ATX power supply and random reset issue (Solved)

[arvinfx]

Hi,
I have random reset problem with 550w green power supply.

I have investigated and found out that IC PS223 that is over-under voltage protection is the guilty one.

But with more checking i found something weird , the PS223 is doing well, and my 12v has some spikes more than 4 VOLTS!!! That total voltage could be 16v on 12v !
More investigate and I end up on 5vsb, when PSU is off, there is some noise on all voltage outputs include 5v,12v,3,3v also! You can see the Oscop image blow.
I totaly confused! Which place i have to check?

 

[arvinfx]

I have found that noise is on the ground, but how i can find the source of it?

 

[ChrisP58]

Voltage, including noise, is measured between two points. Usually you connect the scope ground clip to you circuit ground,

As such, there shouldn't be any noise, or voltage, because the probe tip and the ground clip are at the same point. If there is, you're not probing it right or there is a problem with the scope or the probe.

 

[arvinfx]

Voltage, including noise, is measured between two points. Usually you connect the scope ground clip to you circuit ground,

As such, there shouldn't be any noise, or voltage, because the probe tip and the ground clip are at the same point. If there is, you're not probing it right or there is a problem with the scope or the probe.

Nice to hear your reply,

Yes you right, there is something wrong. I checked scope and prob , i have change the prob with a brand new and same result, also I try 3 different SMPS to be sure about scope measuring. what noise was not on others except the one i had problem.

So , I try to connect both tip and ground of prob to ground as you can see in the pictures.

both hot side and cold side has this noise!!!!!!!

Cold side Ground place:



Cold side wave sample:


hot side earth place:



hot side earth wave sample:



what is it!!!! HOW CAN I SOLVE IT?

 

[Nigel Goodwin]

Nice to hear your reply,

Yes you right, there is something wrong. I checked scope and prob , i have change the prob with a brand new and same result, also I try 3 different SMPS to be sure about scope measuring. what noise was not on others except the one i had problem.

So , I try to connect both tip and ground of prob to ground as you can see in the pictures.

both hot side and cold side has this noise!!!!!!!

Cold side Ground place:

View attachment 129996

Cold side wave sample:
View attachment 129997

hot side earth place:

View attachment 129998

hot side earth wave sample:
View attachment 129999


what is it!!!! HOW CAN I SOLVE IT?

Why do you want to solve it? - the issue is most likely pickup on the scope, not actually ripple on the supply. It's VERY, VERY difficult to get noise free readings with a scope in such cases.

 

[arvinfx]

Because under heavy process pc will reset randomly, I bypass the protection circuit and it never reset, but motherboard show me some warning about abnormal voltage . So something is not correct thought hum?

 

[ChrisP58]

Sounds to me like your heavy process needs more power than the power supply can give.

The protection circuit is there for a reason, and it's not just to annoy the user. And you may have affected the voltage regulation when you bypassed it.

I think you need a higher power supply.

 

[Pommie]

Recently I've had problems with my 2 data drives randomly disappearing. Last weekend this also happened with external (USB) drives. I concluded that it was probably a power supply problem. Fitted a new 650W supply today and, so far, no problems. Try a bigger (more watts) power supply.

Mike.

 

[arvinfx]

I have checked another 4 seconad hand PSU exactly like this with near or same power. (all main controller IC are fan4800 with Active-PFC)
3 of them had same problem on scop !!! and one of them was fine! some thing is wrong!

Look spikes is exactly on PWM switching:

 

[Nigel Goodwin]

Might be high-ESR electrolytics? - check the ripple on the smoothing electrolytics on the secondaries.

 

[arvinfx]

Might be high-ESR electrolytics? - check the ripple on the smoothing electrolytics on the secondaries.

check file names, I mention the prob tip place and also on/off stage of PSU.


The appearance of all electrolytic caps are normal.
I checked 12V cap out of the board capacity was fine and ESR was 0.1 ohm

SOMETHING really strong is messing up the ground level and PSU stability.

 

[arvinfx]

I have found the problem.

The frequency of spikes is around 166mhz, and 1/4 of wave length is really near to my ground clip of my scope prob!

I use a short ground spring and the noise disappear totally:

So the noise was on the air around the SMPS.