I generally don't buy anything but Panasonic branded capacitors. If Panasonic doesn't make it, I usually source from other Japanese companies. In one case I landed up using Vishay branded axial capacitors.

At this point, I don't even trust Rubycon to build a decent capacitor as I've seen way too many of them fail prematurely. It also doesn't help that some manufacturers use 85C rated parts.

Old Thrashbarg wrote:

My mistake, I read 'Capxon' instead of 'Samxon'. Still prefer the Rubys anytime.

After replacing enough bad ones in monitors and modems/routers, I really don't trust Samxon in the least bit either. I know the lot over at badcaps.net is enamored with 'em, but I fully expect that's gonna come back to bite them in the ass eventually (not that you'd ever hear about it, though... there don't seem to be many in that community who'd ever admit they were wrong).

Well I've never seen a bad Samxon cap. In fact, I've never seen a Samxon cap on anything in my entire life until I started using them for recapping.

What do you suggest is used instead? All you offer is criticism.

You have never seen a bad one, therefore they couldn't possibly be bad. Right then.

And what do I suggest instead? How about any of the numerous ones I just fucking listed in an earlier post in this very thread? If you would bother to read, you would see that I have offered much more than just criticism. Unless by 'criticism' you mean, 'doesn't 100% agree with everything you think'.

Half-Saint wrote:

... What do you suggest is used instead? All you offer is criticism.

My apologies, that wasn't my intention 😀 .

A quote from Phaedrus, well known character from the JonnyGURU forum:

"Why japanese electrolitycs are the best: Japanese electrolytics are still cheaper than most polymer caps, and polymer caps aren […]
Show full quote

"Why japanese electrolitycs are the best: Japanese electrolytics are still cheaper than most polymer caps, and polymer caps aren't available in as high of capacitance/voltage as electrolytics are, for instance you can't get polymer caps that will work as primary caps for an ATX PSU.

Tantalum > Solid > Japanese Electrolytic > Taiwanese Electrolytic >>>>> Chinese Electrolytic

In addition to the whole electrolyte formula fiasco, the Japanese caps are built to tighter tolerances than the Taiwanese and Chinese caps... Taiwan caps are built to higher standards than mainland Chinese caps, and many of the Chinese caps on the low end (Fuhjyyu, Canicon, ChengX) have false specifications (inflated capacitance, exaggerated ripple tolerance, etc.)."

If I can choose, I prefer lytics from one of these jap brands: Fujitsu, Hitachi, Nichicon, Panasonic/Matsushita, Rubycon, Sanyo, United Chemicon.

This list helps to recognize caps from the logo and the vent cuts pattern: http://capacitor.web.fc2.com/

Panasonics and Rubycons onboard a recapped PSU:

Hm, a little input from me.

As an owner of two notorious for their bad caps boards - the ECS K7S5A (which had two recaps up till now), and an Epox 8rda3+, my opinion is:

Never use desoldered (from another board) capacitors. (welll. if it's something expensive, or you cant find it ... but it's another story)

Aways get japanese capacitors for replacement. It doesn't matter (for me) if it's a Panasonic, Nichicon, Rubycon or anything else (japanese), but keep in mind the following characteristics:
- 105C
- Low ESR
- Life hours (they do differ for different series)
- Size !!! 😀 (diameter)

You could make compromises with capacitance and voltage - I usually replace 6.3V with higher voltage, cause I can't find 6.3 ones 😀 And I usually try to keep the capacitance, but when I recap my soundcards, for example, I put as big (size and thus capacitance) as possible. Don't know what's it in theory, but slightly bigger capacitance won't hurt the board (I tried on that K7S5A, when I hoped for more stable overclock - it's still working).

Get a dentist probe or something similar - stainless steel. Sometimes you can't vacuum all the soldering material from the hole (or probably my dealextreme's soldering vacuum pump is crappy). But you can heat, then gently put the dentist probe in (heat from the one side, put the probe from the opposite side), and when it's stainless steel, it won't get soldered - you can twist it, pull it out and have a nice little hole. Be carefull, though, cause there are tiny little copper rings in that holes - if you tear them out, you have to solder the capacitor straight on the line of the board (you have to clean the polish first).

Ah, and I'm a rude guy, so I desolder at max temp of my iron (450C) - haven't killed anything yet. I aim to heat the legs (and arround) of the capacitors. Do not use force (because of the tiny little copper rings) to pull them out - heat more, pull out gently.

Well, I think I already said too much, hope nobody gets bored 😀

Old Thrashbarg wrote:

You have never seen a bad one, therefore they couldn't possibly be bad. Right then.

Not only have I never seen a bad one... I've never actually seen them used by a motherboard manufacturer at all.

Old Thrashbarg wrote:

And what do I suggest instead? How about any of the numerous ones I just fucking listed in an earlier post in this very thread? If you would bother to read, you would see that I have offered much more than just criticism. Unless by 'criticism' you mean, 'doesn't 100% agree with everything you think'.

I just happen to have read the entire thread but I don't neccessarily check who's the author of each and every post as I read it. Sorry about that.

I went back and re-read your post with the recommendations. I've seen my share of bad Nichions and Chemicons although I couldn't tell you the exact type. Do you see me trashing the brand because of that?

Got an AOpen motherboard today. Slot 1 with a ton of large caps around the CPU area. They are all fine. Brand is Sanyo, I take it that's also Japanese.

I looked at the websites that Old Thrashbarg posted, thank you by the way, and I'll go with Panasonic because Element14 has most of their caps from that company. I like Element14 because it has a really good filter search that allows you to filter for things such as temperature, life expectancy, dimensions. And free shipping for orders over $40 or so.

Btw, Mau1wurf1977,

I remember you did admire those "silver" ones capacitors a few posts earlier. I found the photo of my main rig videocard 😀

I was in the next room, when two of them popped. I though it's a car engine exhaust somewhere outside.

The last one popped in my legs, and the BANG! was so hard, that I almost shit my pants 😀. The card was working, though - but with crashes every now and then ...

I replaced them with cheap HITANO Low ESR's (taiwanese), 'cause they were available in the neighborhood, and ... hm, that card runs so hot (the passive cooling runs it even hotter), that I expect to replace capacitors once more before I finally retire it.

What I mean is: these "full metal jacket" (I like guns 😀 ) capacitors are not as good, as their marketing is.

I see...

Here is an image of the large caps on the AOpen Slot1 BX440 motherboard. I remember Anandtech mentioning that they are japanese and of good quality.

http://www.anandtech.com/show/225

Quote:

The prominent green heatsink on the 440BX chipset proudly bares the AOpen logo as it blends in with the green of the 11 large, and 4 mid-sized Sanyo tantalum capacitors. The capacitors are concentrated primarily around the slot-1 interface, although they are also dispersed in the area between the CPU slot and the chipset, and between the chipset and DRAM banks. AOpen's extreme attention to detail from an engineer's perspective is well worth it as the stability of the board (which you will learn later) is top notch, and brings the AX6BC to a point above and beyond that of the competition, even that of the award winning BH6.

When you read this and they mention stability, makes me wonder. Was stability a feature? These days it's expected 😀

What about tanatalum caps? I hear them being mentioned for video cards. I think Anandtech got it wrong and they are standard electrolyte caps.

Coming back to the board with faulty caps. I noticed there are heaps of empty "spots" for caps:

Considering this, should go with larger ratings for the caps that I'm replacing?

A member on overclockers australia has written up this fantastic guide!

http://forums.overclockers.com.au/showthread.php?t=1102189

Empty cap spots are normal, manufacturers change the hardware revision but still use the same pcb. So some caps are not used, others are changed with other values and so on...
Personally I'd just use the same capacitance caps as the original design and the same number of caps on the same spots... 😀 no need to go brave on such stuff 😊

Also that "dental" tool trick is more trouble than it should IMHO, desoldering wick works wonders in this situation

Space Cowboy wrote:

... that card runs so hot (the passive cooling runs it even hotter), that I expect to replace capacitors once more before I finally retire it ...

I'd ghetto-rig a decent fan and airflow the heck out of card and caps till they smile => end of problem 😈 .

Mau1wurf1977 wrote:

Coming back to the board with faulty caps. I noticed there are heaps of empty "spots" for caps: […]
Show full quote

Coming back to the board with faulty caps. I noticed there are heaps of empty "spots" for caps:

Maker penny-pinching. You'll find the empty spots are in parallel with the inmediate phisically present neighbours, for instance the empties at the top of the pic are in parallel with the present 6.3V vcore output smoothing caps. It you fill those blanks you'll double the smoothing capacitance, which can only help stability and redundancy.

Mau1wurf1977 wrote:

Considering this, should go with larger ratings for the caps that I'm replacing?

At the risk of openind the Pandora's box 😀 , I'd say yes for input/output smoothing.

While you're at it, consider placing Al or Cu heatsinks on the VRMs:

That would ease thermal stress both from the VRMs and the surrounding caps.

Mau1wurf1977 wrote:

A member on overclockers australia has written up this fantastic guide!

http://forums.overclockers.com.au/showthread.php?t=1102189

Some remarks about that tutorial:

- He's using a soldering gun. These things have no temperature control, and can easily fry heat sensitive components (like... electrolytic caps). A good temperature range for soldering is about 300-350°C. For lead-free solder you'll have to turn it up to 400°C or so. Soldering guns and cheap irons usually carry some voltage on the tip, which can, again, damage sensitive components.

- Resin/flux is sometimes based on hydrochloric acid, which is highly corrosive! It can eat away copper traces and your soldering iron's tip. Good 60/40 solder for electronics always has a resin core, so using extra resin should not be necessary in the first place.

- His dental pick method is dangerous and unnecessary. To open a closed hole, the easiest way is to hold the circuit board upright (vertically), put the solder pump on the component side, and heat up the solder from the other side. To create a good vacuum you can put some silicone grease on the vacuum pump tip.

That said, you can get some good results following this tutorial, but it's really easy to mess it up this way.

I've seen my share of bad Nichions and Chemicons although I couldn't tell you the exact type. Do you see me trashing the brand because of that?

I just call it like I see it, and what I've seen is that most of the Samxon caps I've encountered have been bad. That covers multiple different series (GF being the most troublesome and most commonly encountered, but also GD, RS, LP, and KM, with the LP being primarily in audio equipment and the KM being in LCD TV/monitor power supplies) and over a fair range of date codes. Granted, I haven't seen all that many of them, but percentage-wise, it's enough of an issue for me to mistrust the brand.

True, Nichicon and Chemicon have had some problems, but with those it was just a couple specific series. With Nichicon, it was the HM and HN series, and it was only for a short period of time... they corrected the issues around 2006. For Chemicon it was the KZG/KZH, and I still stay away from those particular ones as they haven't proven themselves to me yet (I think the primary problem is that they're overly sensitive to heat), but the rest of their lineup has been reliable. Also note that all of those are super-low ESR types, whereas my issues with Samxon include other types... low-ESR caps are a lot harder to get right, so I can somewhat forgive issues there (not completely, mind you), but when a company fucks up a general-purpose capacitor, something is seriously wrong.

Edit: It's probably also worth pointing out that the Samxons do not tend to bulge when they go bad, so issues with them are not always readily apparent.

Mau1wurf1977 wrote:

Got an AOpen motherboard today. Slot 1 with a ton of large caps around the CPU area. They are all fine. Brand is Sanyo, I take it that's also Japanese.

Which AOpen board? I used to have an AOpen AX6Bplus, and miss it dearly. It died one day in 2005 and I landed up tossing it. I don't recall any visibly budging or leaking capacitors on it. I loved that board because it had 3 ISA slots, 4 DIMM slots (could take 1GB RAM), built in UltraWide SCSI, ridiculously high CPU multiplier (something like 10x), and could run at 133FSB.

Mau1wurf1977 wrote:

... the AOpen Slot1 BX440 motherboard. I remember Anandtech mentioning that they are japanese and of good quality ... ... What about tanatalum caps? I hear them being mentioned for video cards. I think Anandtech got it wrong and they are standard electrolyte caps.

I only see lytics on AOpen AX6BC mobos. This example has Rubys all around the vcore (click to enlarge):

In this other only the small one is a Ruby, the blue larger ones could be Chemicons:

. . .

NJRoadfan wrote:

Mau1wurf1977 wrote:

Which AOpen board? I used to have an AOpen AX6Bplus, and miss it dearly. It died one day in 2005 and I landed up tossing it. I don't recall any visibly budging or leaking capacitors on it. I loved that board because it had 3 ISA slots, 4 DIMM slots (could take 1GB RAM), built in UltraWide SCSI, ridiculously high CPU multiplier (something like 10x), and could run at 133FSB.

AOpen had a few version of this board. The Plus like you mentioned. There was also a II and a Millennium.

Here some images of the board:

Full resolution: http://i.imgur.com/OffHOHm.jpg

TELVM wrote:

I only see lytics on AOpen AX6BC mobos.

I'm getting another two of these and will let you know what caps they come with. On this board at least they are all from Sanyo.

^ The small black one by the coil in your pic is a Ruby, there is a subtle difference in the legs of the 'K' vent pattern with Sanyos. These AOpen boards look very good.

Ok I checked the AOpen website and they had these boards / revisions:

AX6BC (Review http://www.anandtech.com/show/225)
AX6BZ EZ (440ZX)
AX6BC Pro
AX6BC Pro Gold (review http://www.anandtech.com/show/334)
AX6BC Pro II
AX6BC Pro II Millennium (review http://www.anandtech.com/show/433)

Careful. The Pro II and Pro II Millennium have NO ISA slots

TELVM wrote:

^ The small black one by the coil in your pic is a Ruby, there is a subtle difference in the legs of the 'K' vent pattern with Sanyos. These AOpen boards look very good.

Yea from my research AOpen boards stand out for:

- Weird floppy port location behind the CPU
- Super stable