SX280 full poly

[PCTechNY23]

Is it safe to use a 330uF poly to replace the 1500uF lytic? I have seen a lot of people using 1/2 but less than 1/4 seems like it might be a problem.

The 3x 330uF 16v Polymers replace the 1500uF 16v Lytics

[Dgtech]

Re: SX280 full poly

Thats more like 1/5 the total capacitance. You can try anything for kicks, but I'd advise not to.

[Topcat]

Re: SX280 full poly

The 3x 330uF 16v Polymers replace the 1500uF 16v Lytics

Works fine, I've already tried it.

https://www.badcaps.net/forum/showthread.php?t=9504

[Scenic]

Re: SX280 full poly

i guess those are the 12V input caps..?
if so, i haven't seen any newer boards that came with poly's from the factory with anything higher than 4x 470uF 16V in there. most use 330uF caps for 12V VRM in..

[Pyr0Beast]

Re: SX280 full poly

I'd advise against it unless terribly needed. Bulk capacitance IS needed.

[Topcat]

Re: SX280 full poly

I'd advise against it unless terribly needed. Bulk capacitance IS needed.

Incorrect. All the test subjects are rock solid with these (2x SX280's and 2x GX620 USFF), going on 6 months now of running 24/7 @ 100% CPU load crunching F@H. Never had one crash or become unstable. If the bulk capacitance were truly needed, it wouldn't have stood up to a week of this kind of abuse, let alone 6 months.

FWIW, the 1500uF 16v lytics rarely fail on these unless they're nichicon or UCC. Never seen these 3 caps fail if they were rubycon or panasonic.....so if that's what you have, you don't have to change them.

[larrymoencurly]

Re: SX280 full poly

If the bulk capacitance were truly needed, it wouldn't have stood up to a week of this kind of abuse, let alone 6 months.

Or the mobos were built with a lot of margin?

How much can the capacitance be decreased? It seems that most factory installed polys are about half the capacitance of conventional electrolytics that would go in the same place, but 330uF is only a fourth. Is that still OK?

[Pyr0Beast]

Re: SX280 full poly

Incorrect. All the test subjects are rock solid with these (2x SX280's and 2x GX620 USFF), going on 6 months now of running 24/7 @ 100% CPU load crunching F@H. Never had one crash or become unstable. If the bulk capacitance were truly needed, it wouldn't have stood up to a week of this kind of abuse, let alone 6 months.
What about undershot or overshot events ? Have you even put a scope to check it out ?

Yes, I know it works, how well and for how long is the real question.

I've recapped a mobo from 1500uF taepos to 1000uF Nichicon UHD solid caps.
Esr is less that 1 tenth of the original value. But system did crash, and mouse worked sporadically.

Once I attached 3 2200uF panny's at the bottom it worked flawlessly.


This is how it looked at load transient. It is horrible.

[Topcat]

Re: SX280 full poly

[I]Have you even put a scope to check it out ?

Yes, I know it works, how well and for how long is the real question.

No scope, just the real world acid test. 6 months and counting of 24/7 F@H, which is a hell of a stress test. It has been rock solid. As for how long, the polymers will outlive the lytics, so that should be for how long. I've also put a few of them into the real world, none have come back or have been reported as unstable, and some of these have been in service close to 2 years. I didn't mention these, as I don't have them in my posession. However, I still receive repair work from this corporation, had they failed or become unstable, they would have returned them. I know their admin quite well, he's personally using one of them.

I haven't used my scope in years, it probably doesn't even work anymore (older tektronix). The trusty oscope came in handy for old CRT television troubleshooting and repair. However in the PC field, I've found that in the world of polymodding, trial & error and stress testing utilities were just as accurate (if not more so) of indicators than a scope. I'm certainly not saying you're wrong, but the real-world testing hasn't revealed any problems by using this mod.

[Pyr0Beast]

Re: SX280 full poly

2 years is a fair trial. It would break much sooner, even in a day or so if something would be horribly wrong.

However, I would still advise against it unless absolutely necessary. I have found out in many situations, fiddling with the design can make it worse, much worse, like sticking low esr caps in PSU's. Some live just fine, but some whine horribly about it. Others don't even start.

[kc8adu]

Re: SX280 full poly

330@16 sanyo sep in every one i do.
and mine come from the worst possible environment.factories and machine shops.
scope check shows everything perfectly clean.
one of my first 280 poly mods is at a neighbors place.they run seti@home on it and never turn it off.2 years and counting.been rebooted 4 times in 2 years for updates.i have 60 280 boards going out today and that customer has over 200 that i polymodded.the mod is well proven.