triangulation made easy with a Berhardine Registrierungschrieb (print-out) from two ground stations on a single, extra wide paper strip
(source: ref. 15)

To support triangulation, and provide coverage over a large part of western Europe, a chain of ground stations was constructed (ref. 14, 16, 21, 22). Note that quite a number of Bernhard "Anlagen" ground stations were still under construction by the end of the war, and never entered into service. Obviously, they were installed on the highest point in the area, typically a hill top.

Each Bernhard ground station had a "Be" identifier number. In addition, the (operational) ground stations had a single-letter identifier ("Kennbuchstabe") that was used in the Hellschreiber data-link messages (e.g., "R" and "M" in the double print-out strip immediately above). I do not know the mapping between these identifier letters and the corresponding Bernhard stations. Ref. 27 suggests that "K" stood for the Bredstedt station in Schleswig-Holstein (the reference actually mentions the town of Leck as the location, 14 km (9 mi) further north).

The ground station locations are referred to by the name of a town or village nearby or in the region. In the listing below, the most accurate location name is provided first. The most commonly used reference is in bold face.

	Be 0 - Glau (ca. 1 km northwest of village center; ca. 5 km northeast of Trebbin; ca. 35 km south of Berlin; also near ), Germany. This is the first installation that became operational (September 1941, p. 85 in ref. 14)
	Be 2 - Mont-Saint-Michel-de-Braspart (ca. 23 km east of Brest; also near Sizun, Morlaix, Ménez-Mikel), France.
	Be 4 - Le-Bois-Julien (ca. 40 km south of Calais; ca. 18 km southeast of Boulogne-sur-Mer)), France.
	Be 6 - Marlemont (ca. 70 km north of Reims), France.
	Be 7 - La Pernelle (ca. 23 km east-southeast of Cherbourg), France.
	Be 8 - Schoorl (ca. 8 km northwest of Alkmaar; also near Bergen), The Netherlands.
	Be 9 - "Stollberg" hill near Bredstedt (ca. 32 km southwest of Flensburg, 20 km north of Husum), Germany. The circular "Bernhard" track lies within a predating "Knickebein" circle (the latter has a much larger diameter: ≈90 m)
	Be 10 - Hundborg (ca. 12 km southwest of Thisted, ca. 71 km west of Aalborg), northwestern coast of Denmark.
	Be 11 - Trzebnica (also referred to as Trebnitz, Breslau; ca. 25 km north of Wroclaw (Wrazlaw), Poland.
	Be 12 - Mirošov (near Nevid, ca. 20 km east of Plzen (Pilzen), now Czech Republic.
	Be 13 - unknown.
	Be 14 - La Teste-de-Buch (ca. 4 km south-southwest of Arcachon (ca. 50 km southwest of Bordeaux), France.
	Be 15 - Wieżyca/Szymbark (frmr. Schönberg; 2 km southeast of village center; ca. 40 km southwest of Gdańsk, half way between Gdańsk Bytów (Butow)), Poland.
	Be 16? - "Sonnenberg" hill near Hornstein (ca. 40 km south of Wien/Vienna), Austria; Construction of the installation was started in January of 1945, but never completed beyond the foundations of the antenna ring and of several buildings.
	Be ?? - St.-Michel-Mont-Mercure (near Pouzauges, ca. 75 km southwest of Nantes), France.
	Be ?? - "Keimberg" hill just north of Buke, Germany (in the Teutoburger Forest, ca. 13 km east of Paderborn); construction was completed just before arrival of US troops. See notes further below.
	Be ?? - Favières (ca. 20 km northwest of Chartres), France.
	Be ?? - "Venusberg" hill near Aidlingen (ca. 21 km southwest of Stuttgart), Germany.

The map below shows the location of these stations. The above bold-faced location names are used. Special thanks goes to Gérard Chatry, for helping me precisely locate some of the stations. A file with station coordinates (both in decimal degrees and degrees-minutes-seconds) is here.

Location of "Bernard" FuSAn 724/725 ground stations, with 400 km range rings
(range rings are not corrected for projection of the map - they should be slightly oval north-south)

Below is an interactive map with the same station locations. Locations marked with are exact. In practically all cases, you can fully zoom-in the satellite image map, and see the actual the remains of the station structures. Click on any marker icon, to get the associated information.

For the station marked with , the exact location could not (yet) be determined. You can click-and-drag the map with your mouse, and zoom in & out with your mouse-wheel (or use the buttons in the top left-hand corner of the map).

Click here for a larger view of this map in a separate window.

The station locations appear to form a loose grid:

Ref. 27 provides interesting details about the history of the station at Buke. Actually three Bernhard stations were shipped to Buke:

	a new one, from the Telefunken factory.
	the station from Arcachon, dismantled by the retreating Germans after the allied invasion of France.
	the station from Breslau (Be11; Trebnitz/Wroclaw), dismantled by the retreating Germans, ahead of the advancing Russian troops.

The fact that three stations were shipped to Buke is attributed to administrative and logistic hiccups during the latter days of the war. The two superfluous stations were stored at a furniture factory in nearby Leopoldstal (11 km / 7 mi north of Buke). Construction of the station at Buke started in May of 1944 and was completed by the end of March 1945. Entry into service was planned for May of 1945, after calibration flights. The equipment was scuttled by the Germans when the American troops approached at the very beginning of May 1945. After the war, many parts were re-used locally. The rail tracks, cross-ties, and antenna structure were used in the reconstruction of war-damaged homes and buildings, and in agricultural machines. The motors of the electro-locomotives were used to power grain threshing machines. Buke ended up in the British occupation zone, and the British destroyed what was left of the installation structures.

Location of "Bernard" station at Buke, with the associated bunkers - 1945 (ref. 27, 28)
(the original photo is upside-down (south-side up); above, it has been turned 180º)

The circles in the photo above mark the following structures::

	Nr. 1, 4, 5 (at the center of the photo): concrete ring with two bunkers of unequal size; a "Gerätebunker" for equipment (the transmitters?), and a "Manschaftsbunker" for personnel
	Nr. 2 (550 m / 600 yds to the northwest of Nr. 1): a building with a footprint of 3x3m (10x10 ft) and a "Diodenmast" (antenna). Most likely, this bunker housed a radio receiver and Hellschreiber printer for monitoring the transmitted signals.
	Nr. 3 (400 m / 450 yds to the east of Nr. 1): a building of ca. 5x6m (15x20 ft). This was a "Hochspannungsbunker", i.e., a high-voltage bunker.

2011 satellite image of the same location
 

The Organisation Todt (OT) started construction of the installations at Aidlingen (Be-number unknown)  in 1944. The location is a small wooded area on the Venusberg hill, at that time property of Mr. Wilhelm Breitling of Aidlingen. OT, named after its founder Fritz Todt, was responsible for numerous engineering and construction projects in Germany and occupied territories (the Autobahn system, the Westwall and the Atlantikwall ("Siegfried Line"), roads, railways, airports, U-boat shelters, bomb shelters, V1 and V2 missile launch sites, concentration camps, etc.). Construction at Aidlingen was never completed, and the Germans destroyed the installations before the end of the war. Ref. 47.

British aerial reconnaissance photo of Aidlingen/Venusberg - 26 December 1944

(source: ref. 47)

Recent satellite image of the same location

Remnants of some of the other installations are also still visible to this day (e.g., here, here, and here). Mid-May 2011, I visited the site at St.-Michel-Mont-Mercure (near Pouzauges). See my photos below. It is located at the highest point (≈290 m ASL; ≈950 ft) in the Vendée region of France, on the Haute Bocage ridge. From there, one has a beautiful free view in all directions, including the northeast (to the Atlantic, and beyond to England). About 60 mtrs to the northwest of the antenna base is a concrete revetment that may have been for a Fliegerabwehrkanone unit (a.k.a. "flak", anti-aircraft guns). After the war, the site was used by American forces as a wireless telecom site. These days, there is a cell phone installation.

Visitor sign at the Bernhard site near St.-Michel-Mont-Mercure

(mistakenly referred to as a radar installation; source: atlantikwall.superforum.fr, © G. Chatry)

North side of the circular concrete base at St.-Michel-Mont-Mercure

The ties of the rail track at St.-Michel-Mont-Mercure / Pouzauges

The circular concrete base is about 1.5 m wide (≈ 5 ft). At some sites, the top of the ring is flat, at others it is clearly rounded.

At this particular site, practically all rail ties (UK: sleepers) are still in place, some 120 in total. Here, only two out of every three ties have bolts sticking up for the track fasteners. The bolts for the inside track are taller than those of the outside track. Also, one in three ties does not have mounting bolts, but mounting support plates. Again, the one for the inside track is taller than the one for the outside track. These plates do not have mounting bolts welded on top of them. As the radius of the track is small, a narrow-gauge track is used. In this case, a little smaller than "meter gauge". The diameter of the track-ring is relatively small for a rail track. This causes a non-negligible difference in the speed between the inside and outside wheels of the locomotive bogies. This causes problems if the bogie axles are rigid and the four wheels cannot turn independently. One way to solve this, is to use smaller wheels on the inside track - which in turn requires that the inside track be raised slightly. This approach may have been used here, and at installations, such as Le-Bois-Julien and Marlemont. However, at Marlemont, all ties have a support plate on the inside. See photo below.

The ties at Marlemont

At other installations (e.g., at Arcachon) all ties are identical, without support plates, and all have 2x4 mounting bolts. See photo below (though again, the bolts of the outside track are shorter than those of the inside track...).

The ties at Arcachon

(©2008 R. Lemenicier)

There are also several types of ties. At many sites, the ties are simply sections of a steel I-beam (H-beam, "Doppel-T-Träger") that are partly or completely set into the concrete.

The ties at Schoorl - I-beams and rounded concrete top

(©2008 D. van Lunsen)

At other sites (e.g., Butow, Plzen/Nevid, Trebnitz), the ties were not set into the concrete, but mounted on top of the concrete. In these cases, there are square or rectangular holes in the top of the concrete ring. There are either two rings of holes (inside & outside), or three (inside, center, outside) - as shown below at the installation near Butow. It is unclear what the center holes were for. A third (electrical) rail track?

Holes for the ties at Butow

Patent office abbreviation: RP = Reichspatentamt (Patent Office of the Reich), DP = deutsches Patentamt (German Patent Office)

Patent source: DEPATISnet

If you have any additional information or documentation on this system, please contact me!

	Ref. 10: "Bernhard and Bernhardine", p. 24 in "Some historical and technical aspects of radio navigation, in Germany, over the period 1907 to 1945", A.O. Bauer, PAØAOB, December 2004, 28 pp.
	Ref. 11: p. 18 in "Blitz!: Germany's Arado Ar 234 Jet Bomber", J.R. Smith, E.J. Creek, Merriam Press, 1997.
	Ref. 12: "Empfangsseitige Schreibvorrichtung zur Durchführung eines Verfahrens zur Richtungsbestimmung" Patent 767536A, Adalbert Lohmann, March 1936, published 1952 (!); addendum to patent 767345.
	Ref. 13: p. 291 and 350 in "Geschichte der deutschen Nachtjagd: 1917-1945", Gebhard Aders, 1977, 391 pp., Motorbuch Verlag, ISBN-10: 387943509X. Also appeared as translation with revisions: "History of the German night fighter force, 1917-1945", Jane's Information Group, 1979, ISBN 0867205814, 360 pp.
	Ref. 14: pp. 76-110, 224 in "Die deutschen Funkführungsverfahren bis 1945", Fritz Trenkle, Alfred Hüthig Verlag, 1987, ISBN 3778516477, 215 pp.
	Ref. 15: pp. 96-102 in “Die deutschen Funk-, Navigations-, und Funk-Funkführungsverfahren bis 1945“, Fritz Trenkle, Motor Buch Verlag, ISBN 3879436150,1st ed., 1979, 208 pp.
	Ref. 16: pp. xx in "Die deutschen Funklenkverfahren bis 1945", Fritz Trenkle, 215 pp., A. Hüthig Verlag, 1987, 215 pp., ISBN-10: 3778514652
	Ref. 17: Bild 16 in "Deutsche Funkmesstechnik 1944“, Leo Brandt, Verkehrs- und Wirtschafts-Verlag, 1956, 29 pp.
	Ref. 18: "A survey of continuous-wave short-distance navigation and landing aids for aircraft", C. Williams, Journal of the Institution of Electrical Engineers - Part IIIA: Radiocommunication, Volume 94, Issue 11, March-April 1947, pp. 255 - 266 ["....The aircraft component, called "Bernhardine," comprised an azimuth writing ... German Hellschreiber system....."]
	Ref. 19: "Some historical and technical aspects of radio navigation, in Germany, over the period 1907 to 1945", Arthur O. Bauer, 28 pp.
	Ref. 20: "Beschreibung und Betriebsvorschrift für Funklande-Empfangsanlage Fu Bl 1 Ex", DTA 140, C. Lorenz AG, 1940, 59 pp.
	Ref. 21: "Blind landing and airborne communications equipment. Radio and Radar Equipment in the Luftwaffe - II", S.D. Felkin, Air Ministry Directorate of Intelligence, London/UK, A.D.I. (K) report no. 343/1945, July 1945, 6 pp.
	Ref. 22: "No. 80 Wing Royal Air Force Historical Report 1940-1945", Narrative AIR41/46, Air Ministry and Ministry of Defence, Air Historical Branch, 1946
	Ref. 23: "Flugnavigation und Flugsicherung", Teil 3B "Flugsicherungs-Verfahren und -Technik: Grundsätzliches über die Technik für Flugsicherung und Betrachtung der wichtigsten technischen Hilfsmittel außer Radar", proceedings of the Internationale Jahrestagung 1958, Berlin; Verkehrs- und Wirtschafts-Verlag, 1959, 107 pp.
	Ref. 24: "Organisation des systèmes de radionavigation de la Luftwaffe en Normandie en 1944" [The Organization of the Radionavigation Systems of the Luftwaffe in Normandy in 1944], Jean-François Salles, Revue historique des armées, No. 198, March 1995, pp. 77-88
	Ref. 25: pp. 631-632 in "The Warplanes of the Third Reich" [by far the most complete reference work on German WWII aircraft], William Green, Gallahad Books, 1970, 672 pp., ISBN: 88365-666-3.
	Ref. 26: "Beschreibung und Betriebsvorschrift für Funklande-Empfangsanlage Fu BL 1 Ex", C. Lorenz A.G., Berlin-Tempelhof, DTA-14061, 59 pp. Source: http://www.cockpitinstrumente.de
	Ref. 27: "Völlig in Vergessenheit geraten: Funkanlage bei Buke", Hans-Walter Wichert, "Die Warte - Heimatzeitschrift für die Kreise Paderborn und Höxter". Nr. 24, December 1979, pp. 14-15, ISSN 0939-8686 (courtesy Mr. R.Gellhaus and H.-W. Wichert)
	Ref. 28: pp. 12-13 in "Interpretation of Wireless Installations", Vol. VII of "Development of Photographic Intelligence", Directorate of Intelligence, US Air Force Headquarters, Washington D.C. , 1945
	Ref. 29: "Instruments of Darkness: The History of Electronic Warfare, 1939-1945", new ed., Alfred Price, Greenhill Books, 2005, 272 pp., ISBN-10: 1853676160
	Ref. 30: "Drehfunkfeuer – System Telefunken; Teil 1: Verfahrensbeschreibung", A. Lohmann, October1942
	Ref. 31: "FuSAn, Peil- und Funkanlagen", Harry Lippmann, 2007, Deutsches Atlantikwall-Archiv
	Ref. 32: Air Ministry Directorate of Intelligence, London/UK, A.D.I. (K) Report no. 125/1945, January 1945
	Ref. 33: Protokoll 8 - 262 Nr. 55, geh.Kdos.ES/V/5506, 7 December 1944, Messerschmitt AG, Augsburg (courtesy R.T. Eger www.stormbirds.com)
	Ref. 34: Nachtjäger 8-262 B 2, Zeichnung Nr. S8-0084, DLH, 7.2.45, NA(PRO) MR 1-820 (1)
	Ref. 35: A.I.2.(g) Report No. 2370, dated 23rd August, 1945
	Ref. 36: E Bl 3-H schematic (courtesy Erich Werner)
	Ref. 37: "Blindlandeempf. EBl 3H", Erich Werner
	Ref. 38: "Der EBl 3 im UKW-Empfänger", C. Möller, Funk-Technik, Nr. 20, 1952, pp.556 -558 (courtesy Erich Werner, www.mw1977.de/radioecke/funktechnik)
	Ref. 39: manuscript notes of Fritz Trenkle for "Verzeichnis deutscher Bordfunkgeräte aller Art (einschl. der mit FuG-Nr. belegten akustischen und UR-Geräte) 1935-1945" and "Entwurf zu Funkgeräte Katalog Deutschland 1908-1918 und 1919-1945"; items I.2.060 C, Signatur 07815-07817 in the archives of the Technik Museum in Berlin
	Ref. 40: "He 111 H-6, Flugzeug-Handbuch (Stand August 1942)", Teil 9D, Bordfunkanlage mit: FuGX, PeilGV, FuBl I bzw. FuBl 2 und FuG 25", D.(Luft) T.2111 H-6 Teil 9D, Ausgabe September 1942, 65 pp.
	Ref. 41: "Do 217 N-1, N-2 Fleugzeughandbuch", Teil 9D, Bordfunkanlage (stand Oktober 1943), Ausgabe Dezember 1943, Dornier-Werke G.m.b.H., 94 pp.
	Ref. 42: pp. 59-63 of "Richt- und Drehfunkfeuer", Chapter 3 of “Leitfaden der Funkortung: Eine systematische Zusammenstellung der Verfahren und Anlagen der Funkortung“ ("Lehrbücherei der Funkortung: Band 1"), Walter Stanner, 4th ed., Deutsche RADAR-Verlagsgesellschaft m.b.H., 1957, 160 pp.
	Ref. 43: "Drehfunkverfahren", pp. 119-130 in “Bordfunkgeräte - vom Funkensender zum Bordradar“, Fritz Trenkle, Bernard und Graefe Verlag, 1986, 283 pp., ISBN 3-7637-5289-7
	Ref. 44: "C.W. Radio Aids to Homing and Blind Approach of Naval Aircraft", D. Quinn, R.D. Holland, J. of the IEE, Part IIIA: Radiocommunication, Vol. 94, Issue 16, March-April 1947, pp. 953-960
	Ref. 45a: "C.W. Radio Aids to Approach and Landing", M. Birchall, J. of the IEE, Part IIIA: Radiocommunication, Vol. 94, Issue 16, March-April 1947, pp. 943-952
	Ref. 45b: "Discussion on "C.W. Navigational Aids" at the Radiocommunication Convention, 2nd April 1947", J. of the IEE, Part IIIA: Radiocommunication, Vol. 94, Issue 16, March-April 1947, pp. 1022-1028
	Ref. 45c: "Discussion on "C.W. Navigational Aids" - The Author's Replies to the Above Discussion", J. of the IEE, Part IIIA: Radiocommunication, Vol. 94, Issue 16, March-April 1947, pp. 1029-1030
	Ref. 46: p. 13 "Telecommunications in War", S.A. Angwin, J. of the IEE, Part IIIA: Radiocommunication, Vol. 94, Issue 11, Nov. 1947, pp. 7-15
	Ref. 47: p. 695 in "Aidlingen, Lehenweiler, Dachtel und Deufringen: Beiträge zur Ortsgeschichte", D. Ade-Rademacher (ed.), Gemeinde Aidlingen (publ.), January 1999, 848 pp., ISBN-10: 300004521X, ISBN-13: 9783000045219
	Ref. 48: "Beiträge der Firma Siemens zur Flugsicherungstechnik und Luftfahrt-Elektronik in den Jahren 1930 bis 1945 (Teil 1 & 2)", H.J. Zetzmann, in "Frequenz - Zeitschrift für Schwingungs- und Schwachstromtechnik"    Ref. 48a: Part 1: Vol. 9, Nr. 10, 1955, pp. 351-360 Ref. 48b: Part 2: Vol. 9, Nr. 11, 1955, pp. 386-395
	Ref. 49: "Die Funklandeanlagen", pp. 45-49 in "Die deutschen Funk-Navigation und Funk-Führungsverfahren bis 1945", Fritz Trenkle, Motorbuch Verlag, 1995, 208 pp., ISBN-10: 3879436150
	Ref. 50: pp. 94-102 in "Die deutschen Funk-Navigation und Funk-Führungsverfahren bis 1945", Fritz Trenkle, Motorbuch Verlag, 1995, 208 pp., ISBN-10: 3879436150
	Ref. 51: "Die Bordfunkgeräte FuG 16 und FuG 17", pp. 28-49 in "Berühmte Bordfunkgeräte - ein Beitrag zur Geschichte der Elektrotechnik" [FuG10, FuG16, FuG17, FuG25a, FuG101a, ...], H. Sarkowski, Expert Verlag, 1983, 80 pp.
	Ref. 52: p. 15 in "The Junkers Ju 88 Night Fighters - Profile Number 148", Alfred Price, Profile Publications Ltd., 1967, 15 pp.
	Ref. 53: p. 97-103 of "Leitstrahl-Verfahren", in “Bordfunkgeräte - vom Funkensender zum Bordradar“, Fritz Trenkle, Bernard und Graefe  Publ., 1986, 283 pp., ISBN 3-7637-5289-7
	Ref. 54: "History of radio flight navigation systems", translated into English by M. Hollmann, P. Aichner, 15 pp. Source: radarworld.org
	Ref. 55: p. 61 in “Bordfunkgeräte - vom Funkensender zum Bordradar“, Fritz Trenkle, Bernard und Graefe Verlag, 1986, 283 pp., ISBN 3-7637-5289-7
	Ref. 56: page in "Do217-317-417 An Operational Records", Manfred Griehl, Smithsonian, 1992, 237 pp.
	Ref. 57: p. 195, 197 in "Confound and destroy: 100 Group and the bomber support campaign", Martin Streetly, 2nd ed., Jane's Publishing Co., 1985, 279 pp.

Disclaimer: the owner of this website is in no way responsible for the content of the websites for which a link is provided on this page.

©2009-2011 F. Dörenberg

   top of page