Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

[Behemot]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Never heard of SF before so this is first time I see it at all

[manthi]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

it's not green, it's black
serie sf, date, i don't know, monitor was built in 01 2008
working hours: well let's say 5 hours a day for 7 years
t°: normal operating temperature for a monitor in a living room

[Wester547]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

I'm posting this long list of quotes but have added an addition I found from Big Pope where he claims to have asked Samxon about this and their reply is that they never made a "GL" brand:

Thanks for that quote. I've seen that thread before, I don't know how I missed that. The only conclusion I can safely come to is that Samxon GLs are probably resleeved Samxon GFs from an unauthorized distributor. After all, some have even been desperate enough to resleeve 470uF 200V Teapo SKs as 680uF 200V Teapo "SWs"... I think this happens with many unauthorized distributors. They take many capacitors from crap manufactures, resleeve them in colorful and shiny sleeves and maybe add some varnish to the cans, etc... maybe trying to impress the idea upon people that all capacitors are created equal (obviously not true).

[ReeceyBurger123]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Series, manufacturing date, working hours and temperature?

Its a green Samxon cap.... So its either a Gf or Gk would surprise me.

[Per Hansson]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Per Hansson: I think they have their own aluminium foils and bought the electrolyte. Just look at the GC, GD - they have 100% same specs as Nichicon caps. Everybody can make aluminium foils (though not everybody pure enough) but the elctrolyte is what chinese makers have been failing at.

To clarify again what I mean is where they source the aluminium, it makes a big difference if it's "regular grade" bought in mainland China.
Compared to "regular grade" bought in Japan.
That is to say in my own opinion a large reason for so many crap caps existing is that they don't have good control over the quality of the aluminium they purcache.
And just for what it's worth Samxon owns two aluminium foil factories:
http://www.manyue.com/facilities.html#aluminum

My stock of Samxon caps is old, and the caps I buy nowdays are mostly from Elfa Sweden and of the major Japanese brands.

That's... odd. Not that I'm doubting your skills, but rather the luck you've had with the Samxons.

You're getting your stock from electronic component houses, like we do here with Mouser, Digikey, etc, yet those caps have been fine for you. That's what I'm puzzled by. With the Samxon failures and "smiliarities" they share with crapxon, I almost suspect your Samxons are not the same as the other "samxons" myself and others have seen.

Could there both be *actual* Samxon caps (the ones you use), and really good fakes from someone like Crapxon?

That's what I meant by not being sure of what you're getting...

The Samxon caps I bought have come directly from the Samxon factory.
Because back then I bought them from Big Pope who made some orders directly from the factory and sold them at very good prices compared to what I could get elsewhere.
Nowdays things have changed, back then the only "motherboard grade" capacitors Elfa stocked where a very small selection of Sanyo WX caps.
So in several of my older repairs it's been a mix of Sanyo (when I needed to complement my stock very quickly) or Samxon when I could wait a bit for the shipping from Hong Kong.
Nowdays Elfa has expanded their selection of low ESR capacitors significantly and I can buy most of the Japanese brands directly from them at good prices.
So that is what I do, just to clarify Elfa has never carried Samxon caps.

As for the possibility of fake Samxons yes that is certainly so, as I mentioned in my previous post and also provided a link to an old post with some examples.
Pretty much all of the large brands have copycats that try to ride on their success and to me the first time this happened for Samxon actually made me smile in a way.
Because it meant that the brand had gotten such a recognition that it was worth to copy it
Here is another example: https://www.badcaps.net/forum/showthread.php?t=3099

I'm posting this long list of quotes but have added an addition I found from Big Pope where he claims to have asked Samxon about this and their reply is that they never made a "GL" brand:

"beside couple bad series as a whole (GF, GK)"

And GL... unless, of course, GL is a counterfeit series.

I see what you did. You are not shooting from the hip by suggesting GLs are fakes-of-fakes.

Actually there do exist several companies that make counterfiet Samxon caps.
Well I guess it could be said it's just a coincidence in the naming and model scheme (Samson has for example a X-CON brand which is Samxon's Polymer series).
But just like "Rulycon" it has to be a pretty big "coincidence"
Take a look here to see what Samxon actually manufactures:
http://www.manyue.com/samxon_series_chart.html

Thank you for your information. I have asked Man Yue directly, they never produce GL series in their histroy, so i trust GL is fake Samxon.

[Behemot]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Series, manufacturing date, working hours and temperature?

[manthi]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

hi guys
i just replaced a samxon 2200uf from an acer x223w monitor

and 3 capxon and 1 teapo from a viewsonic vx2025wm

[kaboom]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Yes, Hermei/Asiacon, Chhssi, Wendell, CapXon, Choyo, Jackcon, Teapos, and pre-Teapo merger G-Luxon are all known to do that. In other news, Fuhjyyu is also known to do that.

You know, that is funny, but antecPimp was trying to tell everyone that "increasing output ripple is normal with age- it means you need a new power supply."

"Suuurre!" We all ran out and bought new antecs, complete with junk-in-the-box fuju-joos, didn't we?

Remember that?

KZGs/KZJs are also known for that, so their electrolyte must be extremely messed up or something really critical snuck past NCC during the QC testing phase.

Except for those series, it's rather unlike them. KZGs/KZJs seem to still be suspect, even recently, yet the HM/HN problem only existed for ~3 years.

How many boards were junked because of "unfixable oddities," when it was really those KZGs/KZJs?

As for why it happens, the crap aluminum is amorphous and the highly aqueous electrolyte without the proper neutralizers and cathode depolarizers will attack and destroy the aluminum oxide film on the shelf, converting the aluminum into hydroxide and causing the resulting formation of hydrogen gas...

See PDFs. Or get some lye and aluminum foil.

Yes, I think the output capacitors would like a more consistent duty cycle. I'm pretty sure the voltage spikes (or fluctuating input voltage) don't help their lifespan. Still, Samxon GF seem to go bad and hatch whether they run toasty or chilly.

Even as I read that, I saw "Capxon GF." Well, what's the difference?

[Wester547]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Real-world use involves many thousands of hours with the electrolyte in contact with the aluminum, as well as power-on-hours. IOW, crapxon may actually test for 4,000 hours in one shot, with minimal failures. But, just as with the wendells, cchissis, hermeis, etc, side-reactions with impurities present will kill a cap, without it even soldered to a board.

Yes, Hermei/Asiacon, Chhssi, Wendell, YEC, CapXon, Choyo, Jackcon, Teapos, and pre-Teapo merger G-Luxon are all known to do that. In other news, Fuhjyyu is also known to do that. KZGs/KZJs are also known for that, so their electrolyte must be extremely messed up or something really critical snuck past NCC during the QC testing phase. As for why it happens, the crap aluminum is amorphous and the highly aqueous electrolyte without the proper neutralizers and cathode depolarizers will attack and destroy the aluminum oxide film on the shelf, converting the aluminum into hydroxide and causing the resulting formation of hydrogen gas...

With transformer/inductor rations "just right," duty cycle can be increased for the same power output; rather than high-amplitude, short duration current pulses (integrated into regulated output voltage),with medium-amplitude, medium duration being preferred. With a PFC booster, this is even more practical; the switcher is fed a regulated DC voltage, thus removing it from the effects of low line voltage. Which actually helps those output caps; the SMPS, by way of nearly constant input voltage, can now run at nearly constant duty cycle. Allowing for losses plus some margin for regulation, we can now increase duty cycle even further- even more ideal for the output caps.

Yes, I think the output capacitors would like a more consistent duty cycle. I'm pretty sure the voltage spikes (or fluctuating input voltage) don't help their lifespan. Still, Samxon GF seem to go bad and hatch whether they run toasty or chilly.

[kaboom]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Well, they are at the very least highly exaggerating about the capability of their caps so as to look reputable. But there is more to it than that - dreck quality aluminum and electrolyte can easily cause outgassing on the shelf, operating temperature, voltage, and ripple current need not be anything but an accelerating factor.

Like those boxes of unused motherboards that all blew their caps?

So I think they also get away with those "endurance" tests because they are only done for thousands of hours at the most, not hundreds, and tens and hundreds of thousands of hours is more than enough time for poor electrolyte to break down by its lonesome.

Real-world use involves many thousands of hours with the electrolyte in contact with the aluminum, as well as power-on-hours. IOW, crapxon may actually test for 4,000 hours in one shot, with minimal failures. But, just as with the wendells, cchissis, hermeis, etc, side-reactions with impurities present will kill a cap, without it even soldered to a board.

Also, I think that LTEC is MUCH better quality than CapXon, as Delta are capable of prolonging the life of LTEC but not CapXon.

With transformer/inductor rations "just right," duty cycle can be increased for the same power output; rather than high-amplitude, short duration current pulses (integrated into regulated output voltage),with medium-amplitude, medium duration being preferred. With a PFC booster, this is even more practical; the switcher is fed a regulated DC voltage, thus removing it from the effects of low line voltage. Which actually helps those output caps; the SMPS, by way of nearly constant input voltage, can now run at nearly constant duty cycle. Allowing for losses plus some margin for regulation, we can now increase duty cycle even further- even more ideal for the output caps.

[Wester547]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

Could there both be *actual* Samxon caps (the ones you use), and really good fakes from someone like Crapxon?

"Fake" Samxons are more likely from Samson (AKA Samcon?). Samxon did release a statement some time back cautioning buyers against fakes. As did Nichicon, Rubycon, and Chemi-con. The matter is more complicated than that, though, because all these Chinese brands are so inconsistent to begin with. They don't do any QC testing and buy whatever's available on the market, and the grade of chinese aluminum and electrolyte will vary from batch to batch. So there is no determining how long crapacitors will last - only that they are not worth the risk. To make matters even more complicated, there always exists the possibility that early HMs and HNs were fake because Nichicon is an very high quality manufacturer. But then what about KZGs and KZJs? We may never know without sending samples to the manufacturers. You would think that OEMs get their capacitors and components from authorized distributors or straight from the manufacturer, but that may not always be true for whatever reason.

And don't forget the bean-counters (*cough* Samscum), who proclaim "both series have the same specs, so use the cheaper ones."

On paper, and only on paper, is that true.

As I've stated previously, Nichicon VRs worked while testing devices, when either waiting for the correct caps or to determine if said device was even worth fixing. I've done this.

Now consider two supposedly equal series- one from capxon/ltec, and the other Nichicon. Supposedly equal, meaning can size, ripple current/ESR, capacitance/voltage, as well as endurance. Taken under the same ambient conditions, the junkers blow after a year (hermeis in LCD monitors), yet the "equalivalent" Nichicons outlast the computer that LCD is connected to. If the were equal, they'd also merely last a year.

The datasheets are LYING.

Well, they are at the very least highly exaggerating about the capability of their caps so as to look reputable. But there is more to it than that - dreck quality aluminum and electrolyte can easily cause outgassing on the shelf, operating temperature, voltage, and ripple current need not be anything but an accelerating factor. So I think they also get away with those "endurance" tests because they are only done for thousands of hours at the most, not hundreds, and tens and hundreds of thousands of hours is more than enough time for poor electrolyte to break down by its lonesome. Also, I think that LTEC is MUCH better quality than CapXon, as Delta are capable of prolonging the life of LTEC but not CapXon.

[kaboom]

Re: Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

This I have to say I take a bit offense at, I have never had a customer come back with a system which I used Samxon's to replace whatever was there before.

That's... odd. Not that I'm doubting your skills, but rather the luck you've had with the Samxons.

You're getting your stock from electronic component houses, like we do here with Mouser, Digikey, etc, yet those caps have been fine for you. That's what I'm puzzled by. With the Samxon failures and "smiliarities" they share with crapxon, I almost suspect your Samxons are not the same as the other "samxons" myself and others have seen.

Could there both be *actual* Samxon caps (the ones you use), and really good fakes from someone like Crapxon?

That's what I meant by not being sure of what you're getting...

You make it sound like if a customer came back with a problem I would not help them because I don't take pride in my work?
I certainly would and I think I can speak for Topcat here as well and say that he also takes the same pride in his work as I do!

I didn't mean to insinuate that you'd refuse to fix call-backs or tough-dogs.

I based that comment on my experiences with samxons. But based on the luck you've been having, it's like there are two different worlds when it comes to samxons.

I've even mistaken samxon for capxon several times, until I saw the entire sleeve; those particular ones are quite bad, label wise.

As for my suggestion- let's say there are legit samxons along with capxon-samxons. That could explain why we both have greatly different "opinions" about them...

I don't think it means Nichicon sold their electrolyte to Samxon. Nichicon's HM series is rated for a different temperature range than Samxon's GD series (-40C to +105C vs -25C to +105C). Samxon does make of note on their website that they are the only non-Japanese company to do true and thorough QC testing, but nothing is mentioned about sourcing their materials from Japan (and I think that would be too expensive for them).

And don't forget the bean-counters (*cough* Samscum), who proclaim "both series have the same specs, so use the cheaper ones."

On paper, and only on paper, is that true.

As I've stated previously, Nichicon VRs worked while testing devices, when either waiting for the correct caps or to determine if said device was even worth fixing. I've done this.

Now consider two supposedly equal series- one from capxon/ltec, and the other Nichicon. Supposedly equal, meaning can size, ripple current/ESR, capacitance/voltage, as well as endurance. Taken under the same ambient conditions, the junkers blow after a year (hermeis in LCD monitors), yet the "equalivalent" Nichicons outlast the computer that LCD is connected to. If the were equal, they'd also merely last a year.

The datasheets are LYING.

[Wester547]

Samxon capacitors

Some boards, don't remember, I don't really post everything I find here for a long time. I wouldn't do anything else you know Some were bloated, more of them had capacity through the roof - internal leakage.

Do you recall the actual date codes?

Yeah but GC/GD are (close to) 100% identical IIRC.

I don't think it means Nichicon sold their electrolyte to Samxon. Nichicon's HM series is rated for a different temperature range than Samxon's GD series (-40C to +105C vs -25C to +105C). Samxon does make of note on their website that they are the only non-Japanese company to do true and thorough QC testing, but nothing is mentioned about sourcing their materials from Japan (and I think that would be too expensive for them).

[Behemot]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

You said before it was bulging in the thread in question.

Yeah than it was bad. A single one I think

Where have you seen HMs and HNs fail after 2006 then?

Some boards, don't remember, I don't really post everything I find here for a long time. I wouldn't do anything else you know Some were bloated, more of them had capacity through the roof - internal leakage.

I remember you mentioning somewhere that you recommended GT, and I haven't heard of any prematurely failed GTs and RSes (or RLs for that matter, and GY seems to be the long life version of RS). That's why I made mention of them.

Yeah but those are very old series too and nowhere to find. Besides my RS I rarely run into anything else than green GF/GK crap.

As to GD and GC having the same specifications as HM and HN? Doesn't say much - look at KZE and ZL's datasheet and notice how their numbers are practically identical.

Yeah but GC/GD are (close to) 100% identical IIRC.

[Wester547]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

Not sure if this one has actually been bad - most likely not. If it was not bulging though. I learned by that time the ESR Micro has about 0.4 ohm resistance on the leads and needs to be calibrated.

You said before it was bulging in the thread in question.

Left me think what after means. Umm, maybe, LATER, not sooner?

Where have you seen HMs and HNs fail after 2006 then?

Yeah, and why are you talking about RS and GT? Do I have GT somewhere on my stock? How many RS caps can you see there, in % from all caps?

I remember you mentioning somewhere that you recommended GT, and I haven't heard of any prematurely failed GTs and RSes (or RLs for that matter, and GY seems to be the long life version of RS). That's why I made mention of them.

I only brought up the capacitor plague because bad cap brands lied about where they source their raw materials at the time. Samxon could be doing that as well. As to specific shifts only being bad... this is debatable. Capacitors fail in many different manners and by themselves are only rated with 60% confidence, so there's no way that they're all going to fail at once (unless a Nichicon HM/HN style debacle happens). KZGs and KZJs have a high enough failure rate to be avoided or replaced on sight. And even if they haven't failed yet, with such unstable electrolyte, there's no telling when they will for no reason. Also, GL could be a counterfeit series, or it could simply be a series only sold to OEMs.

As to GD and GC having the same specifications as HM and HN? Doesn't say much - look at KZE and ZL's datasheet and notice how their numbers are practically identical.

[Behemot]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

https://www.badcaps.net/forum/showpo...10&postcount=1
Never seen a bad KM? Samxon does have a GS series, with blue sleeves.

Not sure if this one has actually been bad - most likely not. If it was not bulging though. I learned by that time the ESR Micro has about 0.4 ohm resistance on the leads and needs to be calibrated.

Are you saying you've seen older date codes fail after that, or that you've seen many post-2006 HMs and HNs fail?

Left me think what after means. Umm, maybe, LATER, not sooner?

Not sure how HE, FR, and PW have shorter lifetimes than RS and GT - HE is rated up to 10,000 hours and PW 8,000 hours (GT up to 10,000 and RS 7,000). And Panasonic FR is rated up to 10,000 hours. Larger case sizes yes.

Yeah, and why are you talking about RS and GT? Do I have GT somewhere on my stock? How many RS caps can you see there, in % from all caps?

If I am looking right, the PW ends at 1500/16 with D10. The D12.5 3300/16 sausage has life of only 7000 hours, my KZN have lower ESR, higher current and 10k hours in D10. Same goes for HE.

Best of such caps is NCC KYB (why you still use obsolete series?), digi-key carries them for 0.96 USD. My custom D10 KZN are 0.92.

Per Hansson: I think they have their own aluminium foils and bought the electrolyte. Just look at the GC, GD - they have 100% same specs as Nichicon caps. Everybody can make aluminium foils (though not everybody pure enough) but the elctrolyte is what chinese makers have been failing at.

[Per Hansson]

Samxon capacitor quality discussion (spin-off from OCZ Bug Zapper thread)

As you see in the small comment I added to the OP I spun off the thread here.
I found the discussion highly interesting and worthy of it's own thread so we would not hijack SgtRock's original thread.

Not since late '05/06, they fixed those two series.

Yes I know, I should maybe have worded my post differently.
See for example this thread here for a pretty good example that the HM & HN issue only affected specific production lines.
So even caps produced during the "known bad years" can be perfectly fine...
And if you have the time to read that thread you will see that I later found another system but that time with KZG caps and the exact same behavior:
Caps on the same board with different datecodes.
And where only one datecode=bad caps...

How can you be sure the samxons of tomorrow are the same as those 9-year old ones that've held up in that antec?

I can not, and I don't really claim anything either.
My stock of Samxon caps is old, and the caps I buy nowdays are mostly from Elfa Sweden and of the major Japanese brands.
But I used Samxon to "build up" my stock, because they where cheap and back then it was very difficult to get quality caps in Sweden at a decent price.

I also want to clarify that when I said: "They (Samxon) source their raw materials from Japan but have the factories in China."
I specifically meant the aluminium, because that is what I have heard.
The stories about the stolen electrolyte formula are blown way out of proportion and was not what I intended to make it sound like I meant.

Until I find out otherwise, both samxon and capxon are the same thing- will never use them. What happened to pride in your work?

This I have to say I take a bit offense at, I have never had a customer come back with a system which I used Samxon's to replace whatever was there before.
You make it sound like if a customer came back with a problem I would not help them because I don't take pride in my work?
I certainly would and I think I can speak for Topcat here as well and say that he also takes the same pride in his work as I do!

"beside couple bad series as a whole (GF, GK)"

And GL... unless, of course, GL is a counterfeit series.

I see what you did. You are not shooting from the hip by suggesting GLs are fakes-of-fakes.

Actually there do exist several companies that make counterfiet Samxon caps.
Well I guess it could be said it's just a coincidence in the naming and model scheme (Samson has for example a X-CON brand which is Samxon's Polymer series).
But just like "Rulycon" it has to be a pretty big "coincidence"
Take a look here to see what Samxon actually manufactures:
http://www.manyue.com/samxon_series_chart.html

[Sparkey55]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

Never seen bad KM, LP or HP. And I never seen Samxon GS ever, only those blue capxon craps.

I've seen shitloads of bad HN/HM after 05/06.

Just that all of them are bigger, have shorter lifetime and are more expensive than my D10 caps. But yeah, otherwise they are good.

There is one and only reason why they are expensive: they make almost everything in Japan. That is also why Rubycons are in very little products made in China - as there are high import taxes, everybody is forced to set up a plant in China. Others did, Rubycon not.

Most of the Chemi-Con caps I source are from Indonesia or China, only the custom KZN are from Japan plant.


When you run out, my offer to supply you is still valid. As long as I have something to supply

I have seen plenty of bad 2004 and earlier Nichicon HM but good HM with date code of 0453.

[Rob Northen]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

If you have not been fool you would at least ask when you do not know (which is not bad, but than you ask and not say BS) that Man Yue has been often sourcing electrolyte from Nichicon for many of their series. And because you are fool, you not only don't ask but also fail to understand "beside couple bad series as a whole (GF, GK)" and fail to recognize what do you actually see.

As for ultra-low ESR caps, I have seen shitloads of failed Nichicon, Chemi-Con, even Rubycon caps and only couple Samxon so far (and only on images). Nichicon, Chemi-Con, Rubycon are very good, very very good..... at making crap caps oh yeah…hard logic this is

I rest my case, you have said it all......

[Wester547]

Re: OCZ 700 Watt Bug Zapper Recap & Repair

Never seen bad KM, LP or HP. And I never seen Samxon GS ever, only those blue capxon craps.

https://www.badcaps.net/forum/showpo...10&postcount=1
Never seen a bad KM? Samxon does have a GS series, with blue sleeves.

I've seen shitloads of bad HN/HM after 05/06.

Are you saying you've seen older date codes fail after that, or that you've seen many post-2006 HMs and HNs fail?

Just that all of them are bigger, have shorter lifetime and are more expensive than my D10 caps. But yeah, otherwise they are good.

Not sure how HE, FR, and PW have shorter lifetimes than RS and GT - HE is rated up to 10,000 hours and PW 8,000 hours (GT up to 10,000 and RS 7,000). And Panasonic FR is rated up to 10,000 hours. Larger case sizes yes.

There is one and only reason why they are expensive: they make almost everything in Japan. That is also why Rubycons are in very little products made in China - as there are high import taxes, everybody is forced to set up a plant in China. Others did, Rubycon not.

IIRC, Rubycons with the T, H, and M prefixes on the datecodes are out of Japan, but the ones with A prefixes are out of Singapore and S prefixes out of Indonesia. Also, Nichicon process 90% of their aluminum foil in-house.

Most of the Chemi-Con caps I source are from Indonesia or China, only the custom KZN are from Japan plant.

Well, many KZGs and KZJs are straight out of Japan, but yield a very high failure rate. ALL the bad HMs and HNs came out of Japan in accordance to their sleeves.

Goodpsusearch, I knew VR has a 85*C temperature rating, that's why I specified 105*C low ESR capacitors.