Teapo

[Stefan Payne]

Thing is that you probably won't see OKish capacitors in good designs and more often than not in crappy designs that absolutely need high quality capacitors. But because its cheap crap, you see only cheaper caps here...

And so they fail more often because the capacitors are the wrong ones for the design (and maybe the datasheet isn't entirely correct. As the KF series seem to have rather highish ESR if I remember correctly, around 30% higher than Teapo SC).

So I'd rate the KF a bit under the Teapo SC series due to higher ESR...
And the GL Series higher than KF...


But the price at the Shenzhen Market is the only thing that could give us information about the quality of the capacitors, as i'd expect the KF to be a bit cheaper than Teapo SC...

Another thing you miss:
You rate by Company. And that in general is bullshit. As everyone has different quality grades.
The thing with the "Japanese Companys" is that they do not have such "low end" Series as Teapo SC or something like that. And even if they fail (like in iMAC G5 PSU), they don't get flamed the same as others for whatever reason...


As for the ltec caps:
You're talking about the LYZ series most of the time, right?
Maybe an LZP here and there
But not LZG or LZD series???

http://www.ltec.com.tw/product_e.html

The Series Chart of Manufacturers is also important to check as it tells you where the manufacturer sees the capacitor, if it is a better one or not.

You've probably not seen a Capxon GH or FH capacitor, have you?

[ashiekh]

And even if they fail (like in iMAC G5 PSU), they don't get flamed the same as others for whatever reason...

The G5 Mac power supplies had cheap capacitors, some Ltec LYZ I believe; I rebuilt quite a few of those; it was the motherboards that were hard to recap, as the ground planes sunk a lot of heat.

From my failed collection
* Ltec LYZ
* CapXon GL, KF, GF
* Nichicon NH(M)

the Nichicon came from an iMac G5 mother board.

From my still working collection (capacitors I removed during a rebuild)
* Ltec TX, TK
* CapXon KM
* Teapo SC

I still find it hard to believe the capacitor plague legend that an incomplete formula was stolen, for that does not explain how the well established manufacturers were also hit.

[Rexper]

What do you guys think of Rubycon's Electrolytic life calculation formula?

‹Snap-in type›

②Products with a rated voltage of 160 V or more

*If heat rise (Tj) exceeds 30ºC by applying ripple current, please contact us.

K is uknown.
Internal heat rise is unknown without inquiry. Or it could be calaculated if we had the ESR value...
But I haven't seen a datasheet that even describes the ESR value? Can we calculate through information we do have, such as dissipation factor or leakage current?

[Stefan Payne]

The G5 Mac power supplies had cheap capacitors, some Ltec LYZ I believe;

Sorry, misremember, seems like it was the Mainboard:
https://www.instructables.com/id/Ima...citors-repair/

I rebuilt quite a few of those; it was the motherboards that were hard to recap, as the ground planes sunk a lot of heat.

From my failed collection
* Ltec LYZ
* CapXon GL, KF, GF
* Nichicon NH(M)

1. LYZ -> ULTRA low ESR, according to ltec, 2-3k Lifetime, so lower grade caps.
2. Those are mid to lower grade Capxon (GL, KF and GF are below that).
3. another ULTRA Low ESR type, also 2000h lifetime.

Those ULTRA low ESR are in general garbage, across the board. So replace on sight with Polymer caps. Even crappier ones like KEMET A750 are better...

the Nichicon came from an iMac G5 mother board.

Yeah, probably CPU VRM, where you have high ripple current and temperatures.
This is where you should replace the caps, wherever possible, with polymers. Wet lyrics are the wrong choice here - Polys weren't widely available at the time or way more expensive...

From my still working collection (capacitors I removed during a rebuild)
* Ltec TX, TK
* CapXon KM
* Teapo SC

1) couldn't find TX Series, only TK and its general purpose
2) CapXon KM is general Purpose (for Power Supplys)
3) the only low ESR type here...

I still find it hard to believe the capacitor plague legend that an incomplete formula was stolen, for that does not explain how the well established manufacturers were also hit.

The electrolyte was too instable for the lowest ESR types.
Its always the same type to fail, across the board and manufacturers and that's always the ULTRA Low ESR types like Teapo SZ, Nippon Chemicon KZG, KZJ, Panasonic FJ, the above mentioned Nichicon HM and so on.

All have problems because of the high water count. And Water is a thing that likes to react with other stuff.
And the ULTRA low ESR ones have more water than the low ESR ones, wich also have minor realibility issues.

Thus they were phased out everywhere. The latest ULTRA low ESR ones I've seen was Teapo SZ in Corsair CS and be quiet POWER ZONE units...


As for CapXon KF: Ever seen a Rigol scope with failed caps? Because they seem to like to use CapXon as well...
And the CapXon GL worked rather well in my Philipps 200P6IS (wich I should have ordered a week later *ARGH*)...

[ashiekh]

1) couldn't find TX Series, only TK and its general purpose

I looked again and must have misread the TX (sorry about that)

The question now, for people like me who like to replace old capacitors is where can we get polymers of the right range and physical size? (those with an asterisk I have found)

Things like

6.3V 4700uF

10V 2200uF
10V 3300uF

16V 1000uF * (8mm, 10mm dia)
16V 1200uF
16V 2200uF * (10mm dia)

for now I just keep using low ESR electrolytics for power supply rebuilds (polymer lifetimes don't seem so different from those of electrolytic); I also like to get the same diameter for the replacements.

[ashiekh]

The thing with the "Japanese Companys" is that they do not have such "low end" Series as Teapo SC or something like that.

I think the trick is if one doesn't want low end offerings to tarnish ones reputation, sell them under another name, as APC does with Conext UPS units

[ashiekh]

What do you guys think of Rubycon's Electrolytic life calculation formula?






K is uknown.
Internal heat rise is unknown without inquiry. Or it could be calaculated if we had the ESR value...
But I haven't seen a datasheet that even describes the ESR value? Can we calculate through information we do have, such as dissipation factor or leakage current?

Since K is unknow, might as well just go for Arrhenius's law, where life doubles for each 10C cooler (which is the first part of the above equation).

So a 105C capacitor run at 55C will last around 32 times longer; a 3 month lifetime becomes 8 years, more for a machine that is not running round the clock. If my repairs last a couple of decades I'll be happy enough.

[gdjacobs]

Thing is that you probably won't see OKish capacitors in good designs and more often than not in crappy designs that absolutely need high quality capacitors. But because its cheap crap, you see only cheaper caps here...

Delta, for instance, uses mid-tier caps (like Teapo) in their designs all the time. However, because of their combination of experience and engineering standards, they seem to be able to use them appropriately and extract good part lifetimes. Conversely, even good designs with good components have a hard time when they're up against the varsity (see HP SFF systems with Lite-on PSUs and insufficient airflow).

As for the ULESR NCC and Nichicon caps, yes they're running at the limit of what can be achieved with Al-El caps. However, series issues with HM/HN caps were eventually resolved (not so with KZG or KZJ). Polymers are now cost effective enough to be a blanket replacement for those product ranges while delivering superior performance and reliability.

There was nothing wrong with HM/HN series if they were from a good batch -- poly caps are simply better.

[ashiekh]

Then maybe I will not worry too much about the Teapo capacitors in some of my power supplies.

[Jon Gerow]

Then maybe I will not worry too much about the Teapo capacitors in some of my power supplies.

Unless they're 10 years old, you shouldn't.