Nachtfee revival


Page initiated: 7 January 2021

Current status:   5 May 2021


Contribution: A (8 January 2021)    including YouTube films


Contribution B (II)  (12 March 2021)    including YouTube films


Contribution C (IIb) (19 March 2021)


Contribution D (23 April 2021)


Contribution E  (III) (5 May 2021)


Contribution F  (24 May 2021)




Hans Goulooze started first with removing the rotary converter

This motor has two functions: to supply the 135 Hz AC of ca 16 V, which is feeding the wobbling Ferraris motor at the front-end of the receiver, as well as supplying the ac voltage for driving the HT transformer; which actually is being fed onto an air-tight transformer.

On the left-hand side we notice the transmitter (TX) and the receiver (RX) sections.

On top on the right-hand-side we see the mechanical coding module; where eventually two metal keys can be inserted.



Down of the coding-module (right-hand side) we recognise a vertical shaft with a small slit; which is receiving its mechanical rotation from the 'hook-gear' attached onto the rotary converter

However, there are two open wires visible of which function we yet have no idea.  Maybe it is only to be connected onto a capacitor.  The valve top right of the mechanical coding module is type RG 12D60, and is used for the HT of the receiver as well as the driver stages of the transmitter, though excluding the anode voltage of the LS 50 transmitter stage.

In our perception,  it would be practical to compare the situation existing in our functioning FuG 25a.



Four wires: number 1 and 2 being used for the supplying 24 V dc; the other two are providing ca 16 V 135 Hz ac

The small shaft up is feeding the coding mechanism. 



The transparent sheet has to be used for re-establishing (remake) the various Al cover-plates; as it is always quite delicate to determine exactly where the many 3 mm screw-hole should be drilled



Hans is demounting the hook-gear arrangement, as to get access to the 'hampering' tiny driving (feeding) shaft




When you see it, those used with detaching such a tiny clutch will acknowledge that it is quite likely that he motor-shaft-end will likely be bended!

Please look carefully: maybe you recognise that there exist two dots. Of which one is a tiny conical rot fitting the clutch onto the driving shaft.

However, the second dot is ceiling off the fixation of the clutch onto the tiny driving shaft.

The nuisance we encounter here, is, that inside the motor collector is reaching inside the Bakelite top plate, keeping the bearing inaccessible. 

Hans found, nevertheless a means to get some lubrication inside the invisible bearing.


With a bended 'injection-needle' he sprayed 'molycoat' (graphite like lubrication) and benzine (view the plastic bottle) between the rotor and the bakelite plate.

Afterwards, he managed by using a hot blower to let vapour the benzine and thereafter compared the current consumption of the motor, where the current increased a little bit. Which might be an indication that the concerned bearing being lubricated.



Maybe the hampering dilemma is better visible this time



I hope you understand in which way the coupling clutch is functioning

According Hans, it is rather astonishing that after >75 years have been passed, that the lubrication found inside the bearing wasn't 'sticky' or solid.



Re-assemblage, demands some care, but isn't a real problem



The mechanical driving power enters via the slit on top of the vertical shaft


YouTube films

Film 00063:    Recording date was: 7 January 2021. Operating voltage 24 V at 1.3 A. The AC loading resistor lays on the right-hand side. Viewing on the CRT screen the output voltage 16 volt ac at 135 Hz. Down on the right-hand side the dual-key coding module.


Film 00064:    Viewing briefly the mounting place of the rotary converter of the FuG 25a transponder.




Contirbution B (II)

On 11 March, even preliminary partially on 18th February, we continued with progressing the revival of our, FuG 25a project, in the context of our Nachtfee revival Survey.


Our second FuG 25a (spare) module is undergoing its first electrical tests

For it we use so-called extension cables, which do fit onto most German Air Force electronic gear; home made for decades; though the connectors on both cable ends are genuine.



Viewing it from a different perspective

The black cable is only matching - because we neglect the antenna connections and are exploiting the fact that the two adjacent pins aren't used.

Purely electrically - this is just now sufficient. 



Maybe this set-up explains wy we operate the two sets of extension cables

However, lacking a link onto the antenna.



Hans had to find a solution for the 'bipolar' electrolytic capacitor, which has to supply the 90° phase-shifted current onto our Ferraris wobbling motor, and proved to possess insufficient capacity

The purpose of the wobbling motor is quite essential, as it wobbles the 1 oscillator frequency as to increase its system capture spectrum.

It sweeps the 1st local oscillator for ca. 8 MHz around 124 MHz; say between 120 - 128 MHz (receiving spectrum).

Please bear in mind, all accomplished in the era where phase-lock was hardly existent at VHF, in the early 1940s.

AOB, additional notice: Hans Goulooze used afterwards our Hallicrafters S-27 receiver and found - that the wobbling sweep reached instead ca. 3 MHz; consequently the receiver capture is then, when we still think of 124 MHz as being our centre frequency, which isn't certain - between 122.5 - 125.5 MHz. A bit small in my perception.

We have to go deeper in the true state of affairs later.



The insufficient capacity which hardly reached 4 ą 5 µF - instead of 10 µF;  hermetically sealed-off type - has been just replaced and already build-in down on the left-hand side

On top, just visible we read: Betriebstemp. - 40 + 70   °C

Luckily, Hans bought somewhere in the past some bipolar electrolytic capacitors, which he built-in the original housing; visible just down on the left-hand side.

In case he did not have had one at his disposal, in our current corona days of spring 2021: he could have taken instead: - two electrolytic capacitors in series: but either in the centre + + together in series, consequently both ends constituting 2 times minus poles;  or  -  - connected together in the centre and at the connection sides two times + + poles. 

After all, it does make sense to consider its actual circuitry  


Mw 1 constitutes the Ferraris motor; please notice YouTube Film    00074

 The bipolar capacitor shown in the foregoing photo is C 62 connected onto the Ferraris motor Mw 1

It is phase-shifting against the supplied ca. 116 Hz ac voltage 90°.

Its function is to wobble the 1st local oscillator frequency in such a way - that the receiver sweeps the frequency range of approx. (120 - 128 MHz)  (maybe 122.5 - 125.5 MHz?).

The wobble rhythm is ca. 8 Hz.

Please notice that IFF signals are consisting of short pulses and its duration is such - that during the receiver signal sweeps through the bandwidth of the receiver - still a recognisable signal can be detected.

The 'Schwing-Röhre' is triode type LD 1, a valve which can operate up to about 800 MHz. 

Rö 1 is the front-end valve of the receiver of the FuG 25a transponder.



Please notice the lose wire just below the test-connector

Which should belong to test-pin pin 9 down on the far right-hand side. Simply lacking the according connector.

Luckily we fond a scrap unit with a sound female connector.

Nearly all GAF electronic gear being fit with an equal test facility; employing the standard PV 10 (PV 62)

We discovered that our unit must have been 'accepted' towards the end of 1944; because, exceptionally, both bipolar capacitor being stamped for production week: 39 and 42, 1944.


YouTube films


Film    00066:    Progressing the restoration project, we start today with noticing what we are reviving with the FuG 25a transponder. The frame now mounted the the genuine frame, which we operate for about 9 years already. The mounted unit is electrically connected onto the vertical antenna at the upper right-hand side. Power supply adjusted at 25 V and current consumption 3.5 A.


Film    00067:    Viewing the mechanical driving mechanism to the coding unit; which for this experiment is not essential for our current experiments.


Film    00068:    The detector output clearly is concerning a kind of irregular voice spectrum; of which we are only viewing a part of its content. The 1 local oscillator near the receiver front-end is fit with a wobbling motor, which allows instead of just a signal tuned signal a spectrum of 2 x 4 MHz up and down the fixed tuned receiver frequency. In those days, airborne VHF electronics had to stand extreme temperature changes. As to overcome this downside they wobbled over a spectrum about 4 MHz below and ca 4 MHz up the nominal receiver frequency about 8 times a second.  


Film    00069:   For this occasion we would like to monitor the RF signal generated by the wobbling 1 oscillator of the FuG 25a receiver. The wobbling oscillator generates a buzzing sound in the Hallicrafters S-27 loudspeaker, which disappears just when Hans manually stops the rotation of the Ferraris wobbling motor.  Please notice the way the wobbling motor stops and is starting up to rotate. 


Film    00074 :    Watching now closely the wobble motor, which doesn't rotate yet. Because no power (24 V dc) is being fed onto the FuG 25a set. Then Hans switches the 24 V DC on; and the dynamotor starts generating the ca 16 V 116 Hz ac onto the Ferraris-motor which is forcing the 1 local-oscillator to wobble around a certain frequency. To what I have learned  - with about 8 Hz.   When Hans switched on the 24 V DC, due to the warming up time of the valve cathode, it takes some time before the Hallicrafters receiver will pickup the wobbling oscillator signal, of which it only captures a tiny portion, due to the limited receiver bandwidth and the wobbled spectrum between, say, 120 and 128 MHz.             





On: 19 March 2021

Step: 1      Hans Goulooze had to make a substitute for the lacking, quite complicated, cover-plate

For it he had to make a wooded model first




One need to possess some experience, call it: craftsmanship, which skill he apparently does possess



Hans Goulooze's substitute is for 95 % is fitting perfectly, some, already foreseen, has still to be adapted



Hans' comment: it would have been easier when I would have used 1 mm Al plate instead of 0.5 mm

Which on the other hand might have obstructed the free space between this module and the outside cover. 



Later the screws Hans thought should match (3 mm) proved to stuck after a few turns

We found out - that it should be 2.6 mm screws instead.

Typically for those days, where the sizes were: 1.9 mm - 2.3 mm - 2.6 mm - 3 mm.

Luckily we possess some stocks of these 'classical'  screws necessary.



Last week we noticed already the discrepancy between what is said and the reality in our FuG 25a transponder

What is told: that the wobble sweep is 8 MHz:

We measured, however: 129.6 MHz ... 133.2 MHz → 3.6 MHz

The IF frequency is 7 MHz.

hence, when we would like to operate at 124 MHz, like of main FuG 25a does, the 1ste local oscillator should at least operate at 124 + 7 = 131 MHz

The centre frequency can be easily adjusted by means of trimmer C 37; it is, however, advisable to use for this occasion a special insulated screwdriver.



I stopped the Ferraris wobble-motor rotating and we noticed the two outside frequencies up and down; as we foregoing noticed already

I also doubt about the rotation frequency, as it seems that it runs quite fast.

Maybe we should we should use a stroboscope as to determine its rpm.



Our second task was to adjust the sweeping oscillator such that 124 MHz being well covered

For it we use our R&S SMS digital signal generator.

Modulation 1000 Hz; as an un-modulated carrier would only generate a dc component.



Second picture

Same condition as before.



The foregoing screen picture signal has been taken from R 25 at the detector output



The trimmer in the upper compartment is for aligning the receiver front-end circuit connected onto the receiver-antenna input

Viewing the Bakelite shaft of  the wobble motor, it was running; as you otherwise would have noticed the shaft slit.

The two trimmer on the left-hand side are for tuning the first IF band-filter stage.


What might have confused people: is, that they knew the Gemse data, which wobbling system operates at 50 Hz (mains), whereas the FuG 25a, does operate at ca. 116 Hz 




For your memory

Today 23 April 2021


For your memory - shown are two signals - one the Nachtfee pulses monitored at 124 MHz, one being the Nachtfee signal, the broader signal representing: either EGON or the Nachtfee signal

I don't know at which signal the oscilloscope has just triggered upon the Egon pulses of 502 Hz or the Nachtfee pulses of 500 Hz



To be continued in due course

YouTube films


Film 00075:    Viewing the signal output at 124 MHz, fed onto a miniature dummy load (50 Ω). The fixed standing signal either is the Nachtfee signal at a prf of 500 Hz or the EGON signal at a prf of 502 Hz. I have particularly regulated the parameters of the two pulse-generators. It is evident, that the pulses are sound. The green shape of the pulses visible are constituting the 124 MHz carrier. As the ca. 500 Hz pulses are gating the 124 MHz carrier.   


Film 00077:    Viewing now at just the output of the FuG 25a detector. What we see: are two pulses separated 2 ms, just as is expected, but curious the FuG 25a system is in some way or another blocking for ca 5 pulses and then allowing again 2 pulses appearing at the detector output.  It isn't yet determined whether this phenomenon inflicts with the signal shown at the simulated aircraft display.    




On 5 May 2021

I made a new approach, finding out why after a while, think of 30 to 40 minutes, the receiving signal at the detector output is changing - and the spot on the painted first order Lissajous circle is, over the time, nearly fading away.




The CRT screen photographed when signals aren't yet inflicted



This is what happens after a while

Resulting in fading away of the Z-modulated spot at the virtual aircraft display.

What is causing the mall-functioning after a while?


Today we measure mainly at the anode of Rö 4 the detector valve; in particular at the point where W 25 and C 32 and the RV 12P2000 anode join, this is where our probe has been connected onto


After some experiments I suspected that Rö 3 on the left-hand side, which likely is causing the nuisance.

And after having changed it, it started operating quite favourable since. 



The getter deposited at the inner-side of the glass envelope is quite strange


We consider the silver-like mirror, is resulting from a kind of vacuum defect; as the getter isn't regularly depositing this way on the glass envelope




YouTube films


Film 00083:    We are still bothering with a down-side - that after a while, think of 30 to 45 minutes, that the signal level reduces, and therefore the Z-modulate spot or dot painted at the circular Lissajous figure projected at the virtual aircraft CRT screen.  Probe-signal derived straight from the detector (valve) output, at resistor W 25.

Film 00084:    Continuing: I changed now the third IF valve Rö 3 (RV 12P200), causing a signal which I have never noticed before - especially at the time base-line. Still monitoring at the detector signal output.

Film 00085:    Viewing first the virtual simulated aircraft display. The spot shown 'south' is being caused by operating a special switch at the Nachtfee consol designated 'Freya-Polwender'.

Film 00086:    Please notice the pulse pointing south is the returning from the Nachtfee transponder, but which originated (sent) from the Nachtfee system. That the blip is pointing south is being caused by operating the 'Freya-Polwender' switch (mode). Nachtfee is generating at its output a 'sine-wave' signal. The Freya-Polwender switch is only interchanging the two-output wires. Please remember: Nachtfee relied upon a 'coherent signal system' and when you interchange the output-wires you, consequently, change the signal phase over 180°; resulting in jumping the signal blip over 180 degrees. Thus: when a blip is shown south, when the Polwender-Switch is activated, then the command- or Order blip is visible just opposite to what is given at the real Command-Order Compass.  The weak pulse at say 6 minutes past the hour originates from the time-base-reference pulse; informing the ground operator what the actual time-base-status in the aircraft system is. This essential information is the only means to adjust the correct instant of arrival of the Nachtfee ground-signal at the aircraft display.



 To day 24 September 2021

Hans Goulooze has modified this quite genuine Gemse type receiver.




In my perception, seemingly brand new; the main chassis is made of nice Al casting, I suppose no traces of a Zinc alloy



Everything is so neatly constructed that it looks like a dream. The eccentric drive mechanism is to determine the the swept part of the  I.F.F. reception spectrum

From my visible inspection, nothing had ever been touched since, say, 1944.

As usually, one hardly will find an indication of its manufacturing date. The two 'video valves; , up on the right-hand side, are both of type LV 1, a versatile high slope valve of 9.5 mA/V; in those days a quite exceptional value.



He chassis / frame being rotated among its longitudinal axis

The Germans call this: eine Augenweide!



The connector on the most right-hand side constitutes the video

Gerät-Nr. 124 -121-A-2    actually constitutes the design / drawing - number on which it once had been manufactured.

Hersteller (manufacturer code)


'which once stood for:

Siemens & Halske

Berlin Siemensstadt

für sämtliche Werke

Really, top-class manufacturing!


YouTube film


Film 00091:   Viewing first the Nachtfee LB 2 control CRT. Displayed: the weaker signal is the original Nachtfee command signal which returns through the simulated aircraft transponder FuG 25a. Thereafter viewing the opened ground transponder receiver. monitoring then the video output of the Gemse receiver fed at the input of the Nachtfee consol. Thereafter viewing the EGON reference pulse which is to be used by the EGON guiding station. Remember: Nachtfee constituted a secret command system; whereas the EGON system controlled the path of the Pathfinder aircraft (Ju88-S). EGON signals, by means of Nachtfee which action should be taken or to be prepared for. We also will see the 180° signal jumps, caused by the "Freya-Polwender" selector switch. This is simply inter-changing the two symmetrical output wires (operating with a sine-wave signal), causing a 180° phase shift. Causing on the simulated aircraft display also a spot jump of 180 degrees. All is only valid as long as the system remains in a coherent status. The Phase control-knob on the Nachtfee-consol, is causing a delay or a positive phase change; which is causing the Nachtfee order pulse to arrived retarded or in an advanced manner. The Phase control (knob) should be kept in a certain angle at the LB 2 circular scale. As long as the returning aircraft time-base pulse is kept controlled, in our case, at say 8 minutes past the hour, then the Nachtfee Order signal arrives just in time at the simulated aircraft display. Astonishing? Yes, indeed!
 Interesting is - that that albeit that the EGON signal differs only 2 Hz from the coherent Nachtfee signal, both actually do not interfere at all. When EGON would differ, say, 4 Hz you would even hardly notice that there is another signal being monitors.   


To be continued in due course



By Arthur O. Bauer