NOX Krypton 800W - Higher voltages, need recap?

[goodpsusearch]

Re: NOX Krypton 800W - Higher voltages, need recap?

Lazy mode doesn't work with projects like this. You have to make a list of the original capacitors, volts, uF brand, series and diameter and then find the suitable replacement caps.

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Hmm... Making a list then, searching through the datasheets.
Btw, that website was really useful.

I've already bought the replacements, but at this point it seems some won't get replaced.
I'm making a list of the old -> new capacitors to ask to you all if the thing can be done (with the ones I've ordered. obviously new ones will be ordered again).

In the meanwhile, I've started with the most important (IMHO) capacitor.
Old -> JPCE-TUL 16V 3300uF - Ripple: 2460 - ESR: 0.023
New -> Panasonic EEUFR1E332B - Ripple: 3300 - ESR: 0.017

The ripple is higher and the ESR is slightly lower.... are those values okay or have I got even less tolerance on the choice?

[goodpsusearch]

Re: NOX Krypton 800W - Higher voltages, need recap?

New -> Panasonic EEUFR1E332B - Ripple: 3300 - ESR: 0.017

Why 25V?????

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Because I really can't find anything that's from a trusted brand and a trusted series that is rated at 16V.
That's the only good one I could find.... and then I've always read that the voltage on there is just a limit, so it's only about wasting money if I take an higher-rated capacitor.
Isn't that true?


P.S.: Obviously, the values are expressed in mA r.m.s. for Ripple and milliOhm for ESR.

[goodpsusearch]

Re: NOX Krypton 800W - Higher voltages, need recap?

The esr is a bit too low and the dimensions of the cap would make it difficult to fit there.

http://gr.rsdelivers.com/product/pan...0/7083686.aspx

I use those for psus without any problem.

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Ah yeah, RS Components! Wants 20 euros for shipping here. I'm buying the caps from Conrad, free shipping!

So the ESR is too low (dimensions aren't a problem, it can be fitted, trust me)?! :| Crap... coming with the other caps.

[goodpsusearch]

Re: NOX Krypton 800W - Higher voltages, need recap?

Did you try http://www.farnell.com ?

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Yeah, it was the first one I've tried. Here in Italy they only ship to business... so I can't get things from there :\



EDIT:
Okay, well... it's time to open up the PSU - AGAIN. I'm about to better examine the caps, as I wrote only the values, without specifying brand and model.
I'll be away for a couple of minutes/hours... or I may get back with my phone. Who knows.


EDIT2:
OOOOOUUUCH! I found the bad guy!!
There is an "hidden" (under a wires forest) capacitor, by "su'scon" (WTF?), 3300uf 16v. It's leaking!!

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Sorry, I've had to run and had no time to write here.

@goodpsusearch
Yup, absolutely. It's the same thing


Anyway. That Su'Scon capacitor that was leaking (just a little though) is really strange...
I mean, it's practically the only really-low-ESR capacitor in a forest of same capacitance and voltage with higher ESR. I really wonder why.

Apparently, it looks like being an higher quality thing compared to the other ones but, well, as I said, differently from the others, that one was bad when I found it... and comparing the datasheets, one of the Panasonic FR (3300uF 24V, physically bigger) has an higher ripple current but a really similar ESR (lower by 0.001).
I think it works just okay... so I replaced only this cap.


Obviously, nothing REALLY changed... the voltages are still the same, apart the 5V that now is lower (5.16 instead of 5.20), but still fluctuates.
Oh, it's normal, I didn't expect any big change in the situation.

As for the other ones, I don't know how much I'll have to wait. Sincerely, now I'm somehow undecided whether to replace 'em or buy a new PSU, just because my soldering iron died right after replacing that cap.

Tomorrow I'll take and upload a photo of the bad cap I replaced (don't worry I can resize now lol), also because it didn't explode. It's like it did IMPLODE instead O_o

[Behemot]

Re: NOX Krypton 800W - Higher voltages, need recap?

I have plenty of Samxon RS 3300 uF/16 V D10x30 (nothing better exists in D10x30 mm ATM) and some samples of NCC KYA 3300 uF/6,3 V (more being manufactured ATM).

The primary cap is Chemi-Con, the polymer is also Chemi-Con, there is no reason to expect they are bad. If you want to be sure and will be doing more with capacitors in future, I can send you the newest version ESR Micro v4.0s with the caps with some discount…

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Sorry, I've had absolutely no time to post the photo I promised, so I'm posting it now.

At the end of all, I decided to buy a new soldering iron: got a discount, it's still something chinese, but 60W at 10 euros is a good price IMHO!
I also bought the new capacitors that now are compliant to the old ones' ESR, just higher ripple, but from what I've read the ripple can be higher no problem.

Today I should get the new caps, but I don't know if I'll replace 'em today or tomorrow. I'll anyway shoot a photo after replacing the caps


@Behemot
Well as you can see I've already bought 'em, but thanks anyway for the offer!

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Can't edit the previous post... so I'm writing a new one.
I've "successfully" changed almost all of the caps, apart one 68uF that is really impossible to replace unless I desolder practically everything - and - the 400V 390uF cap (that is by nippon chemicon, so, well, as already said, useless to replace).

Fact is.... I verified the PSU health after the operation and the voltages were obviously high (no load on the PSU), but this time not REALLY high like when I verified with the old caps.
With the old caps it was like 3.6V on the 3.3V rail without load (without even motherboard I mean). Now it was showing 3.4, but perhaps it's just a coincidence and I shouldn't really think about that.

I've reattached everything because the voltage looked like it was good, so I booted up the system and I've opened the voltage monitoring in the BIOS, and the voltages are EXACTLY like they were PRIOR to replace the caps.

Obviously people says you can't trust the software. I say I can because my motherboard is really -- really -- good at this, but I anyway verified that with my multimeter (both on mainboard probe points and directly on PSU) and AGAIN the BIOS was right.


+12V: 12.288 --- +5V: 5.160-5.200 (fluctuating) --- +3.3V: 3.440
SAME voltages.

Any idea?

[Behemot]

Re: NOX Krypton 800W - Higher voltages, need recap?

So you did everything on the basis of BIOS voltage monitoring? You serious? It's there mostly for, well, don't even know, but it's useless. If you want voltages, grab multimeter and measure on Main ATX connector.

[goodpsusearch]

Re: NOX Krypton 800W - Higher voltages, need recap?

I have to agree with Behemot on this.

I don't get it actually. Do you want to fix the psu or you'd rather sabotage yourself and then say "see? I told you so!"

If you really want to get things done you should learn to follow some general principles. One of them is that the bios voltages are inaccurate and can't be used for psu troubleshooting and repairing.

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Hey. Have a better read at my post.

I said that I've measured with BOTH BIOS and multimeter and the BIOS readings were exactly the same of my MULTIMETER!

[Per Hansson]

Re: NOX Krypton 800W - Higher voltages, need recap?

Behemot: Please read the members post before bashing him.

kholk: Well your job was not in vain, you have found a bulging/leaking Su'scon capacitor and fixed that, obviously the PSU is much better after this.

In some PSU's there are potentiometers you can adjust the voltage with.
But 3.3v in particular often has a feedback cable.
Look on the ATX cable, does two cables sit on the same 3.3v pin in one of the ATX contacts?
If so follow that back to the PSU and make sure it is soldered in properly.
And that you can read a voltage there.

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

All of the 3.3V cables are apparently on the same rail... it's useless to say that there are two cables that are grouped per pin, because on the rear they're soldered on the same rail anyway.

The cables... yeah, they're soldered in properly, if I wanted I could easily have lifted the whole PSU taking only the 3.3V cables in my hands... even only one lol.
The voltage can be read on all of the 3.3V pins on the ATX connector AND on the CPU 4+4 connector.

Btw, even if that su'scon capacitor was fine, I would have anyway prolonged the life of the PSU (since it's old)... so well, I did no really useless job

[Behemot]

Re: NOX Krypton 800W - Higher voltages, need recap?

My bad.

As for +3,3, no rail, there probably has never been a PSU with multiple +3,3V rails. There is just feedback wire from one +3,3V pin to control the voltage and boost it if neccessary. There always is.

[Per Hansson]

Re: NOX Krypton 800W - Higher voltages, need recap?

All of the 3.3V cables are apparently on the same rail... it's useless to say that there are two cables that are grouped per pin, because on the rear they're soldered on the same rail anyway.

That's not what I meant, I said on the ATX connector side.
There should be a feedback wire there going back to the PSU, so it can compensate for the voltage drop that occurs in the cables etc.

[kholk]

Re: NOX Krypton 800W - Higher voltages, need recap?

Yup, there's a sense wire.
In my case it's brown, and.. yup, it's properly soldered on both PCB and connector.