Antec EA-430D 5VSB issue

[Rgyver2]

I have an Antec ea-430d that will not turn on due to a problem with the 5VSB circuit. the TNY278PN has a burn "drain" pin (pin 4) and the two diodes (D901 and ZD901) near it look burnt with black on the board. (they were covered in that white goop which i removed).
The TNY278PN needs replaced and likely the 2 diodes as well. I found the TNY278PN, but need to know what to order to replace diode D901 and the zener diode ZD901.
From reading other posts, it seems wise to also replace the large 450V 270uf cap.

The diode info and any other advice of what else to check for the 5VSB issue would be much appreciated. Thanks in advance.

[linuxguru]

Re: Antec EA-430D 5VSB issue

Funky - seems to be Delta OEM, or at least shares the main trafo part number with Delta. If it had been built by Delta, it should not have failed easily, but of course, things do fail occasionally on monolithic flyback switchers exposed to spikes, surges, etc.

[RJARRRPCGP]

Re: Antec EA-430D 5VSB issue

Looks like Kentucky-fried diodes, LOL.

[Rgyver2]

Re: Antec EA-430D 5VSB issue

From what i have seen, the "D" in the model number - EA-430(D) - means it is a Delta version.

From what i can tell, the rest of the Power supply seems to be Ok. But will not truly know until i can fix this 5VSB and further test.

Anyone have an idea yet on what i need to order to replace the two diodes?

[R_J]

Re: Antec EA-430D 5VSB issue

What is wrong with D901 and ZD901? They are likely ok and just black from the ic's blowing. ZD901 is likely a tvs diode (similar to a P6KE200A) and would check open. D901 is likely a Fr107. The diodes should have numbers on them. Replace the electrolytic on the ic's vcc pin C901?

[Rgyver2]

Re: Antec EA-430D 5VSB issue

What is wrong with D901 and ZD901? They are likely ok and just black from the ic's blowing. ZD901 is likely a tvs diode (similar to a P6KE200A) and would check open. D901 is likely a Fr107. The diodes should have numbers on them. Replace the electrolytic on the ic's vcc pin C901?

Thanks.
The Diodes had numbers but do not appear legible now. The zener may be ok but the D901 shows some heavy heat damage on the side next to the ICs D pin that burned off. D901 may be ok, but with it showing heat damage, i figured best to replace it.

The C901 is a Nichicon 220uf 35V PW(M) with no physical sign of damage. But i will test it to see if it needs replaced.

I just was not sure if the the root cause was a failed IC or associated component.

[linuxguru]

Re: Antec EA-430D 5VSB issue

Avalanche protection/clamping diodes can encounter cumulative damage over time, and fail eventually. If it fails short, the switcher will fail almost immediately, since the inductance of the flyback trafo is shunted, thus there is no current limit. If it fails open, the switcher may still fail after a while, since there is no effective snubbing of the turn-off spike from the leakage inductance.

[Rgyver2]

Re: Antec EA-430D 5VSB issue

update
both of the diodes seems fine.
D901 read OL c to a and 270k ohms a to c - diode checker shows 0.515v drop fwr bias. only visible markings are "26" possibly B or D first, K or A second, 26 then 5 or E too much outer damage to figure it out. but seems to read fine, so will keep for now.
ZD901 - read OL c to a and 299k ohms a to c - diode checker shows 0.545v drop fwr bias. R J's post helped me realize i mistakenly called it a zener where I believe it actually is a TVS diode.
C901 - Nichicon 220uf cap read 208uf and 0.18 esr - seems good as well.

now waiting to receive the replacement IC to install and test. still no idea what caused the failure. With the IC removed, i tried to check C902 (connects to "BP/M" pin - on the back) but it reads 29K ohms both ways. And my cheap cap meter said unknown. But hard to know with it still connected.

Also with power applied, i read 167Vdc across the large 270uf 450V cap.

Any other suggesting on what to check while waiting on the new IC to arrive?

[momaka]

Re: Antec EA-430D 5VSB issue

Also with power applied, i read 167Vdc across the large 270uf 450V cap.

That's normal for 120V AC mains power and when the APFC circuit is not running.

Any other suggesting on what to check while waiting on the new IC to arrive?

Check output filter electrolytic caps on the 5VSB circuit - should be two of them, most likely. High ESR on either of these can cause the standby IC to "panic" when it tries to regulate the voltage.

[Rgyver2]

Re: Antec EA-430D 5VSB issue

Update

i checked the 5vsb filter cap (Nichicon 10V 1000uf PW(M)) it read 990uf and esr 0.25 - 0.30 ohms. (esr read with cheap Chinese meter - Data sheet for PW list 0.068 ohms (z) at 20C) i only saw one filter cap for the standby power.

i replaced the TNY278PN with TNY278PG (278PN out of stock until Oct maybe) from the spec sheet, the only difference i saw between N and G was the lead finish.

put all the original tested components in with the new IC (TNY278PG). power it on for one second (to make sure nothing goes pop) and nothing happened as expected.

Then connected power supply tested to the 24 pin connector. Powered it on, and it seemed to work, Fan spun, meter beeped. But after about 2 seconds (if that long) the TNY278PG IC poped and blew off a piece of it plastic top.


Not really sure where to go from here. With the way it exploded, did not know if i should focus on the middle section or downstream section where the purple 5vsb lead is connected.
Would the slightly higher esr on the 5vsb filter cap be enough to cause the IC to panic? if it went into panic, would that cause it to explode the way it did?
Should i consider replacing DZ901? i tested it and it does conduct forward bias (didnt conduct revers bias with fluke diode check). But being that it is connected to D901 anode to anode, i did not have a way to test to see if it conducted as a TVS (i assume higher voltage in reverse bias)? or does that diode only come into play during voltage spikes or during power down for the xfmr?

Thought about replacing the IC (i have 2 more) and the 10V 1000uf filter cap for the 5vstb. then apply power (without the power supply tester - such that it doesn't turn on) and then read the voltage on the purple 5vstb wire? I may have a replacement on hand for the C901 (35V 220uf) if someone thinks i should change it as well even though it read fine.

Lastly, should i go back and look at the C902 smd cap connected to the BP/M pin? I believe it controls the ICs current limit. didnt know how likely it was that it could cause this failure.

thanks

[linuxguru]

Re: Antec EA-430D 5VSB issue

I'd guess that it's an APFC circuit failure - the APFC-boosted primary voltage may be out-of-spec and too high. You may also want to replace the avalanche diode in the snubber circuit of the TNY278PN - something like a P6KE200 or similar should work.

The probable reason it didn't pop when initially powered up is that the APFC boost becomes active only after PS-ON is activated. Until then, it had a lower primary voltage and the +5Vsb auxiliary flyback switcher was perfectly happy with it.

[Rgyver2]

Re: Antec EA-430D 5VSB issue

I'd guess that it's an APFC circuit failure - the APFC-boosted primary voltage may be out-of-spec and too high. You may also want to replace the avalanche diode in the snubber circuit of the TNY278PN - something like a P6KE200 or similar should work.

The probable reason it didn't pop when initially powered up is that the APFC boost becomes active only after PS-ON is activated. Until then, it had a lower primary voltage and the +5Vsb auxiliary flyback switcher was perfectly happy with it.

Thanks
the avalanche diode would be the DZ901 correct? -- also i mistakenly said in a previous post that the diodes were anode to anode. they are actually cathode to cathode. (i confused the orientation from front to back)
What components would make up to APFC boost circuit? And what would you suggest i check first?

also does anyone have a schematic for this power supply? or even a generic ones that would apply to this model?

[sam_sam_sam]

Re: Antec EA-430D 5VSB issue

You need to check for short circuit in the primary side of the switching power supply and also check the secondary side of the switching power supply for short circuits as well and use a incandescent light bulb 100 watt instead of the fuse so minimize the chance of blowing the TYN device

Consider any reading of 1k ohm or lower as a possible short

[linuxguru]

Re: Antec EA-430D 5VSB issue

What components would make up to APFC boost circuit? And what would you suggest i check first?

also does anyone have a schematic for this power supply? or even a generic ones that would apply to this model?

If it's a CM6800 or similar APFC/PWM controller on the primary side, here's a generic schematic that's similar to those used in a variety of APFC models, including Corsair CX430, etc.

https://www.badcaps.net/forum/showpo...9&postcount=10

(It may not be identical, just broadly similar.)

[Rgyver2]

Re: Antec EA-430D 5VSB issue

yes i had previously for shorts but didnt see anything unusual. but also not having a schematic, it was not exactly precise checks.

Thanks to Linuxguru's link about APFC, i believe i figured out the issue, or at least main issue with the power supply.

Now that I have a basic understanding of the role of the main cap, i thought that must be it. If the main cap (C1) Aishi 450V 270uf has greatly reduced or no capacity, then it will not absorb the discharge of the coil. Therefore allowing the voltage to increase to the point that it could blow the TNY278. And the cap is bad only reading 11 nf with 17% voltage loss ( read 80 nf via fluke 87).
Previously, since i ready an expected voltage of 167vdc across it with power applied but power supply off, i mistakenly assumed it was good.

that being said, I have not found a good data sheet for this Aishi cap to look up impedance and/or esr and ripple values as i do when replacing caps. Any suggestions for what to use to replace this cap?
As Linuxguru suggested, i will also replace the avalanche diode.

for this scenario where a failed cap allowed voltage to increase to which the TNY278 blew, is there anything else this event would indicate possibly failed? should i suspect an issue with the thermistor neat the boost coil, for this to occur? or other protective device like a snubber or such that if working would have prevented the TNY278 from blowing even with a bad cap?

[R_J]

Re: Antec EA-430D 5VSB issue

It seems like this cap is basically open, so when the pfc circuit started it likely caused the standby ic to blow due to the high frequency which was not filtered and could have exceeded the ic's scpec. Or even the high ripple from the a/c line not being filtered could have caused the ic to blow. Replacing the main filter should fix the problem.

[momaka]

Re: Antec EA-430D 5VSB issue

... and use a incandescent light bulb 100 watt instead of the fuse so minimize the chance of blowing the TYN device

+1

Here's a visual diagram for that:
https://www.badcaps.net/forum/showpo...4&postcount=70

Note that the bulb needs to be either incandescent (which are not now produced anymore) or halogen type. 60 to 100 Watts rating will do, though in some cases, even 40W or 50W bulb will do the "trick". (And you can also put 2x 40W bulbs in parallel for a total of 80W, and then put that in series with the PSU.)

Also note that the incandescent/halogen bulb trick only work for testing out the 5VSB circuit. Since this PSU has an APFC circuit, testing the main PS (jumping PS-ON to ground) with the series bulb may not work. However, if you still want to check the main PS and prevent parts from blowing up again, you can put a 250-500 Watt appliance with a heating element (toaster, hair dryer, and etc.) in place of the incandescent/halogen bulb, and that may be able to let the PSU turn on with a very light load (maybe an HDD or just a few 12V car/auto light bulbs.) Actually, your PSU load tester should be light enough to allow the PSU to turn ON with maybe 2-3x 60W bulbs in parallel on the input (as a series limiting device.)

i checked the 5vsb filter cap (Nichicon 10V 1000uf PW(M)) it read 990uf and esr 0.25 - 0.30 ohms. (esr read with cheap Chinese meter - Data sheet for PW list 0.068 ohms (z) at 20C) i only saw one filter cap for the standby power.

It's probably fine. PW series from Nichicon are reliable.

Those cheap testers aren't always 100% accurate, depending on firmware. Mine (an older GM328) reads a little higher ESR than normal when testing caps between pins 1 and 3 (which is how the manual suggests to test caps for ESR.) But between pins 1 and 2, I get normal ESR (if not lower than normal.)

Would the slightly higher esr on the 5vsb filter cap be enough to cause the IC to panic? if it went into panic, would that cause it to explode the way it did?

Unlikely.
From what I have seen at least, ESR can be as high as several Ohms before the circuit really starts to act up. Of course, under high load, it may loose regulation a lot sooner. But just running with no load, it should be able to take a lot higher ESR from the output caps before things go haywire.

Should i consider replacing DZ901? i tested it and it does conduct forward bias (didnt conduct revers bias with fluke diode check). But being that it is connected to D901 anode to anode, i did not have a way to test to see if it conducted as a TVS (i assume higher voltage in reverse bias)? or does that diode only come into play during voltage spikes or during power down for the xfmr?

Jury's still out on that one.
Most likely ZD901 is OK, now that we know that the TNY IC probably blew up because of the primary cap loosing capacitance.
ZD901 is used for clamping down inductive kickback from the 5VSB transformer - that is, when the TNY IC pulses ON and then OFF, current will still try to continue to flow through the 5VSB transformer's primary side, due to inductive properties (just like with the APFC coil circuit.) This current has to go somewhere, otherwise the voltage will rise really high and blow the IC up. So that's where the current goes - through D901 and then ZD901. Some circuits use a resistor and cap parallel network in place of TVS diode, which is another acceptable method.

... then apply power (without the power supply tester - such that it doesn't turn on) and then read the voltage on the purple 5vstb wire?

That would be the best way to do it indeed. Check the 5VSB first by connecting the PSU through a series incandescent/halogen bulb and applying power. If 5VSB reads OK and steady for at least 5-10 minutes, you may give the PSU a go with your load tester to see if that works too now.

If the main cap (C1) Aishi 450V 270uf has greatly reduced or no capacity, then it will not absorb the discharge of the coil. Therefore allowing the voltage to increase to the point that it could blow the TNY278. And the cap is bad only reading 11 nf with 17% voltage loss ( read 80 nf via fluke 87).

Yup, that cap is very close to being an open circuit and probably what's causing the issue here.

that being said, I have not found a good data sheet for this Aishi cap to look up impedance and/or esr and ripple values as i do when replacing caps. Any suggestions for what to use to replace this cap?

No need for a datasheet. Generally, you won't find big caps like this in low-ESR offerings (or not easily, at least.) I don't see a temperature rating on that Aishi cap, but if it's rated for 85C, it is most definitely a general purpose low-frequency cap.

That being said, APFC circuits don't really care much about the ESR. The ripple current spec, on the other hand, is rather important, because APFC circuits tend to put a lot of ripple on the primary caps. So here, you should just go with the highest ripple current rating you can cram into the physical space available for that cap. This means picking the biggest capacitor you can fit in the PSU.

Also, from my own calculations I did in that Enermax Pro 82+ EPR525AWT thread linked above, a cap with a voltage rating of only 400V is a bit too low, IMO. So go with a cap rated for at least 420V (or preferably even 450V.) Because of the higher voltage rating, the cap will tend to be bigger, so it will also tend to have better ripple current rating.

Next, the capacitance: many of these "budget" PSUs seem to use just the minimum functionally adequate rating for the input cap, as these parts tend to be expensive. So upgrading to 330 uF or 390 uF would be a lot better. If you can, maybe even fit 470 uF in there (though 390 uF should be reasonable enough for a ~400 Watt PSU.)

Finally, the temperature rating. Generally speaking, try to go with 105C-rated caps here, as even a 105C-rated cap for 1000 hours endurance should technically last twice as long as a 85C-rated cap for 2000 hours endurance (it is an acceptable approximation that a capacitor's service lifetime doubles, more or less, for every 10C drop in the operating temperature below the maximum.) However, if you can't find a 105C cap that will fit the above requirements for voltage and capacitance, then stepping down to an 85C cap is acceptable too.

With all of that said, see what you can find for options and let us know if you have any questions about deciding on what to use. As a temporary fix, you could probably even take a cap from another PSU, just to see if the APFC circuit works and whatnot. But if you want reliable long-term operation, then it is best to go with a primary cap rated for at least 420V and higher capacitance than the original - highest you can fit in the space available. Or as an alternative, you may also put 2x smaller caps in parallel instead of one big one.

[Rgyver2]

Re: Antec EA-430D 5VSB issue

Thank you Momaka for answering my questions. and I greatly appreciate the advice on the replacement cap. the original cap was a 450V so I already planned on 450 to 500V.

Unfortunately 25mm diameter is the max without running into other components in that space. event the original cap was 22mm and had some yellow tape around the bottom. I am guessing as extra insulation. the cap had 85 on it so i believe it was a 85C. I always go with 105c replacements caps when possible.

Thanks for the info on the incandescent assisted testing, it is a good idea. I already understood how to include it and why it would help. The diagram and explanation will help others that come across this post. I feel confident that to cap is my issue, so i think i will give it a shot. If by chance i blow another TNY, i will definitely use it to protect my final one.

Also thanks for the cap meter/ esr meter info. Mine has been fairly decent for my needs and mostly accurate farad reading. Any time i question it, i use my fluke 87 5 for comparison. although the fluke normally reads just a little higher than most cap meter - in my experience. When i go to upgrade to a better esr meter, i will keep yours in mind.

Once I get the parts and get it working, i will post an update.

Thank you to everyone that has taken time to help and provided insight and knowledge.

[ChaosLegionnaire]

Re: Antec EA-430D 5VSB issue

wow a failed 450v cap on an apfc circuit! i guess that was more due to the cap being of junk quality rather than the voltage rating being inadequate? if not, i'd be shitting in my pants everytime i powered on a system with an apfc psu!

[stj]

Re: Antec EA-430D 5VSB issue

maybe the way apfc bs works just isnt good with electrolytics in general.
maybe it should have something faster in parallel - like a polyester brick.