Trust PW-5150 370W

[goodpsusearch]

I don't know if anyone of you is familiar with that name. I got that power supply from a friend, he told me his computer stopped working.

By the time I opened the case it looked familiar. Either Macron Power is the OEM or they copied them. Just compare the add on board of MPT-350 and Trust, the layout of the PCB, the spot of the fan connector, the 5vsb output caps and choke, etc.

Even the psu case is the same, apart from the black paint.

"XHY-300AB", the PCB code is located at the same place where the MPT-301 and MPT-350 have "ATX991 2P Macron Power".


But is it a decent power supply?

-The input filter is ok but there are no MOVs.

-3A diodes instead of a bridge rectifier. I could forgive that.

-2x470uF 200V CS caps. Real capacitance is 352uF and 323uF. I call those 330uF caps.

-2x 13007 switching transistors. This is more like a 250W (300W peak) unit, not 370W.

-Good ERL35 transformer

-3.3V: S20C45C (20A), 2x2200uF 10V BH caps. Good to see 2200uF caps on a psu like this.

-5V: S20C45C (20A), 2x2200uF 10V BH. The 30A the label claims is just a lie.

-12V: U20C20 (20A) 1x2200uF 16V BH

All the caps are BH, except a Sapcon 100uF 50V and the CS primaries. 2 10uF 50V BH caps are bad with esr up to 20ohm when the maximum allowed is 3.5ohm. There is severe heat discoloration on that area. Maybe that's the reason it stopped working.

There is something really nasty on the back side of the PCB, something like glue that is very difficult to clean...

One of the two 13007 gives me a short between Base and Collector. Is this normal or it needs to be replaced?

[tom66]

Re: Trust PW-5150 370W

I have seen them a lot in the UK; I would say they make a lot of stuff, but more than certainly they just import re-branded junk.

[Wester547]

Re: Trust PW-5150 370W

A number of BH capacitors lie about their capacitance (primaries, AFAIK and as stated, but that could extend to secondaries) so those '2200uF/10V' capacitors on the secondary may be 1500uF/6.3V/10V. Some parts in this PSU look like they have merit in keeping and it may be worth a repair but it generally looks like junk...

Looking at how this PSU measures up it looks more like a 200W PSU continuous, 250W peak... at least that's how I'd rate it. And that glue on the PCB is conductive.

[goodpsusearch]

Re: Trust PW-5150 370W

One of the two 13007 gives me a short between Base and Collector. Is this normal or it needs to be replaced?

I took it out to test it... It's semi-shorted. Time to find a replacement...

[lti]

Re: Trust PW-5150 370W

BH capacitors lie about their capacitance so those '2200uF/10V' capacitors on the secondary are probably 1500uF/6.3V/10V.

I have a BH 470uF 35V cap that measures 479uF, and it has had 36V applied to it without any problems.

[goodpsusearch]

Re: Trust PW-5150 370W

I found a replacement transistor from a dead psu, it's a E13007F. It is in TO-220F package and the collector dissipation is only 40W, when the original was 13007 in TO-220 package with 80W dissipation...

1) What impact this could have on the psu performance?

2) Is there a problem that the two 13007 transistors on the power supply are not be identical?

[momaka]

Re: Trust PW-5150 370W

Yeah, looks more like a 200W continuous, 250W peak PSU to me. Not worth new caps IMO (although old but tested-good Teapo and Ost would make a nice upgrade for this PSU).

The 5VSB looks like it's about to go south. Check the 5VSB voltage under all loads. If it hisses, whines, or makes weird noises with a 1 to 2A of load, it's time to replace the output filter caps on the 5VSB along with the controller supply filter caps, even if they all look good. If the 5VSB still makes noises afterwards, those ceramic caps on the primary of the 5VSB circuit are the next suspects. At that point, though, I would just say it's not worth it unless you're really desperate to get a working PSU .

I found a replacement transistor from a dead psu, it's a E13007F. It is in TO-220F package and the collector dissipation is only 40W, when the original was 13007 in TO-220 package with 80W dissipation...

1) What impact this could have on the psu performance?

2) Is there a problem that the two 13007 transistors on the power supply are not be identical?

1) Probably won't be able to pull as much power out of the PSU - if it works at all. I never tried it.
2) Not sure, but I don't think it's a good idea. Best to replace them in pairs. While you did find the shorted one, the other one could have failed open.

A few other things worth mentioning:
- The E13007F is in an insulated TO-220 package, which won't allow it dissipate as much heat on the heat sink as the regular metal tab TO-220.
- Check the snubber diodes across those switching transistors - they may also be shorted or intermittent. If in doubt, replace them. They need to be of the ultra-fast recovery type, rated for up to 1000V. They usually are 1 or 1.5A on these generic PSUs.

[goodpsusearch]

Re: Trust PW-5150 370W

First of all thanks for the detailed answer!!!

I checked all the capacitors on that psu and only the 2 10uF 50V BH were found bad and I replaced them with G-Luxon I am not going to waste good caps on that thing.

I checked the diodes and the non shorted 13007 and found them good.

Time for AC test

[momaka]

Re: Trust PW-5150 370W

You're welcome.

Hm, I wonder what caused all of the darkening on the 5VSB then. Probably the ceramic caps going out of spec or just the circuit was designed badly.

Time for AC test

[tom66]

Re: Trust PW-5150 370W

If ceramic caps get hot YOU'RE DOING IT WRONG!

Seriously, they have nearly zero ESR. Their ESR is as close to a metal conductor as you can get; they are just layers of metal and ceramic insulator. It's close to 0.5mohm for 1206 size devices.

More than likely it's simply the MOSFET getting hot... remember it's mostly in use during standby, when the power supply fan is off so there's limited cooling from the heatsink and a lot is dumped into the PCB.

[goodpsusearch]

Re: Trust PW-5150 370W

It works

More than likely it's simply the MOSFET getting hot... remember it's mostly in use during standby, when the power supply fan is off so there's limited cooling from the heatsink and a lot is dumped into the PCB.

You are right. The NPN transistor used for 5vsb is BU3150, with collector current 3A continuous at 25C, not so good. And the tiny primary heatsink doesn't help.

You're welcome.

Hm, I wonder what caused all of the darkening on the 5VSB then. Probably the ceramic caps going out of spec or just the circuit was designed badly.

It all makes sense now that I powered that thing. The fan controller sensor is attached to the secondary heatsink that stays fairly cool, while the primary heatsink gets too hot to touch even without any load! The fan nearly moves any air...


I am going to mod the fan-controller. Then it is going to be a quite decent 230W-250W psu.

[Behemot]

Re: Trust PW-5150 370W

You may wanna try to swap the 13007's for some more decent transistors. They will likely produce less heat too.

[momaka]

Re: Trust PW-5150 370W

It works

Nice! . I like it when kludge repairs unexpectedly work.

I am going to mod the fan-controller. Then it is going to be a quite decent 230W-250W psu.

Hm. I don't mean to be negative, but I don't think you'll make it even to 200W with that insulated-tab TO-220 E13007F transistor - not for too long anyways. More like 100W to 150W max.

[goodpsusearch]

Re: Trust PW-5150 370W

Hm. I don't mean to be negative, but I don't think you'll make it even to 200W with that insulated-tab TO-220 E13007F transistor - not for too long anyways. More like 100W to 150W max.

I put Arctic Silver 5 between the E13007F and the heatsink to improve contact.

But the E13007 needs to be insulated with insulation tape, and this has also an impact in the heat transfer from the E13007 to the heatsink. This makes me believe that E13007F is not that much worse. Please correct me if my thought is wrong.

Another good thing is that I took the E13007F from a dead Enlight 250W and those Enlight psus seemed honest about their ratings.

About the fan mod, I put a 5.6Kohm in parallel with the thermistor and the fan gets 10V-12V instead of 6V-7V. It is not a silent power supply anymore, but the air is still warm, due to the primary that gets hot.

Overall, I don't think it will do 250W. The primary heatsink gets too hot, even with psu off and 3W load at 5vsb......

I would replace the primary caps with 680uF and the transistors with E13009 in TO-3P package, maybe replace the primary heatsink as well and do a full recap on this thing, but it is not worth it when I have some Jou Jye 400Watt psus waiting for recap.

Thank you all for your help

[momaka]

Re: Trust PW-5150 370W

I put Arctic Silver 5 between the E13007F and the heatsink to improve contact.

But the E13007 needs to be insulated with insulation tape, and this has also an impact in the heat transfer from the E13007 to the heatsink. This makes me believe that E13007F is not that much worse. Please correct me if my thought is wrong.

Another good thing is that I took the E13007F from a dead Enlight 250W and those Enlight psus seemed honest about their ratings.

You may be right then.

Overall, I don't think it will do 250W. The primary heatsink gets too hot, even with psu off and 3W load at 5vsb......

IMO, something is definitely wrong with the 5VSB then. To get the primary heat sink that hot, the 5VSB must be dissipating at least 3-5 watts. So it's probably doing less than 50% efficiency!
I know 5VSB circuits aren't generally too efficient to begin with, but I think this is just crap!
I wouldn't put it in a computer (that I cared about) if I were you.

By the way, why not pull parts from this PSU instead and fix the Enlight PSUs?

[goodpsusearch]

Re: Trust PW-5150 370W

IMO, something is definitely wrong with the 5VSB then. To get the primary heat sink that hot, the 5VSB must be dissipating at least 3-5 watts. So it's probably doing less than 50% efficiency!
I know 5VSB circuits aren't generally too efficient to begin with, but I think this is just crap!

Is is usual for me to see power supplies that get too hot when off. That's why I advice all my friends and people I know to turn off theis computer completely.

On that psu they should have used a better 5vsb transistor and better heatsink..

I wouldn't put it in a computer (that I cared about) if I were you.

A friend gave me an Athlon 64, gforce 7300 with a dead K-Mex and he doesn't want to pay for a decent recapped psu. Guess what I am giving him for free

By the way, why not pull parts from this PSU instead and fix the Enlight PSUs?

I have 5 or 6 Enlight psus. 2 of them are used for parts, the other ones are keepers. They are decent power supplies but they are clearly designed for 5V PCs: 10A rectifier on 12V, 1A label rating for 5vsb, they even miss the ATX 12V connector.


Due to the financial crisis here in Greece no one is interested in good power supplies anymore. When their old psu fails they just buy the cheaper they find. Then they buy another one...

[ben7]

Re: Trust PW-5150 370W

IMO, something is definitely wrong with the 5VSB then. To get the primary heat sink that hot, the 5VSB must be dissipating at least 3-5 watts. So it's probably doing less than 50% efficiency!
I know 5VSB circuits aren't generally too efficient to begin with, but I think this is just crap!

Yes, those two transistor designs aren't that efficient.

It would have to be pumping out at least 10W of heat to make that heatsink burning hot... with only a 3 watt load.

[momaka]

Re: Trust PW-5150 370W

Is is usual for me to see power supplies that get too hot when off. That's why I advice all my friends and people I know to turn off theis computer completely.

Most 2-transistor 5VSB PSUs do run hot indeed, though not THAT hot. The IC-based ones run fairly cool, however. And they have very decent efficiency. I have only a few of these so I leave them on. The 2-transistor ones I turn off from the back when I'm not using.

On that psu they should have used a better 5vsb transistor and better heatsink..

It wouldn't have helped much. If the circuit is still as inefficient as it is now, there would still be that much heat to dissipate regardless. The only difference is that the heat wouldn't be as "concentrated" on the components and would be dissipated better with the bigger heat sink. That might save the PCB from burning, but it still doesn't fix the original problem.

I actually have an old Morex 235W PSU that has its 5VSB transistor mounted on a tiny heat sink - similar to the Enlight PSU you showed above. With a 1.1A load (PSU is rated for 1A on 5VSB), it runs hot - but not burning-hot. And that 2-transistor 5VSB doesn't have any caps on the primary either. So apparently it was just designed well.

A friend gave me an Athlon 64, gforce 7300 with a dead K-Mex and he doesn't want to pay for a decent recapped psu. Guess what I am giving him for free

Well, in that case, what you're giving him looks A LOT better than what he had .
I think I will stop with my complaints about the 5VSB of that PSU now .

It would have to be pumping out at least 10W of heat to make that heatsink burning hot... with only a 3 watt load.

Probably true. I didn't want to sound too pessimistic, though .

[spleenharvester]

Re: Trust PW-5150 370W

Hehe, old thread I know, but I have one of these that's 8 years old and still running perfectly. It's been on an 8800GT with E2140 for some time now. Just goes to show the odd crap brand box can surprise you!

[Pentium4]

Re: Trust PW-5150 370W

Hehe, old thread I know, but I have one of these that's 8 years old and still running perfectly. It's been on an 8800GT with E2140 for some time now. Just goes to show the odd crap brand box can surprise you!

If you haven't replaced the caps yet, you should! I think this thing has potential, usually if a PSU has a 35 transformer, it's worth upgrading. Bigger input caps, 13009's, 20A schottky's in parallel for the 12V and you've got a nice *inefficient* 350W PSU Although the main toroid is a little on the small side