PCI digitizer shows short on voltage rails

[Guille]

Hi everyone,

I recently was lucky enough to get my hands on a PCI digitizer Agilent U1068A Acqiris DP211, basicaly an osci that lives inside your PC, neat! However the device is untested by the owner and I do not have PCI ports on my motherboard so I have to find an adapter to test it. Before that though I decided to check for any obvious faults that could short my PC power supply and while measuring the resistance between ground and power rails I found the following:

• +12V: 3KΩ
• -12V: 20KΩ
• +5V: 189Ω
• +3.3V: 32MΩ
• +V I/O: 11KΩ

That 5V rail seems shorted to me but I couldn't find any obvious short from visual inspection. Does it seem right to you? The device is supposed to drain 25W when switched on, I understand the load will change when it's not on operation but a load of 189Ω, although low, would only source:

P = (V^2)/R = 25/189 = 132mW

Far below a fraction of those 25W. I guess my first question is Does 189Ω seem abnormally low for you? Would it indicate a faulty component somewhere?

In case the answer is affirmative I guess second order of business is finding and replacing said component. Visually everything seems right so all I can think of is hooking 5V to it's input with some current limiting and touching the components until I find one getting too hot, (I got a thermometer to verify the temp is abnormal for regular operation). Second question then is Is this a safe way of testing this 1000€ board? Do you recommend something else?

I'll attach some pictures and resources below:

Resources

• Datasheet:  https://www.keysight.com/us/en/assets/7018-08316/technical-overviews-archived/5989-7113.pdf
• Owner's manual: https://www.artisantg.com/info/Agilent_Acqiris_DP111_Manual_2018726163254.pdf

PS: You'll see that I removed a couple of caps on the photos but the short persisted, disassembling the front end PCBs changed the reading to a higher value of 357Ω.

[Swake]

189 Ohm is certainly not a short.
1 Ohm is not a short on a 5V 'computer' bus. This only represents 5A. (and 25W, can't be a coincidence  )

This method is ok to detect a short, however is not very good to evaluate how much a board might consume. On startup/boot lots of things will happen in that board before those big chips start consuming big power.

If you chicken out I suggest you to:
- Find an old PC with PCI slots, there is enough cheap stuff out there to not risk anything with your main computer.
- Slam the card in it and power it on!
- If something smokes it was already broken and now you know what is broken, (don't switch it off till you could see what component was smoking).
- If it is not smoking, well it might be working after all ! Install the drivers and find out.

[DaJMasta]

+1

That's definitely not a short and there shouldn't be any problem in powering it up to see.  Yes the card may cost something, but the vast majority of issues you can run into will only be evident with symptoms of a problem, so unless there's physical damage from handling or storage (like scraped off SMD parts), I wouldn't hesitate to plug it in and power it up as a first course of action.  If there's an issue from there, then you can try to find out what's wrong and address it.

[magic]

These measurement was likely taken with a DMM. All it says is that with 0.1mA or 1mA or similar test current, the circuit dropped 18.9mV or 189mV, respectively. This is not unusual on a board full of semiconductors. It says nothing about what the behavior will be with 5V input. Chips are not linear devices like resistors, their current doesn't scale proportionally with voltage.

It is also not a dead short, of the sort that would draw many amps from the PSU and set the board on fire.

[Guille]

This makes so much sense.

[fzabkar]

This makes so much sense.

To see this for yourself, take a working HDD or SSD and check the rating on the label. Then measure the resistance of the 5V and 12V rails at the SATA power connector.