Huntkey HK340-72FP SMPS shut on load above 2 amps

[abhishekkumar1902]

I have this Huntkey HK340-72FP PC SMPS where I am facing a strange problem where the SMPS shuts down when the load on any of its 12v output rails goes above 2 amps while the rating of each 12v rail is above 10amps capable.

I have checked the following:
1. All output caps are OK, and all output rails read correct voltages at no load
2. All output diodes are OK
3. NTC is ok
4. Hi voltage cap on input correctly reads 150uF on LCR meter per the spec, also the bridge rectifier is OK
5. No short in main switching MOSFET
6. All other components visually look OK, no signs of burns or anything
7. Standby 5V supply is good, which is why also the SMPS is able to switch ON when green wire (PS_ON) is shorted to ground

When I connect let's say a 5 amp load on any of the output 12 v rail, it shuts, and it works fine till a 2 amp load with a stable 12v reading.

Surprisingly the SMPS doesn't turn off if I directly connect the load before the rectification of the 12vrail that is directly at the secondary of the HI-POT transformer. What could be the issue? any help is appreciated.

[TheMG]

I'm assuming you're trying to use it as a 12v power supply and not to power an actual PC?

Looks like an older power supply design, where all the rails are derived from different windings on the main transformer with no post-regulation.

Do you have any load on the 3.3v and 5v outputs? If not, there's you problem. On this type of power supply if you put load only on the 12v and no load on the other rails, the unloaded rail voltages will rise outside of specifications and the supervisor IC will shut down the power supply.

Solution is to put one or two amps load on both the 3.3v and 5v.

Alternatively, use a more modern power supply that has buck regulators running off of the 12v rail to produce the 3.3v and 5v. Such power supplies you can fully load the 12v rail without any load on the 3.3v and 5v without any problems.

[abhishekkumar1902]

Thanks for the response.

Yes I am trying to use it as a 12V supply for some project.

But it has only four output rails:
1. 12v, 12 amps
2. 12v, 16 amps
3. -12v, 0.2 amps
4. 5v, 2.5amps (this one comes from standby supply section)

I have tried putting 2 amp load on the 5v rail, and also tried putting 2 amps on one of the 12v rails alternatively but it follows the same strange shutting down issue. I see that the 12v rails don't see any drop in voltages also, and they are perfectly maintained at 12.2v when I load them with a 2amp load.

I have also tried changing the main rectification diodes on the output section and checked each and every cap but no pointers 

[TheMG]

What type of load are you connecting? Is it just a resistive load?

Reason why I ask is that the overcurrent protection on PC power supplies is typically very fast acting, it usually does not tolerate the inrush current of charging up medium to large capacitors. Even incandescent/halogen bulbs, which has a lower resistance when the filament is cold, will often trip a PC power supply.

If you connect your load prior to to turning on the power supply, it might work.

If you're using resistive loads only, and still experiencing it shutting down, then in this case I'd say there probably is something wrong with it. If you've ruled out the usual capacitors, maybe turn your attention to the over-current protection circuitry. For instance, a dodgy soldering joint on the current sensing shunts might cause the OCP circuit to think there is more current than in reality.

[abhishekkumar1902]

  Thank you so very much for responding. Yesterday, I was testing it and found out the exact same thing that you just mentioned. I actually bought these supplies from a friend of mine and planned to fix them so that I can use them in any future project.

For a quick high-load capability test, I usually connect a 12v, 60-watt car halogen bulb on the 12 rails. And with this specific supply only, it tripped when I connected this 60-watt bulb to the 12v rail when the supply was ALREADY SWITCHED ON. Surprisingly, when I did it the other way, that is when left the load connected, and then switched it ON, magically it started and kept the load running without any issues at all. I then even abruptly connected another 60w bulb and then it didn't trip at all.

So yes, the SMPS seems to be working fine. It seems like it's just with this SPECIFIC SMPS design, they designed it NOT TO TOLERATE high INRUSH CURRENTS. I usually connect this same 60w load to other SMPS supplies I have even when they are already switched ON, and they take and run this load easily, maybe something specific to this Huntkey SMPS only in terms of inrush currents tolerance that it trips.

But can you please tell, why doesn't it TRIPS when I switch the SMPS post connecting the load? How does the SMPS handle inrush currents in such cases when the load is connected prior to turning the SMPS ON?

[TheMG]

But can you please tell, why doesn't it TRIPS when I switch the SMPS post connecting the load? How does the SMPS handle inrush currents in such cases when the load is connected prior to turning the SMPS ON?

Several things could be happening here.

One is that there is a time delay during power-up before the supervisory IC begins to do its thing. Several of the normal protections are temporarily inhibited during this time as the rails come up and stabilize to proper voltage. This allows just enough time for the filament in the halogen bulb to warm up so its resistance is high enough it no longer draws excess current to trip the OCP.

Measure the resistance of one of your bulbs when cold - you'd be surprised how low the resistance is. I happen to have a 25W 12V halogen bulb sitting here, this bulb would normally draw only about 2A in operation. The cold resistance is 0.5 ohms, meaning for a brief moment when power is applied to the bulb it would draw 24A!!! That's more than 10X the normal operating current for that bulb!

Another reason might be the very sensitive short-circuit protection incorporated into most PC power supplies. It might be interpreting the extremely sudden high current as a short-circuit and thus tripping the SCP. It's deliberately sensitive as most of the wires in a typical PC are way too small to carry anywhere near the full current that many power supplies can output, so it needs these protections for fire safety, in case for instance an individual wire gets pinches or chafes on the metal edges of the PC case and shorts out.

Lastly, the sudden load of the cold halogen applied to the power supply when it is already turned on without other loads, might just be too much for the feedback loop to respond quick enough to avoid the rail dropping out of tolerance, thus shutting down the power supply on under-voltage protection (UVP).

I would suggest looking up the datasheet for whatever supervisor IC is used in your power supply. Lots of good information to learn from in there about all the various protections inherent to a PC power supply. Also, halogen bulbs make lousy power supply test loads because of their non-linear resistance characteristics.

[abhishekkumar1902]

Sure, agree to all your points, this is great learning, thank you so much again for taking your precious time to reply. I highly appreciate that.