E97. Power supplies for German WW2 equipment

See the normal dc-dc inverters on page 28a and 28b, also NA6a and NTG2 on 28c
Build your own mains power supply, see page
e96, and for 15W.S.E.a on page 31a

Helge LA6NCA's 12V DC-DC mobile power supply unit for receivers and smaller transceivers

Output voltages:
Filament: +1.2....10VDC (regulated)
Anode voltage +60....150V (regulated) - galvanically insulated from ground
Grid bias voltage 0....-10VDC
(high impedance)
The PCB measures 100x43mm.

Earlier version of the DC/DC-converter. We have forgotten the transformer winding details, and this version is now discontinued.

Helge's inverter mounted in a box - the way I want it - with some extra noise filters and 12V input polarity protection relay (the so-called PYE trick). It might be a good idea to add some fuses, but you know, the constructions are usually decided during the mechanical attempts, and another box was attempted first, but was found too small. It is also an LED to indicate that the relay is operated. Also connected a 10k resistor from minus-side of 100V supply to ground, and 100k from positive side to ground to decharge the capacitors. Minimum current for a receiver like Torn.E.b is 10mA (on 90v) with -3V bias, so a 10K resistor hardly changes the bias voltage.

The circuit diagram showing the external components. The relay is any cheap type, but it is important that you check that it will operate safe with 8V over the coil (to make room for supply voltage variation).

The problem with this construction is that it is used SMD components, and after assembling a douzen boards you are fed up, it seems easier to use normal components, another problem is interference. So a new version is under experiments, but it seems not to be any problems with the filament supply, so the old construction will be kept for this part. A 100Hz inverter using conventional 50Hz transformers will be used in the next version, with voltage control, and it seems to operate ufb.

With some more experience it was found desirable to add a 12V DC output fed just from the input wire, to connect 12V heaters on auxialliary equipment which could operate with the same HT+ as the other equipment.

In one occassion it was problem with interference via the filament voltage. It is quite easy to make improvement. Since mainly dc current flows in the coil toroid cores may have little effect for chokes where magnetizing current may be somewhat high, ferrite cores with air-gap or straight cores must be used here. For the 90V circuit toroid cores may work provided that the current from positive and negative side of the supply voltage have opposite magnetic fields. A simple choke may be wound on a short piece of ferrite core from ferrite antenna.

Winding 25 turns in 2 layers of 1.5mm enamelled copper wire on a 35mm long piece of 10mm diameter ferrite core (ferrite antenna) is quite useful as choke, it has 27µH, and it has 17 ohm impedance on 100kHz. So it is quite useful for filament circuit with this inverter. But since you don't need so thick wire, 1mm thick would be far more than you need, the impedance could easily be increased to 35-50 ohm with the more proper wire.

2nd DC/DC-inverter (+1.5...10V filament supply, +60...220V anode supply, 0...-10V grid bias).

The latest LA6NCA power supply mounted in Pac Tec CM6-300 Kit Electronic Enclosure.

To avoid winding the transformer, a standard type 110+110V to 9+9V AC transformer for pcb mount is operated on 100Hz from push-pull mosfets with a 12V input DC-regulator. With +12,2V input the output voltages may be varied in the region of 60...120V and 120...240V, dependent on whether one or two 110V windings are chosen. The voltage regulation may be poor close to the upper limit. I hooked up the polarity protection relay (RLA1) and extra 12V input filter of choke (75µH - 10 turns - on toroid core with air gap) and 22µF capacitor. The minimum current seems to be around 300mA.

Idiot protection. Everybody becomes idots, particularly in the field when it is dark. Used a 12V 1,3kW relay for the input polarity proction, 4x2 contacts, and to lower the operation voltage one section was disabled, the operation voltage dropped to 9V. Since the relay shall operate before current is drawn a small relay is sufficient, and all systems are connected in paralell. An LED is mounted on the front to indicate that the relay is operated.

Cooling arrangement. With 12,2V DC input and 30mA load on 97V HT only a smaller cooling fin (3x5cm) is necessary for T3, but mounting a cooling fin for this transistor should be considered when mounting the pcb, it might prove neccessary later.

Mode selector. It was found convenient to mount a rotary switch (S1) on the front (instead of using the jumper on the pcb) to choose between open and grounded HT- (A-), the switch ought to be rotary type, and I found one which is only operated with a screwdriver, but should have clear markings on the front which position it is in. It is a 4k7 resistor on the pcb and the regulation depends on this resistor, it limits the negative swing, but for German receivers where the maximum voltage drop is -3V over a smaller resistor, this is not important. It applies for KwEa, LwEa, Torn.E.b, Ukw.E.e, Ukw.E.h and many more.

Current limit. It is no real short-circuit protection, and I am not aware of what the maximum ratings are. But for my own applications typical fuse values could be 50mA for HT+ and 3A for filament voltage. To increase pulselength of a short-circuit a smaller resistor is added on the +HT, the value is chosen such that the voltage drop is not important for the operation.

Extra negative bias voltage. Some equipment need an extra negative voltage, usually -3V, this applies for Torn.Fu.b1, Torn.Fu.c, Torn.Fu.f, Torn.Fu.d2, E381H, E381S. Note: E381 also needs -1,5V, but since it is no current drain, a simple voltage divider could be used. Torn.Fu.b1 is specified to have -4,5V, but it is not critical, probably is -3V more correct since every other comparable equipment use this voltage.

Output connections. Although provided for separate connectors for the different voltages, it is a bad idea to use them for standard connection. Fixed cable should be used as often as possible. It only cause few problem, and it is mentioned below how avoid the problem with Torn.Fu.d2.
It is some individual opinions about colours and voltages, my suggestion is
RED for A+ (or HT+), BLUE for A-, GREEN for H+ (filament), BLACK for H- or ground, and YELLOW for -G (neg.bias). Normally the colour for negative voltage would have been blue, but it seems not to match here, but the connection is rare.
With some more experience it was found desirable to add a 12V DC output fed just from the input wire, to connect 12V heaters on auxialliary equipment which could operate with the same HT+ as the other equipment.

The solution so far.

The DC/DC-inverter under test

You may wish to see the originator's comments about the
dc-dc inverter

Suggested mains power supply for Kw.E.a, Lw.E.a, Fu.H.E.c, Torn.E.b etc.

Apart from the problem of finding a suitable box to house the power supply unit......, the worst problem is to find a suitable transformer without paying too much, but many transformers are found as surplus. So if you start with a transformer to say, 14V and buy one to transform back to 100-120V for the anode supply, then LA6NCA's inverter will solve the heater problem. It may also provide an additional negative bias voltage of 0...-10V. The original supply is discontinued because we have some interference problems to solve, but do not experience such problem with the half part which delivers filament voltage.

See Norwegian language articles about mains power supply units for these receivers on page e96

Got some interesting data sheets from Hans Jürgen Keller / DH1AB; Thomson (ST) VB408 high voltage regulator which may be used in series regulator from +1.5....370V 40mA output (max input +400V), 89W max power for the TO220 version. It should be quite suitable
for most Wehrmacht receivers.

Suggested values for 90V output. D=1N4004, R1= 1kW, R2= 68kW, C1= 1µF, C2= 10µF 150V, estimate 120-150VDC input. 130V output for R2=100kW, with respectively higher input voltage.

Another idea for making supplies from +12V is found in TNC2 or MFJ1270/1270B, it uses an NE556 type dual monoflop device to produce a dual phase signal which is rectified in a two-way bridge, or fed to an amplifier stage. Note that 1N4001 or 4002 works, I've tested it on 10 and 20kHz switch frequencies and couldn't see any problems with output current up to 50mA

Regulated mains power supply for Luftwaffe receivers (EZ6, E10L, E10K)

Power supply for EZ6, E10K, E10L and similar Luftwaffe receivers

EZ6, EL10, and EK10 require 25.2VDC and +210VDC. Perhaps not too critical. They will work with 12.6V heater - without any hum problems, but the motor in EZ6 will of course not operate with AC supply. It isn't really neccessary to use stabilized anode supply, but the problem is to find suitable mains transformers,
so I decided to use a regulator. Have already built supplies with EL86 type shunt regulators, but with the available components it seems better idea to use a series regulator. LA4OE runs his EZ-6 with 150V supply voltage and it still works fine.With +200V supply EZ6 draws 38-40mA anode current, 2mA more with BFO, but less for strong signals.
EZ6 motor supply isn't solved, and I haven't investigated into running it, it might operate with a simple rectifier to +24V, but it is only a cosmetic problem - as long as you don't have any of the DF antenna system to go with it, you have no need for installing the motor.

Considered some different regulators. A valve is better if the voltage drop over the regulator is large, but today you can always find some mosfets to use. Problem with high voltage BJT's is that they have very low current gain, and only few have in the region of HFE=10 (except MJE340), so you need to make a darlington circuit. I will eventually build such regulator later, but this was made 8 years ago dependent on available components. Philips EL86 has proven to be a very useful valve since it will draw 100mA triode connected with only 100V, it is difficult to find another valve to do the same, with similar size and heater current. Other valves requires higher screen voltage, but this is always a problem when limited supply voltages are available..

In my construction the voltage from the rectifier varies between 250-290V, it means that the regulator have minimum 40V across it, and maximum 80v. So unless you don't shortcuit the output it is no real danger with a 100V device, you may also use a zener diode across it to protect it.
I used a Radio Spare device which stands at least 200V, but I have no data.
Another problem with shortcircuit is that the small transistors Q3, 4, 5 are very likely to have secondary breakdown. Even the old type 2N398 will work satisfactorily. The 1k resistor in series protects against such, and it is always better to cascade lower voltage devices instead of using for a 300V type. I have lots of suitable HV-devices, but find it more interesting to prove that low cost devices will work just as good provided you take proper steps.

It is very important for valves to have soft start in heater supply, it may increase the life-time a lot, so I still hope to have the over 60 years old valves for many years. The relay was measured and operated with the current shown. You just look for a higher voltage relay and make some tests to find the optimum relay (which draw as little current as possible, often AC relays will give best results in such respect).
In order to avoid damages if output is shortcircuited a smaller value resistor is inserted in series with the fuse, it will slow down the transient so the fuse have time to blow before the power supply.

See more notes about these receivers on page 23a

Power supply for Ukw.E.e/Ukw.E.h (+120...130V 30mA, and 12,6VAC 2A)
Netzgerät für Ukw.E.e/Ukw.E.h (+120...130V 30mA, and 12,6V 2A)

Got a power supply with the UkwEe many years ago, and when I opened it it was quite a surprise that nothing had been shortcircuited, and another surprise was that the 120V DC was 210V which is quite a lot for this kind of receiver. It is always a problem to find boxes so I decided to keep this ugly box - even the selenium rectifiers since they seem to work properly, but I needed a regulator to drop the voltage. It is an advantage with cardboard or wooden boxes, but it was really not much space inside. The best solution seems to mount the voltage regulator on the back. Wanted some VB408, but it would take time, so I decided to try the insulated type BUT12AF which don't need any mica dish and one avoids the problem to consider if the mica is good enough for the voltage.

Since many of the UkwEe's have bad paper capacitors which may shortcircuit, a current protection seems to important. It is set to 40mA. In spite that the power may not exceed 1W a large cooling fin was put on the rear, for practical reasons this is -A plane which means it may swing to -3V and as such is not a problem (see explanation further down the page). It Is a a problem with BUT12AF or BU508 that the HFE is rather low, only 10. This was measured before starting the construction. Probably would an MJE340 be better in this respect, but again it needs mica dish with the problem of flash-over to ground. Zener diodes were taken from what was available to make an output voltage around 130V and stand 5mA current. Since zener diodes are quite noisy a 0,47µF decoupling capacitor is connected across them, but to avoid peak current in the base circuit if the output is shortcircuited a diode gate is used toward the gate. Voltage regulation is not important for this type of regulator, and UkwEe receiver has its own 95V neon type regulator for the local oscillator. The voltage drop resistors were calculated to have the minimum current neccessary, and still not burn with shortcircuited output. Although not planned, all components could be mounted on the rear late, with a square hole so the power transistor could be screwed to the cooling plate - remember thermal compound.
Used pieces of pcb laminate to support the construction - for soldering points - with at least one screw and araldite glue.

A more ideal power supply for checking the state of old receivers.

Having built the unit shown above I soon discovered how it would ideally have been. As already mentioned the paper capacitors are likely to be shortcircuited, and in the next receiver to be tested two capacitors were shorted, one for +130V supply voltage and another for audio amplifier screen voltage. If a 40mA lamp is connected in series with the power transistor it would light up when current approaches this value, and for normal current which should be 30mA it will only glow. The lamp voltage is not important. But this was not the only short, the on/off switch was wired incorrect, shorting the heater supply, and the transformer got very hot. I didn't find the fault before I connected another 12VDC supply with a 10ohm series resistor. Such shortcircuit indication could have been done with a smaller car lamp switched in series with the heater voltage.

Considerations for using the same power supply connection/cable with different equipment (note the different heater current).

+2V 0,8A
+90V 10mA

KwEa: +2V 2,2A
..+90V 20...25mA
LwEa: +2V 1,8A
.......+90V 20mA

+2V... 1,9A

+2V 1,14 / 2,25A
+135V 20...25/30...45mA
-G: -4,5V (Sollwert)

+2V 0,89 / 1,14A
+135V 25 / 40mA
-G = -4,5V ?


Some silicon diodes are added (modification) for Torn.Fu.d2 (unfortunately not equal to Torn.Fu.b1 / c / f.

It was considered some different use of pin 3 on the power supply. When it is called +Umf it is supposed to switch a relay or other in the power supply when the receiver is in use, see figure. I am not sure if I should have seen such application mentioned. It may still bet available, but if a diode is added, the same pin 3 can be used for negative voltage Torn.Fu.d2 without causing any damage when different units are connected using the same power supply and cables. When it is used for negative voltage it is no current consumption, so it can be kept relatively high impedanced, say 1000W or higher, and and the circuit requiring positive voltage from the equipment may use an isolating diode. The returned voltage from KwEa, LwEa and other are reduced to 1,3V, and it may be difficult to find a suitable relay, so a transistor may be added to switch a relay, but now one may use +8V from the low voltage rectifier.

For NA6 the anode voltage is limited to 90...100V, but it is not so bad, the transmitters will still work. See also notes for NA6/NA6a on page 28c . For LA6NCA's power supplies this is no problem, it is possible to achieve stabilized voltage up to +200V with his new circuit.

Some more power requirement for equipment:

Lo1UK35 (SE 42444/c):
Empfang .........+2V 1,3A, 130V 40mA, -3V Sollwert
Telefonie: ....... +2V 1,7A, 130V 50mA, -3V
Telegr. tönend: +2V 1,7A, 130V 50mA, -3V
Telegr. tonlos:. +2V 1,7A, 130V 60mA, -3V

E381S (E381H) power requirements:

+3,8V 0,35A
100V 20mA
-1,5V zero current
-3V zero current

The power supply mentioned above is suitable for this receiver (since filament voltage may be adjusted up towards +12V).

The same cables will also work for Ukw.E.e and Ukw.E.h and other equipment.

See list of mains adapters for possible used to transform back from LT to HT supply voltage on page p31

Some RV2,4P700 just bought from the local surplus shop.

Socket connections. Note that suppressor grid for RV2P800 is not connected to the filament inside the glass tube, but inside the socket/tube, so it is possible - according to SM6HYG Carl-Gustaf to reconnect it to a free pin on the socket, suppose the same applies for RV12P4000 (?).

German power connectors  
Male and female type
for +90 (+130V),
+2 (12V) and -3V
1) Another type Wehrmacht
cable connector
2) Wehrmach connector
with larger holes
3) AC mains power connector
for Lo6K39, Lo6L39

Back to index for German receivers (pg11a)

Last update: 2003-06-14