How to measure max. DC current of a desktop SMPS

[hobbyiclearner]

Hi,

My 3 year old desktop SMPS stopped working properly ie. it was able to supply power to motherboard but not to HDD and DVD writer. When I connected a brand new SMPS, the system is working fine. So I am guessing that the components have aged.

While trying to debug it, when I switched on the SMPS, all the DC voltages were getting generated (+5, +12, +3.3, -12). Since the SMPS was not able to supply to all the load, my guess is that the current delivered from the SMPS to the load is less than the max. limit. My question is how do I check the max. DC output current of the SMPS? For eg. In case of +5v, the max DC current is 29A. So basically the load would be practically nothing. Can I simply insert the terminals of a multimeter in between the terminals of +5v supply to check the current delivered (I am preferring to ask before risking with the SMPS).

Thanks,
Hobbyiclearner

 

[FvM]

No, you can't. Variable load resistors would be required.

Realistically, I see two option:

- Dump the unit into the electronic waste container
- Replace all primary and secondary electrolytic capacitors

 

[hobbyiclearner]

Pls. enumerate the variable load resistor check concept.

Thanks.

 

[Pjdd]

Pls. enumerate the variable load resistor check concept.

I don't mean to be rude - really - but if you have to ask about the concept rather than about some detail, then that approach may not be for you. For one thing, it will need a very high-power resistor or combination of resistors whose value can be varied from a few ohms down to less than 0.2 ohm. At the same time, it must be able to withstand up to 150W without burning up. Getting hold if such a resistor will not be easy.

Anyway, the idea is to connect such a resistor to the output of the SMPS and monitor the voltage while the resistor value (and thus the load on the SMPS) is varied.

You could go the route of FvM's second alternative by opening the SMPS and checking for bulging, burst or leaking capacitors.

 

[d123]

Hi,

Something along these lines:

**broken link removed**

 

[hobbyiclearner]

You could go the route of FvM's second alternative by opening the SMPS and checking for bulging, burst or leaking capacitors.

I had done that check... none of the capacitors seem to be faulty. I will go ahead and replace all the electrolytes and directly plug it into the desktop. Creating an resistor array as per d123 's post will cost the same/more than replacing the electrolytes.

Hobbyiclearner.

 

[Pjdd]

Yes, there's always the question of whether it's worth spending the time and money for just a one-time use. Several years ago, I designed a variable active load for testing computer SMPSes using high-power transistors, active cooling, and control circuits to vary the effective load current. But I left it unfinished because I had stopped working regularly on other people's computers and it wasn't worth the time and money needed for very occasional use.

The project in d123's link represents the idea, but will have to be modified to test an SMPS. For example, all 18 resistors will have to be connected in parallel to present a meaningful load to the +5V output.

 

[hobbyiclearner]

- Replace all primary and secondary electrolytic capacitors

No joy... I replaced all the electrolytes with new ones... still the SMPS could not supply to the desktop.

 

[schmitt trigger]

In the early 90s, I used to repair computer "silver boxes".
If the electrolytic capacitors were not at fault, usually it was one of the several output schotky diodes becoming faulty.

 

[hobbyiclearner]

In the early 90s, I used to repair computer "silver boxes".
If the electrolytic capacitors were not at fault, usually it was one of the several output schotky diodes becoming faulty.

Maybe I have something.... There are three components on a vertical heat sink...

1. STPS3045CW -- google search reveals it is a SMPS based 3 terminal schottky diode. I checked the forward and reverse resistance between A1 & K and A2 and K. All four were of similar value. Maybe that should not be the case (though I did not de-solder it from the PCB)

2. U7020 - ST make; 3 terminal. could not find its datasheet (in the PCB it was marked as D21 ; assuming D stands for diode)

3. B69322 - ST make; 3 terminal. could not find its datasheet (in the PCB it was marked as D22 ; assuming D stands for diode)

Any advice pls.

Thanks,
Hobbyiclearner.

 

[schmitt trigger]

The problem with schottkys is that they start becoming leaky, specially if the maximum reverse voltage is too close to its maximum ratings and coupled with high temperature.

 

[hobbyiclearner]

OK... but how do I find out if the schottky diode is OK or not? And any idea of the other two diodes?

Thanks,
Hobbyiclearner