Dead HEC-250LR-T PSU

[socketa]

Hello,
Hopefully a quick one.
Anyone got any pointers on where to look, for this fault condition?

This was working OK, then next time i tried it, it didn't start.
No noises or smells. No visible signs of damaged components/overheating.

No voltage on any rails, including standby
No fan twitch (so no short?)
Have 160v DC voltage at primary capacitors

Can't find any shorted diodes.
All of the transformer windings are continuous.
TOP222Y standby? switch start-up cap, 25uf/25v tests OK (24uf /1.3ohm ESR)
Have 160v across all legs of this topswitch
Optoisolator tests good (removed and tested).

Found one resistor that appears to be shorted (tested in-circuit) (see photo)
It's connected to the transistor? that is mounted on the heatsink.
Would this prevent the PSU from starting up?
Could this be a start-up resistor?

[stj]

Re: Dead HEC-250LR-T PSU

you have 230v mains - you should see about 300v on those caps

[momaka]

Re: Dead HEC-250LR-T PSU

you have 230v mains - you should see about 300v on those caps

+1
Anywhere between 300 and 340 Volts is acceptable.
Check input fuse, thermistor TH-1, and bridge rectifier.

[socketa]

Re: Dead HEC-250LR-T PSU

Yep, there is 320v across both capacitors
(160v across each of them)

[momaka]

Re: Dead HEC-250LR-T PSU

Ok then.

Next check voltage on the auxiliary rail that powers the TOP222Y IC (i.e. check the voltage on the "Control" pin of the TOP222Y IC with respect to primary-side ground). The 5VSB transformer should have two windings on its primary side: the main winding (which is what powers all of the other windings on the 5VSB transformer), and the primary-side auxiliary winding. To distinguish which is which, the main winding should have one of its legs directly connected to the positive bus (320V) on the main capacitor and the other to the Drain of the TOP222Y IC. The primary-side auxiliary winding should have one of its legs connected to primary-side ground (the negative bus). The other leg is usually rectified by a diode and is filtered by a cap - this is the primary-side auxiliary power rail that powers the TOP222Y IC. According to the TOP222Y datasheet, you should have somewhere around 4.7 to 5.7V, if I am not mistaken. Under 4.7V will trigger UVLO mode (under-voltage lock-out).

Let us know what you find. The first step is to get the 5VSB running.

[socketa]

Re: Dead HEC-250LR-T PSU

Thanks for that.

There is 0V on control pin of TOP222Y, with respect to primary-side ground.
(my primary side ground, is the position of the -ve test lead when i place the multimeter across the capacitors and measure +320V)
Yes, the main winding has one leg connected to the Drain, and the other leg is connected to 320 via a bigger 4.7ohm resistor (you can see it the picture yellow/purple/gold) - except that it measures 1376ohms with +ve probe on winding, and megohms or greater when i reverse the leads.
Is this a fuseable resistor gone bad?

Primary-side auxiliary checks traces out OK, just as you say.

Is the operation of the primary-side auxiliary dependent upon current in the main winding?

[Longbow]

Re: Dead HEC-250LR-T PSU

the other leg is connected to 320 via a bigger 4.7ohm resistor (you can see it the picture yellow/purple/gold) - except that it measures 1376ohms with +ve probe on winding, and megohms or greater when i reverse the leads.
Is this a fuseable resistor gone bad?

That would make sense, since the high side of the winding should be AT the full V+ during operation. 1376 /= 4.7.

Is the operation of the primary-side auxiliary dependent upon current in the main winding?

Why would it be? You need the +5 (typically) whether the main supply is up or not. Replace the blown resistor, then replace your 222. It is shorted Source to Drain.

[momaka]

Re: Dead HEC-250LR-T PSU

(my primary side ground, is the position of the -ve test lead when i place the multimeter across the capacitors and measure +320V)

Good.

Yes, the main winding has one leg connected to the Drain, and the other leg is connected to 320 via a bigger 4.7ohm resistor (you can see it the picture yellow/purple/gold) - except that it measures 1376ohms with +ve probe on winding, and megohms or greater when i reverse the leads.
Is this a fuseable resistor gone bad?

Yes, most certainly it is if it measures that high.

The reason you likely get a reading of 1376 Ohms one way and much higher resistance the other could indeed be because the TOP222Y has a short to ground and you are reading the two diode drops on the bridge rectifier. So definitely check for a short-circuit between Source and Drain on the TOP222Y IC.

Also, check the snubber network across the main winding on the 5VSB transformer. Typically, it consists of a diode in series with a resistor-capacitor parallel network (or alternatively a diode and Zener diode in series). See the sample schematic diagrams on pages 6 and 7 of the TOP222Y datasheet to see what I mean. Most importantly, check for any shorted diodes and ceramic capacitors or open resistors.

Is the operation of the primary-side auxiliary dependent upon current in the main winding?

Yes.

Basically, the TOP222Y IC has a low-power internal current source that uses the high-voltage present on its Drain pin to charge a capacitor on its Control pin. Once the cap is charged to the Auto-Restart voltage, the TOP222Y fires so that Source-Drain is closed. This causes a large current to travel through the primary-side main winding on the 5VSB transformer, which also causes a current to appear on the primary-side auxiliary winding (and also the secondary-side windings, such as the 5VSB output itself and possibly a secondary-side aux. winding, seeing how this is a half-bridge PSU). This current on the aux. winding is then rectified with a diode and passes through the opto-coupler for feedback. It is used both for feedback (i.e. voltage regulation) and to power the TOP222Y IC.

After replacing the faulty fusible resistor and TOP222Y IC, use the series incandescent bulb trick when applying power to the PSU. This should prevent your new TOP222Y IC from blowing up if there is something else wrong.

[stj]

Re: Dead HEC-250LR-T PSU

if the topswitch is blown then you also need to replace the startup capacitor.
a bad startup cap can blow a topswitch.

[socketa]

Re: Dead HEC-250LR-T PSU

if the topswitch is blown then you also need to replace the startup capacitor.
a bad startup cap can blow a topswitch.

That was checked in the first post. And appears to test OK.
Could've that open resistor caused the blown topswitch if the start-up cap is good?

Thanks momaka.
i was wondering why the reading across the resistor changed when reversing the leads, and thought that it seemed a bit odd that the resistor should behave like diode.
And yes, that's close enough to the what the diode drops measure across the bridge rectifier.
So, that resistor has failed open, and i've desoldered the drain pin and determined that there is a 30ohm short on the topswitch, between drain and source.
Have found a replacement resistor and will order another TOP222Y
Then will post the result back here, when done.
Will spend some time absorbing the rest of what you have said.
Really appreciate the insight, clarity, and forward thinking.

[socketa]

Re: Dead HEC-250LR-T PSU

Replaced parts, including start up cap
Runs well
Thought might as well have look at ripple voltage
The first photo is a well defined trace of the 5VSB, with just the hard drive connected
The other pics are of the 12, 5, 3, and 5VSB with just the motherboard connected (can't get a well defined trace)
All appears to be good, except perhaps for the 12v rail which show 100mV peak to peak.
How much ripple, on this rail, is classed as acceptable?

[Per Hansson]

Re: Dead HEC-250LR-T PSU

The ATX spec is max 120mv for 12v and 50mv for 5v & 3.3v.
That said the spec also calls for a small cap to be used on the probe while testing.
(Don't remember the exact specifications for this).

[momaka]

Re: Dead HEC-250LR-T PSU

Good to see you got that PSU going.

Like Per said, you need to have a small cap between the probes to get more accurate results, so that might be why you are seeing high ripple. I remember there was an article on this somewhere, but I forgot where (it was a very long time ago that I saw it).

[socketa]

Re: Dead HEC-250LR-T PSU

Thanks.
Sounds good enough then.
And could also compare the ripple with another PSU