Mind-Blown: Orange Drops have Polarity?

[The New Steve H]

I was Youtubing to kill a few minutes, and I came across a video which instantly changed my worldview. The guy in the video said non-electrolytic capacitors like Orange Drops have a polarity, and that if you ignore it, you can get hum and oscillation.

True, or is he smoking old carbon resistors?

https://www.youtube.com/watch?v=BnR_DLd1PDI

I really do not want to go into my amps and turn capacitors around.

[pdf64]

Any difference is probably marginal, if the amp is stable and seems ok, don't worry about it.
Plus, as Aiken notes, some oroange drop types don't have an outer foil http://www.aikenamps.com/index.php/wher ... capacitors

Perhaps bear it in mind for future builds though.

[Malcolm Irving]

If the cap has a recognizable 'outer foil', by connecting this to the side which has lowest AC impedance to ground, you can screen the inner foil from picking up hum and interference to some extent. This can be important for radio frequency circuits and if you delve into old point-to-point wired tube radios you will often find that non-electrolytic caps have one end marked to show it is the outer foil.

For audio circuits it is debatable whether it makes any noticeable difference. But it can't do any harm, and if you like that kind of thing you can find (on the internet) ways of detecting which is the outer foil, if it is not marked.

[Theashe]

I copied his experiment quickly. My scope only goes down to 20mV/div.

Here are the peak-to-peak hum figures for 5 different caps, 3 different brands, 4 different values:

Orange Drop 715P 473K hum peak-to-peak:
Direction 1: 5.6mV
Direction 2: 14.4mV

Mallory 150 - 0.1uF
Direction 1: 5.6mV
Direction 2: 15.9mV

Suntan - 0.1uF
Direction 1: 4.8mV
Direction 2: 10.4mV

Mallory 150 - 0.0022uF
Direction 1: 78.4mV
Direction 2: 148mV

Mallory 150 - 0.01uF
Direction 1: 18mV
Direction 2: 56mV

Looks like there's some weight to what he says. Putting your non-polar caps in the wrong way can make them more sensitive to hum. Even Orange Drops or Mallory's. Hell, the Suntan (Taiwanese) caps I tested picked up the lowest hum per value.

Whether that's significant or not, I'll leave that up to you to decide. Smaller caps have the problem worse, so it may be worth making sure those go in correctly.

[JazzGuitarGimp]

and if you like that kind of thing you can find (on the internet) ways of detecting which is the outer foil, if it is not marked.

Be aware though, that some current production caps have the stripe on one end, but it may not really be the outer foil end.

[John_P_WI]

Welcome to Dumbleville...

[norburybrook]

I went to the trouble of doing this and it made no audible difference to me.


Marcus

[10thTx]

Here is a somewhat lengthy thread with photos and such about this:

http://el34world.com/Forum/index.php?topic=11427.0

It certainly has made a significant difference in "idle" floor noise in my amps to check polarity. I realize that might not be the case for others.

You can sure hear a difference in hum/noise when you are "testing" certain values. I can't tell much difference smaller then .0022 or greater then .047. But with values like .01 thru .03, I hear a very noticeable difference with Orange Drop PS series.

With respect, 10thtx

[R.G.]

Here's the deal:

IF a cap has a recognizable outer layer in the way it's constructed,
AND
the cap is metallic foil, not metallized film
AND
the cap is NOT constructed with low-ESR techniques
AND
the circuit it's connected has a high impedance point feeding one end of the cap,
AND
the other end of the cap in the circuit feeds a low impedance point, much lower than the other side
AND
the impedance of the cap over the relevant audio range of frequencies in the application is more than roughly 10% of the high impedance side's impedance,
THEN
it is possible that there is some difference in what happens when you swap the cap.

This whole rigamarole is based on the idea that the "outer foil" can shield the inner foils somehow.

This would be most prominent in a situation where the cap feeds a signal to a grid with a high grid leak resistor, and the driving impedance is quite low, like perhaps a cathode or MOSFET follower and there are high voltage signal/noise sources physically nearby the cap so their voltage swings can be coupled into the outer cap layer if it's turned backwards, and all the above stuff is true.

I'm guessing that there are a good ten or twelve such situations in all the amplifiers in the world, and someone noticed one of them.

EDIT: I should mention, there is a reason behind each of the item above.

Got questions?

[The New Steve H]

It sounds like anyone who has a dead-quiet amp that doesn't oscillate should ignore this.

[Scumback Speakers]

Welcome to Dumbleville...

I hate to add to the aura, since we no longer speak, but the truth is I heard this out of HAD's mouth 16-17 years ago. And not surprisingly, certain caps work better turned one way than another. And yes it had to do with ESR, how far apart the caps were, what design they used in construction, and a whole bunch of other stuff that went right over my head.

But you need a cap meter, multimeter and then you can test the caps you have to see which way they should be set to sound better, with less hum.

I am not one to stroke Dumble's ego, as I said we're no longer on speaking terms. But he was right about this.

Don't ask me specifics, though, it was a long time ago. I just know that it's true.

[R.G.]

It is very, very difficult to say that something like this is generically true. That is one reason that I qualified my post so extensively.

What I did with that post was to conjure up a set of circumstances based on extensive circuit design experience such that I could reasonably expect to see some signal change detectable with laboratory instruments, repeatably.

That is the litmus test. It is true that human minds can do wonderful work in digging subtle differences out of very noisy situations.

The dark side of that is the the human mind can conjure up amazing "false positives" out of raw randomness, too. The best illustration for this is no longer available. TVs before the digital age could be turned to a vacant channel and produce a full screen of random video noise - "snow". Humans readily perceived patterns in the "snow", when in fact there were none. Digital TV largely did away with that example.

There is another human psyche feature at work. Humans tend to think that anything the personally do is "progress". Amp tinkerers will often perceive any change as good as long as it's not grossly bad. And they will hear "differences" where there are none in blind tests so often that the subjective audio lunies will no longer participate in blind tests - they got embarassed too often.

The result of this bias is that psychological researchers have to go to extreme lengths to keep tests so nebulous that the participants can't guess what is being tested.

It's largely not possible for a human to do a fair listening test if they know when an actual component is changed, and which way.

It's even more murky when there is a slight chance, however rare, that there is a measurable difference there.

[bluesfendermanblues]

[quote="R.G."............... Amp tinkerers will often perceive any change as good as long as it's not grossly bad. And they will hear "differences" where there are none in blind tests so often that the subjective audio lunies will no longer participate in blind tests - they got embarassed too often.
[/quote]

I think you should add HiFi amp tinkerers. IMO guitar amp builders (usually) tend to be quite grounded in regards to perceiving circuit changes.

I've build 3 dumble clones and on the way tried all the designs in the "Dumble File" section. One (there are many) of the ingredienses in building a Dumble clone is making sure the signal caps mounted correctly in regards to the outer foil on the 6ps Orange drops. The noise level in the clean channel in those three Dumble clones is so low, that i sometimes wonder if an amp is actually switched on from standby. Pretty impressive for a tube amp. However, low noise levels is a minor thing, more importantly the amps sound great. Take a look at a gut shot of any Two Rock amp and you'll that the mark all signal caps as well.

BTW the Aiken tech note was very helpfull in regards to finding the polarity on so called non-polarity caps. I marked all my caps in my drawers with a black sharpie.

Just my two cents.

[bal704]

For those of us without an o-scope, is there anyway other way to find the polarity of these caps? I normally use Mallory caps for what it's worth.

[guitarmike2107]

Yes you can,

1. make up a short cable with a 1/4 jack Plug on one end and two croc clips on the other end.
2. Plug it into an amp volume turned up a bit and connect the cap between the croc clips, hold the cap body , not touching any metal parts, it will hum.
3 Now flip the cap around on the croc clips, it will still hum when you hold it but it will be louder or quieter
4. When it is loudest the outer shield is connected to the tip of the jack and visa versa.

This does work on high gain amps, I had one that I changed the caps due to noise and it made a small difference. I now also use polystyrene caps for small cap values as they can be orientated too and are much quieter compared to mica caps.

Polystyrene caps come marked as do many boutique caps like Sozo, it is easy to do and costs nothing.