Hipro D2537F3H - Teapo output capacitor replacemnt

[socketa]

Thought that it would be worth rescuing this one, but it's got one bulged Teapo that is in front of the big coil.
Would it be perfectly OK to replace this 4700uf SC (which currently measures 600uf and 1.2ohm ESR) 5V output cap with XYJ 3300uf? More specifically, was the original 4700uf Teapo cap overkill?

i also (inline) tested the 3300uf SC (which is parallel with a 2200uf SC cap, that is right next to it) on the 12V rail, and get a measurement of 5500uf and around 0.1 ohms ESR - So those caps look like they are still healthy?

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

teapo are pretty good if kept cool, i wouldnt drop to 3300uf unless you have no choice.
what diameter is it?

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

13mm

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

no problem finding a 4700uf panasonic or rubycon then.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

of course - but not so possible if ebay is one's only pragmatic option

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

hmmm.....

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

how about a 6.3V (original was 10V) ZL, or is that too low ESR for this PSU?

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

6.3v on the 5v rail is too close really.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

Ok.
How about these Lelon RZW's? They aren't as low ESR as ZL's and they are a lot closer to PW's, but the datasheet doesn't list for these 30mm (only 35mm)
https://www.ebay.com.au/itm/5pcs-470...MAAOSwlndZHn5~

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

if height isnt a problem then try them.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

Done. Thanks bud

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

Now the 3.3 rail
The tester shows 5822uf on the 3.3v rail, but i can only find 2 X 2200uf Teapo caps that are connected to that rail, so where could the extra 1422uf be coming from?

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

leakage - the caps are bad

[momaka]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

I suppose I'll chime in with my 2-cents worth of PSU recapping.
In particular, I think it's better to go with the Japanese caps, even if they have slightly lower capacitance, than with the Lelon caps. I don't trust Lelon caps much. And L-tec? If you see them in these HiPro PSUs, get rid of them ASAP. They tend to have nearly 100% failure rate. Considering they do work for at least 5 years in these PSUs, they are not terrible. But nevertheless, I like to recap my PSUs only once and then not worry I have to touch them again.

Would it be perfectly OK to replace this 4700uf SC (which currently measures 600uf and 1.2ohm ESR) 5V output cap with XYJ 3300uf?

Should be OK.
The lower capacitance would just increase the ripple and noise when PSU is near max load on the 5V rail - not a big deal if you are going to using this PSU on a 12V-heavy PC (just about any modern PC.)

More specifically, was the original 4700uf Teapo cap overkill?

A little.
If I remember correctly, Johnny Guru did a "basement round-up" review a while back, and one of these PSUs, even with some bad output caps, was still outputting all power in spec - just barely, but in spec, nonetheless.

6.3v on the 5v rail is too close really.

Nah, not really.
I've done it plenty of times. 6.3V caps on the 5V rail are fine.
Only when you use cheap cap brands it can be a problem, because they rarely can stand the ripple current close to their rated voltage.

how about a 6.3V (original was 10V) ZL, or is that too low ESR for this PSU?

With these HiPro PSUs, I've gone as low as Rubycon ZLH on the 5V rail and Nichicon HZ (the lowest ESR e-lytic cap out there) on the 3.3V rail. Both recaps were okay. These PSUs are happy, as long as you keep the capacitance reasonably close to the original or no less than 1-2 notches below.
And for the 5VSB rail, Chemicon KZE or Panasonic FM/FR/FS -grade caps are not only fine, but recommended. The 300W and 350W versions based on this platform sometimes have a 16V, 2200 uF KZE on the 5VSB.

Now the 3.3 rail
The tester shows 5822uf on the 3.3v rail, but i can only find 2 X 2200uf Teapo caps that are connected to that rail, so where could the extra 1422uf be coming from?

I can't remember off top of my head what arrangement this PSU has for the 3.3V rail, but I think it was 1x 2200 uF on the 3.3V rail after the rectifier (but before the MOSFET for the linear regulator) and 1x 2200 uF + 1x 1000 uF on the output of the 3.3V rail with a coil in between them. Need to double-check, though.

Either way, I don't recommend you check caps in circuit, as you can sometimes get erroneous readings, especially if there are ceramic caps in parallel with the electrolytics.

Also, I think this PSU has a 100 or 120 Ohm resistor on the 12V rail (usually located right next to the big toroid coil) that tends to cook the first 3.3V rail cap. So it may not be a bad idea to remove it. It's just a minimum load resistor for the 12V rail, so it's not really needed. If you're worried about making modifications like that, you can replace it with a 270-Ohm or higher resistor. I usually put 2x or 3x 1000-Ohm SMD resistors in parallel on the bottom of the PCB on the 12V rail for a total parallel resistance of 500 to 333 Ohms. Haven't had any issues with running that setup at all.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

Thanks for that good info
The caps measure 2480uf and 2676uf, (and 5268uf combined) out of circuit.
Have got a lot of 5V boards (not to mention older PSUs), so will likely stick wth the Lelon (after testing it)
Am assuming that that load resistor is the big beige-coloured one, that is mounted vertically (also seems to have lost a bit of it's casing).
Looks like it's 27 ohms (and i get 27 ohms, measuring across it in-circuit) - which, after reading your comment, sounds far from ideal.
There is 200 ohms resistance between the 12V cable plugs, and ground (so am assuning that somehow more resistance gets added to the rail after the 27ohm load resistor)

What type of resistor (non-SMD) should i look for, and what rating. The 5v rail is rated for 12V @ 14A; so is that any resistor greater than 270 ohms that can tolerate greater than 168W?

[stj]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

get rid of any more of that brown glue.
it turns conductive.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

started removing that glue and some of the insulation came of the leg of a coil on the primary side - what's a good substance to recoat it with.
I tested the glue on Mohms range and dont get any reading - so looks like it's not conductive yet.

Just realized that i was talking nonsense regarding the wattage for the minimum load resistor.
270ohm resistor on SUPPLY to the 12V rail. (i.e, why do i get 200 ohms on the actual 12V rail, and not 47ohms)
I=V/R=12/270=0.45A
P=IV=.45 X 12 = 5.4W

but since all of the resistors that i can find that look the original are 5W or less...

for a 500ohm resistor 12/500=0.24A
0.24 X 12 = 2.88W

So would it be a Metal oxide film Resistor, or a wire wound resistor flat wire heat dissipation power resistor, or either?

And how does the load resistor, that is on the supply to the 12V rail, cook the first cap in the 3.3V rail? Is the first part of the 3.3V rail in series with that load resistor?

[momaka]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

get rid of any more of that brown glue.
it turns conductive.

+1
And even when it doesn't, it can eat through PCB tracks and corrode terminals and leads.
I really hate this brown crap glue, though sometimes, some of it is less aggressive.

started removing that glue and some of the insulation came of the leg of a coil on the primary side - what's a good substance to recoat it with.

Any kind of varnish or lacquer should do. I often use nail polish (just don't use anything with metallic paint in it.) Hot glue should work in a pinch too (or any other glue that is not conductive, corrosive, and won't get damaged from heat and heat-cycling.)

Am assuming that that load resistor is the big beige-coloured one, that is mounted vertically (also seems to have lost a bit of it's casing).
Looks like it's 27 ohms (and i get 27 ohms, measuring across it in-circuit) - which, after reading your comment, sounds far from ideal.

Yeah, that's not the resistor I'm taking about.

I think I got confused with the part numbers of your HiPro, as I have a slightly different 250W version. Could you post a few more pictures of the PSU? Also maybe a PCB pic? That should help us a bit.

Looking at my own pictures, it's the 300W and 350W models that have the resistor by the toroid inductor, but it's not at the edge of the PCB - it's near the center of the PCB, next to the toroid. Your PSU looks a bit different though, so it might not be there.

With that said, don't remove that 27-Ohm resistor yet, as we haven't determine what it's for.

There is 200 ohms resistance between the 12V cable plugs, and ground (so am assuning that somehow more resistance gets added to the rail after the 27ohm load resistor)

No, shouldn't be that way.
You probably have either 220 or 270-Ohm resistor on the 12V rail, which should be OK then. Reason for the lower resistance reading is because of the capacitors in the circuit (which, when they are large, will take some time to charge up, thus lowering the overall resistance) as well as any other logic circuits connected to the 12V rail.

Might want to check the 5V rail resistance to ground as well. Anything less than what would seem to be a 47-Ohm resistor, and I'd suggest you increase that a bit (especially if you plan to use the PSU with a 5V-heavy PC, so you won't really need much of minimum load on the 5V then, as the PC will provide it.) Though, I don't think any of the HiPro PSUs I know to have a problem with the 5V rail load resistors. It's usually FSP and Deer that do.

What type of resistor (non-SMD) should i look for, and what rating. The 5v rail is rated for 12V @ 14A; so is that any resistor greater than 270 ohms that can tolerate greater than 168W?

You don't need to use SMD resistors. Anything will do. The power rating of the resistor will depend on the voltage rail that it going on. The current ratings of the PSU don't concern us with this.

More on this...

Just realized that i was talking nonsense regarding the wattage for the minimum load resistor.
270ohm resistor on SUPPLY to the 12V rail. (i.e, why do i get 200 ohms on the actual 12V rail, and not 47ohms)
I=V/R=12/270=0.45A
P=IV=.45 X 12 = 5.4W

Your math is a little off there.

I = V/R = 12 / 270 = 0.0444 Amps
P = I * V = 0.0444 * 12 = 0.533 Watts
or you can simply do:
P = (V^2) / R = (12^2) / 270 = 144 / 270 = 0.533 Watts

So with that power, if there is indeed a 270-Ohm resistor, it should ideally be rated for at least roughly twice that... or about 1 to 2 Watts, just so it doesn't run very hot. Also, when doing minimum load resistors, it's usually a good idea to keep dissipation of each resistor at less than 0.5 Watts if possible. This will slightly increase the PSU efficiency at low load and also make it produce less heat inside its case.

for a 500ohm resistor 12/500=0.24A
0.24 X 12 = 2.88W

Again, you're missing a zero before the 24 on your current calculation.
Should be 0.024 Amps.
or
And P = (12^2)/500 = 144 / 500 = 0.288 Watts.

So would it be a Metal oxide film Resistor, or a wire wound resistor flat wire heat dissipation power resistor, or either?

Doesn't really matter here.

And how does the load resistor, that is on the supply to the 12V rail, cook the first cap in the 3.3V rail? Is the first part of the 3.3V rail in series with that load resistor?

That might not apply to your PSU. I will have to see more pictures to determine. On the latter 300W and 350W models, it does, because they have that resistor right where I stated by the inductor and next to the 3.3V rail input cap.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

Your'e right! There is 210 ohm directly across the 12V output.
i was fixated with looking for a large resistor, which was reinforced because of the wrong high wattage that i had wrongly calculated.
Looks like the top of it has suffered from some heating, and subsequent decay
The designer must have desiginated the resistor there for some reason - what are the consequences of removing it and not replacing it at all?

Might want to check the 5V rail resistance to ground as well

The 5V rail resistance to ground is 206 ohms, (and the 3.3V rail resistance is 520 ohms, and 12V is 209 ohms)

So what's that larger 27 ohm resistor, that's on the other side of the torodial coil, for? One side is connecetd to ground and the other side goes through the brown ceramic cap that is next to it, and then goes underneath the large torodial coil base? Looks like that could do with being replaced.

[socketa]

Re: Hipro D2537F3H - Teapo output capacitor replacemnt

(was suspicious of 520 ohms on 3.3V rail, and just realized that makes a difference which way the test leads are, so black to ground = 285 ohms. Is there some diode being added to that when the meter leads are swapped? )
And is this the reason why the physical size of the 220 ohm (load?) resistor, that is on the 5V rail, is so small compared to the 220 ohm load resistor that is on the 12V rail?:
P=25/220=0.11W
P=144/220=0.65W



Getting back from that tangent, to the previous question... why does that 220 ohm load resistor, that is on the 12V rail, cook the first cap on the 3.3V rail?