PSU OCZ700MXSP no +5VSB

[vladnl]

Hi Everyone!

So my PSU decided to give up on me after trying to power up my PC 4 weeks after vacation. I’ve heard small bang, smelled fumes and thought it has to be a cap 😊
After opening the PSU, except the bulged bulk cap (TAEPO 470uF 400V 85C A3) I haven’t seen anything else rare, burn marks etc. I checked the cap and it was indeed toast. I’ve ordered replacement cap (TAEPO 470uF 400V 105C A3) with similar values except it was 35mm diameter in place of 30mm, so naturally it wouldn’t fit in the space of the original. I did some obscure wiring, just to check whether I can wake up the PSU so I can order correct part nr. No Luck. I couldn’t get the +5VSB, however the bulk cap was charging correctly at around 330V.
After educating myself a little bit more on the yt and this forum, I started tracing the problem, checking diodes and caps, all of it seemed fine. (was checking around wrong transformer btw. &#128515. Finally I located the correct section where stb power is generated from, and discovered shorted diode, as well as scorched 22ohm resistor.
My problem is that I can’t find anything about this diode on the web, if I try read markings correctly, it says t9250 200A… Perhaps someone can help me out finding replacement for this diode? At home I only have rectifying diodes IN4001 and 5401. Not sure whether I can at least try those before ordering parts.
P.S. My knowledge in electronics is pretty limited, looking forward to expand it.

Thanks in advance.

[Milio034]

A hint: lookup the power IC for the 5V SB. It could be fried. There are data sheets in which you can find the standard layout of such a power swich. http://www.datasheet-pdf.info/entry/TNY279PN (figuur 14).
Am also dealing with a blown safety resistor in my PSU and most likely also the power IC. I ordered a new one with some caps and hope to solve the problem. I will also check my diodes.
Succes and keep in touch!

[momaka]

I agree with Milio. Look up the part number for the 5VSB generator IC first. The 5VSB IC should be the 8-pin DIP one next to that small electrolytic capacitor, C503, as seen in this picture:
https://www.badcaps.net/forum/attach...1&d=1631256100
Post what part number you read on it here.

After this, could you draw on your picture of the solder side of the PSU where the 22-Ohm resistor and that bad diode connect to? That should give us a good clue as to more specific ratings for these two parts. The diode is most likely rated for at least 600V... so I don't suggest trying to replace it with 1N4001 or 1N5401. Besides, it also needs to be fast-recovery (FR) type, due to high frequency switching of the circuit. A regular diode like 1N400x and 1N540x series might quickly overheat in such application. *EDIT* That "22-Ohm" resistor appears to have color bands of red, red, gold and gold... so it most likely is a 2.2 Ohm fusible resistor. Can't say exactly if it's rated for 1 or 2 Watts, based on the pictures, but it's probably close to that. Since that resistor is blown, the standby IC is likely bad too. In terms of specs, that fusible resistor needs to be flame-proof/flame-resistant... so basically go with a metal oxide type of resistor, as those are flame-proof/resistant. Do not use carbon film type of resistor here.

As for the big 400V Teapo cap that blew - that's because OCZ decided to cut corners here and use a cheaper primary cap in a PSU with Active PFC circuit. That cap really should have been rated for at least 420V and preferrably for 450V, because APFC boost voltage can sometimes get very close to 400V, and caps don't like it when they often start seeing voltages very near their maximum rating - especially cheaper brands like Teapo. Also, for a 700-Watt -rated PSU, that primary cap should really be at least 650 uF or higher (i.e. 2x 330 uF caps in parallel or 1x 680 uF cap) for getting proper hold-up time. Unfortunately, this kind of corner-cutting is frequent with just about every "budget" high-power PSU. That said, the replacement Teapo cap you have right now should be OK for testing the PSU. Once you do manage to get the PSU fixed, I would highly suggest to go with a better-rated primary cap, and preferably a Japanese brand like Rubycon, Nichicon, United Chemicon, or Panasonic. Otherwise, this kind of failure will probably happen again after some years.

[vladnl]

Thanks guys, not sure how I couldn't see the your response earlier.
It's super difficult to read the chip letters, but it appears to be TNY278PN.
Not sure how to test it though?
I couldn't detect short between common s pins and d or bp/m or en/uv.


As for the resistor, it didn't fail in its body, just one of the connecting wires blew, so I could read and confirm the value red red gold brown 22 ohm.
Thanks for the tip, else I would have installed common 1W resistor.

As for the diode, I'm still not sure which one, placement you can see on the attempted drawing ZD501.

Sorry for the quality of the photo.

[vladnl]

Ok I've found the way to use magnifier on my phone, chip is definitely TNY278PN.

[momaka]

Originally Posted by vladnl

Ok I've found the way to use magnifier on my phone, chip is definitely TNY278PN.

Awesome!
So basically just download the TNY278 datasheet and use it as a reference. It will have an application circuit inside that may look very similar to the one in your PSU's 5VSB circuit.

Originally Posted by vladnl

Not sure how to test it though?
I couldn't detect short between common s pins and d or bp/m or en/uv.

No need to test it. Given the high-power components that blew around it, it's probably bad already. Of course, you could just replace the bad diode and resistor and hope that it's not blown... but in that case, using a series incandescent light bulb with the PSU's power input is a must, if you want to prevent blowing up more parts. In fact, using a series incandescent bulb is always recommended any time work is done to the primary side of a PSU. This is how to connect it:
https://www.badcaps.net/forum/showpo...4&postcount=70

Originally Posted by vladnl

As for the resistor, it didn't fail in its body, just one of the connecting wires blew, so I could read and confirm the value red red gold brown 22 ohm.

Red Red Gold Brown would still be a 2.2 Ohm resistor (just 1% tolerance instead of 5% if the last band is gold). It's possible that the resistor has gone high-resistance from the over-power and just coincidentally happens to measure around 20 Ohms. In that case, put 5V through it with a small charger/adapter and let it sit for 30 seconds to a minute. Then disconnect and measure its resistance. Has it changed significantly? If yes, it's certainly bad.

Originally Posted by vladnl

As for the diode, I'm still not sure which one, placement you can see on the attempted drawing ZD501.

Thanks. The drawing helped figure things out, actually.
Going by the application circuit in the TNY278 datasheet on page 8, there is a "Zener" diode in the similar arrangement on the snubber clamp circuit... and it's actually not a Zener diode, but rather a TVS diode (part number P6KE150A). Your snubber clamp circuit looks slightly different, though, so the TVS diode in position ZD501 may have slightly different specs than the P6KE150A part suggested in the datasheet. That said, looking up the words "TVS diode" with the string "9250" after it didn't yield any results. But looking up "TVS diode" with the string "200A" after it suggests your diode may be a 1.5KE200A, which is similar to the P6KE150A part, but rated slightly higher on the clamping voltage. Datasheet for 1.5KE series TVS diodes can be found here:
https://www.littelfuse.com/products/...1_5ke200a.aspx

Since both the resistor and that TVS diode are components related to handling high power in that circuit, I suggest you carefully check many of the smaller components around the TNY278 IC after you remove the TNY278 from the circuit. That way, if something else/more is blown, you can all get it in one parts order, hopefully.

[vladnl]

Momaka,

Thank you for your time and energy!

Red Red Gold Brown would still be a 2.2 Ohm resistor (just 1% tolerance instead of 5% if the last band is gold). It's possible that the resistor has gone high-resistance from the over-power and just coincidentally happens to measure around 20 Ohms. In that case, put 5V through it with a small charger/adapter and let it sit for 30 seconds to a minute. Then disconnect and measure its resistance. Has it changed significantly? If yes, it's certainly bad.

Indeed, you're right, I think I've missed that dot with my first post and stayed faithful to my typo So it's 2.2 Ohm, by size, I'd say 1W, but I'll order 2W metal oxide.

For the TNY278, as well as for the good quality and size caps, it seems impossible to source that in Europe nowdays. So if the IC is indeed toast, I'll have to wait some time before attempting to replace it.
In meanwhile, I'll get the recomended diode and resistor, test the PSU with lamp method. Even the incandescent bulb can't be found in the Eu so I'll be probably wiring halogen bulbs.

Thanks again!

[lotas]

ZD501 LT9250 P6KE200

https://www.eevblog.com/forum/repair...o-600w-atx-ps/

[momaka]

Originally Posted by lotas

ZD501 LT9250 P6KE200

https://www.eevblog.com/forum/repair...o-600w-atx-ps/

^ Excellent find!
So looks like these OCZ PSUs fail quite often, judging by that thread. No surprise though - cheapo Teapo caps simply can't last too long on the primary of an APFC circuit. Heck, even the Japanese grands struggle there sometimes if the manufacturer didn't use an appropriately sized cap.

Originally Posted by vladnl

Indeed, you're right, I think I've missed that dot with my first post and stayed faithful to my typo So it's 2.2 Ohm, by size, I'd say 1W, but I'll order 2W metal oxide.

Yup, 1W looks about right, but 2W won't hurt anything. Just make sure it's flame proof/resistant type metal oxide (some aren't).

Originally Posted by vladnl

For the TNY278, as well as for the good quality and size caps, it seems impossible to source that in Europe nowdays. So if the IC is indeed toast, I'll have to wait some time before attempting to replace it.

Yeah, most likely you will need to find NOS somewhere or order from Ebay or AliExpress... so indeed it probably would be a while before you get them.

Originally Posted by vladnl

In meanwhile, I'll get the recomended diode and resistor, test the PSU with lamp method. Even the incandescent bulb can't be found in the Eu so I'll be probably wiring halogen bulbs.

Yeah, I should probably specify that halogen is also an acceptable substitute for incandescent. Actually, they are pretty much the same thing, except halogen is just a more "improved" design of the regular incandescent bulb.

I know EU banned incandescent bulbs, but aren't 40W and lower-rated ones still available for sale? Here in North America, they are, as they are considered for "decorative" purposes... so you can still find incandescent bulbs under 40W. That said, I also traveled to SE Europe this summer and was very surprised to find that some hardware stores still carry 60, 80, and 100W incandescent bulbs in E27 socket. I think they're able to skirt around EU regulations, because they were labeled as "Emergency Light Replacement", or something along those lines. They're quite cheap too - fraction of the cost of even the cheapest LED bulbs. Hence, people still buy them and use them. I had a harder time finding halogens in GU10 and MR16 sockets than those incandescents. But they really are pushing the cheap LED garbage everywhere.

[sam_sam_sam]

I can still find incandescent light bulbs but they are to find I have a few to be able do testing on devices

[vladnl]

Originally Posted by momaka

^ Excellent find!
So looks like these OCZ PSUs fail quite often, judging by that thread. No surprise though - cheapo Teapo caps simply can't last too long on the primary of an APFC circuit. Heck, even the Japanese grands struggle there sometimes if the manufacturer didn't use an appropriately sized cap.

Luckily my order was cancelled by the shop, as I could shop there only as B2B.
I guess I can use 1.5KE200A in place of P6KE200? Latter is only available via Ali or ebay.. The 1.2KE200A is readily available.

Thanks for the tip about flame proof resistors, I assumed all of the metal oxide are.

For the light bulbs, I really need to go shopping and looking, most of the hardware stores and grocery stores carry the LED's only, unfortunatelly.

[Per Hansson]

Originally Posted by vladnl

For the TNY278, as well as for the good quality and size caps, it seems impossible to source that in Europe nowdays. So if the IC is indeed toast, I'll have to wait some time before attempting to replace it.
In meanwhile, I'll get the recomended diode and resistor, test the PSU with lamp method. Even the incandescent bulb can't be found in the Eu so I'll be probably wiring halogen bulbs.

Thanks again!

I considered this a challenge so here you go
300w incandescent light bulb: https://www.lamportillallt.se/sv/lam...-e27-300w.html

TNY279PN: https://www.tme.eu/ie/en/details/tny...-integrations/