Tektronix TDS7404 passes self-test but not SPC

[swperk]

I am trying to revive a Tek TDS7404 that has been in storage for a couple of years. It was working properly and passing its self-test and SPC at the time it was stored away.

On powering it up, I discovered that it had a bad Dallas DS1245YP-100 NVRAM. The PPC log reported a failed walking 1's test. I replaced the NVRAM and battery, and now the scope boots properly and passes all self-tests.

The channels all work, and I can trigger from all of the channels. Vertical and horizontal calibration look good, and all of the features are working properly. The only issue is that channel 1 has a slight negative offset of a few millivolts, easily adjustable by the vertical position control.

After warmup the SPC fails with the following messages from the PPC console port:

-> 0x364cdf0 (CalWatch): Temperature transition into SPC-supporting range.
0x365d9b0 (Ch1 Cal): >> Error at step ChanCalPart1.0: Measurement of 5.120 > limit of +/- 4.80during ibnull1X
0x365d9b0 (Ch1 Cal): >> Error at step ChanCalPart1.1: Measurement of 5.047 > limit of +/- 4.80during ibnull1X
0x365d9b0 (Ch1 Cal): >> Error at step ChanCalPart1.4: 1X Ibnull cal failed with dac value 2.614302during ibnull1X

These errors are reported almost immediately after starting the SPC, but the SPC process continues on for several minutes afterwards and completes with "Fail" displayed on the SPC screen. I ran the SPC five times in a row with no change.

Does the fact that the SPC completes without reporting any other errors mean that these are the only SPC errors? What is likely to be causing this failure and how can it be repaired?

Thanks,
Stan

[mikehank]

Ibnull failure can mean a bad preamp.  Set ch1 I put to less than 50mv/div and see if there is any voltage on the input.  That is install  at tca-bnc and measre volts with a meter.  Anything more than 8mv is a bad preamp.

[swperk]

I measured the DC offset on all four channels when set to 50 mV/div. Here's what I found:

Channel 1: 4.8 mV
Channel 2: 0.2 mV
Channel 3: 0.8 mV
Channel 4: 0.4 mV

So Channel 1 is clearly very different from the other three, but not above the 8 mV threshold. Does this still mean it has a defective preamp, or is there another possible cause?

If the preamp is defective, is it a part of the input hybrid, or is it another chip on the acquisition board? In any case, are replacements available?

Thanks for your help,
Stan

[mikehank]

Hi swperk,

It is likely a bad preamp.  Sadly for the TDS7404 it is not a hybrid its a BGA ASIC HFD144 Tek name.  This product is far out of support and Tek will not have them.  The only way to fix will be to cannibalize   a BGA from another board.  Sorry its not easier.

[swperk]

Thanks for the info, even if it was bad news. It will save me all the additional time I would have spent disassembling and troubleshooting the scope.

[manicdoc]

Do you have a schematic?  I suggest checking if any control voltages are running into the preamp from the main board and compare across channels for anything out on channel 1 alone. It could be fault upstream, i.e. bad opamp, bad connection etc.

Worth dismissing the simple stuff.

[swperk]

I have never found an actual schematic for the TDS7404, only a general block diagram. That being said, I can still swap hybrid modules pretty easily and see if the problem follows the hybrid.

[mikehank]

50mv is a direct shot into the preamp.  The ibnull failures like this one usually mean static damage which is erosional and not catastrophic like an overpower situation.  The fact that its channel one is also a clue to the failure.  Most folk use channel one to start. 

[mikehank]

Also you mentioned hybrid swapping.  The TDS7104 does have hybrids but they are limited to 1ghz BW.  The TDS7404 is a 4ghz scope and does not use hybrids but a fixed preamp.