DIYHiFi.org

This is an old DSO card that has worked well for me. It is the Link Instruments DSO-28264. It samples at 200MSa/s to an internal buffer, and has an 8 channel logic analyzer and very good software.

It has recently stopped working. So far I have located a short on the -5v rail, it is reading 0.0 ohms to ground. First I removed all tantalum caps on that rail. That didn´t fix it. Then I cut the ground leg of the switching regulator, on the lower left. The rail still reads 0.0 ohms to ground.

Does anyone have an idea of what is wrong?

Thanks
Alex

You can cut the power trace near each chip and go measuring.
Probably one chip went kaput.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

A more subtle way is feeling each chip with your finger. The one with a short circuit will be hot.
Also a good idea is cleaning the solder side of the PCB with some PCB cleaner and old tooth-brush. Especially if the tracks are very close a minute conducting particle can make a short circuit. Been there, done that.

Thanks for the suggestions. I hope it´s not an expensive opamp that is shorted.

I have powered the rail with a small supply at 1.2v and the current maxed out at more than 2 amps. I only did this briefly and couldn´t find any hot spots on the board. Since the short reads 0.02 ohms it will dissipate much less than this small supply.

Can ceramic bypass caps short at 0.02 ohms? What components have this failure mode?

Thanks
Alex

Philips contact cleaner, red can.
Miraculous stuff, no joking.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

I definitely need that.. is there a generic?

IPA, isopropyl alcohol is great stuff. Even better can be to wash a board in hot (clean) water, then dry/clean it with a dip in IPA. (yes I am a compleat amateur)

Meanwhile - do you have away of applying a really low voltage at >1A, say 0.2- 0.3v and measuring current - just to check you are looking for a dead short, not just something with a forward-biased junction ?

If you need a contact cleaner, this is it.
Not that you really need it for this particular case, but otherwise... it just solves problems that others don't, IME for many, many years.
Actually, my father used it too. More than 20 years at Grundig, since the late 60's.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Hi Martin,
Thanks for your suggestion. Good idea. I´ll take the card to the lab with me, sometime this week.

Best
Alex

Finally got around to this. I have applied current and the AD9617JN opamp is heating up a bit, I hadn´t noticed this before. Looking at the circuit, the non-inverting input is grounded. What might be happening is a short inside the opamp, between the non-inverting input and the -V leg. I shall desolder the opamp and see.

Best
Alex

The short was in the small ceramic bypass next to the ad9617, the upper one. The cap near the smd ferrite.

Now I have to replace one opamp and a lot of caps. Well, this should fix this nice instrument. It even works in Win 7, in my P4 3GHz w/ isa slot.

Best
Alex

Now I can answer my own question and the answer is YES. The short was in the cap itself, I haven´t seen any short external to it.

Best
Alex

This is not yet fixed. There is still some problem in the analog section or in the A/D converters, as the traces are stuck at the maximum value.

I checked the datasheet of the AD9012 converters, absolute maximum ratings:

Analog Input Voltage . . . . . . . . . . . . . . . . . . . . –VS to +0.5 V

Where -VS is -5v. This input might have been exceeded when the -5v supply to the opamps was shorted to ground. The -VS supply to the A/D was shorted to ground as well, it is the same supply.

I guess the next step is testing these A/D converters with a direct input from a potentiometer connected to ground and Vref at -2v. The analog input to the A/D is from -2v to 0v.

Best
Alex

I did more tests and narrowed it down to the data latches after the AD converters. Looks like both latches blew up when I ran the hardware without converters (I removed the ADs from their sockets while troubleshooting the analog front end).

The latches are 74fct2574atp (dip 20). All input signals going to the latches are switching normally, including the clock (latch) input and the output enable. But the outputs stay at zero (actually 100mV). Looks like the inputs failed.

I did a quick search but failed to find the dip package at any of the popular vendors or ebay.

Best
Alex

I was mistaken about this. The inputs of these data registers must be fine. The real reason why it´s not working is a mystery.

(Took one 74FCT2574 out of the board and tested it, works good. Put it back in: does not work, even though all signals are switching. I put the board to work at low sampling frequencies like 10K and it still doesn´t work.)

I´m sure this thing can be fixed and I would like to, but it is a PAIN.

Best,
Alex