TDS540 shift button action (ongoing repair)

[manicdoc]

Hi,

Got a TDS540 to fix, the previous owner has replaced some elec caps but not all, I'm going on a blitz as nothing comes good on ESR so far, amazing it boots up.

Regarding the shift button, is that meant to latch? Tried doing the shift and then Utility buttons get to the diagnostics but it wasn't having it. the shift LED lit up but went straight out. I've just replaced the caps on the display board and found a SM cap with brown snot under it measuring 0.3ohm, cleaned that up and it went to a proper resistance... There are a few signs and cat pee smells of capacitor leakage but not too bad.

Just want to know what is the correct behavior for the shift button..  had a look at various videos but didn't see anyone using it...  thanks

Update (in case anyone comes across the same problem):

Did a teardown of the Front Panel, found some liquid under the pads, gently cleaned those, and put back together - now the shift button does latch with the LED on and I can get into the Utility menu...

[manicdoc]

A quick update:

With cap replacing now down to just the acquisition board throwing mem test errors for two addresses 0x730000 and 0x732000. From what I can google, this looks like something up around DIG A. I'm going to follow HFADJB as others have indicated possible unity gain opamp failure is possible with maybe a bad OCLBKB and ~OCLKB to boot.

Got the whole thing behind an isolation transformer, so free to probe without fear...

Don't have a GPIB to USB, yet, might see if an Arduino can be utilized for that, can't warrant myself throwing north of $200 on just that.

[wn1fju]

I had exactly the same Shift key problem and solution as you on my TDS520.  The leaking electrolyte finds its way where it wants to go.

It is false economy to replace some of the surface mount electrolytic caps.  Replace them all.  Scrub the board several times.  Look with good lighting and a magnifying glass at all the traces, pads, IC pins, etc, anywhere near the caps.

Some "experts" on the TDS-series scopes warn that one shouldn't even power up the scope before the caps have been replaced, for fear of conducting electrolyte shorting components.  But who can resist turning on an old piece when you first get it?

Good luck with the repairs.

[manicdoc]

I had exactly the same Shift key problem and solution as you on my TDS520.  The leaking electrolyte finds its way where it wants to go.

It is false economy to replace some of the surface mount electrolytic caps.  Replace them all.  Scrub the board several times.  Look with good lighting and a magnifying glass at all the traces, pads, IC pins, etc, anywhere near the caps.

Some "experts" on the TDS-series scopes warn that one shouldn't even power up the scope before the caps have been replaced, for fear of conducting electrolyte shorting components.  But who can resist turning on an old piece when you first get it?

Good luck with the repairs.

The previous owner had turned it on, so there is a possibility of zapped components for sure, one op amp had been replaced for sure; I suspect some others have also drowned in electrolyte.

Yep, I'm trusting nothing cap wise atm, the ones that were replaced measure good, so I'm leaving them for now.

[manicdoc]

Getting somewhere..  the HFADJA is garbage. Looks like something up in the Analog Multiplexer, incoming +/- 1.71v is fine, just gets dropped in the multiplex circuit. The power to the chips looks good. I suspect the decoding control circuit side... So have it cornered. More probing in the afternoon.

[james_s]

It's not so much parts getting zapped, it's rapid galvanic corrosion that literally eats traces, vias, and other parts. Dip DC energized wires into salt water and watch how rapidly they corrode. Electrolyte has a similar effect.

[manicdoc]

Chased the DAC through the multiplexers, but nothing not controlling voltages.

Had another look across the board and saw around R1613 signs of residue, so took that off with the transistor next to it and gave the area a good clean, then put them back. Seems the DAC issue has now cleared, might be a fluke, but something has changed as the DAC error came up every time.

Getting the following now.



Yep, it's upside down, that's why the lighting looks weird.

I'll go after the Trig error next, lots of op amps and levels I can check.  Actually, based on a quick search, might be likely corrosion is to blame... I shall find out.

[manicdoc]

Decided to check around where an op-amp had been previously remounted. hmmm U508 is meant to buffer a 10v ref, but it's coming out as a sawtooth between 8.5v and 11.1v and a following inverting unity op-amp on the same chip is showing -4.72v and not -10v. hmmm

I've got some TL072's on their way, I'll remove U508 and probe around for passives not behaving. the area has signs of corrosion...

[manicdoc]

Took the opportunity to do some more checks around the suspect area last night. U508A & B are not behaving as expected, see attached.



The A op-amp isn't tracking on the -/+ inputs, a 4.35Mhz 8.62v to 11.1v sawtooth has come out of nowhere (aliens?). The +/- 15v is solid, the ripple there is certainly not causing a 2.48v jitter.  I did notice between pins 2 and 3 a 10k reading, which seems somewhat suss, this should be in the meg range. There is a 5-way fan out from pin 1 - there might be some crud somewhere interfering though nothing I've probed around it and nothing has the same character so far.  I also checked all the passives in the drawn box and replaced C527 as it was a bit too high on ESR for my liking (this had been previously replaced) and nothing really changed - all the other passives seem good.

I'm suspecting either U508 is bad or some short/crud interfering. Just waiting on the parts delivery now... hurry up UPS!

[manicdoc]

Update: Replaced U508 with a TL072 - now the buffered +10VREF is not jittering, nice and solid. The inverted output leading to the VSUB inputs on U501/502 is now -11.7v where it was before -4.72v so getting closer to the required -10v. Took some more measurements around U508, see attached.



What's puzzling is how pin 7 of U508B sitting at -12.4v below the -11.7v on the other side of 100-ohm resistor. I simulated the circuit and pin 7 should be around -9.5v...  Either we have faulty passive somewhere or more gunk hiding somewhere...   I'm going to remove all surface mounts around U508 and check out of the circuit, too many parallel loops via earth in a DC circuit to check in the circuit otherwise.

I also did some work around U400, I did some additional cleaning and two address pins came loose - I'd previously checked connections to the ram chips, and all buzzed out fine. This was following on from the mem errors for 0x7300000 and 0x7320000. Soldered those back and confirmed working. I've also changed the caps on that bank.

Question: Has anyone got captures for good +/-ACQCLK to U501/502 and +/-CLKA and +/-CLKB ? What I'm seeing doesn't make sense to me, looks too noisy. Just want to work out if I need to go down the signal path some more.

[manicdoc]

Update:

I checked the resistance on the Dig A memory banks back to U400 - was getting 30 to 40 ohm on some of them. Resoldered those and found one marginal pin that came loose with ease, fixed that. Also working through replacing all the caps in the memory banks, found one with white pee under it, so they all need to go.

Once it warms up I'm down to the STRIGA expected 1 got 0 error atm.

Also went back to U506/508, took them off and the components around them, much gunk was found under the TL074C, the rest was reasonable, just cleaning and putting back. The TL074C I'll replace with a new one.

[manicdoc]

Update: the DAC issue came back...   Now, i did do a clean around that part of the board, given it was looking pretty nasty compared to the rest (someone had cleaned it before but not well to me). Long story short, I think some muck moved under one of the multiplex chips where the buffered DAC output line runs under the chip and via's to the 7 multiplexer chips to drive the group, and is shorting to some clearly sine wave line - not what should be going on at all. I think I'll carefully remove those multiplexer chips where I cannot see the common DAC line come out, as they must be under the chip and see if my gunk theory is correct. Luckily the chips are in a tight group so the common line isn't running all over the board. I have also done some additional cleaning and measured resistances on neighbouring pins to be sure - but I think the chips need to come off to be really sure.

I've got a set of multiplexer chips in the post just in case we have a bad one. I think the DAC multiplexer circuit was marginal to begin with as it used to crop up as a fault occasionally - the DAC itself is rock solid. I've replaced the buffer op-amp as I think that suffered, the signal improved a bit but the problem remains... I might have to remove all 7 multiplexers and the op-amp and check if the bus is open circuit with reference to anything else. hmmm

Fixing this is proving to be a bit of a mission... but better than NetFlix 

[manicdoc]

Removed and checked all the 1 to 8 analog switch chips - found one that was bad and replaced it, put them all back, and still got the DAC fault error. The DAC signal is getting to all the multiplexers okay...

I put together a spreadsheet to allow me to easily trace out connectivity and signal paths for all the places the DAC signals will go (see attached). Established connectivity is good from the switchers to the numerous hold and buffer circuits, next I'll check all signals in and out of the buffers. I'll also check the -10v ref generation circuit again...

Does anyone know what specifically triggers the DAC fault error?  i.e. how does it know there is a problem... If I know that can I trace back from it. thanks.

[manicdoc]

The spreadsheet made it easy to determine that 2 of the 1 to 8 switching chips were not working (U940 and U941), replaced those, and confirmed the right voltages are going where they should do and holding steady on my scope. The DAC range error still persists...  Something further down the line must be out.

The adventure continues, with many twisty passages...

[manicdoc]

I think I have worked out how the DAC gets range tested, on sheet 23 for the acquisition board schematic, you have a VARHO input from the DAC (via an analogue switch), this feeds into a pair of comparators which ultimately then feed into the CPU board. There looks to be a lot of CPU controlled lines around this circuit and its not tied to any one channel, with variable holdoff - so ideal to be configured to perform the DAC range testing...

Everything else off the DAC nest of switches goes into per channel adjustments pretty much, so this must be it... I noticed at power up the DAC output does a series of +/- max/min voltage transitions then goes into the discrete voltages for driving the hold circuits. I think it does this cycle twice.

I'll probe around it this evening.

[manicdoc]

Interesting, the DAC voltage is getting through, I see +/- 10.1v at start-up being cycled through to pin 3 on U1500A, then a voltage divider and offset pull down with ref to -15v gets that down to around +/-4.5 v - but on the U1502 comparator, the - input is pegged to -5.0v and seems to stay there. I used single shot on its output to see if I could get a zero transition - didn't get it..  the area around U1502 has suffered cap leak - I've cleaned up the chip was some gunk under it..

I'll do the maths on the line between U1500A and U1502 and see if the range is what it should be when presenting to the comparator.

The rest of circuit beyond that I see digital transitions via the output of the other comparator. it could be that I have a bad comparator or weak input or the - input is wrong or the transition is too fast, but I doubt that.

getting there...

[manicdoc]

Interesting, the DAC voltage is getting through, I see +/- 10.1v at start-up being cycled through to pin 3 on U1500A, then a voltage divider and offset pull down with ref to -15v gets that down to around +/-4.5 v - but on the U1502 comparator, the - input is pegged to -5.0v and seems to stay there. I used single shot on its output to see if I could get a zero transition - didn't get it..  the area around U1502 has suffered cap leak - I've cleaned up the chip was some gunk under it..

I'll do the maths on the line between U1500A and U1502 and see if the range is what it should be when presenting to the comparator.

The rest of circuit beyond that I see digital transitions via the output of the other comparator. it could be that I have a bad comparator or weak input or the - input is wrong or the transition is too fast, but I doubt that.

getting there...

Quick update, replaced the op-amp and comparator - the voltage is slightly stronger into the comparator but its not passing over the - input to transition the output. I think something around the flip flop driving Q1543 isn't right - as when thats conducting its forces the - input of the comparator down to -5v effectively. Did the logic table analysis on the flip flop and something ain't adding up., either it is bad or the S input never pulses. I'll do some more probing tonight. Hopefully I'm not chasing a dead end, as it should be wiggling the line during diagnostics for a good reason...