How long we should test Samxon caps?

[yanz]

You may already know that our fellow, Big Pope, or I rather comfort to call his original name, Joe, (please don't edit his email address, Topcat, read my entire post), is supplying what he call the "king of capacitor".

The spec of his caps is pretty good, although that's on paper. I will not saying that the reality is different, indeed his caps is excellent on duty as some of us have already try it for - still - a short periodic. (Me, trodas, davmax, chris1992, and other that I can't remember, it seems I late in rejoining this forum),

I think Topcat is right that Joe too aggressive in promoting his Samxon on this forum board, because....

- Samxon could be an excellent product that only works for short period, example 2-3 year, and that would be disaster. Building a brand image is so hard\difficult. No shortcut. Testing reliability in a long term for one product is something that only an expert (or they who care) can do. You may research a product in year A, and marketed it in year B. You promise that it will last to year E. I would call that bullshit (for al electrolyte caps, though). Nothing wrong with the marketing/data sheet words, I would trust that product only for 75% (of my trust) if it made by good known brand like Chemi-Con, Rubycon, Sanyo and else. But my trust for Samxon only 50% even though I really satisfied with Samxon caps (if it used for overclocking project, the result is similar/slightly better than using Rubycon MBZ/Sanyo WG <- only my experiences for short periods).

BUT, I like if there are a new challenger beside of Japan manufacturer, add to that a cheap price (although I can get Chemicon KZG or Rubycon MBZ for really really cheap in where I live).

I really disagree with what Topcat do of banning Joe from this Forum. Although I think that's because he (Topcat) can't successfully advice Joe to stop his promoting of Samxon caps.

My opinion is:

- Let the world -- at least the people who visiting badcaps forum -- know that Samxon caps exist; the performance is excellent, the reliability is unknown. (Take note to the word I bold).

- Let people, who understand that, and may want to try Samxon caps, to know where/who he can get Samxon caps from. In this matter, it is Joe (censored_temporarily at gmail.com).

- We try Samxon. We see how it does. We tell what the reality based on our experiences. But we shouldn't take Samxon to a level degree of Japanese brand/manufacturer. The time is not enough to tell. For us, overclockers/enthusiast, we have no problem (we know the risk) if our hardware broken after one or several year of trial using Samxon. But what will happen after some period of time if someone who visits the forum now tries Samxon in his equipment which is use for task in field like medical, aeronautical, or other critical mission? I am a man whose job is in science field, that's the way of my thinking: base everything on white and black, although my love is often on the grey area (). Hell, for oc, performance for only one two years is not bad at all. But this is only my opinion.


I think this is what I can seriously think about this matter. I still a bit feeling fuzzy after driving about 200km, and then watching Football World Cup matches and still too much adrenalin in my blood to get me sleep. I hope I made my points and everyone understand it. I'm not meaning to offense anyone, especially Topcat, yes I defense Joe but that merely because I interested in Samxon caps, and I will be disappointed if we 100% banned it (Samxon caps) from this forum. English is not my mother tongue so please have patience if something looks offensive I can guarantee that it is not my meaning.

I expect everyone here to throw one - two opinions about this.

Thanks.

[PeteS in CA]

Well, from a databook perspective, one year is almost 9000 hours. Thus a PW or LXZ is guaranteed to last nearly a year at maximum ripple current and temperature; an HE or KY for somewhat more than a year at those conditions. Given that MCZ, HV, KZG, etc. at 2000-hour-rated parts, one hopes that they are not being used at maximum current and temp.

A more practical approach would be to think in terms of the timeframe in which the defective HMs and HNs, and crap-caps usually fail. E.G., if they fail after 6 months of running 24x7, then running 24x7x52 w/o failure or sign of impending failure (i.e. bulging) could be reasonably interpreted as passing the test.

[Topcat]

Once again, "joe" aka bigpope was not banned because of Samxon caps, he was banned because of his immature and childish attitude. I did edit out his email address AGAIN. There are many other roads to Samxon than him, I'll find one. Do not post his email again. I'm so sick of that moron, I almost delted the entire thread, but being it has some technical merit, I changed my mind. Bigpoop is banned from here, NOT Samxon. I wish he would give you the real and true story before you start crying for him. This is not going to turn into a boo-hoo, bigpope got banned thread either. Of the 1500 members on this board, bigpope (and his 40 duplicate names) is the ONLY active member to ever be banned from this forum!! That should tell you somehting right there.

I would like to see some hardcore testing spanning between 1~2 years, with ESR and ripples posted atleast once a month, so their deterioration can be recorded. I'd still love to get my hands on enough samples to recap an Abit KT7 with them. I have 3 of these boards. I would like to recap one in Samxon, one in Rubycon MBZ, and another in Nichicon PM's, and set them up in identical configs and conditions and see how each fares over an extended period of time.

[kc8adu]

i got some samples from bigpope 5 months ago.i stuffed a evilmachines imperial gl with them.
running seti at 100%load there was 17mv p-p on vcore.
checked it yesterday while adding some more ram and it was 135mv p-p
i told the owner to call me at the first sign of trouble.they are a willing participant in this test.
will look again in another month or 2 or at the first sign of trouble.

[davmax]

KC8ADU. That is certainly a big increase in ripple. It would be ideal if all conditions remain the same. What caps are in the power supply? This can have an impact.

Need to measure ripple into VRM and output. Very interested.

I will now try to do same and post results. Others please try also.

[davmax]

Well I have created some test points for VRM In & Out

Test conditions. C: in defrag operation. Largest excursions are when disk accessing.

Results. VRM In 157mVpeak worst case.
VRM Out 27mVpeak worst case.

Equipment.Velleman 5 MHz bandwidth digital battery scope with peak detection.

See my rig below. Samxon caps 3 months operation typically 8 hrs/day. Cap temp 45 -50C

Lets see results from others.

[Topcat]

KC8's results are a bit troubling, but like davmax said, that could be attributed to the PSU. I'm sure he'll fill in the blanks when he can take a look.

If anyone has some spare Samxon's around (enough to do a full recap on a KT7), let me know. I'd like to set up my test with the 3 KT7's I have... It would be Samxon versus Rubycon versus Nichicon, in an identical and controlled environment.

[davmax]

How about my results. They look good without any extra added ceramics.. Power supply has good caps too. I will now keep track about every three months.

[Topcat]

Originally Posted by davmax

How about my results. They look good without any extra added ceramics.. Power supply has good caps too. I will now keep track about every three months.

So far so good. Try something more CPU-intensive, defrag really isnt, its just HDD intensive, which doesn't really tax the motherboard. Try it running SETI or F@H 24/7 and note the results, it would be more accurate as to their long-term lifespan. It's also harder on them if they're allowed to run hot. I'm not saying to overheat them, any cap will fail if it overheats, but if your case has a dozen fans in it, and its sitting next to the air conditioner, thats not too accurate either. That's why I want a more controlled testing environment, to really see how they hold up.

[davmax]

Sorry please expand on F&H. I am not doing acclerated testing which would be great for 24/7 guys.
Yes defrag is disk intensive but I checked to see what is causing high ripple peaks, rather than running cpu intensive. This should give a good indication of working ESR over time. Scope bandwidth is limited for the very short pulses, but I feel that looking for the larger disturbances and keeping all conditions constant will be revealing enough.

[davmax]

Just wanted add that the 5 Mhz bandwidth on the scope actually covers the frequency spectrum where the electros are effective, what we are testing. The higher frequencies are handled by the ceramics and they are not at issue.

[davmax]

TC just to check I ran at 100% CPU and checked ripple on VRM Out. Worst value was 34mV p-p with defrag 33mV p-p. (after all there are six 3300u Samxon GC, a lot of bulk capacitance and 9mohm/6 =1.5mohm ESR). Notice that I am now quoting peak to peak ripple. Originally I quoted Vp, this was a misquote, has been Vp-p all along after checking scope manual.

My conditions are very controlled for relative ESR with time, the conditions you quote are understandably focused on stress.

[willawake]

please don't edit his email address, Topcat, read my entire post

hmm is that what its all about then? getting his email address in the forum...what is the motivation? free caps in exchange for promotion?

[davmax]

Willlawake. That seems a very unfair view. Others are asking for some people to have athe chance try Samxon, as the post states. The only known source right now is Big Pope. He is not pushing this issue right now, others are. Suspicion and exaggeration is not a good way to go.

[Topcat]

Lets keep on topic. As stated before, bigpope has the entire internet to plaster his email address on, it won't be plastered here. If he wants to send me some samples to test with, great (but very unlikely). If not, I'll get them elsewhere. I will give Samxon a very thorough and very unbiased review. My beef is NOT with Samxon.

[kc8adu]

the psu is a fortron 250 with all panasonic fm.the bestec was burned up and was used as a rifle target.
ripple is in the single mv.
cooling is slightly inproved with a added fan in a blank grille at the rear of the case.
otherwise just a common crap evilmachines .

[PeteS in CA]

... nother in Nichicon PM's ...

Is this a typo? The PM series is between 15 and 20 years old (it's the lead-free version of the PL series). PW is the current low-Z, non-aqueous part from Nichicon; HE is the long-life aqueous series, and HN & HZ are the ultra-low Z, aqueous, 2000-hour series.

[yanz]

Originally Posted by Topcat

Once again, "joe" aka bigpope was not banned because of Samxon caps, he was banned because of his immature and childish attitude. I did edit out his email address AGAIN. There are many other roads to Samxon than him, I'll find one. Do not post his email again. I'm so sick of that moron, I almost delted the entire thread, but being it has some technical merit, I changed my mind. Bigpoop is banned from here, NOT Samxon. I wish he would give you the real and true story before you start crying for him. This is not going to turn into a boo-hoo, bigpope got banned thread either. Of the 1500 members on this board, bigpope (and his 40 duplicate names) is the ONLY active member to ever be banned from this forum!! That should tell you somehting right there.

I would like to see some hardcore testing spanning between 1~2 years, with ESR and ripples posted atleast once a month, so their deterioration can be recorded. I'd still love to get my hands on enough samples to recap an Abit KT7 with them. I have 3 of these boards. I would like to recap one in Samxon, one in Rubycon MBZ, and another in Nichicon PM's, and set them up in identical configs and conditions and see how each fares over an extended period of time.

The thread is going like what I expected. Thanks guys for your attention and effort in testing Samxon caps. I want to participate but I have no such equipments.


Topcat, I believe that Joe will be glad to send you some samples. Aside from your personal problem between you and him, I guess we should have an achievement. That is to find out what is the true quality of Samxon caps. If you want, I will tell Joe that we are conducting a test for Samxon caps and you need some samples for your KT7.

I guess it can't be helped if that email address isn't allowed, I will pointing out to XtremeSystems Forums then for people who searching Joe (Big Pope).

[yanz]

Originally Posted by willawake

hmm is that what its all about then? getting his email address in the forum...what is the motivation? free caps in exchange for promotion?

No, Will...entirely no. I have no agreement whatsoever with Joe to promote his caps. He sends samples to everyone who asks him, why I should help him that far. My reason is exactly like what davmax said. Beside, I'm not trying to promote Samxon caps, I even criticize it -> It's not easy for us to recommend Samxon caps yet we still examine/monitoring its reliability/endurance, and it's not wise to let the forum reader think that Samxon caps are as good as Rubycon, Sanyo, Panasonic, Chemicon, or Nichicon.

If you are still prejudiced me, than you're not like Willawake I know. But I think you just ask the question out of curiosity. I guess that's unavoidable, someone has to ask it. So, thanks..

[Topcat]

Originally Posted by yanz

Topcat, I believe that Joe will be glad to send you some samples. Aside from your personal problem between you and him, I guess we should have an achievement. That is to find out what is the true quality of Samxon caps. If you want, I will tell Joe that we are conducting a test for Samxon caps and you need some samples for your KT7.

I guess it can't be helped if that email address isn't allowed, I will pointing out to XtremeSystems Forums then for people who searching Joe (Big Pope).

If he wants to contribute to the testing, and donate some caps, thats cool, and I will give him full credit for his contribution in the results thread. For folks in the USA, here's the info to get Samxon caps factory direct:

USA
Samxon Electronic Components LLC
1621 A, South Main Street, Milpitas
California 95035, USA
Tel: (408) 956-9738
Fax: (408) 956-1798

Thanks Yanz.