KZG - KZJ - Same Electrolyte

[PCBONEZ]

I found this snippet in Chemicon's KZJ datasheet.

>>>>>>>>>>>

The KZJ series is a new ultra low impedance series from United Chemi-Con. These capacitors are different from the standard low impedance capacitors, as they use a new low resistivity electrolyte. Compared to our KZG series that also uses this advanced electrolyte technology, the KZJ series has lower ESR/impedance and higher ripple current capability, making them ideal for use in computer motherboard circuits where very low impedance at high frequencies is required.

<<<<<<<<<<

KZG don't like heat.
KZJ may turn out to have the same problem.

.

[linuxguru]

Re: KZG - KZJ - Same Electrolyte

Lower ESR and higher ripple current usually means more water in the aqueous electrolyte. It may fail even quicker than KZG under thermal stress. What's the rated endurance of KZJ? If it's 2000 hours or lower, I'd suggest giving them a pass and sticking with Rubycon MCZ or Panasonic FM, both proven workhorses. Maybe even Nichicon HE/HD/HM.

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

I know that. - It wasn't a question it was a warning.
Been seeing people look for KZJ lately.

MBZ, MBZ, KZG, KZJ all have the same 2000hr Endurance.
Also all have same -40C, which suggests about the same water content.

Whatever the difference is, those aren't it.

.

[yanz]

Re: KZG - KZJ - Same Electrolyte

if kzj is almost the same with kzg, then people should avoid using it..

[enduser]

Re: KZG - KZJ - Same Electrolyte

Lower ESR and higher ripple current usually means more water in the aqueous electrolyte. It may fail even quicker than KZG under thermal stress. What's the rated endurance of KZJ? If it's 2000 hours or lower, I'd suggest giving them a pass and sticking with Rubycon MCZ or Panasonic FM, both proven workhorses. Maybe even Nichicon HE/HD/HM.

I just Goggled the KZG series and this post came up. Wow you are right on the money with your prediction. A little over two years from this post, both the KZG and KZJ failed on a Dell Mobo. What is interesting is that the Mobo only had one of each and they were the only ones that failed. No other caps are faulty. It looks like United Chemi-con has now discontinued both those series.

[Rulycat]

Re: KZG - KZJ - Same Electrolyte

What's going on at UCC? Were they having manufacturing issues?

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

No, it's just the particular electrolyte used in KZG and KZJ is crap.

[mockingbird]

Re: KZG - KZJ - Same Electrolyte

I have no idea at all whatsoever what they put in the electrolyte, but let me hypothesize that they need to put some sort of chelated metal in it, and instead of using a more expensive alloy, they use iron

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

It's some kind of Aluminum solution because the foil it is 'electrolyticing' with is Aluminum.

I think they made a bad choice in one of their stabilizer additives and the electrolyte simply breaks down.

.

[mockingbird]

Re: KZG - KZJ - Same Electrolyte

I wonder how gold would factor in with capacitor electrolyte. Is it possible to create really high quality stuff with it? It is a highly conductive metal and it doesn't oxidize, unless you throw it in highly concentrated sulphuric acid.

[Pyr0Beast]

Re: KZG - KZJ - Same Electrolyte

You have silver mica capacitors.

But ordinary lytics need to oxidise to work properly. Aluminium oxide works as an insulator.

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

In Aluminum Electrolytics the Aluminum Oxide layer -IS- the Dielectric.
The separator paper is only in there to act as a carrier [flow path sort'a] so the electrolyte can get to the Aluminum to keep it oxidized.

Ripple currents blast chunks [figuratively] of the Oxide [or Aluminum depending on instantaneous polarity] into solution where it disolves.
The DC voltage present restores the damaged areas similar to how electroplating works.
- That's why electrolytics used to be called 'self healing capacitors', although the term has fallen out of use over the years.
.

[mockingbird]

Re: KZG - KZJ - Same Electrolyte

Interesting. But if this is the case, why do capacitors instantaneously provide their rating, if it took voltage to "stimulate" them, wouldn't you have to wait a minute or two for the voltage to do it's work?

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

You lost me.. ??

""instantaneously provide their rating"" ??

What which rating??

[mockingbird]

Re: KZG - KZJ - Same Electrolyte

... You said that the electrolyte oxidizes the aluminium when DC voltage is applied, and when it ceases, it "self heals". Surely this process would take a minute or two, so when the current is initially applied, the capacitance would be significantly out of spec?

[Pyr0Beast]

Re: KZG - KZJ - Same Electrolyte

Nope. Oxide layer does not dissolve that easily.

Old caps that have gathered dust for past 5 years, yes. They need to be regenerated.

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

... You said that the electrolyte oxidizes the aluminium when DC voltage is applied, and when it ceases, it "self heals". Surely this process would take a minute or two, so when the current is initially applied, the capacitance would be significantly out of spec?

No, misinterpretation,,,

Ripple damages the -existing- oxide layer and the DC repairs the damage.
- Bear in mind the ripple is in mV and the DC is in Volts. Small local damaged areas, big global [so to speak] healing.
- Also once the oxide layer gets to a certain thickness [based on the applied DC] the healing will stop because the oxide layer is an insulator.
- Also a thin area [damaged] of the oxide layer will get the most healing effect simply because the insulation is thinner there.

The aluminum foil is etched and pre-oxidized to some extent [usually chemically] before the foil pack is rolled.
Then they use a process called 'forming' to build up the oxide layer.
Forming is basically applying DC at a low voltage then slowing increasing it to build up the oxide layer's thickness to what it should be for the rated voltage.

In other words caps come from the factory with an oxide layer.

The oxide layer doesn get thinner with no DC applied but it takes years.
-The oxide dissolving back into the electrolyte on unused caps is one major things [if not THE major thing] that determines shelf-life.
That's why they recommend 're-forming' old caps before use.
.

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

Yeah,,, what he said...

[Pyr0Beast]

Re: KZG - KZJ - Same Electrolyte

I doubt that ripple would eat oxide layer. That wouldn't happen as long as cap doesn't go into reverse polarity.

[PCBONEZ]

Re: KZG - KZJ - Same Electrolyte

I doubt that ripple would eat oxide layer. That wouldn't happen as long as cap doesn't go into reverse polarity.

Ripple DOES go in reverse polarity.
It is a small AC ridding on a DC.
.
If you think it won't eat the layer stick a DC [polarized] cap on AC with no DC and tell me what happens.
.