G-Schreiber T52d

headache 2


Page initiated on 6 January 2014

Status: 7 January 2014




During the special exhibitions on Secret Communications in November and a single Sunday in December, I encountered serious troubles again in respect to the correct functioning of our T52d Geheimschreiber.

For it my instant conclusion was - that this time the failure goes back to the mall functioning polarity relays.

Please consider also the dedicated webpage: Geheimschreiber Polar-Relais

Giving attention onto the correct functioning of the polarity relay type 43a has to be proceed from the understanding why these might fail. My suspicion is that it might originate, or at least being enhanced, by the lack of sufficient magnetism of the two internal magnets sections. These being mounted in a way that they reject oneanother. In my understanding, this force is demanding energy where else can deliver this power - right - the magnets and this should having an effect over decades. However, the type T43a are at least say 70 years old and its permanent magnetism may well have been diminished. 

I decided to give it a try by replacing them experimentally (temporally) by the more sophisticated type T64xx. We got some and also possess some.

As to be sure that I do understand the interactions between the 5 polarity relay (for each Baudot bit a single one) correctly. I first have to findingout what the actual pin interconnections of the type T43a are. This proved being more difficult than expected as nowhere is this been described. The only means possible was to view the wiring layout underneath the Geheimschreiber wiring schematic. With some imagination, because the figures are hardly readable. However, its has been solved.


The pin layout is shown from above the chassis and not as usually from the pin connection side

Actually, in contrast to what is explained in the Siemens T36/Si manual is the mutual ground (MLB) not connected onto pin 5 but onto pin 6. No other pin showed any measurable resistance against ground. However, pin 6 measured 17 ohms against MLB. Following the wiring the common neutral line runs trough the highly complex Secret/plain text selector (Geheim/Klar). As to minimise the problem a short cut was made between the common relay pin 6 and ground of the G-Schreiber chassis. Instantly this resulted in a kind of functioning, but still most unreliable.

It was found that some of the polarity relays failed. Replacing them caused another problem, which all the time could be determined by interpreting the 'Baudot True Table'.

In my perception this is the right moment trying peu peu replacing experimentally a T43a relay by a suitable type T64.


On 13/14 January 2013



A new approach is undertaken.

The question still remains, why is our machine showing problems again and again?


Before approaching the main chassis from underneath, I would like to be sure that when the polarity relays being fixed wired that no faults occur.


Which actually is the case

The previously experimentally adopted T64/-36 relay replacing temporarily the original T43a types was first set (programmed according the Baudot table) for the character 'c' and there after changing the setting of bit 'one' (R1) from T to a Z connection. The character K was being printed failure free. Thus a fault in the + and - 60 V supply is not likely; and, the encoder is also performing in this setup correctly.


With some luck one can see that on the left the 'c' characters and on the right 'k' characters being printed without failures

This result is promising, but I need to access the heavy chassis from below and it is for me impossible laying the very heavy chassis fram on its back without assistance. Thus I have to wait until someone is helping me.

My first plan is to replace experimentally the 5 loading capacitors C 1 - C5 by modern types temporarily. Also checking the wiring and soldering at various places. Thereafter, when the chassis is put in its horizontal position again, fitting in a single relay first, like in stead of R1. Leaving the other four relays (R2 - R3 - R4 - R5) being fixed wired.


To be continued in due course 




By Arthur O. Bauer