Tone Wheel Recap - Voltage Choice of Capacitors ?

[Randall_Dibble]

Ok, While researching recapping my tone wheel I discovered that the voltage of the old caps to be replaced are valued at 200volts. Give that the cap are in fact are handling millivolts what are alternate voltage choices do I have ? Or is it simply getting caps of the correct uf and long enough leads to span the distance between the terminals ?

[tiredoldgeezer]

Oh boy! "Cap Talk."

Ordinarily, a capacitor in series blocks DC. But where is all this DC coming from? It's a sine wave.

I could offer some meaningless speculation, but I'm sure someone has a better story.

Like you, ain't seeing the 200V either. But there it is.

[tpappano]

Or is it simply getting caps of the correct uf and long enough leads to span the distance between the terminals ?

You got it! Back in the day, 200v was about the lowest voltage rating available for that type capacitor. Most any polyester (Mylar) or polypropylene caps will work great. For example, Panasonic 'ECQ' series are inexpensive and compact. Be careful when ordering because various lead lengths are available, I was in a hurry one night and wound up with short "pcb mount" leads 8). The .255 are not real common, I used two caps in parallel for those filters.

[Randall_Dibble]

SO, other then getting the correct uF value, lead length and enough voltage to cover the signal strength are anything to selecting capacitor material for a good sound ? Are orange drop caps something to consider ? or just about any type will be better then the original waxes?

[tpappano]

Pretty much any polyester or polypropylene is good. I read that many seem to be fans of the old "orange drops" but in my humble opinion as a manufacturer, they should have no advantage over any other polyester cap. If one was really wanting to split hairs, some capacitors can have lower (better) ESR (effective series resistance) which provides higher theoretical performance of the capacitor but since they are connected to coils that already have substantial resistance, ESR in this circuit is pretty much not a factor.

[David Anderson]

ESR in this circuit is pretty much not a factor.

With respect, in this particular instance, this is not exactly correct.

I would strongly suggest using Mylar and not polypropylene here. Usually, I prefer polypropylene capacitors, but this is a rather unusual application. The DF/ESR of the capacitor used does affect the filter's bandwidth. This is a case where you have to analyze the circuit and do the math, which I have done on a spreadsheet. Paper and Mylar have approximately the same dissipation factor, but polypropylene has about 1/10th the DF of paper or Mylar. Polypropylene will cause the filter to have a narrower bandwidth (sharper slope) and make the tuning of the filter more critical; in other words, as you move away from the capacitance/inductance combination that peaks the filter, the slope will fall off faster with polypropylene, magnifying any mismatch. Mylar should get you closer to the original design parameters.

It is, in fact, the increasing DF of the paper capacitors as they age that widens the bandwidth and makes it so the filters still work at all, considering the drift in capacitance value.

In most audio coupling applications, the DF/ESR of a film capacitor is not a significant parameter but it matters here.

[David Anderson]

And keep in mind that Hammond did tune the filters via selecting capacitors based on capacitance value. They reportedly tested the capacitors, separated them into several ranges based on capacitance, and chose a value that best tuned that filter to its associated frequency.

[Randall_Dibble]

So, if I under you, David, correctly the choice of a Mylar cap will better reflect the character of the original caps having a gentler slop and wider bandwidth which results in less adjustment needed in re-calibration ? Will this also result in a wider range or effect in the use of the draw bars ?

Do you recommend a particular Mylar cap ?

The Mylar caps I've found are expensive !

[tpappano]

There should be some inexpensive mylars, for example:

https://www.mouser.com/ProductDetail...A%2FbghBod8%3D

$0.145 each if you buy 100 pcs.

I'll defer to David's judgement on the polypropylene/polyester issue, he tested both. I used polyester on 3 TGs and polypropylene on one, but I have not compared their relative bandwidths.

[tiredoldgeezer]

Polypropylene will cause the filter to have a narrower bandwidth (sharper slope) and make the tuning of the filter more critical; in other words, as you move away from the capacitance/inductance combination that peaks the filter, the slope will fall off faster with polypropylene, magnifying any mismatch.

So maybe help me out here, just what does all this do for the circuit?

I mean slope, peak, bandwidth, and similar terms I understand about filters. But the input to this circuit is a sine wave created by a rotating wheel. So the frequency is fixed with or without the filter. It's a sine wave, so there's nothing to speak of in harmonics.

What you're saying makes sense in something like a VCF or a full range filter, where slope and bandwidth would matter, but these filters really don't filter anything, do they? And if so, what could it be? And if nothing is being filtered, what is achieved with a selection of one type over another?

I suppose the need would be to achieve resonance via the tank circuit and in an ULZ circuit (16 ohms?) prevent crosstalk or back feeding. But that wouldn't be affected by mylar/poly/wax/etc. would it?

[tpappano]

I think one issue is that the wheel produces close to a sine wave, but not a perfect one. I think the intent of the filters is to clean up the tone by reducing unwanted harmonics and also perhaps reduce some crosstalk from other wheels.

edit: and I think what David is getting at is if a polypropylene cap produces a narrower bandpass curve, there is less forgiveness for variations in the actual L and C values, which in turn could make the signal levels coming out of the filters more variable from note to note. A wider filter response would tend to reduce these variations.

[Randall_Dibble]

So, does a Polyester Cap = Mylar Cap ?

Mylar being a brand name to describe a Polyester Cap ?

[David Anderson]

Mylar is the original trade name for polyester made into a thin film -- as opposed to threads or fabrics.

[tpappano]

You are correct!

[Randall_Dibble]

So, the bottom line here is that using a Mylar/Polyester Cap will help reduce variations in signal strength and require less adjustment in re-calibration ? or even many not require re-calibration ?

[tpappano]

Yes, that would be true. On my TGs, the three with the polyester caps sounded good without any adjustments. On the TG with the polypropylene caps (my first), I did adjust some of the pickups to get the levels within 1 dB of their neighbors but that was probably me just being a little OCD 8) the other three I just left alone because they tended to be within 2 to 3 dB, which to my ears sounded fine. I did a ‘waterfall' plot on my living room sound system which showed as good as ones speakers may be, what the actual room acoustics do to sounds once they leave the speaker is pure mayhem! I think the same is true with an organ 8)

edit: I'm curious enough about the bandwidth issue that I think I'll do some measurements on an actual TG filter and see what the two types of actually do

[David Anderson]

edit: and I think what David is getting at is if a polypropylene cap produces a narrower bandpass curve, there is less forgiveness for variations in the actual L and C values, which in turn could make the signal levels coming out of the filters more variable from note to note. A wider filter response would tend to reduce these variations.

This is correct.

The "problem" with modern capacitors is that they tend to be pretty close to the nominal value. In the old days, Hammond took advantage of the range of values (centered around 105nF) to select capacitance values to tune the filter and center it on the filtered frequency. If you have to err, it's better to err on the higher side of capacitance. I did tests with a decade capacitor substitution box, and if the peak value is 105nF, but you only have 100nF and 110nF, 110nF will sound better. That's what my ears told me listening to a single tone at a time. As you drop under the capacitance value that peaks the filter, the tone starts to sound thinner. You can believe me or not on that. I'm not going to take the time to get into a long argument about it.