Electrolytic Caps in a Series Connection

[kc3ase]

Hi all,

I've heard when you are unsure if these caps might undergo a reverse voltage condition you can place these caps in a series connection and reverse voltage conditions wouldn't damage the caps in this case.

1) Why is this?
2) I'm assuming the caps don't actually provide their rated capacitance in this case.

I've seen a high reliability cap network of let's say four or six electro cap in a bank with balancing resistors. Two caps are always in series with the additional caps in a bank.

I think this keeps the aRMS currents lower than usual keeping the lifetime high and also should one of the electro's short on a failure mode you still have the cap in series below it which can continue.

3) Do I have this mostly correct?

[Simon]

I have heard that if you connect two electrolytics in series with oposing polarites you get an AC capacitor, the only other reason to connect capacitors in series is to increase the voltage rating but the capacitance halves

[Andy Watson]

The connection between the two caps needs to be biased to a voltage that ensures neither cap will be operating with reverse voltage across it. Otherwise it is guaranteed that at least one of the capacitors is operating with reverse voltage - leading to reduced capacity - leading to more reverse voltage - etc., and early failure.

[kc3ase]

I have heard that if you connect two electrolytics in series with oposing polarites you get an AC capacitor, the only other reason to connect capacitors in series is to increase the voltage rating but the capacitance halves

I've seen the drawing as in a series:

+V   +C- +C-    -V

Does this mean opposing? I'm thinking opposing would be:

+V  +C- -C+     -V

[Simon]

Your first is in series with the same polarity to increase voltage rating. As suggested balancing resistors are needed. The second one is with oposing polarity and I beleive would make it an AC electrolytic.

[edavid]

The connection between the two caps needs to be biased to a voltage that ensures neither cap will be operating with reverse voltage across it. Otherwise it is guaranteed that at least one of the capacitors is operating with reverse voltage - leading to reduced capacity - leading to more reverse voltage - etc., and early failure.

This is demonstrably untrue.  Nonpolar electrolytics are just 2 polar caps in series, with no balancing resistors or biasing.  They don't have a particularly high failure rate.

[f5r5e5d]

no, its a common misunderstanding but nonpolar/bipolar are fundamentally different from series "back-to-back" polar Electros


nonpolar/bipolar Electrolytics have full thickness, full rated V blocking strength oxide layers grown on both foils and there are no intermediate unformed foil with only the spontaneous oxide that only blocks a V or two

[Zero999]

I believe an electrolytic capacitor simply acts as a rectifier with an enormous capacitance. Under normal conditions, it's reverse biased so it doesn't conduct. When the capacitor is connected in reverse it will be like a forward biased diode and conduct DC. If a significant current is allowed to flow, for a long period of time, the capacitor may leak and explode. However, when two capacitors are connected back-to-back, no DC current can flow so nothing bad will happen.

If you're really paranoid about this, you could connect a diode in reverse parallel with each capacitor but it should be fine without it.

[retrolefty]

 I've seen on some magazine schematics, back to back (series, with reversed polarity +- -+) electrolytic caps used as a non-polar high capacitance caps as used say to drive the start winding on a AC motor. However in decades of component level troubleshooting/repair I never saw a commercial product use such an arrangement over rather a proper non-polar cap which are certainly much larger per farad and probably more costly.

 I've never come across a definitive artical about using electrolytic caps instead of a (proper?) non-polar cap? Can anyone add light to this subject?

[f5r5e5d]

nonpolar/bipolar Al Oxide dielectric, Electrolytics are used in motor caps - but nonpolar dielectric polyethylene caps are often seen for longer life

nonpolar Electrolytic cap /= "nonpolar dielectric" cap (which are most often plastic film dielectrics such as Mylar, polyethylene)

[edavid]

no, its a common misunderstanding but nonpolar/bipolar are fundamentally different from series "back-to-back" polar Electros

nonpolar/bipolar Electrolytics have full thickness, full rated V blocking strength oxide layers grown on both foils and there are no intermediate unformed foil with only the spontaneous oxide that only blocks a V or two

http://www.digikey.com/Web%20Export/Supplier%20Content/CDE_338/PDF/CDE_AEappGuide.pdf

Aluminum Electrolytic Capacitor Application Guide

If two, same-value, aluminum electrolytic capacitors are
connected in series with the positive terminals or the negative
terminals connected together, the resulting single capacitor
is a non-polar capacitor with half the capacitance. The two
capacitors rectify the applied voltage and act as if they had
been bypassed by diodes.  When voltage is applied, the correct-
polarity capacitor gets the full voltage. In non-polar aluminum
electrolytic capacitors and motor-start aluminum electrolyte
capacitors a second anode foil substitutes for the cathode foil to
achieve a non-polar capacitor in a single case.

[f5r5e5d]

are you objecting or confirming? - you might want to read your own quote more carefully if you think it says I'm wrong

If two, same-value, aluminum electrolytic capacitors are
connected in series...

...In non-polar aluminum
electrolytic capacitors and motor-start aluminum electrolyte
capacitors a second anode foil substitutes for the cathode foil to
achieve a non-polar capacitor in a single case.

 Bateman was employed by a cap manufacturer, made extensive measurement, wrote up articles on Electolytics for use in Audio

http://linearaudio.nl/cyril-batemans-capacitor-sound-articles

major cap manu http://www.nichicon.co.jp/english/products/pdf/aluminum.pdf

[edavid]

are you objecting or confirming? - you might want to read your own quote more carefully if you think it says I'm wrong

Both... they do optimize away the extra foils, but it doesn't make them fundamentally different.

[f5r5e5d]

lack of the unformed spontaneous oxide layer in nonpolar/bipolar single package Electrolytic shows up clearly in Bateman's measurements

if low nonlinear distortion is important for your application then no combination of polar electros in series, even with bias V to the midpoint will be able to match the lower distortion of 2 full oxide thickness foils in a non/bipolar electro