I noticed that you started the thread in 2017! What do you think is wrong with your calibrator?
Spurred on by the very helpful replies here gave me the confidence to tackle this repair. Thanks for your shared experiences.
So my earlier thought that the HV MOSFET string was the problem as it was getting very hot turned out to be only a symptom of the actual failures.
After replacing the entire HV transistor set TR6-TR12 with no fix, and observing that 1600 V was being applied across the HV transistor string it was no wonder they were getting hot.
Also RLJ was cycling on and off every 1 to 2 seconds.
So the next tests were static measurements for any obvious component shorts wrong values etc from the HV rectifier output to the output control CCT. Nothing stood out as failed.
Replacing the chopper IC5 made no difference. The reference cct measured okay too.
Moved onto IC1 and opto IC2. replaced both with again no fix
.
Okay obviously being a bit slow on the up take! , I then checked the HV switcher transistors TR201 and TR202 out of CCT as trying to decipher the readings otherwise was not conclusive.
Upshot was TR202 had gone rather leaky C-E (not a full short or open CCT any terminal), and TR201 seemed okay. All other components in that CCT measured okay.
Note that TR201 was an MJE13005 with higher VCE ratings, not the 2N6499 specified in the parts list.
Finding the 2N6543 used for TR202 proved fruitless from the usual trusted suppliers, so not wanting to take the risk on ebay etc I looked at suitable sub's. There are a number of possibles including the BUX80 which I had one to hand so I tried that plus a new MJE13005 just in case!.
Result was no fix
, still overheating TR7-12 with maxed out HV across them and RLJ slow cycling on off.
Measuring the output of IC1 showed no switching just ~ -13V, loop not functioning at all.
I removed IC 2 and put a shorting link across pins 4 and 5 to simulate the opto transistor turned full on. The theory was it would turn TR201 and 202 completely off.
The result was such that the HV output dropped from max ~1600V down to just under 100V with no clicking from RLJ or heat issues and now the output could be controlled from 0 up to 80 odd volts
.
Proving the chopper control was working.
All components around IC1 tested okay so why no feedback ??. I tested on the bench the opto coupler with the conditions found in the 2701 and found the opto was not turning on properly !, so upping the LED current a bit improved the switching and was still within the LF353 specs so I reduced R23 to 2K and used a CNY117-3 opto which has better CTR than the 4N25.
The result was a now functioning HV switcher with a steady 230V across the HV transistor string and output controllable from 0 up to 1200 V.
In the end the culprit was TR202 which had failed for whatever reason and the hot output transistors and hot bleeder resistors R209-216 as well as RLJ cycling on/off were a result of the HV transformer being fully on.
The one thing that puzzles is the fact the opto feedback required such a change to the LED drive current to get the switcher working again. Must be a marginal design to start with. Also not what you would call fail safe with MAX HV output
.
I am in the process of measuring the HV switcher with a MicSig HV diff probe and will post some pictures showing various conditions of TR202 C-E wave forms with no snubber then with the original snubber, and some further experiments with added snubbers across the rectifiers and HV transformer PRI. Inspiration to tinker taken from Electrole's write up (see link he posted earlier).
I will also do some of the suggested mods and use my HP3400 to measure the output RMS noise.
Will update thread as these things are done.