Summit tla 100

[pucho812]

This is an odd one. Luckily we have schematics in the tech docs and calibration instructions.

When metering output meter jumps far right off the scale.

I tracked it down to an opamp ic10 which was outputting -12vdc in ref to ground.

I replaced opamp 10, an lf347 chip. And it was working fine for a second, then back to thevoutput metering pegged far right.

The unit is passing signal and all other functions work as they should.

Per calibration instructions in the tech documents, the psu is measuring correctly. W other chips seem to be ok.

The only other thing I have noticed is pin 14, 13, 12, and 8, 9 , 10 all measure continuity to ground when switched to meter output.

With the chip out, I get 1.4vdc at pin 13 when set to meter output.

 

[gyraf]

What is IC11? Some CMOS switch? I'd suspect that one - or perhaps an err in the +/-5V supplies..

 

[pucho812]

Ic10 is a lf347. I had thought it failed as it was pushing out dc.
Ic11 is a multiplexer cd4052be

The only thing unusual with ic11 that I see so far is that my +5v at pin 16 is reading 4v

 

[gyraf]

..analogue multiplexers are known to to fail in mysterious ways - and this one "sees" the output transformer, so there's a good reason it has seen something it can't unsee..

 

[pucho812]

..analogue multiplexers are known to to fail in mysterious ways - and this one "sees" the output transformer, so there's a good reason it has seen something it can't unsee..

Well even more the head scratcher.
This morning it worked without issue. Meter did not peg when no signal was present. It lasted a whole 2 minutes before going back to a pegged meter.

 

[Ike Zimbel]

Well even more the head scratcher.
This morning it worked without issue. Meter did not peg when no signal was present. It lasted a whole 2 minutes before going back to a pegged meter.

"Well it's intermittent alright, just not all the time!" Me. This sounds like a cracked trace to me, or a similar connectivity issue. IC sockets? Hairline crack in a trace?

 

[pucho812]

"Well it's intermittent alright, just not all the time!" Me. This sounds like a cracked trace to me, or a similar connectivity issue. IC sockets? Hairline crack in a trace?

Maybe a cracked trace, except I didn't move a thing when the unit was working and it just went back to, full pegged meter. The IC's are directly soldered to the PCB, except for the one I replaced. when ever I replace an IC, I always include a socket as there is a good chance I will have to replace the same IC down the road.

 

[Ike Zimbel]

Maybe a cracked trace, except I didn't move a thing when the unit was working and it just went back to, full pegged meter. The IC's are directly soldered to the PCB, except for the one I replaced. when ever I replace an IC, I always include a socket as there is a good chance I will have to replace the same IC down the road.

It doesn't take much heat to open up a hairline crack in a trace. I suppose this also applies inside the devices in question, like the logic chip that Jakob mentioned. I haven't owned cold spray for decades, but you might try sticking it in the fridge for 1/2 hour and see if it works when powered up. If it does, that definitely points to heat being the trigger for the failure mode. If you have a small, accurate heat source, you could then try applying that to various parts of the board in hopes of triggering the failure. I have an ITI MEP-30 on the bench right now that works fine for the first 30-40 minutes after you turn it on...then it loses the positive half of the test waveform. I have tracked this down to one of the potted amp modules as I can trigger the failure very quickly if I play a heat gun on that module.