Tektronix TDS754D Repair - Acquisition Board Self-Test Failure & Funky Channel 2

[Snekers]

Hello all, I have a TDS754D with a rather strange issue, and could use any help on diagnosing I can get.

Symptoms:
-Fails acquisition board self test
-Channel 2 is funky, will expand on that
-All other channels work perfectly

Extras:
-Original Dallas NVRAM, still holding on but needs replacing
-Electrolytics are still on the serial board, also needs replaced
-Dim display(seems like a common thing with these)

Channel 2 doesn't respond to inputs, and appears to have lots of glitches in the trace. An acquaintance said that it looks like it's "dropping bits", which makes sense to me but I have no idea on how to diagnose it. What I find odd is that the glitch changes the way it looks seemingly at random at each startup, the images attached are the three different states it goes into.

Any input would be great, thanks!

[Snekers]

Update:

just powered it up again, now it doesn't even get through the self test. Just sits with all of the lights on. Think it might be a PSU issue?

[james_s]

I had a similar problem on my 784C which turned out to be bad soldering on some of the large square ASICs, fixing that is something that requires expertise in soldering though. Checking the power supply voltages won't hurt either but your mention of missing bits jogged a memory.

Your screen looks really good in the pictures but it's hard to tell how dim it really is. These things never were anywhere near as bright as modern LCDs, the shutter is nifty but absorbs a lot of light and they run the tube really hard to make up for it, so worn CRTs are common. Yours is still nice and sharp at least.

[Snekers]

Thanks for the reply, at this point I'm most likely going to get a refund. Bought it with the listing clearly showing it passing all self tests, and I'll bet the extremely lackluster packing job might've broken something.

[james_s]

Yeah if it wasn't packed well it probably bounced around and cracked the solder or damaged something else. These are big heavy beasts and are quite delicate. They're lovely instruments when working properly though.

[Snekers]

Here's the "wonderful" packing job, absolutely no protection. I'm surprised the front panel wasn't damaged with it only being protected with a piece of paper.

[james_s]

Wow that's terrible! I would not even think of trying to ship one of those without several inches of foam all around it. These things are delicate precision instruments, new they cost about $40k in the late 90s. You should definitely request a refund, it's very likely that it was damaged and is now a parts unit.

[Snekers]

In the case that the seller doesn’t want me to ship it back, do you think there’s any hope of repair? I’ve heard the ceramic attenuator boards break easy, and many more things could’ve been jostled around. It’s a damn shame, there’s only so many of these out there.

[james_s]

Yes there's certainly hope. I would start by very closely examining the soldering on the ASICs on the acquisition board, I had to resolder a couple of those on mine to resolve a similar problem with missing bits. The attenuators are delicate but reasonably well protected within the scope so they're probably ok. Another good thing to check which is nice and easy is the interconnect board that goes on the side to join the stack together.

[Snekers]

What would be causing this new issue where it doesn’t appear to do anything? Could that be some bad solder joints on the processor board, or would an issue and the acquisition board cause that as well?

On another note, what’s the best way to check power supply voltages? I don’t see any test points, or any available pinouts, not to mention you need the boards in place for the power button to work.

[james_s]

That could be any number of things, these have some built in diagnostics though, does the display on the CPU board do anything? There is also a debug serial port inside that you can connect to and see diagnostic messages as it boots. Checking the voltages across decoupling capacitors is a good start. IIRC power is distributed via a ribbon cable bus that runs up the right front side. Did you already check that the bus interconnect board and all other connectors are fully seated? It's very hard to troubleshoot something like this remotely, there is no substitute for having the scope in front of you.

[Snekers]

Well... darn. UPS wants me to ship it back exactly how it arrived for investigation, no doubt causing even more damage. Guess it’s the end for this thing

[james_s]

You shouldn't have to do that, it was clearly packaged inadequately, if you bought it on ebay put in a claim through ebay, request a refund or at least a partial refund. You can point the seller to the UPS packaging requirements which very clearly were not met here, they are not going to cover the damage.

[Johnny10]

That is a shame but don't waste your time.

That thing is always going to have problems.

[james_s]

I don't know why you say that. Mine had problems (that I was aware of) when I purchased it, I fixed those problems and it has been fine. I think it's pretty likely that this one could be fixed without too much trouble.

[Snekers]

Sorry for the delay, everyone. Sorted out the issue with UPS and the seller, I’ll be keeping the scope. The psu voltages appear to be fine, every cap I measure is either 5V, -5V, 15V, or -15V(haven’t proved the crt driver Board yet, so can’t verify the 25V). How should I go about looking for faults on the processor board that would prevent it from starting up?

[Snekers]

Tried reseating the acquisition board to processor board interconnect and it booted again! But now there’s a processor board error in addition to the original one. Hmm.

Edit: power cycling solved this, now it’s back to the original dead channel 2 and failed acquisition board problem.

[shakalnokturn]

What does the error log have to say?

[Snekers]

Surprisingly there’s only one after clearing, it was the same one over and over. Presumably that’s from all of the startups.

[Snekers]

After a few power cycles, it seems that there’s two distinct error codes. Both have to do with this “diaHFStepDiag”.

[Johnny10]

You may have already found this page,

https://forum.tek.com/viewtopic.php?f=568&t=142036

[james_s]

Isn't U400 they mention one of those large ASICs on the acquisition board? If so I already mentioned that fixing soldering on those fixed a very similar problem on my scope.

[Snekers]

I somehow missed that post, it’s absolutely identical to my issue in every way. Thanks for finding that!

[james_s]

If you resolder those chips yourself, be very, very careful. They have a lot of pins and if you short something out and pop one of the ICs the only way to get one is from another acquisition board. Use flux and only enough solder to wet the tip, then use a magnifier to inspect very carefully for bridges.

[Snekers]

Honestly I’m not experienced with smd at all, so I’ll probably find someone much more skilled to solder them, I don’t want to take any risks with these ASICS.

No idea who in their right minds would want to reflow these, though.