Low value capacitor voltage

[poppydog]

Why is it so difficult to find low value/low voltage electrolytic capacitors such as a 1uf? I have been doing a fair bit with the far eastern transistor radios lately and have been having real difficulty finding values such as the 1uf under 50v.

Regards
Poppydog

[MrBungle]

The voltage doesn’t matter. That’s just the maximum rating. A 63v electrolytic is a fine substitute for a 10v one for example.

Voltage is rated as in “will probably explode if you put more than this many volts across it”. It is pretty much just dielectric voltage withstand.

[G8UWM-MildMartin]

And ceramic and polyester capacitors of that capacitance are now available relatively cheaply.

[Diabolical Artificer]

"The voltage doesn’t matter." I thought it did having read in several threads that V max should not be exceeded by too much so the cap remains polarised.

Looked on Farnell there were only two 1u leaded radial electrolytic caps under/@25v most being 63v rated. I suppose this is down to the fact that through hole components are passe' ; no one makes radio's anymore like they did in the 70's, if at all. A component manufacturer will only make stuff he can sell and make profit on; a few old fellas in sheds wanting the odd one or two caps is not a good base to expand your bizniss.

Andy.

[stevehertz]

To say that the voltage doesn't matter so long as it is the same as or higher than the original is not wholly true. I posted on here some time ago on the subject of how electrolytics 'like' to be used at a voltage that is towards the upper end of their rating. As is often the case when you get down to molecular level, it's a complex issue, but as I recall the general consensus from our in-house experts, chemists and physicists was that (what I just said) was the case. In practice however it doesn't seem to make a lot of difference. Capacitor technology has improved over the years, and they have also got smaller. Hence manufacturers have stopped making low voltage caps when there's plenty of space for higher voltage ones, it means they don't have to make so many different combinations of capacitance and voltage.

[Radio Wrangler]

Operating at low bias voltage (compared to the rating) is a known cause of increased probability of failure due to de-polarisatio of aluminium electrolytic capacitors.

When they are made, the oxide fil on one electrode is reduced to metal, while the oxide on the other is built up. This process is driven electrically, and the DC component of the bias maintains the state created. The general recommendation varied, but it was somewhere around 10% of rated voltage. Also, it was known that the max voltage rating of parts would fall in response to their being used for prolonged periods at lower voltages..

However, this information may be (not necessarily is) dated. There have been interesting changes in electrolyte design which have reduced ESR and increased longevity. Is anyone aware of any newer research on aluminium electrolytic capacitor behaviour? Much of the problems have been recently due to fakes and the theft of an incomplete electrolyte formula and its spread across a number of dodgy makers. But how have the parts from the real reputable makers been changing?

David

[MrBungle]

In this case the data sheet is king. Look at the lifespan, ripple current, temperature derating and plan for it. Other failure modes are statistically insignificant within those boundaries. You can pay more for more guarantees. All of which have been backed up by testing and sampling.

Personally I only use them for filter caps and bulk decoupling. Coupling caps are film, ceramic or tantalum. Timing are film. Everything else, they are last choice.

[Craig Sawyers]

OK here are some exception to the voltage rule.

Coupling electrolytic capacitors in multi op-amp circuits to remove cumulative offsets in the tens to hundreds of mV level. Typically 10-100uF.

Or non-polar electrolytics, electrically equivalent to two back to back polar electrolytics, used without any DC bias in loudspeaker crossovers. For example http://www.nichicon.co.jp/english/pr...fs/e-db_gb.pdf . From 1u to 68u less than a quid from mouser.

Or in signal level circuits (so not intended for high ac currents) http://www.nichicon.co.jp/english/pr.../pdfs/e-es.pdf non-polar too . 0.47u to 1000u.

Cyril Bateman, who was a capacitor developer with (I think) Erie, wrote a series of papers about capacitor harmonic distortion in the early 2000's. Most of his tests were initially with zero bias, including electrolytics. The zero bias condition was always the lowest harmonic distortion. Applying bias of varying amounts always increased distortion, with figures typically many tens of ppm. The lowest distortion capacitors were non-polar, and the very best construction being two non-polar (which are cheap) back to back, with zero bias - which gave typically 1-3ppm THD.

Basically there seems to be no problems in using electrolytic capacitors with zero bias voltage at typical line level audio signal level (no more than 1Vrms - beyond that the peak negative voltage starts to damage the etched foils). If the signal level is more than 1Vrms a small bias is necessary - a potential divider to the centre point has been used by some designers with back to back polar electrolytics.

[Neutrino]

Nichicon have a General Descriptions of Aluminum Electrolytic Capacitors.
http://www.nichicon.co.jp/english/pr...f/aluminum.pdf

“2-9-4 Applied Voltage and Life
The degree that applied voltage effects the life of the capacitor when used below the rated voltage is small, compared to the degree that ambient temperature and ripple current effects life. Therefore, when estimating the life of a capacitor, the voltage coefficient to the applied voltage (Fu) is calculated as 1. An example of the test results is shown in Fig.2-16.”

I cannot see any difference in the test results for a 50V capacitor tested at 50 V, 40 V and 30 V.

David

[stevehertz]

But what about a 100V rated cap that is being used to replace a 10V rated one and is being run at say 8v? That's more like the situation we find ourselves in these days.

[Craig Sawyers]

Nichicon also say "Aluminum electrolytic capacitors are polarized.
Make sure that no reverse voltage or AC voltage is
applied to the capacitors. Please use bi-polar
capacitors for a circuit that can possibly see reversed
polarity.
Note: Even bi-polar capacitors can not be used for
AC voltage application."

But they do not say how much reverse voltage for an AC signal. And the restriction on bipolar electrolytics is clearly misleading "Please use bipolar if there is reverse voltage", and then that you can't use these for ac applications. And some such capacitors are designed for loudspeaker crossovers, clearly at zero bias and potentially rather high AC voltage.

Craig

[Craig Sawyers]

The question regarding using (say) a 100V cap at 10V shouldn't cause a problem. I'd do that substitution in a heartbeat.

PS just need to clarify the zero bias comment I made earlier when I said "less than 1V rms", what I meant was that this condition was at the lowest audio frequency for that equipment.

[G8HQP Dave]

There may be two different issues here:
1. Under what conditions does an electrolytic retain its stated characteristics?
2. Under what conditions does an electrolytic remain as a reliable capacitor, but perhaps with modified capacitance value and somewhat increased leakage?
Datasheets are mostly interested in 1; we are mostly interested in 2. That may explain how people can often get away with conditions which the datasheet forbids.

[MrBungle]

Regarding point 2; avoiding those conditions is where the engineering bit comes in thus only relying on point 1.

[Terry_VK5TM]

And point 1 - only in a lab under controlled conditions

[PJL]

The oxide layer is reported to dissolve back into the electrolyte over time and capacitors left on the shelf used to need 'reforming' to rebuild the oxide layer.

Capacitors used at lower voltages would also need reforming if they were re-purposed for higher voltage use.

The extreme of this was a 1930's article I read relating to wet electrolytics and their ability to provide over voltage protection (I think the article called it regulation).

Difficult to say whether any of this applies to modern capacitors. Circuit design issues are the main cause of faults in modern equipment where the ESR is not low enough and they overheat and leak.

[GrimJosef]

I've read somewhere that from the 1960's onwards manufacturers changed their initial formation process to produce a much more durable oxide layer (I recall the word 'anodising' being used) with the result that later capacitors are much less prone to degradation and in much less need of reforming than before. The problems then became drying out and, as with other capacitors, dielectric breakdown.

Cheers,

GJ

[Neutrino]

Originally Posted by PJL

Capacitors used at lower voltages would also need reforming if they were re-purposed for higher voltage use.

Agreed, but I would have to be marooned with a non-working radio before I repurposed a used electrolytic capacitor.

David

[FrankB]

New 'lytics are all I use, due to the dismal failure rate of used 'lytics. Not worth my time to deal with faulty parts. (yes, I do test for ESR, voltage, & capacitance before installing too., as I do for every new part. QC for some mfg. is horrible).
I seldom order any 'lytics below 50 or 63V. & 105C. They are basically the same price and physical size. Why stock several voltages when one will handle 99% of the needs?
One mfg, here bragged in their QC literature that "They test every capacitor with 2X the stated working voltage to weed out any defective ones". (Check's in the mail...).

Remember the problems with the "fish oil" electrolyte? Made a nice income from all those defective 'lytics when I was in business.
You never know on used caps which had that problem. It ate up the traces, smelled to high heaven, and was nasty to get off the circuit boards. Good solid reasons not to re-use 'lytics.

[Argus25]

Originally Posted by Craig Sawyers

Nichicon also say "Aluminum electrolytic capacitors are polarized.
Make sure that no reverse voltage or AC voltage is
applied to the capacitors. Please use bi-polar
capacitors for a circuit that can possibly see reversed
polarity.

One interesting thing I have noticed about vintage transistor radios and the very old low voltage electrolytic caps they use is the most common one to fail in the radio: It is invariably the AGC filter electrolytic.

The reason for this is that typically, for most transistor superhet circuit designs, there is a small correct polarity bias (off station) across the AGC capacitor (derived from the transistor bias networks) which sets the gain controlled stage's base currents. However, when a reasonable strength station is tuned in, the AGC voltage results in reverse voltage being applied to the AGC capacitor.

It turns out to be the fastest way to get a small electrolytic cap to degrade & fail.

Most transistor radios (not all) have this design dilemma and it is easy to check with a meter across the AGC capacitor on & off station. The occasional radio has the capacitor reversed and ends up with the correct voltage on station. In some designs, the problem doesn't occur.

For this reason most/all vintage transistor radios should have a bipolar electrolytic cap for the AGC filter.