the gutless, bloated, and fried power supply hall of shame

[Dan81]

Originally Posted by ruky con

I feel like something is missing from this power supply.

It claims it's maximum output is 4.5-9.5v 800mA.

Main cap?

[ruky con]

Originally Posted by Dan81

Main cap?

Yup that's exactly what's missing.

[PeteS in CA]

Looks like its "beef" came from Taco Bell. (Rimshot! Crash!)

[momaka]

Originally Posted by goodpsusearch

They're here!

Force SL-8600EPS (MAX 600W)

The only notable difference it has from this unit..
..is the fan, that is different. Here, we have a sleeve bearing Globe fan that seized and cooked the output inductor toroid coil.

LOL.
By the way, do you still fix/save the fans after they have seized?
I certainly do. And the only reason why is because it looks like sleeve bearing fans are becoming a thing of the past now. Many manufacturers have started switching to these shitty "long life" bearing fans - i.e. *sealed* sleeve bearing fans. I HATE THOSE THINGS!

At least the cheap (non-sealed) sleeve bearing fans are serviceable and actually last quite a while with a proper cleaning and good lubricant. My oldest sleeve bearing fan is going on a record for almost more than 10 years now. Granted I haven't used it 24/7 at full blast, but rather 20 to 30 minutes per day (on average per its entire life). Yet, it still spins like new.

Originally Posted by goodpsusearch

Powertech ATX-450W

Oh man. Now the PSU you just posted before this one looks like a 10 Gigawatt powerstation. This one is like a cell phone charger / power adapter in a PSU case.

Oh, and even the fan doesn't inspire confidence with that label. Speaking of which, I've never seen H.X.S. brand before? Perhaps that was meant to foreshadow what would happen with the PSU if you use it - i.e. lots of thick black smoke pour out and pollute everything. .

In case anyone didn't catch what I meant above about the fan brand: I was making a reference to Hexxus from the kids movie "Ferngully".
https://www.youtube.com/watch?v=hr4knvNNgtU

Originally Posted by goodpsusearch

-Funny scissors drawings appear on PCB. They tell you what to do with this pos.

LOL
Or perhaps also meant to tell the slave children/workers assembling these PSUs which components to remove in case the PSU needs "fine tuning" to work stable with the undersized components.

Originally Posted by goodpsusearch

Hantol HPSU550

Same model as this, but different platform.

Hmm.. no fake passive PFC coil in this one? I guess the black PCB mask and case paint job ate the budget for the fake PFC.

On that note, I HATE BLACK PCBs! Sorry if that sounds racist, but I really do. Same goes for any dark color PCB, actually. I really don't understand why people think it looks neat. To me it looks like the PCB burned itself to a crisp. I could easily achieve the same by dumping a bunch of coal on an old motherboard or PSU PCB and lighting it on fire.

Seriously, this black PCB fad needs to go away like 5 years ago now.

Originally Posted by goodpsusearch

12A rectifier for 12V and 16A for 3.3V and 5V. This thing is still being sold and people are buying it.

Because people have stayed informed and learned about the new nano-cosmicon particle energy field that is only built into new electronics. It basically allows you to squeeze up to 10 times the power from any one power component and easily overclock your CPU to over 10 GHz. So those rectifiers above are good for 120 and 160 Amps respectively, because of that technology. Hence the 550 Watt power rating, which I am sure is real. You just haven't read up about it and simple don't understand.

Originally Posted by goodpsusearch

PAN STAR PSA-235
...
Any comments?

While still pretty overrated, at least it wasn't that terrible of a PSU for its time period. After all, PCs in 1998 and up until the Pentium 3 era ended were just not heavy power users. Unless you had over 3 or 4 HDDs and 2 ODDs connected to a PSU like that, there's no way you could blow one of these up under load.
... though that 10 Amp rectifier for the 5V rail is cutting it very close, and thus limiting the power on the 5V rail to 50 Watts max. I suppose that was enough back in the day for just about any single CPU PC. As far as I remember, GPU power consumption really didn't take off until the GeForce 4 TI / Radeon 9700 generation of GPUs came about.

And a dual CPU (Pentium Pro / Pentium II) setup in 1998? HA! Ain't nobody (regular) got money for that. (Back then)

[momaka]

Originally Posted by ruky con

A dead Athena Power "Hercules" Model: AP-MPS3ATX35

Yeah, I gotta say I mostly agree with you: the soldering is pretty bad and the secondary side filtering is pretty bad. But at least the unit is not that heavily overrated, other than the 12V rail. And I do see some spots for PI coils. So this *could* *potentially* *maybe* *perhaps* be turned into a decent 150-200 Watt (continuous) power supply with the right components, if one really wanted to bother with it.

Originally Posted by ruky con

The fan still works but the bearing might be getting a bit dry.

Clean and oil it before it becomes loose, and it will actually last a pretty long time. The only cheap sleeve bearing fans I can never properly fix are the ones that have too much wobble in the fan blade assembly / rotor, as that tends to eat the sleeve bearing away much faster and then they start to loose oil very fast (typically less than 1 year with regular use.)

Originally Posted by dmill89

I was able to get an oscilloscope (hurray for < $30 analogue scopes on e-bay) on this thing with the same load and the results are pretty bad. The scale is set to 50mv per division, looks like we're looking at 200mv+ (ATX spec is 120mv on the 12v rail and even that is pretty bad) spikes and a very solid 100mv+ ripple with a 110-120w 12V heavy load
...
While bad in general it does look like this PSU does handle a 5V/3.3V heavy load better than a 12V heavy one.

Probably because it only has a single small cap on the 12V rail and no PI coil. A single 2200 uF cap on the 12V rail really is the minimum for a medium load, and more for heavier load - at least for these older, slow-switching h-bridge PSUs.

Originally Posted by dmill89

I realize it isn't completely "fair" to test a partially re-capped PSU either since there may be other caps that are failed but didn't bloat

That's quite possible too.
But like you said, why bother with this PSU (other than having fun with it). Mine is also sitting in the closet, case-less and still partially recapped. Every once in a while when I need 320V DC to test an adapter, I use the voltage-doubler side of things on that crappy PSU to generate the 320V DC bus for me.

Originally Posted by ruky con

I feel like something is missing from this power supply.
https://www.badcaps.net/forum/attach...1&d=1519374106
It claims it's maximum output is 4.5-9.5v 800mA.

A hammer to the case?
Seriously, did this thing come with no input filter cap like that? What a POS! Perhaps the manufacturer only meant to say it's capable of 800 uA (microAmps)

Originally Posted by PeteS in CA

Looks like its "beef" came from Taco Bell. (Rimshot! Crash!)

You mean,
ba dum tss

[ruky con]

Rhino PSU, Model: SP-450Y 450w PC Power Supply.
I think this unit may have overheated.

Label.............

Looks cheap.

Top Side..................



Bottom Side.............................

Soldering looks bad.

Primary side..........................................


There are no common mode chokes, X-class capacitors or Y-class capacitors at all
The primary caps are 2 Cheng LS 470uf 200v.
The primary switching transistors are two KSH13009A's.
The 5VSB Mosfet is an SVF2N60F.
There is a fuse, not sure how good it is though.

Secondary side.............................................. ........

The capacitors are a mix of ChengX CD288H, GR, SF, and Asia'X TNX.
The 3.3v rail and 5v rail both use MBR3045CT diodes.
The 12v rail uses the STTH1602CT diode.
The 3.3v rail has one 3300uf 10v capacitor with no PI coil.
Both the 5v and 12v rails have one 2200uf 16v capacitor with no PI coil.

And the fan............................................... .............................

The fan is seized up and it looks like the motor got really hot, there was also a plastic sheet in front of half the fan but I removed it before taking the photo.

Overall this thing is and will always be garbage.

[Dan81]

Looks like they copied a bunch of crappy brands - Raidmax for the label, Deer for the heatsinks and possibly Linkworld for the PCB design.

[ruky con]

One odd thing i noticed is the (possibly fake/copied) UL number on the label of the Rhino PSU is identical to the UL number on the Athena Power power supply label (a unit i posted here a few months ago).

[PeteS in CA]

With those heatsinks and the 470uF input lytics that might be a credible 250W PSU. 450W? Extra crispy.

[kaboom]

Originally Posted by Dan81

Looks like they copied a bunch of crappy brands - Raidmax for the fail, Deer for the kill...

fify

Originally Posted by ruky con

Rhino PSU, Model: SP-450Y 450w PC Power Supply.
I think this unit may have overheated.
************
Overall this thing is and will always be garbage.

Those always seem to be rated twice what they can barely do.

[momaka]

Originally Posted by ruky con

The fan is seized up and it looks like the motor got really hot

But it's a Yate Loon, so it's probably still okay if you disassemble it and clean the bearing (I would certainly save it ).

Originally Posted by PeteS in CA

With those heatsinks and the 470uF input lytics that might be a credible 250W PSU.

And that's only if you ignore the massive ripple and noise that would be present on the output at that power level. With a single cap per rail, good luck trying to get more than 150 Watts out, if even that much (though the design can impact that a little... but in a cheapo PSU, I doubt it).

Originally Posted by kaboom

Those always seem to be rated twice what they can barely do.

Well said.

[ruky con]

Originally Posted by momaka

But it's a Yate Loon, so it's probably still okay if you disassemble it and clean the bearing (I would certainly save it ).

I tried doing that but the fan is in really bad shape, it also looks like it got water damaged as it's all rusted inside.

[kevin!]

The worst PSU I've had in my hands, a Deer, has burned resistors (with Viva and rulycons capacitors )

[Sparkey55]

Sticker change time> BURNT-IN Test O.K.

[kevin!]

Originally Posted by Sparkey55

Sticker change time> BURNT-IN Test O.K.

LOL, I did not see that sticker, now that you mention it and reading it is causing me to laugh out loud hahaha.
How ironic.

[Dan81]

I once had the chance to buy such a Deer.

Thank god I thought twice and got a pretty beefy K-Mex.

[kevin!]

One more:
L-LINK "500W" I imagine that L-LINK warns the buyer, if you put it at 500w, we are not responsible for the explosions caused by our product.

[PeteS in CA]

OMG! Is that beefless POS good for 200W continuous? And if it "is", which would be worse, the output ripple or the chances of transformer core saturation?

[momaka]

Originally Posted by kevin!

The worst PSU I've had in my hands, a Deer, has burned resistors (with Viva and rulycons capacitors )

Now that's a classic Deer for you!

The burned resistors are a product of a runaway single-transistor oscillator 5VSB design (no optocoupler feedback). One of them is actually used in series with the Vcc power rail for the KA7500 / DBL494 PWM chip. So the voltage there must have been in excess of 30V for that resistor to burn.

And yes, you get Viva and Rulycon caps, haha! And a lovely DOB rectifier for the 12V rail (Diodes-On-a-Bracket)

All things considered, though... I would probably trust that Deer PSU slightly more than the L-Link above. At least the Deer has a 33 mm core transformer, somewhat better heatsinks, and "a lot more" output filtering (relatively speaking, of course ) - not that this means anything with Rulycon caps doing the filtering.

And last but not least.... that CBE fan. These fans are extremely crap quality, but they have surprisingly good sleeve bearings. Once cleaned and oiled properly, they will last a long time. I used one of these as a case fan and later on as a PSU fan for a HiPro PSU that didn't have a shell/case. I haven't oiled that fan in more than 5 years, and it's still running fine (though lately it hasn't seen much use).

Originally Posted by PeteS in CA

OMG! Is that beefless POS good for 200W continuous?

I doubt it. Probably even 150 Watts would be a stretch with that 28 mm transformer, given that this is an old, slow-switching h-bridge design. And that's probably if you load all the rails somewhat equally. The 12V rail appears to have a 12 Amp rectifier, so that limits it to 144 Watts. I seriously doubt you'd be able to pull that much off of it before the output toroid releases the magic smoke. But with the crappy heatsinks of this unit, maybe the 12V rectifier will overheat and short-out to save the output toroid.

And then just imagine the ripple on the 12V rail at that power level. Your mobo will be surfing for sure - no, not surfing the internet, but those huge voltage ripple/noise waveforms from the PSU.

[kevin!]

Originally Posted by PeteS in CA

OMG! Is that beefless POS good for 200W continuous? And if it "is", which would be worse, the output ripple or the chances of transformer core saturation?

I do not think it passes more than 140 watts as Momaka says, I'm afraid to demand power haha, I do not want to explode in the face xD. And just by seeing that there are only 3 capacitors for the 3 most important lines, I can fear the worst, that PSU already in itself must deliver a bestial ripple.
The Deer even in spite of being a garbage of PSU, has better filtering in the lines, and inductors.

Originally Posted by momaka

Now that's a classic Deer for you!

The burned resistors are a product of a runaway single-transistor oscillator 5VSB design (no optocoupler feedback). One of them is actually used in series with the Vcc power rail for the KA7500 / DBL494 PWM chip. So the voltage there must have been in excess of 30V for that resistor to burn.

And yes, you get Viva and Rulycon caps, haha! And a lovely DOB rectifier for the 12V rail (Diodes-On-a-Bracket)

All things considered, though... I would probably trust that Deer PSU slightly more than the L-Link above. At least the Deer has a 33 mm core transformer, somewhat better heatsinks, and "a lot more" output filtering (relatively speaking, of course ) - not that this means anything with Rulycon caps doing the filtering.

And last but not least.... that CBE fan. These fans are extremely crap quality, but they have surprisingly good sleeve bearings. Once cleaned and oiled properly, they will last a long time. I used one of these as a case fan and later on as a PSU fan for a HiPro PSU that didn't have a shell/case. I haven't oiled that fan in more than 5 years, and it's still running fine (though lately it hasn't seen much use).


I doubt it. Probably even 150 Watts would be a stretch with that 28 mm transformer, given that this is an old, slow-switching h-bridge design. And that's probably if you load all the rails somewhat equally. The 12V rail appears to have a 12 Amp rectifier, so that limits it to 144 Watts. I seriously doubt you'd be able to pull that much off of it before the output toroid releases the magic smoke. But with the crappy heatsinks of this unit, maybe the 12V rectifier will overheat and short-out to save the output toroid.

And then just imagine the ripple on the 12V rail at that power level. Your mobo will be surfing for sure - no, not surfing the internet, but those huge voltage ripple/noise waveforms from the PSU.

Strongly agree, the Deer is better than the L-LINK. Although, none of the 2 I trust, are bombs xD.

A pity that I do not have an oscilloscope, I would love to try the ripple, to see if I can one day, I will acquire an oscilloscope, I will save the PSU for when I have the oscilloscope and I will share the results in the new post hall shame ripple.

I'll take advantage of that CBE, the fan barely has use.

Bye