I thought Polystyrene Capacitors weren't polarised?

[trys]

Hiya all,

This is hopefully a simple question that somebody can please answer for me.

I was of the understanding that polystyrene capacitors were not polarised. The ones that I have though have either a coloured band on one side, and others are red at one end. (Photograph below)

I thought initially it was perhaps a colour coding for tolerance or voltage but looking at the datasheet there is no mention of this. http://www.farnell.com/datasheets/1869697.pdf

Any ideas?

Trys

[glarsson]

Foil capacitors often have a ring to mark the wire connected to the outermost foil. Connect this to ground for lower noise in some applications.

[Refrigerator]

Mr. Carlson did a great video regarding this.

[rrinker]

 I watched that video quite some time ago and learned a lot - highly recommend it.

Just because they aren't polarized doesn't mean they don't have a direction!

[TimFox]

Many decades ago, when "outside foil" marks were more common than now, I was taught that the "outside" end of the capacitor should be connected to the node with lower voltage (from ground), as a safety matter.  This was especially true for paper capacitors, coated with wax, where the case insulation was "iffy" compared with modern epoxy-dipped film capacitors.

[schmitt trigger]

Many, many moons ago, I was building a high gain audio preamp with RIAA-equalization, for playing LPs.

RIAA-equalization required of course, RC networks.
For optimal curve fit, the audio press recommended precision tolerance capacitors, which meant of course plastic foil types.

Long story short: one of the channels had significantly higher hum than the other, and that would occur even with the inputs shorted to ground.

In those pre-Google, pre-Siri, pre-Cortana days, the only way to solve the problem was by tedious and time consuming troubleshooting.

I discovered that, when I touched with my finger one of those capacitors, the hum would increase.
I said "aha! a defective capacitor", and proceeded to replace it -unknowingly- in the correct direction, which solved the problem. I strongly believed that it had been a defective capacitor.

Many years later, I read a paper which mentioned the pitfalls of capacitors in high-gain circuits, which mentioned the external foil issue.

Fortunately, I still had the preamp, which I was no longer using. Out came the soldering iron, and removed and placed the capacitor "backwards".
To my astonishment, the hum came back!!!

So yes, place the foil side to the lowest potential, preferably ground.

[floobydust]

The film-cap stripe connects to the output of the stage driving the capacitor, in the case it's a coupling cap, leading to the next stage.
The film-cap stripe connects to ground, in the case it's a filter cap.

Just to prevent noise pickup from nearby radiated E-fields.

[Brumby]

So yes, place the foil side to the lowest potential, preferably ground.

For capacitors that run to ground, then yes - but for those that don't (eg coupling capacitors) you are better to learn the more general rule.

As Mr. Carlson said, the outer foil end goes to the side of the circuit with the lower impedance.  The reasoning for this should be obvious.  A lower impedance means that an induced voltage will have less effect on the signal.

Obviously a ground point would have to represent the lowest impedance in most gear, so for capacitors going to ground that choice is obvious - but with other topographies, you will need to think just that little bit more...

[Zero999]

Mr. Carlson did a great video regarding this.

That's interesting. I wonder when the capacitor manufacturers stopped marking the shield correctly or did they ever do it properly?

I question the use of a 555 timer in the test jig. The CD4013 can be made to oscillate on its own or a quad NAND/NOR gate (CD4011/CD4001) could be used to make an astable which will give alternate output pulses. CMOS IC astables will work with very high values of R in the timing circuit, resulting in much lower current consumption, than a 555. Get rid of the 555 and the circuit will consume around 0.3mA.

I'd also be tempted to go with a more low tech approach and use a DPDT switch, but that might not be as user friendly.

[floobydust]

I think a 555 timer is better than a CMOS RC oscillator in that the 555 uses comparators and Vcc-referenced voltage-divider, so freq. variations are much less than CMOS parts sensitive to Vcc and temperature, those using a Schmitt trigger. Their threshold and hysteresis voltages move around.

[Zero999]

I think a 555 timer is better than a CMOS RC oscillator in that the 555 uses comparators and Vcc-referenced voltage-divider, so freq. variations are much less than CMOS parts sensitive to Vcc and temperature, those using a Schmitt trigger. Their threshold and hysteresis voltages move around.

I haven't done any tests, comparing the two but does it really matter? The timing isn't critical. If it was, then why use a tantalum capacitor, which won't have a very good tolerance? It just has to be slow enough to compare the waveforms, but quicker than changing the connections manually.

The CD4013 works quite well.

[KE5FX]

Interesting video.  I had no idea that those black "outside foil" markings were untrustworthy.  It might explain why I've run into hum problems when recapping vintage radios, which I've always blamed on filament lead dress.   

[trys]

Foil capacitors often have a ring to mark the wire connected to the outermost foil. Connect this to ground for lower noise in some applications.

Glarsson - thank you for answering my question perfectly - it's much appreciated.

I had sent an email enquiry to LCR Capacitors EU Ltd (who manufactured a pair of the capacitors) but they hadn't bothered replying.

Mr. Carlson did a great video regarding this.

A great video. I've learnt lots from that - thank you for pointing it out to me. The strangest thing is the capacitors whose lines bear no relation to the outside foil. The video is pretty drawn out though, but it's well worth watching.

I wonder if an easier test would be to hold the outside of the capacitor using a small metal clamp which is fed with (say) 100KHz, and measure which lead of the capacitor has the highest amplitude? The outside foil side one would be capacitively coupled to the test tone and be shown on its leg I imagine?

Thanks all for your replies!
Trys

[Audioguru]

I used high voltage capacitors like that 50 years ago with vacuum toobs. They were made with foil and plastic film wound around and around and around and around creating a fair amount of inductance that most modern circuits do not want and they were difficult to connect to a pcb. I used modern film capacitors ever since like these ones that fit on a pcb: