Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

[bogart219]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

You already learned the wiring colors according to voltage from ATX standard?

Thank you as well sir, didn't make that connection

[Behemot]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Yeah, if you learned the colors for ATX wiring, it's actually just as easy from there. Just look where the cables of each color are soldered and trace it from there (combining visual, and if necessary, continuity tester tracing) to the rectifiers and check the caps by the way

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Yeah, if you learned the colors for ATX wiring, it's actually just as easy from there. Just look where the cables of each color are soldered and trace it from there (combining visual, and if necessary, continuity tester tracing) to the rectifiers and check the caps by the way

+1

That's how I always do it too. It's trivial, really.

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

After I have read all the info posted here related to Antec EA-500 I decided is time to open my own veteran PS that came with the Antec Sonata III case (somewhere in december 2007).
So, now it has more than 11 years since I bought it and more than 22000 Power-ON hours (based on first HDD that is still in working condition).

What I have discovered inside:
Obviously I found the -12V & +12V OST capacitors swollen and PCB area around Q901 heated.

The power supply didn't cause me any problem so far. But since I planned to increase load by adding a supplementary HDD I said that will not hurt to do some checking&cleaning. Now it seems that I will need to do a full recap.

My question for you all is:
- Do you think it is a good ideea to install two capacitors in paralel on the +12V rail instead of 3300uF/16V?
For example 2x2200uF or 3300uF+1500uF.

Do not worry about space. I intend to remove all components from FAN circuit (Q901, U901, fan connector, etc.) I will use a FAN powered from outside the PS, from MB connector. This way I will have a better control to FAN speed and eliminate one source of heat near those two capacitors.

By the way, replacement capacitors will be Panasonic FM/FR series.
Thank you!

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

This way looks the affected area...

[Behemot]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

If it is one of these older platforms with only single large cap, I usually replace it with single NCC KZN 3300/16, that one is superior enough to be a good upgrade

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

On the +12V rail, after the rectifier it is only one big capacitor OST RLX 3300uF/16V then an split to two PI coils each having an OST RLS 1000uF/16V.

I already placed an order for replacement capacitors.
For this one I found:
Panasonic EEUFM1C332L 3300uF, 16V, 12.5 x 35 mm, 12mΩ/+20°C, 3750mA

I made a quick comparation with the one recommended by you:
NCC KZN 3300uF, 16V, 12.5 x 30 mm, 18mΩ/+20°C, 3660mA

So now I am confident that one capacitor (as designed) will be enough.
Thank you!

[Behemot]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

You have wrong capacitor, I have custom D10x40 mm variant.

If there is full pi filter, than there is no reason for some major changes. I do have 1200/16 D8 caps for some time now though.

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

You have wrong capacitor, I have custom D10x40 mm variant.

I'm pretty sure the 12.5x35 mm 16/3300 Panny FM he selected is not the wrong capacitor for the 12V rail. Should work just fine indeed.

If there is full pi filter, than there is no reason for some major changes. I do have 1200/16 D8 caps for some time now though.

Now those would be real useful for the 16/1000 caps in this unit. I also have a 350W CWT PSU that uses 16/1000 for everything... but that PSU is hardly worth the upgrade to all new caps.

My question for you all is:
- Do you think it is a good ideea to install two capacitors in paralel on the +12V rail instead of 3300uF/16V?
For example 2x2200uF or 3300uF+1500uF.

Considering my temporary recap shown on page 1 was with just a single United Chemicon LXZ 16/2200 worked fine (and it hardly has specs anywhere close to the original OST RLX), I'd say you 2x2200 would certainly work. Of course, as behemoth point out... why? There's no need for re-arrangement on the output side, because the 16/3300 OST RLX is 12.5 mm diameter. So finding a proper replacement shouldn't be that hard.

Do not worry about space. I intend to remove all components from FAN circuit (Q901, U901, fan connector, etc.) I will use a FAN powered from outside the PS, from MB connector. This way I will have a better control to FAN speed and eliminate one source of heat near those two capacitors.

Just be careful that whatever external circuit you use to control the fan doesn't let the PSU overheat. Voltage-wise, running the fan on at least 5V should be its starting point (as that's fairly close to what the original fan circuit outputs for the PSU when cold - actually close to 4.2V, but anyways. )

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

*EDIT*
Sorry, double post.
But while at it, at lest let me say that I have now also recapped my EA-430 PSUs. Did a few things different on that one, but it still works great. Will post it another day. As for the EA-500... it's been powering my OCed Q6600/GTX560 rig for a few months now. Also, the weird noise I mentioned in post #16 that I thought the big cap fixed - nope, it's still there, and usually much louder when the PSU is cold than when it isn't. In fact, the noise, when PSU is cold, is almost as loud as the fan in the PC it is in (and that PC has considerably loud fans.) What are the chances that some of the 1000 uF OST RLS caps on the 12V rail or 5V and 3.3V rails are bad? Who knows! Could also be the APFC coil being a bit loose. On the EA-430, it was very lose and covered in brown conductive glue. I removed it on that PSU, cleaned the glue, and then put it back in and glued back tightly with silicone. So maybe that made the difference. But on the EA-500, I hardly see any conductive glue on the APFC coil. Ah well... will pull it out probably some sunny day in the summer when it gets too hot in my house to run the Q6600/GTX560 rig.

[Behemot]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

He was comparing with standard D12.5 variant:

NCC KZN 3300uF, 16V, 12.5 x 30 mm, 18mΩ/+20°C, 3660mA

Anyways, I carry those 1200/16 and 2200/6,3 D8 for some time now, yo don't watch me thread? https://www.badcaps.net/forum/showthread.php?t=39123

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Anyways, I carry those 1200/16 and 2200/6,3 D8 for some time now, yo don't watch me thread? https://www.badcaps.net/forum/showthread.php?t=39123

I did read through it a year or two... or three ago. Probably even bookmarked it somewhere, but forgot where and then forgot about the thread. I do know you got all the PSU caps that are hard to find, though, so I do keep you in mind if anything is needed.

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Today I succesfully replaced all electrolitic capacitors from my EA-500 PS, with one exception: Hitachi HP3 330uF/400V/85°C.

I've made a small drawing with all capacitors and the coresponding number found on PCB.
I also attached a list with the exact Panasonic Part Number that I used as replacement.

Things that deserve to be Noted:
- successfully installed a 3300uF/16V/12.5x35mm into +5VSB rail;
- the first capacitor on - 12V rail is now better than original OST RLS (Panasonic FM 10mm);
- successfully installed 10mm diameter capacitors on 12V1 and 12V2 outputs;
- 3.3V and 5V outputs have now 1500uF/10V/8mm;
- I removed from PCB: Q901, U901 and Fan connector;
- I installed first capacitors on - 12V and +12V slightly tilted in order not to be in direct contact with the big coil that can get quite warm when the PS is running.
- replaced original ADDA fan with Noctua NF-A8 PWM in order to have better control to fan speed and indication. Original fan seems to me that is spining quite slowly when the PS is working. Not to mention about hot spot around Q901.

Momaka, thanks for the advice!
Maybe I will let Noctua Fan run at full speed. I can barely hear it.

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Wow, very nice work, Doru!
I frankly wasn't sure if D10 capacitors would fit in place of the two 16V/1000u D8 caps on the 12V rail. But you proved it's possible! Also, I really like the diagrams you created. With those and the ones I posted, I think that gives quite a complete picture of how to recap this PSU.

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Also, the weird noise I mentioned in post #16 that I thought the big cap fixed - nope, it's still there, and usually much louder when the PSU is cold than when it isn't. In fact, the noise, when PSU is cold, is almost as loud as the fan in the PC it is in (and that PC has considerably loud fans.) What are the chances that some of the 1000 uF OST RLS caps on the 12V rail or 5V and 3.3V rails are bad? Who knows! Could also be the APFC coil being a bit loose. On the EA-430, it was very lose and covered in brown conductive glue. I removed it on that PSU, cleaned the glue, and then put it back in and glued back tightly with silicone. So maybe that made the difference. But on the EA-500, I hardly see any conductive glue on the APFC coil. Ah well... will pull it out probably some sunny day in the summer when it gets too hot in my house to run the Q6600/GTX560 rig.

It is quite weird with that noise that you described.

What I can say about the output capacitors removed is that they are quite ok regarding their values (after 22000 hours, on my PS):
C406 = 1028 uF (+3.3Vout)
C403 = 1025 uF (+5.0Vout)
C504 = 909 uF (+12V1out)
C503 = 916 uF (+12V2out)
So I will not bet that the weird sound is due to those.
Anyway I intend to test all of them more proprely when my multitester with Atmega328 will arrive.

Regarding the APFC coil, you are refering to the big one? Because I saw a smaller one with two types of wires near that area (see attached image).
Is there any chance that we can find a detailed electronic schematic on this power supply or we have to do it by ourselves?

[momaka]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

It is quite weird with that noise that you described.

What I can say about the output capacitors removed is that they are quite ok regarding their values (after 22000 hours, on my PS):
C406 = 1028 uF (+3.3Vout)
C403 = 1025 uF (+5.0Vout)
C504 = 909 uF (+12V1out)
C503 = 916 uF (+12V2out)
So I will not bet that the weird sound is due to those.

That's what I think as well. But at this point, I don't feel like taking this PSU out of service to check it. Maybe later in July-August, when the weather gets too hot to run that PC and I exchanged it for a less power-hungry PC, I might take it out and check then.

Anyway I intend to test all of them more proprely when my multitester with Atmega328 will arrive.

Sounds good. Keep us posted.
By the way, is that one of those multi-function "transistor testers"? (The ones that also test ESR, have frequency gen, and etc.)

Regarding the APFC coil, you are refering to the big one? Because I saw a smaller one with two types of wires near that area (see attached image).

Yes, it's the big one.
The smaller one is actually a driver transformer for the upper MOSFET in the 2-transistor forward topology of the main PS.

On my EA-430, the APFC coil was quite loose, so I actually unsoldered it, cleaned the conductive glue from its turns and bottom (for some reason, it had a lot, whereas the EA-500 didn't seem to have almost any), then isolated with tape and glued back with silicone glue. So it's possible this actually might be the reason why my EA-430 is quiet and the EA-500 isn't.

Is there any chance that we can find a detailed electronic schematic on this power supply or we have to do it by ourselves?

TBH, I rarely look for PSU schematics, at least for ATX ones. Maybe one might exist somewhere, but for me, the time it takes to search (and find) a schematic is pretty much the same as just observing the circuit and coming up with my own, then comparing to the application circuits of the ICs. (In this case, the EA-430 and EA-500 use the CM6800 PWM+APFC combo IC on the primary side. So obtaining a schematic of the CM6800 and looking at the application circuits given there should more or less closely resemble the circuit in these Antec PSUs. Component values might be different, of course. But with a bit of reading and datasheet "interpolation", it's often possible to understand the circuit if you know at least some basic circuit theory.)

[doru2304]

Re: Antec EarthWatts EA-430 and EA-500 – an easy bad caps fix

Sounds good. Keep us posted.
By the way, is that one of those multi-function "transistor testers"? (The ones that also test ESR, have frequency gen, and etc.)

I'll do that. I will come back with full details. I intend to compare the extracted capacitors with the replacements used (I have some from the same batch).

Yes. It is a multi-function "transistor testers".
What I have ordered is called: "2018 V1.13 English DIY Mega328 Transistor Tester LCR Diode Capacitance ESR meter PWM Square wave Frequency Signal Generator".

[momaka]

Antec EarthWatts EA-430 - build quality

Hmm… looks like I never really posted pictures of the EarthWatts EA-430 power supply that I've been talking about in this thread. OK, here it goes then.

This one was also a cheap eBay purchase from way back when (for like $1 for the PSU and $8 S&H.) Not bad, considering these are pretty decent PSUs. It's essentially an EA-500 with slightly downsized components in some areas… but mostly the same thing inside. On the outside, besides the label (being blue instead of brown/burgundy), the only other notable difference is the lack of a 2nd PCI-E 6-pin power connector, which the EA-500 has.

Case pictures:





Now moving onto the insides… again, it's almost the same as the EA-500 PSU shown in post #1. Probably the biggest difference is the size of the primary capacitor - the EA-500 has a 330 uF cap, while the EA-430 has a 270 uF cap. Both are rated for 400V, though, which isn't optimal IMO due to the APFC circuit's high-running voltage. But I did an experimental add-on/mod on this one that I will get to a little later below. Also, another difference is that the EA-500 has 2x PV218N50 MOSFETs whereas the EA-430 has 2x FOP 13N50G MOSFETs (with the former supporting up to 18 Amps continuous and 72 Amps pulsed vs. the later only being capable of 13 Amps continuous and 52 Amps pulsed.) And lastly, the EA-430 doesn't have that weird factory “flying” cap-resistor mod/fix on the optocoupler of the 5VSB circuit that the EA-500 does. Everything else appears to be pretty much identical. But here are pictures of the EA-430, just for the heck of completion:






Also a picture of the fan, since I didn't seem to include one with the EA-500:

It's an ADDA AD0812HS-A70GL rated for 0.25 Amps @ 12V.

Apart from this, there's nothing new to add for the build quality here. So below is just a parts summary before I jump into the mods I did on this PSU. Actually, there is one typo in post #1, which is why I am including the summary again. It has to do with the 12V rail's recitifiers: I said there's only one 30 Amp rectifier on the 12V rail in post #1, but there are actually 2x 30 Amp rectifiers in parallel.

Primary Side Summary:
- EMI/RFI: four 2.2 nF Y2-class caps; two Carli X2-class certified caps (0.22 uF + 0.47 uF); three common-mode chokes
- GBU806 bridge rectifier (8 Amps @ Tc = 100°C, 600 V reverse voltage)
- PFC section: 1 uF 450V film cap after bridge rect., 2x PV218N50 MOSFETs, one STTH8506d diode, one 1N5406 diode
- Main filter cap: one United Chemi-con SMQ, 400V, 270 uF, 25 x 35 mm (dia. x h)
- Main PS topology/design: double-forward with 2x FOP 13N50g MOSFETs
- Transformers: main with 35 mm core and 5VSB with 19 mm core
- Input wire: 16 AWG, 300V for Live/Neutral and 18 AWG, 600V for ground

ICs:
- CM6800 chip for PWM+PFC function
- ICE2A0565z off-line SMPS IC for 5VSB generation
- HY-510N supervisor IC on the secondary side (OVP on 3.3V, 5V, and 12V; UVP on 3.3V and 5V)
- three NEC 2561 optocouplers
- two 431 reference shunts

Secondary Side Summary:
- T130-52 common-mode output inductor
- 3.3V rail output inductor with -52 type core

3.3 V rail:
- mag-amp regulated
- ST STPS30L30CT (30 Amp, 30V, TO-220) schottky rectifier
- 1x OST RLX, 10V, 2200 uF, 10 x 20 mm before PI coil
- 1x OST RLS, 10V, 1000 uF, 8 x 20 mm after PI coil
- PI coil: 4-turn, 4 mm core, 14 or 16 AWG
- load resistor: 100-Ohm SMD (3015m)

5 V rail:
- ST STPS30L30CT (30 Amp, 30V, TO-220) schottky rectifier
- 1x OST RLX, 10V, 2200 uF, 10 x 20 mm before PI coil
- 1x OST RLS, 10V, 1000 uF, 8 x 20 mm after PI coil
- PI coil: 4-turn, 4 mm core, 14 or 16 AWG
- load resistor: 220-Ohm SMD (3015m)

12 V rail:
- 2x SBR30A60CT (30 Amp, 60V, TO-220) schottky rectifiers in parallel
- 1x OST RLX, 16V, 3300 uF, 12.5 x 25 mm before PI coils
- 2x OST RLS, 16V, 1000 uF, 8 x 20 mm after each PI coil (2 coils for “2” rails)
- PI coils: two 5-turn, 4 mm core, 14 or 16 AWG
- load resistor: 560-Ohm SMD (3015m)

-12 V rail:
- derived from output inductor winding, not main transformer
- 1.5 or 2 Amp (?) diode
- 1x OST RLS, 16V, 1000 uF, 8 x 20 mm before PI coil
- 1x OST RLS, 16V, 220 uF, 6.3 x 11 mm after PI coil

5 VSB rail:
- SB540 schottky rectifier
- 1x OST RLX, 10V, 3300 uF, 10 x 25 mm before PI coil
- 1x OST RLS, 16V, 220 uF, 6.3 x 11 mm after PI coil

[momaka]

Antec EarthWatts EA-430 – recapping and modifications

And now back to how/why I got the EA-430 PSU for cheap: it had the same issue as the EA-500: bad caps on the output. Actually, my EA-430 had only one bad cap - the 12V rail's OST RLX 3300 uF filter cap. Also, IIRC, the 2nd output filter cap on the 5VSB (OST RLS 16V 220 uF) was starting to go slightly high ESR. The other capacitors still appeared OK. And the few that I pulled out for a pre-emptive recap (2x OST RLX 2200 uF) tested OK as well, including the primary cap.

Therefore, the first order of business was to recap the PSU, which I did in a similar fashion as the EA-500: use whatever low ESR caps I had on hand that were “relatively” close enough. As such, I didn't replace the 8 mm OST RLS caps right before the output wires - these all appeared OK and so far have been holding up OK in the EA-500. I don't see a reason why they wouldn't in the EA-430 either. Below is a summary of what was replaced:

12V rail: 1x OST RLX, 16V, 3300 uF ---> 1x UCC KY, 16V, 3300 uF
-12V rail: 1x OST RLS, 16V, 1000 uF, 8mm dia. ---> 1x Rubycon YXJ, 16V, 1000 uF, 10 mm dia.
-12V rail: 1x OST RLS, 16V, 220 uF ---> 1x Rubycon YXJ, 25V, 220 uF
5V rail: 1x OST RLX, 10V, 2200 uF ---> 1x Panasonic FM, 6.3V, 2200 uF
3.3V rail: 1x OST RLX, 10V, 2200 uF ---> 1x Rubycon ZLH, 6.3V, 2200 uF
5VSB rail: 1x OST RLX, 10V, 3300 uF ---> 1x UCC KY, 6.3V, 2700 uF
5VSB rail: 1x OST RLS, 16V, 220 uF ---> 1x UCC KY, 10V, 220 uF

And of course, I also changed all of the small caps on the primary side with equivalents from Nichicon, Panasonic, Rubycon, and/or UCC. Picture of the recap:



Another noteworthy change, as is already shown above: “securely” mounting the series regulating transistor for the fan onto the secondary heatsink to allow it to dissipate heat better/easier. Of course, I still don't know how thermally-overloaded this transistor may be. So just like with the EA-500, I added a small diode between the 5V rail and the positive (+) lead of the fan to feed it 5V when the fan is operating at minimum speed (and in case the series transistor fails open-circuit.) The modification can be seen here with a SOT-23 diode:


This wasn't the only modification I did, however. Another one is when I was changing the small 220 uF output cap on the 5VSB rail - I couldn't bend the leads on the cap enough to make it fit due to using a recovered capacitor from an Xbox 360 motherboard. Therefore, I made my own hole in the PCB to make the cap fit.


While messing with these small “dinky” mods, I noticed something a lot more concerning – the APFC toroid inductor was a bit loose and the tan glue on it had turned quite dark / brown.


After removing it for closer inspection, I found something even more concerning:


The APFC toroid had started to chip away at the insulating coating of its current sense resistor next to it. This could surely result in bad news for the APFC circuit if a winding on the inductor shorts to the resistor.

So this certainly needed addressing. First, I removed as much brown/conductive glue as possible. Next, I unwound several turns on the toroid to remove more glue and added electrical tape between some of the windings before winding them back on. This was done in order to (hopefully) better protect against turns shorting. The crap glue had already started to tarnish the insulating coating slightly on a few windings. After all this was done, I put the APFC inductor back in and secured/glued it with regular household bathroom silicone/caulk both on the bottom and the top (to the case of the APFC MOSFETs.) Obviously regular silicone isn't ideal for this application either (due to slight acidic off-gassing when curing)… but beats cheap crap conductive glue ANY DAY!


Needless to say, I tested the PSU on its first power-up with a 500 Watt heating element in series with the AC input, just in case I goofed up somewhere on the “repair” of the APFC inductor. However, all was well and the PSU worked fine.

I'm not quite done here, though.
The last piece was this:


Hmmm… Now what could that be? I guess let's see the other side of the board:


Something definitely will be installed here. But what?

Well, remember how I mentioned the APFC putting extra stress on the primary cap because of the high voltage and high frequency? I've had this idea that perhaps putting in a film cap could help it deal with some of the HF ripple – or so I'm hoping. Actually, recently I opened a Delta/Newton PSUs for a Precision T7400 tower, and it also had a film cap in parallel with the two primary caps… which I think gives “my idea” more credibility. And so what did I do? Exactly this – I put a Panasonic 450V 2.2 uF polypropylene metal film cap in there.


And here it is all soldered up on the bottom:


As for how much this is going to help ease off ripple current from the primary electrolytic cap – this I don't know. I'm just trying to prevent something like this from happening:
https://www.badcaps.net/forum/showthread.php?t=37779

Moreover, even in the even the primary cap fails open-circuit, I think the paralleled film cap could/should still provide at least some form of a low-impedance path to ground so that the energy stored in the APFC inductor doesn't do a crazy voltage shoot-through and break down primary silicon… or so is my theory, anyways. I guess we will have to see it in practice, which will be who knows when. I don't have this modification/addition in the EA-500, but that one also has a noisy/whining APFC already, which worries me a little. On the other hand, the EA-430 was quiet before and still is quiet after this add-on. So at the very least, I'm sure my add-on / modification is not hurting anything… except possibly my wallet with a few extra cents (or so at the time of buying these Panasonic PP film caps… now they are over $1 and regularly out of stock on Digikey.)

And this concludes just about everything I have done to the EA-430 and EA-500 PSUs. For the EA-500, I might swap the 3300 uF KY cap on the 12V rail with a KZE, just for the slightly more appropriate ESR. The 3300 uF KZE is currently sitting in an old Inno Power PSU (by Macron Casing) that doesn't need anywhere that kind of low ESR on its output. It's been working OK with it… but is just not needed there. This will be work for another day, though. I'm just mentioning it in case anyone wants to recap one of these EarthWatts PSUs, since the KZE is probably closer in terms of ESR to OST's RLX series than KY (though the difference between KZE and KY isn't that big – just 5 milliOhms.)

Other than that, I've already been using the EA-430 in one PC. It sees very infrequent use – actually, it's still in a test configuration stage. But the EA-430 managed to power that PC without any issues, providing power to an Intel Core 2 Quad Q8200 CPU and basic GPU (probably barely getting tickled with such a load.) But it works and appears to spit out stable power, so that's all that matters.