help!! An old tek 7904 that power supply is "tic tic",i think is the multiplier

[MadTux]

My favourite 7904A failed . Had it for about 10 years now and liked it very much. Replaced it with another 7904A from storage, but would really like to fix it.

Symptoms are: Scope starts, fan spins up, all voltages on PSU are present, but slightly low before PSU goes into ticking mode after about 2-3 seconds. It then ticks a few times, restarts and repeats the cycle at faster rate about once every 1.5-2 seconds.
(+/-50V rails => +/-47V, +/-15V rails => +/-14.7V, 5V rail => 4.7V, measured simultaneously with 5x HP-3456A with max/min recall in STAT mode)

Because of good voltages on main rails, I excluded a obvious short somewhere. Upon further investigation, I smelled  electronic smell from HV compartment behind CRT cathode and noted that R14 (20Ohms current limit for HV inverter) gets rather hot.

 Now, since most parts there are floating on -3000V, I'm a bit reluctant/scared to probe in it . Unfortunately, because of insane prices, I also don't have nice HV probes like Tek 6015. I could hack some simple HV divider from 10M resistors however.

Before I start to take it apart any more, is there any good strategy to further isolate faults? Some known bad parts in there? I somewhat suspect the anode HV multiplier to be bad (which would suck hard), it should be easy to disconnect it (1 wire) and I think the scope should run fine without it, in case it's bad(without screen of course). Any ideas?

[tautech]

Wrong schematic, focus on what's providing the 25 KHz HV driver and that it's oscillating. There should be waveforms and amplitudes in the SM.
It should be LV, probably ~30V and if you can't find a reference point for a scope probe you should be able to confirm its operation with a probe just held near picking up the EMI, ahla near field probe style. 

[MadTux]

Thanks, will measure the 25kHz for HV transformer tomorrow. It's late now, here in central europe

[MadTux]

25kHz from PSU for HV inverter is there, nice differential 25kHz square wave on both outer pins of P40.

I removed 25kHz inverter power connector P40 and both cables that go to Z-axis & focus amplifier (P20 & P83) along with the blue connector on the bottom (P146) to isolate the HV board from the rest. The PSU started fine. So there is certainly something wrong in there.

Then I de-soldered the HV multiplier to test it with a few 100V 10kHz sine input from modified audio amplifier (audio amp with small 220V=>12V transformer reversely wired to output) Strangely it seems to work fine (200V AC in yields about 1200V out, stuck a screwdriver into anode plug and increased volume on audio amp while measuring both voltages until the 3456A 1000V range maxed out)

To test the HV board without HV multiplier, I reassembled it back into 7904A and connected everything except the CRT anode plug. PSU also worked this time.

So either the HV multiplier only fails at really high voltage or something is wrong with the CRT anode. (something shorted inside CRTor bad HV cable). Everything was covered with that nasty HV related soot, which I cleaned. Now I will put HV multiplier back onto the board and test it in circuit with audio amp. Then I will probably put the the board back into 7904A with anode plug disconnected and see wether that makes a difference.

[chris_leyson]

Off topic but I think the 7904 used discrete transistors in the power supply whereas all of the other 7000 series scopes used an unobtainable controller IC so a least you don't have the problem of trying to find a replacement IC . I had (still have) a few problems with a 7854 power supply, all the rails checked OK so I assumed it had to be the multiplier, warmed it up with a hair dryer to get rid of any moisture and gave everything a good clean with IPA, it seemed to be OK for a while and then failed again. Turned out to be bad solder joints where someone had soldered PCB mount caps to the aluminium screw terminals salvaged from the original primary smoothing caps. Anyway, fixed that problem but it still won't start up, think I might have to re-cap the low voltage regulator section because if any rail goes out of spec in voltage or current the PSU shuts down, I think there may even be a link or jumper on the PSU PCB to override the shutdown. Overall I think the 7000 series PSU is a nice design and pretty quiet for a switcher but there are a lot of possible faults that will shut it down, it might be a primary current sense problem and the HV supply could be pushing it over the primary current limit because of a problem in the low voltage section. At least you get schematics and waveforms in the service manual, thanks Tektronix.

[MadTux]

That poor HV multiplier is sparking at places, where it shouldn't



Put HV multiplier back onto board, feed it some 100V's AC from audio amp and no HV => Very weird.
Since it worked in desoldered state, I removed both nuts that hold it to board and HV was back again. So I started probing the threads with a grounded screwdiver and saw some small sparks

Only problem, how do I make that HV multiplier happy again? PET foil has some high dielectric strength, so I thought about cutting that thread with a dremel and put several layers of clear PET foil (packaging tape) over it. Any opinions?

[MadTux]

7904A and me are happy again 

Wrapped lots of PET packaging tape around that HV multiplier. Then glued a PMMA plate to the bottom for extra isolation.
Unfortunately the space inside the HV compartment is quite limited, so the multiplier with added plate was too high and I had to rip the plate off again. Zip-tied the multiplier to the board. Space is still pretty limited and I had to slightly bend the PCB to get everything back together.

Scope works fine, get a good trace on CRT. Will put current backup 7904A into storage again and use favourite and now repaired 7904A instead to see, if and how long the repair will last.

[tautech]

 
That'll keep you going for now however if you're feeling flush you might be able to get a replacement from Sphere.
http://www.sphere.bc.ca/test/tek-parts/tekparts5.html

Their 79** multipliers seem out of stock but it'd be a good idea to send them an email anyway.