I just obtained this as a working machine with a fault or two. Very profuse apologies if all this has been covered before. I'm hoping to work through this generator and hope the journey is of some interest. I bought this because it has valves in it. I already have a couple of solid state marconis and hoped this would be an enjoyable ramble through the insides of an earlier set. So far I can see that quite a lot of work has taken place. Many resoldered joints and a few new components; and worryingly one or two absent. My first clue was when running for a couple of hours the carrier setting on the meter faded so that it was well below the set mark.Tired valves?-high resisors or leaky caps? First inspections revealed much a-do with the meter circuits. It takes a rectified signal from the RF stages through a few resistors to the 100microamp meter. Two of the resistors are absent and another inside a shielded switch is a lower value. The meter is effectively wired direct to the source. I suspected that there was a low output from the RF stage but all the vaves check out high and several are new cv types. So I looked at the CS2A rectifier and found it almost completely failed. It seems they are prone to this and are very expensive to buy (£70+) so I picked up rumours that schottky types might do the job. I've ordered some HSMS 282 diodes as they have a very low capacitance and claim to handle small forward voltages - so fingers crossed one of these might get me nearer a functioning meter. I'll post more as soon as this bit is back together. In the meantime any advice or interest will be much appreciated. The pics show the unit as purchased and a general view of the chassis then a close up of the shielded switch mechanism with the diode out of its nylon holder. Next to the holder can be seen a decoupling cap of 820pf (c88) which has been broken away from its screwed earth mount at some time in the past. Also a 470 ohm resistor in place of a 1k original (r71). R70 and R121 both in line with the meter are missing; I'll replace all these! Can anyone shed light on why r70 (2.7k 1/4 watt carbon) and r121 (950 ohm 1 watt wirewound) should be in series in the 100u/amp meter feed? I wonder if the w/w has and inductive role.
The diode in your last picture looks very similar to the Schottky mixer diodes we used in the detector horn of our microwave security alarm sensors back in the 70s. The devices we used were Mullard but the part number escapes me at the moment. It would be worth doing an internet search on mixer diode to see what turns up.
Two things to be aware of with this type of diode. They are very static sensitive and any more than 4 volts or so of reverse polarity will damage them.
Alistair, thanks for the static warning esp in this hot dry weather! I'll take precautions when fitting it. The diode I've ordered has a reverse voltage of 15v. I have 12 on their way so hopefully I'll get it right with a few to spare. Jonathan
While awaiting the diodes I have given some attention to the declared problem of the modulation osc running at 9khz on the 400hz setting. I had also noted that proper readings on the meter were not possible on the AM and FM meter functions - being very low compared with the CW readings. The osc is controlled by switched caps from anode to earth and also in the return from anode transformer to grid of V1 (again to earth) all of which are at ht potential to earth. In the 400hz circuit I found anode cap C101 disconnected at one end and C103 (its corresponding grid cap) absent. This looks fatal in terms of any useful function. C101 was very slightly leaky so has been replaced as too has C103. A few others in the other switched circuits were slightly leaky so they have gone too. Not quite sure why these caps were 'investigated' as very slight leaks would not crash the circuit (the grid being isolated by a further perfect 0.1 uF cap. In looking at low readings on the meter functions I noted a relacement cap 'soldered' in place of 2x 2uF paper caps in parallel feeding the relevant circuits. This cap was marked as C32 the original of which is still in place in the RF section and has not been touched. Confusing ? A little ! This replacement is 0.1uF and is trying to pass a signal designed to be conveyed by 4uF ?! The original 2x 2uf caps have been cut out and to give access to them C105 has been broken out of place at one end and I found it hanging on 1 wire. Anyway all this has been put right and when I receive the diodes I'll try powering it up to see what I am left with to look at. This will be an oportunity to check the regulation of the ht supplies in case the voltage stab valves are tired. The pics show how the board was found including at the extreme top the 0.1uF cap relacement for 4uF cut out. The other pic is the repaired board. The diodes have just arrived and I cannot believe just how tiny they are. I need to find a magnifying glass now.
Well, having seen how small the twin diode package is I had a plan how to put it in circuit. I taped it down onto an alloy surface with its pins in contact with the metal and then taped thin wires on in contact with the two chosen diode connections. Then soldered them into place with a fine nosed iron which was very hot. I hoped that this method would remove the chance of static charges when soldering and it has worked. The diode looked very vunerable to strain if hanging on wires so I encapsulated it in epoxy to give rigidity. I marked it with a red pen and now it's in place I have a stong supply to the meter which remains rock steady even after several hours. I think for now this looks like a fix. Turning to the Modulation osc.-It refused to operate on 400c/s but was perfect on 1k/s and 1.5k/s. The anode voltage is slightly down on spec but I thought to try playing with the new capacitor values as nothing else seemed amiss. By reducing the grid cap to 0.1uF it runs at 510 c/s. Playing with other sizes and doing the same with the anode caps fails to improve things. I either have a higher frequency or a reduced output voltage so for now I'm tempted to leave it running a bit fast as its output is the same level as the other two speeds. I may return to this as and when I sort the rest of the machine. Pics show the diode wired, encapsuated and in place. The meter is now working well. The question of why there was a low wattage resistor in series with a wirewound feeding the meter still remains but so far my two carbon 1/8 watt replacements are doing the job.