EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: AndrewB on October 18, 2022, 10:59:25 pm
-
Hi. Wondering if there are any PSU experts out there who could help. I’m trying to repair a PSU from a Compaq Portable II. I found I had a short on the 5V. Looking over the PSU I found a blown tantalum and replaced it. The 5V short was still there. Tracing back I found that if I removed a bobbin inductor the short on the 5V would go away. But when I tried the PSU I found that it seemed to be cycling – clicking noise and voltage cycling – as well as a resistor started to burn.
Does this make sense that the inductor would cause a short on the 5V or is this a red herring?
Is It an indication the inductor is bad?
Can anyone identify a replacement for it?
Thanks.
-
The circuit probably won't work without the inductor, they're put in series with important things.
An inductor will appear to be a short when tested with a multimeter. It's just a coil of wire, to DC that looks the same as a wire.
The noises are probably from the switch mode power supply (that you took the inductor from) misbehaving.
If you suspect a short: I'd suggest a thermal method of searching. Reinstall the inductor, spray your board down with alcohol and then power it on (preferably with a current limited supply). Watch and see where the alcohol evaporates from due to heat. It will probably disappear on both the voltage regulators AND some shorted chip somewhere. EDIT: Or just the voltage regulators if they are faulty.
-
Thanks for the suggestions. I swapped out the regulators for the 5V and 12V. I'll try the thermal method you suggested. I tried probing around with a temp probe but didn't pick up any warm components. The only component I was not sure about testing was the rectifier (MBR3035PT) - if it was bad could it cause what I'm observing?
-
How do you know you have a short on 5V? Is that just your measurement of the inductor? What exactly are the symptoms of this PSU? Electrolytic caps frequently die in switched mode power supplies, because they take a beating from all of the switching (and if made to a low price the caps may be Chinese junk).
-
Hi. Tested for the short on he connector pins for +5V and Ground. I also removed and tested the electrolytic caps and none were shorted. Their old but not cheap, this is a PSU from 1986. Their ESR values were not high. I've replaced the voltage regulators for the 12V and 5V. So right now the 5V short remains. Plugging in the PSU I get a clicking and cycling.
-
If there are SMD ceramic capacitors on the power rail, check whether one of them is shorting.
-
There are no SMD caps except on some little circuit panels. I'm going to continue and remove and test the ceramics.
-
maybe zener diode
-
Something is shorted before or after the inductor and when you remove it you are isolating that part of the circuit. The inductor itself is just going to appear as a short to DC, but it will attenuate high frequency noise.
-
That's just a choke to filter out high frecuency noise...
Check the rectifier diode, if shorted, that's your issue!
-
That was the one component I was not sure about when I was testing. Can you suggest the best way to test and what to expect to test this? It's an MBR3035PT. When I checked it it did not have a short between the legs. Maybe I was doing it wrong.
-
MBR3035PT is a dual Schottkey diode. Using the diode check function on your meter you should see a diode junction drop of about 0.6V from either side leg to the center leg. If it isn't shorted then it's probably good.
Do you have a schematic? I had one of those Compaq Portables many years ago, it was cool, I don't remember how much documentation was available though.
One other thing, you should never power up a SMPS with any parts removed, especially an old one, that can let out the magic smoke. It also will probably not work properly without a load, 12V automotive lightbulbs make good test loads.
-
@Strawberry: Thanks for the diagram but this a Compaq Portable II. The power supply is different. I have a Portable I and it doesn't use this crazy card edge hookup to the motherboard. Unfortunately it looks like no one has schematics for the Compaq Portable II. The ID numbers on the board are: ASSY No: 000175-002, Diagram No: 00176-000, 000177-001 Rev. F
@James_S: Thanks. I'll test using the info you provided. Ya I know it's not a good idea to power it up without a load and with components missing. It was a short momentary power on to see if anything has changed - maybe pointing to a weak component. Not going to try it again until I'm sure. Ya the load part is going to be a pain because of the card edge interface. But I'll Jerry-rig something and I usually use a HDD to give it a load on the 12V and 5V. Right now I'm continuing to pull parts off to trace back the short. I got the short off the edge connector but still have some on some caps spots. I'm worried it might something on these little circuit boards.
-
For stuff with weird connectors you can tack solder wires to the other side of the PCB to connect a load. 12V bulbs work fine as loads on both 12V and 5V lines.
-
schematic is same
same TL431(SO-8)
same opto isolators
...
they went for ceramic sip modules
-
@Strawberry: Ok I'll take a closer look at that schematic.
@James_S: Thanks. Your confirmation confirms how I had initially tested it. I only got .14V on the legs. I went ahead and ordered a few. Might as well change as many of these 35+ year old components while I'm in here. Ya soldering directly to the pins under the board was the plan for testing going forward once everything is back together.
-
Was that in circuit or out? 0.14V might be ok with a Schottkey, I was going off the maximum Vf from the datasheet. Normally when a diode is bad you would read a direct 0.0V short.
I don't agree with replacing semiconductors as a preventive measure, they do not wear with age so it is irrelevant that they are 35 years old and every time you replace a part you risk creating a new problem where there was not one by installing a part that is wrong in some way or inferior to the original. There are some designs that are picky and rely on characteristics of specific parts and there are some modern parts that have the same part number but are not truly identical to old ones. Understand the fault, test the components, and replace components that test bad.
-
It was out of circuit. I'll compare it to the new part and keep it if they test the same.
-
0.59V@12A
0.3V@0.1A
~0.2V@0.01A
reverse leakage ~1..20mA (depends on temperature and voltage)
multimeter generate 1mA