Tek TDS784A restoration - One last fault to fix !

[snoopy]

Hi

I recently purchased an optioned up 784A from ebay at a really good price. Scope basically worked but failed the SPC. Most of the relays in the attenuators were high in resistance or completely open so I ordered a set of genuine ones and replaced the lot of them. Now SPC is fine but there is still one problem where the Acquisition test fails with the following error "trigNibDiag trigNib A is 0xa7". Not sure what this error code points to but the scope runs fine except Instavu has a major problem on ch2. If I enable ch1,3,or 4 Instavu works ok except the trace goes blank above 10ms/s and reappears at 2Gs/s and above. Now if I enable ch2 with any of the other channels on (still with Instavu on) the screen fills up with garbage. If ch2 is on and the others are off with Instavu enabled there is no trace at all on all time-base settings. Otherwise with Instavu turned off the scope works fine on all channels. Could the acquisition error be linked to the problem with Instavu ?

On another issue the Horizontal position control was so dicky as to be useless. I pulled the front control panel off and sprayed some CRC contact cleaner into the center of the back of the pot and managed to clean the pot. There is a small enough opening to allow a bit of cleaner to get inside the pot. While I was at it I did the rest of them and now they are as right as rain

cheers

[TAMHAN]

Hello,
this is exactly the same issue I had on my TDS754D, which ended up dead as I lifted a trace trying to replace the SOJ RAM IC.

AFAIK, it almost always points to a memory error. trigNibDiag apparently fails as it cannot find something in the RAM due to a error in the memory. Did you already try to do a detailed self check?

Also, in my case, in postmortem it looks like I had a issue with a solder joint gone cold rather than with the actual RAM IS. So, maybe probing around the board a bit and/or reflowing all the relevant ISses in the first place might be workable.

Finally, beware that there are two sizes of the RAM IC. SOJ300 and SOJ400...which are incompatible.

Tam

P.S. Appended a pic of the board area for you...on mine, the problematic IS was U416.

[snoopy]

Hello,
this is exactly the same issue I had on my TDS754D, which ended up dead as I lifted a trace trying to replace the SOJ RAM IC.

AFAIK, it almost always points to a memory error. trigNibDiag apparently fails as it cannot find something in the RAM due to a error in the memory. Did you already try to do a detailed self check?

Also, in my case, in postmortem it looks like I had a issue with a solder joint gone cold rather than with the actual RAM IS. So, maybe probing around the board a bit and/or reflowing all the relevant ISses in the first place might be workable.

Finally, beware that there are two sizes of the RAM IC. SOJ300 and SOJ400...which are incompatible.

Tam

P.S. Appended a pic of the board area for you...on mine, the problematic IS was U416.

Hi Tam

I did find a dry joint on one of the 240 pin QFP demultiplexer IC's for channel 1 but that didn't solve the problem unfortunately. I will investigate the memory chips in detail now that you mention it. It's a bit hard to inspect J leaded packages from the top because you can't see much with a magnifier looking down on top of the package so some sideways inspection will be necessary.

BTW how do you do the detailed self-check ?

regards
david

[TAMHAN]

Hello,
one thing I found to work were special probes which were very thin - you could, barely, reach the PCB with the one and touch the IS with the other. Also, many of the ISses share vias with their neighbor ISses.

As for running the selftest:


Sadly, I never came around to filming the second part with the Dr Hakan errorfinder...

[snoopy]

Yes that's pretty much what I did to get the error message.

I had a prod around again tonight and pressing on the acquisition board didn't seem to make much difference. However I have noticed the following pattern with Instavu switched on.

With ch-1 on only and Instavu the trace disappears between and including 25Ms/s and 1Gs/s.

If you switch on any of the other channels then the ch-1 trace re-appears and of course switching on ch-2 fills the screen with junk while ch-1 trace is normal. I really need to know what that error code stands for. Perhaps I may try my luck with the Tektronix forum to see if anyone can decipher the code.

cheers
dsvid
cheers

[TAMHAN]

Hello,
I would try to run the self test for a few times. Usually, after a few runs, more errors are emitted.

Tam

[snoopy]

I keep getting the same error !!

[snoopy]

OK I have made some progress and fixed it.

Re-soldering all of the ram chips in ch-2 did not make any difference, however I did find some more dry joints on ch-1 demultiplexer (U400) which did not appear bad the first time I inspected them. Sure enough when I soldered them up everything works now

cheers