Use F11 on your keyboard to switch
between "normal view" and "full screen" (without browser toolbars)
The update notifications of all pages of this website are now listed here.
[general]
[mechanical
trouble-shooting] [anode-voltage
trouble-shooting]
[motor-generator and printer-solenoid testing] [motor speed calibration]
[lubrication] [drum cleaning] [miscellaneous hints & tips]
GENERAL
My Feldfernschreiber had to undergo some basic maintenance. I started this (protracted) adventure late 2008. It was a good opportunity to get to know the machine inside & out. Prime reason for the maintenance action was the fact that the motor was slowing down periodically, and the drive-train made a metal-on-metal noise – the source of the problem was unknown. I preferred to fix the underlying problem before operating the machine.
First step in narrowing down the problem was separating the motor-generator from the gear-box in the Bottom Unit. Turning the gear-box input coupling by hand required absolutely no effort, and there were no hard points, nor grinding noise. Obviously some light resistance is encountered at the point where the notches at the ends of the contact drum push out the spring-loaded follower that moves the key-enabling mechanism. Turning the tappet of the pin-coupling at the output shaft of the motor-generator by hand, did produce some "dry" noise, in particular at the top of the motor, in the speed regulator area. This meant disassembling the motor-generator of the Hellschreiber.
As I am not aware of the existence of a Hellschreiber manual that covers this type of maintenance & repair, I have contacted the corporate archive department of the Siemens company headquarters in München. They did not have anything pertinent on file, beyond the 1941 general Operations & Maintenance manual (ref. 1).
 |
It has taken me quite a while to get my courage up, and start the disassembly - with "some" trepidation at each new "layer of the onion", not wanting to do any damage to this museum piece. I captured and illustrated all the steps that I followed to fully disassemble the motor-generator, replace the ball bearings, refurbish capacitors, perform lubrication, re-assemble the motor-generator, and perform final adjustments and test. All of this is now available as a manual: "Overhaul of the Hell Feldfernschreiber Motor-Generator". The document counts about 90 pages plus some 80 photos and diagrams! It is available here (≈ 10 MB pdf file). Special thanks go to Arthur Bauer, PAØAOB, for his expert advice and moral support. His superb collection and displays of German radio equipment (including his Feld-Hell machines) and documentation has recently moved to a new location near Amsterdam; photos of several displays are here (download may be slow due to hi-res of photos of his equipment displays). One of the illustrations in the Motor-Generator Overhaul manual is the diagram below. I made it to illustrate the construction of the motor-generator, and to guide the disassembly and re-assembly. It is available as a high-resolution jpg file here (2.3 MB). |
A companion manual is "Tuning of the Hell Feldfernschreiber Tone-Generator and Filter". The 900 Hz tone generator and the 900 Hz bandpass filter of my machine needed to be re-calibrated (the tone frequency had moved to 720 Hz, due to ageing of the paper capacitors). I finished this manual today (5 April 2009) and have posted it here.
In the process, I have measured the frequency response curve of this filter, using a PC-utility (freeware) that includes a test signal generator (I used Gaussian White Noise) and a spectrum analyzer. See below.
We can determine the following characteristics from this bandpass curve:
 |
Center-frequency: ≈ 923 Hz (2.5% high, but within the ±3% spec tolerances). |
|
|
Bandwidth: ≈ 130 Hz (vs. 100 Hz according to the 1941 Manual (ref. 1), 150 Hz according to Dr. Hell (ref. 3), and 140 Hz that I measured quickly with a DVM). Note that filter bandwidth may originally have been defined as 0.5 Np (4.34 dB) down from the filter's center frequency (ref. 9), rather than today's 3 dB standard. |
|
|
Pass-band damping: ≈ 4 dB (vs. 0.7 spec); this remains to be explained… |
|
|
Pass- and stop-band ripple: could not be determined from the spectrum. |
|
|
Quality factor: Q = 923/130 ≈ 7. |
I have already posted the document "Schematic, Component-Layout and Connector Pin-Outs of the Hell Feldfernschreiber" a while ago. The title speaks for itself. This document has also been updated 15 April 2009. It is available here.
In due time, there will be a third manual: "Overhaul of the Hell Feldfernschreiber Gearbox". The motor of my machine now runs fine - by itself. When mounted back onto the gear-box, the rpm goes down after a couple of minutes of operation. I suspect that this is caused by lack of lubrication and wear of the ball bearings in the gear box. They warm up (like the motor, some of them also turn at 3600 rpm), and can no longer turn freely. Another adventure!
If you have always wondered how ball bearings are made, here is a 5-minute video that explains it all (it is in mpg4 format; here it is in wmv format).
Update May-09: have disassembled the complete Gearbox of the Feld-Hell (all four drive shafts), as well as the printer. Have added some photos on the photo page. Now I have to replace eight ball bearings. This requires that I remove a number of gear wheels. In turn, this requires that I (very carefully) remove the taper pin (conical pin) from the collar of the gears. I am making jigs for this, to guide a pin punch. The corresponding overhaul manual is keeping up with this progress. The current (incomplete) draft version is available here.
Update September-09: with "some" professional help (thanks Wolfgang & Ulrich), the taper pins have been removed from drive shafts of the printer spindle and the output to the character drum. See photos below. The taper pin of the large input gear wheel of the character drum is another issue. It cannot be removed (by design?) without damaging or destroying the hub of that wheel. The ball bearings on the drum shaft are still OK, so there is no urge to replace them. However, the opportunity arose to have this gear wheel removed, the hub to be machined away, and a new hub to be installed. More details to follow.
The drive shaft of
the printer spindle after removal of the taper pins, gears, and ball bearings
The drive shaft of
the character drum after removal of the taper pins, gears, and ball bearings
Perspective
view of the drive shafts installed in the Gearbox
(also
available as red/green anaglyphic stereoscopic ("3D") photo
here)
Purpose of the trouble-shooting was to isolate the source of the worrisome dry metal-on-metal noise somewhere along the drive-train (gears, couplings, ball bearings).
|
|
Update 29-Mar-09: finally done! See the manual above. |
||||||||
|
|
Update 7-Dec-08: while I had the motor-generator separated from the gear-box, I measured and adjusted the speed. See "motor speed calibration" below. For some reason, the motor hardly makes any noise, other than some motor-commutator noise. Not sure what I did !? |
||||||||
|
|
Update 6-Dec-08: finished translation to English of the 1941 manual. See "articles/literature" page or ref. 2 below. |
||||||||
|
|
Update 30-Nov-08:
|
The Feld-Hell on the
operating table - motor/generator detached
(strobe disk attached to the motor output)
The (non-original) motor ball bearings are of type 627: a standard metric miniature deep-groove ball bearing with a bore (inner diameter d) of 7 mm, an outer diameter D of 22 mm, and width B = 7 mm.
Symptoms:
|
|
I have replaced all ball bearings in my Feld-Hell machine, as well as the two capacitors in the base of the Motor-Generator (M-G), and a number of capacitors in the electronics box. After these replacements, the anode-voltage (as indicated by the volt meter on the Feld-Hell's electronics box), was too low: about 140-145 volt, instead of 150-180 volt (165 nominal). |
|
|
same behavior when the four tubes are removed from the electronics box, so the tubes are not the cause. |
|
|
With the M-G disconnected from the electronics box (i.e., no load on the generator), the voltage is close to nominal. |
|
|
I measured the capacitors in the base of the M-G with an ohm-meter (remove the carbon brushes of the generator, measure between each brush holder and ground = metal base of the M-G). This showed normal behavior: the measured resistance exponentially goes to infinity. |
Assuming that the collector of the generator is OK, the possible causes are the following:
|
|
surface of the collector (transition to the carbon brushes) |
|
|
carbon brushes and associated wiring |
|
|
wiring within the M-G (between the carbon brush holders and the pins of the rectangular connector that is plugged into the electronics box) |
|
|
anode-voltage load within the electronics box (capacitors; the tubes have already been eliminated as a possible cause). |
In order to isolate the problem to the M-G vs. the electronics box, I needed a separate power supply for the anode-voltage. The Feld-Hell generator produces 25 mA max at 150-180 Vdc (nominal 165 Vdc). This happens to be about the same as what is required for 4 or 6 Nixie tubes. I searched the web for "nixie power", and found several schematics for simply power supplies. Ref. 6, 7, 8, 10. One of these designs is available commercially (late 2010, kosbo.com, ref. 8). It is a tiny card (2½ x 4 cm, ≈1x1½"), with a 180 Vdc output at 21 mA. The voltage is adjustable with a fixed resistor. It also has a regulated 5 Vdc output. I don't need the latter, so I hooked up an LED "power on" indicator to it, via a 470 Ω resistor.
The tiny 150-180 Vdc
+ 5 Vdc Nixie tube power supply
The generator voltage is applied to the electronics box via one of the two 6-pin rectangular connectors that are plugged into the left-hand side of the electronics box. The 12 Vdc to the M-G also passes via this connector.
Two 6-pin rectangular connectors plug into the left-hand side of the F-H eBox. The rear one (with the green dot) connects the character drum contacts and the printer solenoid to the electronics box. This connector remains plugged in: the anode-voltage powers the solenoid. The second connector (with the red dot) carries the 12 Vdc to, and the 165 Vdc from the Motor-Generator. For stand-alone testing of the electronics box, this connector must be unplugged.
Regular banana plugs can be (loosely) inserted into the connector of the electronics box. To avoid plugging the external anode-voltage into the wrong socket, I decided to make a simple full-proof (and hopefully fool-proof) insert. I used a 2½ x 10 cm (≈1x4") piece of an 8 mm thick plastic kitchen cutting board. M5 screws with self-locking nuts serve as guide pins. Two banana plugs serve as connector pins. I drilled pin holes into the side of the plastic insert, through the banana plugs, and inserted a finishing nail into each hole. This fixes the banana plugs into place, so they don't pop out when plugging in the insert. Note that the nail heads must be sufficiently countersunk, to avoid shorts between them (e.g., via the housing of the electronics box). To avoid an electrically shocking experience, I coated the heads of the banana plugs with a dab of standard liquid electrical tape.
Dimensions of the
plastic insert
The Nixie power-supply with 12 V battery and connector insert
It is important to use a separate 12 volt source for the electronics box and the anode-voltage supply. If the same 12 volt source is used for the electronics box and the anode-voltage supply, the fuse will blow, when switching the main switch back and forth between "Aus", "Bereit", and "Ein". Note that the 165 VDC return and the +12 VDC are interconnected inside the electronics box!
I use a small motor-cycle battery, though a second 12 volt power supply should be fine as well. Also note that the red light on the electronics box will not light up in the "Bereit" position, if the connector to M-G is not plugged in – this is normal.
Now we can run the test!
The standard procedure is followed for powering up the Feld-Hell:
 |
12 volt is hooked up to the electronics box |
|
|
the main switch is switched from "Aus" to "Bereit". |
|
|
Wait 1 minute for the tubes to warm up. |
|
|
The anode-voltage supply is hooked up and powered up. |
|
|
The main switch is switched from "Bereit" to "Ein". |
Now we check the anode-voltage with the voltmeter of the electronics box (push the blue button). The voltmeter indicated 180 volt. This is without the load of the printer solenoid. With the M-G supplying the anode-voltage, the voltmeter indicated 140-145 volt...
I used IZ8BLY Hellschreiber software in "tune" mode to generate a constant 900 Hz tone. The output of my PC's soundcard was hooked up to the "Empfänger" connector of the electronics box (the La-Lb/E connector will also work). With the constant tone, the printer solenoid is energized continuously. With this load, the anode-voltage drops to 140-145 Vdc. Note that my little anode-voltage supply is limited to 21 mA, whereas the M-G can source 25 mA.
Conclusion: the problem is not in the electronics box. Troubleshooting is to be continued...
TESTING OF THE MOTOR-GENERATOR AND PRINTER SOLENOID
I needed a way to test the motor-generator and printer solenoid, without the Feld-Hell's amplifier box. So I basically needed a circuit that mimics the switching action of the motor speed regulator tube and of the solenoid driver tube. This is easy to do with a high-voltage MOSFET.
One side of the motor's governor windings is tied to the +165 Vdc of the generator. The other side is connected to the anode of the motor speed regulator tube. When the tube is turned "on" via the control-grid voltage, the cold side of the governor windings is pulled down to about +12 volt. The printer solenoid is hooked up de same way, and the solenoid driver tube also pulls it down to +12 volt. The tester circuit should replicate this.
The schematic below shows how this can be done. I used the general-purpose MOSFET IRF640 that is readily available and inexpensive. It can switch 200 volt and handle 16 amps. The loads are normally driven by the anode of the associated control tube; max anode current of the RV12P4000 tube is 3 mA. A suppression diode is connected across the windings. It should be able to handle several times the applied voltage (153 volt). So I selected the 1N4007 diode for this. It can handle 1000 volt, and is inexpensive.
The circuit includes switchable loading resistors for the generator (no load, 11 mA and 20 mA load). The volt meter shows the generator voltage under the selected load condition.
Test box for the motor-generator and printer solenoid
(click on image or
here to get full size)
CALIBRATION OF THE MOTOR SPEED
While I had the motor-generator separated from the gear-box, I measured its speed with the cheap-and-simple tachometer (4 components, $2-3). This tachometer is described on my home-built Hell printer page. I attached a simple paper strobe-disk to the tappet of the motor.
At the nominal setting of "5" on the speed adjustment scale, I measured 3712 rpm with an oscilloscope: 247.4 Hz = 247.4 periods with 4-segment strobe disk, which corresponds to 274.4/4 = 61.8 rps = 3712 rpm. This is within the 3600 rpm +/- 180 nominal adjustment range. As I had the top of the motor open for picture taking, I decided to calibrate the speed. Now it is basically at the nominal 3600 rpm, which corresponds to 240 Hz on the scope (see below). This only took very minor adjustment of the adjustment screw - about 1/8 of a turn.
The speed adjustment is done per line item 106 of the manual (see ref. 1 and 2). The speed adjustment screw is installed radially on the speed regulator plate that sits on top of the motor shaft. It is accessed via any one of the four holes just below the speed scale tape around the top of the motor. I turned the pin coupling at the bottom of the motor-generator by hand, so as to fully align the (large round, slotted) head of the adjustment screw with one of these holes. Obviously the actual adjustment takes some trial-and-error, as adjustment can only be done when the motor is not turning! With the motor already removed, the whole procedure took just a couple of minutes. The standard procedure (with motor-generator installed and speed scale set to "5") calls for adjustment with the pause-character key engaged, until 50 pause-characters are sent in 80 +/- 2 sec
The scope indicates
3600 rpm (240 Hz with 4-segment strobe disk)
- scope is set to AC-coupling -
The 1941 Feld-Hell manual (ref. 1) calls for four different kinds of oil and grease to lubricate the drive-train:
|
|
"Wählschienenöl Shell P 37" – (teleprinter and typewriter) selector bar oil |
|
|
"Shell Ambroleum" |
|
|
"Rostschutzfett 40" - rust inhibiting grease nr. 40 |
|
|
"bestes Kugellagerfett" – highest quality ball bearing grease |
According to the manual, "Waffenfett" – gun/rifle grease – may be used instead of Ambroleum. Production of Shell Ambroleum ended decades ago. It was replaced by Shell Retinax G, which is a "Getriebefließfett" – a soft, free-flowing gear grease. Retinax G was manufactured until ca. 2002. It is still used today, by vintage motorcycle and automobile buffs. Siemens also recommended it for teleprinters such as the T100. According to specialists at Shell Lubricants Technical Information Centre, Ambroleum can also be replaced with a 1:1 mixture of Shell Retinax CS 00 (still manufactured) and a monograde engine oil of type Rimula R3+ 30. Apparently Addinol SGA 600 is very similar (it is a sodium-soap gear-grease based on mineral oil).
As a substitute for "Wählschienenöl" I use sewing machine oil. It is a highly refined, light weight mineral oil, without additives, and intended for high speed precision machinery. As it is inexpensive, get a decent brand at your local sewing machine store or fabric store. Do not use WD-40® for lubrication. As nice a product as it may be, it is primarily designed as a penetrating cleaner/degreaser, to loosen stuck parts, prevent corrosion and displace water ("WD-40" stands for "Water Displacement - 40th formula"). It is explicitly not intended for use on parts that move continuously! Some use Custanol-F ("F" stands for "Fernschreiberempfängeröl", teleprinter receiver oil, but it was even used in computer disk drives); it is a stable, non-gumming blend of paraffin oil and neatsfoot oil. But alas, its manufacturer Firma W. Cuypers in Radebeul (near Dresden/Germany) has ceased production in the late 1990s (after having been absorbed as Werk 2 (plant nr. 2) by the VEB Arzneimittelwerk Dresden pharmaceuticals conglomerate).
Both Retinax G and Custanol-F still turn up on eBay once in a while (2009)...
The manual of the Hell-80 machines recommends the following lubricants:
|
|
"Fernschreibmachinenöl" – (teleprinter oil)
|
||
|
|
"Fernschreibmachinenfett" – (teleprinter grease)
|
In September of 2009 I got a hold of a copy of the 1940 Feld-Hell lubrication manual (ref. 3). According to this manual, the (non-shielded) ball bearings are to be greased once a year with "Fernschreiber-Motorenfett" (teleprinter motor grease), after cleaning them with "Benzin" (petrol (UK), gasoline (US)). No particular brand or type of grease is suggested. Grease is to be applied to designated points four times a year, whereas oiling is to be done with "Fernschreibmachinenöl" – (teleprinter oil, see above) every 100 operating hours (but at least once a month).
The required regular/scheduled Feld-Hell maintenance basically normally consists
of:
|
|
cleaning and lubrication of 50+ points identified in the 1941 manual (ref. 1, 2, 3); note that this does not cover the ball bearings of the motor-generator, as that was deemed to be well beyond field maintenance. |
|
|
cleaning the character drum, such that keying of the character pixels has little or no contact noise/crackling/bounce |
|
|
if necessary, adjusting the motor speed regulator relay contacts, such that the motor is running at nominal speed when the 0-10 manual speed adjustment scale is at 5. |
Important:
|
|
read the manual first, and follow it (yes, I know, this is unnatural for engineers and techies). |
|
|
to prevent damage to original equipment, always use screwdrivers that closely match the screw head, are not worn out! Same for wrenches and other tools. |
|
|
only use acid-free lubricants. |
|
|
when replacing an electronic component, leave the original in place (de-soldered on one side) and put the replacement part across it. If necessary, put shrink tube or similar isolation on de-soldered component lead. |
|
|
when de-soldering wires (e.g., as required to get into the electronics unit of the Hell Feldfernschreiber or when disassembling the motor-generator), put a numbered piece of shrink tube on the wire ends (and if necessary, write down where the number is supposed to be reconnected, or - better still - mark up a photo. |
CLEANING OF THE CHARACTER DRUM
As I had the protective cover of the drum already removed (just two knurled-head screws), I had already noted that the conductive patches of the drum looked dirty. Another way to tell whether cleaning is needed, is by listening to the own keyed-tone. It should be clean, without crackling. If the drum is dirty (which happens during normal operation), the own monitoring print will also show signs. Note that the 900 Hz voltage across the contacts is very low (only several tens of millivolts), so it does not take much increased contact resistance to become noticeable. This is different from using a pull-up voltage of several volts from the key-input of a CW transmitter, as is the case when diverting the connection of the keying contacts to the 12-pin round connector on the front of the Amplifier and Interconnect Unit.
Cleaning of the character-drum is described in line item 98 of the manual (ref. 1, 2 above). It is a simple procedure.
Never turn the drum by hand if it is installed on the Bottom Unit and therefore connected to the gear-box! If the keyboard-Drum Unit is separated from the Bottom Unit, then the drum is simply turned by hand. If not, the machine must be powered up such that the motor turns, or the motor must be removed and the input to the gear-box turned by hand.
Simply use a lint-free cloth and some denatured alcohol ("spiritus" in Dutch andGerman), or special contact cleaning fluid. The manual actually calls for kerosene (US) or paraffine oil (UK). Do not use any aggressive cleaning products as this may damage/dissolve the drum.
Quite a bit of crud came off the drum, and afterwards, the contact rings looked nice and clean.
As I perform more maintenance on my Feld-Hell machine, and/or learn from other Feld-Hell owners & operators, I will post miscellaneous hints & tips here.
|
|
Whereas WD-40® should not be used for lubricating Hellschreiber parts, it may be used (lighly) on felt ink rollers, if (oil-based) ink has dried out and the roller has become hard. Thanks for the tip Andrew (SM6MOJ). |
|
|
When not in use, tightly wrap the ink roller in plastic food-wrap film from your kitchen. This will slow down the drying-out of the roller. |
REFERENCES
|
|
Ref. 1: "Der Feldfernschreiber", document D 758/1 of the Oberkommando des Heeres, Heereswaffenamt, Amtsgruppe für Entwicklung und Prüfung, Berlin, 1 April 1941, [this is the official original manual in German for Feld-Hell "T typ 58" (drawing 24a-32 (a1 &a2))] |
|
|
Ref. 2: "The Feldfernschreiber", translation by Frank Dörenberg, N4SPP, of "Der Feldfernschreiber", 2 May 2008. |
|
|
Ref. 3: “Öl- und Schmieranleitung für den Siemens-Hell-Feldfernschreiber“ (lubrication manual), Fernmelde Technik, Siemens & Halske A.G., Wernerwerk, SH 7976 5.6.40 T T1 (C/0921), 7 pp. (courtesy Siemens Corporate Archives, München) added 25-sept-09 |
|
|
Ref. 4: "Die Entwicklung des Hell-Schreibers" by the inventor himself: Rudolf Hell; pp. 2-11 in "Gerätentwicklungen aus den Jahren 1929-1939", Hell - Technische Mitteilungen der Firma Dr.-Ing. Rudolf Hell, Nr. 1, Mai 1940 [in German] |
|
|
Ref. 5: p. 10 in "New methods for the evaluation of the lubricating ability of oils and greases", post-war translation to English by the US Combined Intelligence Objectives Sub-Committee (C.I.O.S). of "Neue Methoden zur Bewertung der Schmierfähigkeit von Ölen und Fetten", by E. Heidebroek, Reichs-Verkehrsministerium, Zwischenbericht Nr. 83, April 1940 (C.I.O.S. ref. no. SF3). |
|
|
Ref. 6: "Nixie HV Switching PSU", Nick de Smith (designer of the Nixie PSU; schematic "borrowed" and commercialized by kosbo.com) |
|
|
Ref. 7: "Build a nixie power supply" [based on an NE555 timer], kit manual from LEDsales |
|
|
schematic of simple 120-180 Vdc switched power supply (based on 555 timer IC) |
|
|
Ref. 8: "NIXIE Tube Power Supply Module V.02", 2-page product description and schematic from kosbo.com; MAX1771 IC. |
|
|
Ref. 9:
pp. 170-172 in "Hochfrequenz-Nachrichtentrechnik
für Elektrizitätswerke", 2nd ed., Gerhard Dreßler, Heinrich-Karl
Podszeck, Springer Verlag, 1952, 183 pp.
|
|
|
Ref. 10: "Nixie power supply" by Mike Harrison [based on the MAX771 IC, the (obsolete) predecessor of the MAX1771]. |
©2008-2009
F. Dörenberg. All rights reserved worldwide. No part of this publication may
be used without permission from the author.
It has taken considerable effort to create these pages. If you "borrow"
content from them, at least reference the source.
top
of page 
