Who makes the Chton caps in my AT PSU?

[Paddan1000]

I was worried that the AT PSU of my irreplaceable 486 contained bad capacitors and I opened it up to take a look. The two big ones were Nippon Chemi-Cons (picture 1) but all the small ones were labeled "Chton" or "Cnton" or something similar (picture 2). I have no idea if these caps are good or bad and I wonder if anyone here could identify them for me. The PSU is made in Taiwan by Ya Hsin.

The PSU also happens to be over 20 years old. Even if all the caps would turn out to be of high quality, should I replace the caps pre-emptively despite none of them showing any outward signs of failure?

The computer is somewhat erratic in that it fails to detect the hard drive if I reboot it more than three times and in that it only accepts two 8 mb sticks of RAM. It becomes unstable if I insert four sticks or use DIMMs of 16 mb or higher. The motherboard has no electrolytics whatsoever, so I figured that perhaps the PSU was at fault...

[mariushm]

Re: Who makes the Chton caps in my AT PSU?

Some small time capacitor manufacturer in an industrial park in China, or some company buying unlabeled capacitors from a chinese manufacturer and labeling them with their brand.

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

Some small time capacitor manufacturer in an industrial park in China, or some company buying unlabeled capacitors from a chinese manufacturer and labeling them with their brand.

I was hoping that much, since it gives me an excuse to start improving, but perhaps they are really labeled "Chcon", as an abbreviation of "Chemi-con"?

[twbranch]

Re: Who makes the Chton caps in my AT PSU?

more than likely to confuse you so you think it is Chemi-con! Replace 'em!

[Heihachi_73]

Re: Who makes the Chton caps in my AT PSU?

Your best bet would be to test them with an ESR/capacitance meter, which if they have failed they most likely will be simply dried out inside rather than the way today's Chinese caps are made (which have a tendency to leak or even explode if left unchecked).

One thing to check though, if it isn't caused by bad caps, are the contacts dirty on the RAM or even the board itself? I had trouble with a 256MB DDR module last year, only to find that the 'gold' contacts on the RAM itself were a dull yellow and wouldn't work in anything until I cleaned it (now it's running fine in a 512MB P4 system).

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

I've ordered new caps now, minus the two big input filter capacitors that may continue to serve. I ordered mostly Rubycon 105 degrees, but I couldn't find any 50V 3.3uF ones. I ordered the 3.3uF as Samxon 85 degrees instead. Will they be fine or are they a weak link that will fail prematurely because of the lower temperature spec and the inferior brand? They are only one or two out of a 17 caps in total.

I've done a lot of experimenting with different RAM-modules in this computer and it's consistent in that it only accepts two 8 MB DIMMs and nothing else. I've never tried a different power supply though.

EDIT:
There are some Rubycon 3.3uF on Ebay that would be perfect, but they are probably counterfeit. The specs of the caps on the picture don't match the specs of the advertised caps. What do you think? Has anyone ever ordered from Goodbuy711?
http://www.ebay.com/itm/20pcs-RUBYCO...item3efdf88b06

There are also a lot of Ebay sellers selling 50V 3.3uF Sanyo caps, but I'm afraid those might be fake too. They are also only speced at 85 degrees so they might not be more reliable than the Samxons.
http://www.ebay.com/itm/4pcs-Sanyo-R...item4aac549b77

Could I use a solid capacitor instead, like the Chinese 50V 3.3uF tantalum capacitors also found on Ebay?
http://www.ebay.com/itm/10-Pcs-x-Tan...item41667d21e7

EDIT 2: I did some googling and yes, Goodbuy711 definitely sell fakes.

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

I've replaced all the caps in the AT PSU now with Rubycons except the primary filter capacitors that were Japanese to begin with.
The computer works just as well as before, but when I measure the voltages they are way out of spec. The +5 rail is 6.45 volts and the +12 rail is 13.44 volts. Before the cap change they were 5.10 and 12.30 volts respectively.
Is this because the new caps could have a different ESR rating than the old caps and how will this affect the computer in the long run? I don't know the AT specifications, but according to the ATX specifications the voltages can't deviate more than 5% when the computer is idle.

[lti]

Re: Who makes the Chton caps in my AT PSU?

I think stuff would have fried if the voltages were that high. Make sure your multimeter is working properly. Some of the cheap ones give unusually high voltage readings when the battery is low.

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

I think stuff would have fried if the voltages were that high. Make sure your multimeter is working properly. Some of the cheap ones give unusually high voltage readings when the battery is low.

You were right about the multimeter!
It's a good quality one, but about 30 to 40 years old and therefore a bit worn. It appeared that the "diode test" button was jammed which added 1,5 volts to all readings.
I managed to unjam the button and get new readings. The new voltages are 5.01 and 12.07 volts, which makes the PSU better than new!
I think I'm going to give it a new Noctua fan while I'm at it.

[Th3_uN1Qu3]

Re: Who makes the Chton caps in my AT PSU?

Just out of curiosity, what do you use that system for? I used to have an overclocked 5x86 at 150MHz with 24MB RAM, lying around just for kicks. I gave that away 3 or 4 years ago IIRC.

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

Just out of curiosity, what do you use that system for? I used to have an overclocked 5x86 at 150MHz with 24MB RAM, lying around just for kicks. I gave that away 3 or 4 years ago IIRC.

I use it mostly for old MS-DOS games. I don't play much anymore, but I like to tweak the computer to be able to play the games at the highest settings. My efforts pay of mostly in the sound department and I have a collection of very rare sound cards and external MIDI modules (LAPC-I, AWE64 Gold, Ensoniq Soundscape Elite, MT-32, SC-55 MkII) that allow me to listen to the in-game music as high quality MIDI instead of regular Adlib FM music.

Here is a recording I made of Monkey Island that shows the difference in sound between my computer and one with a regular sound card:
http://www.youtube.com/watch?v=a6aUs2SW0aE

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

My Youtube link seems broken and it's too late to edit the post. Here is a working one:
http://www.youtube.com/watch?v=a6aUs2SW0aE

Anyway, I got access to a capacitance meter today and I tested some of the 20 years old Chton caps. Their capacitance of all the caps I tested were actually 10-15% higher than their specified rating (The 2200uF were 2400uF and the 470uF were 540uF). Is this a normal sign of aging, that the capacitance increases somewhat, or are these caps actually quite good in that they after 20 years are still in the "breaking in"-phase?

I also tested some 1000uF Sanyo that I removed from a motherboard of about 10-15 years of age and their capacitance was at about 850uF.

[370forlife]

Re: Who makes the Chton caps in my AT PSU?

Most of the time when caps go bad they start to loose capacitance. I'm not sure if it is possible, but they could have retained their capacitance but their ESR may have increased causing excessive ripple.

AT power supplies had much more loose Voltage regulation and Ripple specifications than ATX. I think something like 250mv of ripple on the 12V and 5V rails, and +/- 10% regulation versus 120mv ripple on 12v, 50mv on 5v, 3.3v and 5vsb with +/- 5% regulation on all rails except negative voltages.

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

Oh no! I believe the 2200uF Rubycons I used are fake. I had to buy them separately since they weren't included in the Rubycon set I used for the others.

Still, the supposed fake ones are pretty consistent and the 8 leftovers are showing capacitance values between 2125 and 2175uF and ESR values between 0,05 and 0,02. The capacitance meter I'm using is a cheap chinese one, so these values are probably not very accuate but can be used as an indication of the state of the caps.

Here are some pictures. What do you think? Their dimensions are 12,5x20mm.
If they indeed are counterfeit, will it be possible to identify the "OEM"?

[mariushm]

Re: Who makes the Chton caps in my AT PSU?

They look OK to me, what makes you think they're fake?

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

They look OK to me, what makes you think they're fake?

1) The K-shaped vent is more like a Sanyo vent than a Rubycon vent.
2) The bung looks like an ELNA-bung and not a Rubycon bung.
4) The negative stripe on the sleeve doesn't line up perfectly with the negative lead on some of the caps.
3) The 10V 2200uF ZLH should have the dimensions 10x23mm and not 12,5x20mm according to the data sheet.

Still, no explosions from the PSU so far, the voltage is stable and all the caps were within specifications and deviated at most 2% from each other. If they are fakes, perhaps the cap-pirate made the effort to replicate the properties of the real deal and not just put a ZLH sleeve on any random general purpose cap?

[Heihachi_73]

Re: Who makes the Chton caps in my AT PSU?

The thing is, even counterfeit caps will have a "close enough" rating when brand new, if they didn't you would be able to tell and/or they would cause issues from the start. After all, Chinese PSUs get away with using their own generic brands (names which can't even be identified online, such as 'Wisdom', 'Forever' or two-letter names like DG, TL, ZW, S.I.), as well as "Rulycon" knockoffs.

Also, the T1203 looks like a Nichicon date stamp too. All Rubycon caps I've come across have a code like 1M0123 or similar (although I mainly have older general purpose caps from TVs and the like).

I suppose that's what happens when you try your luck with Fleabay, especially buying direct from China.

[mariushm]

Re: Who makes the Chton caps in my AT PSU?

1) The K-shaped vent is more like a Sanyo vent than a Rubycon vent.
2) The bung looks like an ELNA-bung and not a Rubycon bung.
4) The negative stripe on the sleeve doesn't line up perfectly with the negative lead on some of the caps.
3) The 10V 2200uF ZLH should have the dimensions 10x23mm and not 12,5x20mm according to the data sheet.

Still, no explosions from the PSU so far, the voltage is stable and all the caps were within specifications and deviated at most 2% from each other. If they are fakes, perhaps the cap-pirate made the effort to replicate the properties of the real deal and not just put a ZLH sleeve on any random general purpose cap?

Aye, I guess that's what 18 hours of no sleep does to you. Didn't even realize the vent is different.

Checked the bung, looked like a more professional than the one fake caps used, saw a K instead of that Sanyo specific stuff, the font was ok and the printing seemed clear...

[Paddan1000]

Re: Who makes the Chton caps in my AT PSU?

The thing is, even counterfeit caps will have a "close enough" rating when brand new, if they didn't you would be able to tell and/or they would cause issues from the start. After all, Chinese PSUs get away with using their own generic brands (names which can't even be identified online, such as 'Wisdom', 'Forever' or two-letter names like DG, TL, ZW, S.I.), as well as "Rulycon" knockoffs.

Also, the T1203 looks like a Nichicon date stamp too. All Rubycon caps I've come across have a code like 1M0123 or similar (although I mainly have older general purpose caps from TVs and the like).

I suppose that's what happens when you try your luck with Fleabay, especially buying direct from China.

It's like they took one property from each japanese brand cap and merged them into the perfect cap; Sanyo vent, ELNA bung, Rubycon sleeve and Panasonic date code. Let's hope they stole the electrolytic fluid from Nichicon...

As for my source, it's not like I have any alternatives if I don't want to pay $10 shipping for 2 caps. The other 15 caps I ordered were genuine Rubycon, so I just have to be more attentive in the future and weed out the occasional fake ones.

Checked the bung, looked like a more professional than the one fake caps used, saw a K instead of that Sanyo specific stuff, the font was ok and the printing seemed clear...

I did the same mistake at first. I saw the vent was not X-shaped, that the bung was not a bull's eye and assumed that meant the cap was genuine.
Still, I wonder why the counterfeiters never make perfect copies but always leave these little clues?

I did an experiment where I charged and discharged the three of the caps a couple of times with a battery to break them in. Now their measured capacitances are more or less identical and repeated readings give me values between 2159 and 2199. Not bad for a badcap.

I will leave the 2 fake caps in the power supply for a while and see what happens. I'm not in a hurry to replace them if they are working. If I notice that the voltage changes or that the caps start bulging I will remove them.