Power supply build quality pictorial. part 2

[Behemot]

Re: Power supply build quality pictorial. part 2

As there is not much going on here lately, I'll link my last review if that is not a problem? It's in czech, but I think you can get the images and silicon part numbers I am kinda lazy to re-write it all here

It's Evolve, local brand with Guangzhou AOJIE Technologies as OEM (SAMA). It has 80 PLUS label, but has not been certified under the Evolve brand, though the SAMA PSU's have been certified. http://diit.cz/clanek/evolve-pulse-8...cze-p-500-plus

[Pentium4]

Re: Power supply build quality pictorial. part 2

Not sure if I've already posted this one but found it recently... Paid $7 for it but it isn't fully working. It gets 5VSB but will not power on. I'm thinking maybe the start up cap is bad? All the caps consist of Canicon and Cheng I don't think it could do its rated wattage though judging by the primary heatsink, the complete lack of PFC, small transformer...Those input caps though are 1200uF

[Behemot]

Re: Power supply build quality pictorial. part 2

What silicon is in there? Smaller heatsink on primary does not neccessarily degrade it, depends on silicon. Even average MOSFETs can work at temperatures over 100 °C with full power. What topology and driver does it use?

The transformer seems the same as FSP uses, it may do maybe 500 W with some efficiency?

[Pentium4]

Re: Power supply build quality pictorial. part 2

They're better than I thought....it has two 18N50 FET's. Do you think it's most likely the start up cap causing it not to power on? I can't find any shorts

[Behemot]

Re: Power supply build quality pictorial. part 2

And the driver/topology?

Well, you may want to try

[Pentium4]

Re: Power supply build quality pictorial. part 2

I'm not sure but I think it's single transistor forward. The PWM controller is a PS222S

[Behemot]

Re: Power supply build quality pictorial. part 2

Seems more like two-transistor forward to me, I can see two transistors now, it would impossible to achieve such power with single transistor only.

PS222S is secondary monitor, primary is something else. Maybe it's on the other side, in SMD package, it is very common lately.

ADD// I've checked and this platform is manufactured by ATNG power, very good manufacturer so definitelly it should be able to provide 600 W.

[ben7]

Re: Power supply build quality pictorial. part 2

OMG I think I see a VIPer22A in there driving the 5vsb!

[Behemot]

Re: Power supply build quality pictorial. part 2

How can you tell? You know there is not only VIPer on market, there are hundreds such chips now…so far I've seen something else in almost every PSU I've reviewed so far Every time I learn one new

[ben7]

Re: Power supply build quality pictorial. part 2

How can you tell? You know there is not only VIPer on market, there are hundreds such chips now…so far I've seen something else in almost every PSU I've reviewed so far Every time I learn one new

Uh, dude, in the picture....
It is even made by STmicro!

[Behemot]

Re: Power supply build quality pictorial. part 2

Oh, there is something what MAY be 22A, was not readable until I zoomed in

[Pentium4]

Re: Power supply build quality pictorial. part 2

Indeed it is a Viper22A ATNG makes this? nice. I think there might be a toasted little chip on the bottom of the PCB I will try and upload a picture soon!

[DJduck]

Re: Power supply build quality pictorial. part 2

Delta DPS-460DB A from my main pc. Rated at 460w but will probably do more.

The label:


Inside:


PCB taken out of the chassis:


Very nicely built, very heavy. The only disappointment is that there are a lot of Ltec brand caps. They are currently holding up fine and will soon be replaced. There are two small Rubycons in there and the input cap is a Samxon.

As you can see it says 12V a 12V b and 12V c on the label. It does actually have 3 12V rails, both sides of the heatsinks are populated, but since there are a lot of components in the way i can't see the model/part numbers on there.

Also has active PFC which is very nice. Heatsinks are thick and massive.

I got A LOT more psus upstairs, some of them need work. I am going to repair and recap them and post some pics here.

[Wester547]

Re: Power supply build quality pictorial. part 2

I have a random question, sort of concerning this thread...

I noticed an extra coil on the right end (in this view) of the main transformer in older Dell/Newton Power 200W and 250W PSUS... for example, in one of the pictures of a 250W Newton tom66 posted:



Is that for input filtering? I've never seen a ferrite coil for input or transient filtering in that location in most PSUs.

[Behemot]

Re: Power supply build quality pictorial. part 2

What was the resply No. ?

[DJduck]

Re: Power supply build quality pictorial. part 2

Astec as202-3556-2329




250w.
Filtering consists of 4 x and 4 y caps and an MOV, there is another small coil, might me a second MOV? Also has PFC. For 250w, heatsinks are pretty big, primary caps are 200v 470uf Korea Chemi-cons. Secondary and miscellaneous caps are Rubycon and Nichincon. Looks very well made and heavy, probably going into my P4 machine 14A on 3.3v, 20A on 5v, 10A on 12v, 0.5A on -12v and 2A on 5vSB.

[mariushm]

Re: Power supply build quality pictorial. part 2

It's still a power supply, albeit a linear one...

Dave reviews a $450 programmable linear power supply:

http://www.youtube.com/watch?feature...&v=Ju9Q05y_3V0

The only good thing I see there is

* the beefy transformer and
* the row of optocouplers and
* the idea to use a microcontroller (with pretty good specs) on each channel for ADC/DAC (but no safety checks in case communication between controllers fails for some reason)...
* and the bonus of using thermal trips to disconnect the channels in case of heatsinks overheating (but shutoff is still done by software by disabling the input relay for both channels)

Other than that...

* more than likely fake UCC KMH capacitors on each primary channel (10000uF/63v, poor quality printing)
* Aishi and some other no-name capacitors and
* horrible soldering (partially manual)
* flux remains all over the boards
* shitty 24v fan that may have a bad bearing or is improperly pwm'ed (in first part you can hear it and it's making that particular noise)

I'm not even sure the ftdi chip for serial to output is genuine... looks sketchy to me.

[Pentium4]

Re: Power supply build quality pictorial. part 2

not a bad unit in my opinion...nice heatsinks, good transformer, and i'm pretty sure it has dual 12V rails instead of a beefy single rail. Fuhjyyu and Gold Link caps...Worth keeping after a recap. I think it could do around 400W in a cold room

[cheapie]

Re: Power supply build quality pictorial. part 2

Here's my Astec ATX202-3545 from 1999 that I've been bragging about for a while now. The first thing that I noticed when I opened it was the Nidec fan. That's always a good start (unless it's siezed, which this one isn't). It also has 100% Nichicon caps - LQ on the primary, I forgot what's on the secondary. The PWM controller and supervisory chip are both Astec-branded. This power supply also, interestingly enough, has a fully populated line filter. Now, of course, a power supply this old isn't going to have an ATX12V connector, or any SATA connectors, so I chopped some off of an old dead power supply, and spliced them onto this one. Right now, it's powering my Router/Web Server/Main Computer (This one). I'm sure I'm really pushing it on the +12... Also, the only problem I've ever had with this thing was when the solder joints for the fan's power broke. I simply spliced the fan into the +5V line.

(Sorry about the terrible pictures - they were taken with a $10 camera)

Anyway, I'm no expert on this (yet), but I think it could probably do its rated wattage (200W).

[Behemot]

Re: Power supply build quality pictorial. part 2

Yeah these are good, but the low power degrades it to almost useless these days.

not a bad unit in my opinion...nice heatsinks, good transformer, and i'm pretty sure it has dual 12V rails instead of a beefy single rail. Fuhjyyu and Gold Link caps...Worth keeping after a recap. I think it could do around 400W in a cold room

Depends, what's a rail for you? Usually rail is considered when it has shunts for separate OPP. Sometimes coils are used as shunts (not as precise I think, but seems to work).