|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
150.148.14.6
Yesterday I was obsessed with finding the lead attached to the outer foil for some 10uF film capacitors that I want to install in an amplifier. But none of the published ways of identifying the outer foil would work, which is to say that I was not able to detect any ACV difference between the two leads. And I tried 3 sightly different methods, using variously my ACV meter, my 'scope, and an audio signal generator. I then realized that this is probably because a 10uF capacitor has very little impedance when frequencies even as low as 20Hz are applied. So, to the cognoscenti, does this mean that orientation of a film capacitor is progressively less important, in terms of eliminating noise, as the value of the capacitor increases? Or is it worth it to pursue the issue until I do succeed?
Follow Ups:
All film capacitors are directional in as much as one lead of the capacitor is attached to the innermost foil and the other lead to the outer foil layer. There is a small audible difference to be heard, depending on which direction any film capacitor is oriented in a particular circuit. There is some disagreement as to the preferred orientation from user to user, and manufacturer to manufacturer. The most logical suggestion is to attach the outer foil lead closest to ground, thereby taking advantage of the inherent shielding of the outermost foil, but this may not necessarily sound best to you. The manufacturer of the Hovland capacitors suggest that the outer foil be at the source side in coupling applications or at the highest potential in power supplies, opposite to what one might expect. The manufacturer of the Wonder Cap/InfiniCaps/DynamiCaps has recommended both orientations depending on the particular vintage. Experiment and see which orientation you prefer. In order to determine which lead is outer, and which is inner foil, we use the following procedure for capacitors that are not marked to identify which is which. First, establish a good signal source. We use a signal generator supplying a 1K sine wave at about 15 volts, but you could use almost any combination of equipment that will output some kind of relatively steady low frequency signal. A CD player with a test CD playing a steady tone into a preamp turned to full output probably would work very well. You will also need a voltmeter with reasonably sensitive AC measuring capability. Attach test leads from the plus and minus output of your signal source to the two leads of your capacitor. Next wrap a wide piece of copper braid, any type of conductive material, even multiple turns of wire, around the body of the capacitor, assuring intimate contact over most of the capacitor surface. Take just the hot lead of your voltmeter and attach it to the braid or wire wrapped several turns around the capacitor (voltmeter ground lead is not attached to anything). Note the reading you get on your voltmeter. Now reverse the connection to the capacitor leads from your signal source and note the new reading. Whichever reading was highest indicates that the plus lead from your signal source was attached to the outer foil lead of your capacitor. You are just inductively measuring the strength of the signal on the outer foil of your capacitor. That's it!
I have the same findings. All film capacitors are directional soundwise. In some, the difference isn't small at all. I remember listening to a Mundorf RXT as a coupling capacitor. I considered throwing them because of the horrible signature they had. But after reversing them, they sounded really good.
This week, I listened to a slew of 4 uF caps on my 100+ dB ALTEC home system, and I was easily able to hear which polarity sounded BEST to my ear, in the crossover locations. I never measured inner or outer foils, because what really matters is how it sounds best !!'
Have FUN in 'yer audio journey.
Speakers with sensitivity over 100 dB sure do allow you to make aural determinations easily !!
The above post is what Michael Percy sent me, last year. A good man, is Percy !!
Jeff Medwin
Dear Jeff,
I have the same paragraph from M Percy on polarity and how to find it. Thanks for your good intentions however.
As to choosing the direction based on "how it sounds", that's a lovely idea in principle but messy in the application. Where these capacitors are going, deep into the circuit of a direct-drive Beveridge amplifier that sits inside the base of massive Beveridge speakers, it is simply not possible to use the "how it sounds" criterion. It would take me a couple of days to muster up the time to make the swap, during which, unless the differences were profound, I'd lose the sense of how one way might be different from the other.
My question was about why can I not detect a difference between leads, using M Percy's method and a few other different methods as well. Not about whether it's important or even a good idea to worry about it.
"My question was about why can I not detect a difference between leads"
Lew, first of all, are these being used in the circuit where position of the foil actually matters? If it's difficult to describe the application, maybe you could post a partial schematic showing where they're located in the circuit.
More to the point of determining the construction, can you sacrifice one cap for dissection? That might tell you whether the problem in making the determination is the type of construction or if it's the test methods you've been using. If the ambiguity is due to the type of construction, and the caps are in a circuit location where "polarity" matters, I'd suggest changing to a different type of cap.
--------------------------
Buy Chinese. Bury freedom.
Most 10uF film caps aren't rolled like a cigar. Could be more of a stacked film design. Anyhow I think it of minor importance compared to other things. I do agree as caps gets bigger it is harder to measure.
As TK mentioned it has a lot to do with impedance. The outer foil lead should go to the lower impedance side of the circuit.
Consider a coupling cap. Be defintion (i.e. for it to be coupling) the AC voltage across it must be near zero. That pretty much means your impedance at each end is roughly the same. The part a lot of folks miss is the impedance of the coupling cap itself. The stuff you are concerned about happens when one end is many kilo/mega ohms to ground and other end is at ground.
I suppose the point I'm trying to make is....when you have high impedance signal carrying wires/parts....it is more important to keep them away from unwanted noise signals. Don't think the "outer foil of the cap stuff" makes up for having that cap too near AC heater wiring!
I'll admit I can manage a "thought experiment" where outer foil location matters (and by all means it can't hurt so do as you wish)....but there are many areas where signal purity/intergity is far more important.....be it aerospace/defense/medical imaging/diagnostics/etc.....yet never once have I seen such caps marked, or tech notes suggesting one pay any mind to which lead goes where on a film cap.
In retrospect, I would like to re-phrase my question, because I think my original post is confusing. The heart of the matter is how would you identify the lead connected to the outer foil for a high value of capacitor (e.g., 10uF in this case) that has very low impedance even at very low frequencies, or, to put it yet another way, I am assuming that the reason the usual procedures don't work with this capacitor has to do with the low impedance of a 10uF capacitor; do you agree? If my assumption of cause is correct, then I was wondering out loud if it follows that the higher the value of capacitance, the less would be the theoretical benefit of orienting the capacitor in the orthodox fashion in the first place.
I am not at all expecting that identifying the outer foil lead is some sort of panacea for noise. In fact, in my past experience, I can't recall ever hearing a difference based solely on film capacitor orientation.
Make the cap part of an oscillator circuit....have cap sets its resonate/determining hertz. Place a conductive plate coupled to a very high impedance amp next to it. One lead of said cap must be grounded. If it happens to be the outer foil little signal will be coupled to high impedance amp.
Good luck. Wouldn't quite know how to do this myself with what I have at hand. Probably could get by with AC voltage across cap and foil wrapped around cap body connected to scope for quick and dirty method.
P.S. are you certain beyond all doubt your 10uF cap is a true film and foil construction?
You wrote, "Probably could get by with AC voltage across cap and foil wrapped around cap body connected to scope for quick and dirty method."
I tried that too, last night, but I think it is worth another try. Not quite sure my copper-backed tape (in lieu of foil) was doing the job.
Am I sure beyond any doubt that these are film and foil capacitors? Answer: I was when I bought them from either Mouser or Digikey, both of which seem to be reliable sources. Beyond that, no.
He used a signal gen and scope to determine which lead had the greatest electrostatic coupling to the exterior metal case. That lead should be the one tied to the "outer foil". This method is pretty simple but allows me to discriminate between random 60/120/180Hz stray AC noise and the desired signal.Steve's method I use on k40s and all metal cased caps for finding the outer foil but I haven't tried it on polys.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
Edits: 07/25/16 07/25/16
One method comes to mind, but it requires fairly expensive test equipment. You could ground one end of the capacitor, then drive the other end from a 50 ohm RF generator. Wrap four or five turns of wire around the body of the cap and feed that to a spectrum analyzer. The cap will shunt most of the energy to ground, but not quite all. A spectrum analyzer with a floor of perhaps -100 or -120 dBC will be able to detect the difference between the cap with the foil grounded and with the foil "live." You might be able to accomplish the same thing at 100 kHz using a low-Z output audio generator, a high gain audio amplifier (at least 60 dB gain) and an oscilloscope.
--------------------------
Buy Chinese. Bury freedom.
Yep. That is beyond my capabilities and my available test gear.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
He pronounces solder "sol der" not "sod der". Apparently either is correct depending on who you ask. I found it pronounced both ways on the web.
Nice link Mike, thanks!
Edits: 07/26/16
Mike, Thanks but I have watched this video before posting. I tried his technique and with my 10uF capacitor there is no difference between one lead and the other using my scope. This experience is in part what led me to post here. He seems to be looking at small value ( < <1uF) capacitors.
Great stuff Mike ...thanks sir
It's my pleasure.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
The importance of "polarity" depends on the use of the cap and impedance of the circuit. For example, a coupling cap should conduct well and have very little differential across it at most frequencies. Whatever noise appears on either lead will appear on the other. A bypass cap, however, should always have the outer foil connected to ground. By doing that, the "active" side of the cap where signal is present will be shielded by the grounded foil. Even in this situation, the degree to which polarity is important will vary depending on circuit impedance. Low-Z applications, such as cathode bypass, will probably not benefit much from this technique. As far as determining polarity, you might consider cutting one of the caps open. They're probably all the same in terms of the direction of printing on the body and the foil-connected lead.
--------------------------
Buy Chinese. Bury freedom.
I've only got four of these 10uF/100V polystyrenes, and I need exactly four. They seem to be No Longer Available anywhere that I have looked. (Not merely this exact product but 10uF polystyrenes rated at > 50V in general. I bought them 4-5 years ago.) So, cutting one open is not an option. Like Steve says, I too have found inconsistency in using the direction of the label as a guide.
However, thanks very much for your analysis of when it might be critical to orient the outer foil and when not.
Not necessarily unless specifically stated. Some are random (K40s). Depends on brand.
The lack of "polarity" may have to do with specific construction. I've run across film caps that are apparently 2 caps in series but wound simultaneously. Mostly HV caps. These don't seem to have an outer "foil".
I've got some 10uF metallized film capacitors. I will try the same tests with those, and that will at least answer my own question about whether the problem has to do with high capacitance per se. Should have thought of that last night.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: