NPE Caps

[davepellegrene]

I have been using the Dayton Metalized Polypropylene Capacitors on all my builds basically because I don't know any better. I couldn't help but to notice you can get the NPE caps for SO much cheaper. So my question is why not or is there any part of the circuit that they can be utilized. The only difference I see is they are 100V instead of 250V. If this is really the only difference why not parallel them to match the volt rating.
I bought a bunch of the buy out woofer crossovers for $3 a piece mainly for the 9 mH coils. They also have several 75 and 200 uf caps. I have tried them in my temporary circuits then put the Metalized poly caps in, measured both and see no change in the measurements. I have also checked them with my meter and they seem to be accurately rated.
Hopefully this won't start a cap thread fire. Just curious if I can save some money.


Dave

[carlspeak]

Re: NPE Caps

What many folks seem to do is use a low value film cap bypassing NPE's when a large uF value is needed in woofer and midrange parallel wired circuits. This saves mucho $. I replaced a no-longer available Sprague compulytic 2000 uF cap with a bundled set of 500 uF NPE's in an AR LST with no problem. To do that with film caps you'd have to take a second mortgage out .

I am in the process of writing an article for audioXpress about the measurement of caps. Testing I've done shows NPE's are more variable over the entire 10-20kHz frequency range than film types.

[mzisserson]

Re: NPE Caps

I have been using the Dayton Metalized Polypropylene Capacitors on all my builds basically because I don't know any better. I couldn't help but to notice you can get the NPE caps for SO much cheaper. So my question is why not or is there any part of the circuit that they can be utilized. The only difference I see is they are 100V instead of 250V. If this is really the only difference why not parallel them to match the volt rating.
I bought a bunch of the buy out woofer crossovers for $3 a piece mainly for the 9 mH coils. They also have several 75 and 200 uf caps. I have tried them in my temporary circuits then put the Metalized poly caps in, measured both and see no change in the measurements. I have also checked them with my meter and they seem to be accurately rated.
Hopefully this won't start a cap thread fire. Just curious if I can save some money.

Dave

NPE's can change with temperature, humidity, time (not all that long either). It is why it is best to keep them out of the signal path. That said when need, choose the lowest DF possible. I use the Erse 3% DF caps whenever possible. They cst a whopping $1-$3/ea, but well worth heafty investment.

I also found bypassing with a small value poly like a 0.1uF helps stablize the NPE and seems to give teh benefit of a poly with cost of an NPE.

[davepellegrene]

Re: NPE Caps

So you guys are saying to bundle, or parallel, a low value "quality cap" with the NPE's?
Part of the reason I ask the original question is it seems every time I order parts I end up making changes, then order more parts and on and on. So, my order, arriving today, I ordered a bunch of NPE's so I can bundle them together so I can make changes with more flexibility. Then once I get it the way I want, buy the quality caps. I assume as the sacrificial NPE's change I will just have to measure them to make sure what the values are as time goes on.

Make sense?

Dave

[mzisserson]

Re: NPE Caps

So you guys are saying to bundle, or parallel, a low value "quality cap" with the NPE's?
Part of the reason I ask the original question is it seems every time I order parts I end up making changes, then order more parts and on and on. So, my order, arriving today, I ordered a bunch of NPE's so I can bundle them together so I can make changes with more flexibility. Then once I get it the way I want, buy the quality caps. I assume as the sacrificial NPE's change I will just have to measure them to make sure what the values are as time goes on.

Make sense?

Dave

Dave, that is what I do while prototyping x-overs. There are many times also where I just end up bypassing with a daton film and foil, or poly, and call it done. Other times (mostly in mids and tweeters) I spring for better qualty replacements once honed in.

You may not measure much of a difference. It is when they are asked to work under loads/changes they begin to act up over time.

[ocdSCHACK]

Re: NPE Caps

NPE's can change with temperature, humidity, time (not all that long either). It is why it is best to keep them out of the signal path. That said when need, choose the lowest DF possible. I use the Erse 3% DF caps whenever possible. They cst a whopping $1-$3/ea, but well worth heafty investment.

I also found bypassing with a small value poly like a 0.1uF helps stablize the NPE and seems to give teh benefit of a poly with cost of an NPE.

So how do you implement the cap?
-In the beginning of the circuit?
-before or after the NPE?
-Before or after every NPE?

It seems like a great idea.

I found some good info here for those that know nothing: http://www.seattlerobotics.org/encod...97/basics.html

[50 watt head]

Re: NPE Caps

why not parallel them to match the volt rating.

Voltage is common in parallel circuits. So, for example, the voltage rating for two 100v caps in parallel is 100v. (If you put them in series, though, you would effectively double the rating to 200v.)

Parallel capacitance is additive. So, for example, two 10uf caps in parallel will yield 20uf net capacitance. Another example: a 10uf NPE and a 0.1uf film in parallel give a total of 10.1uf.

[carlspeak]

Re: NPE Caps

So how to you implement the cap?
-In the beginning of the circuit?
-before or after the NPE?
-Before or after every NPE?

It seems like a great idea.

I found some good info here for those that know nothing: http://www.seattlerobotics.org/encod...97/basics.html

Your link was an interesting read. However, much of it doesn't apply to passive loudspeaker xover circuits that run on AC. Only on very rare occasions will DC get to a xover circuit because of some amplifier defect, and I believe, the cap's principle of operation will block any DC from getting to the drivers which could cook the voice coils.
To answer your questions, bypass caps (roughly <=10% of NPE value) are wired right around a large value non-polar electrolytic cap.

[Chris Roemer]

Re: NPE Caps

I have been using the Dayton Metalized Polypropylene Capacitors on all my builds basically because I don't know any better. I couldn't help but to notice you can get the NPE caps for SO much cheaper. So my question is why not or is there any part of the circuit that they can be utilized. The only difference I see is they are 100V instead of 250V. If this is really the only difference why not parallel them to match the volt rating.
I bought a bunch of the buy out woofer crossovers for $3 a piece mainly for the 9 mH coils. They also have several 75 and 200 uf caps. I have tried them in my temporary circuits then put the Metalized poly caps in, measured both and see no change in the measurements. I have also checked them with my meter and they seem to be accurately rated.
Hopefully this won't start a cap thread fire. Just curious if I can save some money.


Dave

Hey Dave, this is absolutely just MY opinion, but . . .

MOST DIYers (especially those who DON'T design/build several speaker pairs every year; AND those who's hearing is not what it used to be - so not applying to those in their 20's or 30's, but definately for those in their 40's, 50's, and beyond; AND those who are not taking out a loan to buy top of the line drivers, i.e. Scan Speak, Seas (higher end), or the latest offerings from morel, or the "exotics" - planars, ribbons, and such) can get by with 99.9% of all their cap needs by buying just 2 series of caps from PE.

The 027-4xx series "Dayton Audio Metallized Polypropylene" (I THINK these are the ones you've been using). In the catalog it states "10% tolerance", BUT it SAYS +/-5% RIGHT ON THE CAP! EVERY one of these I've ever measured has been between its stated value and +TWO PERCENT maximum. So they are very good and make the extra cost for the "Precision 1%" caps seem quite silly to me.

The 027-3xx series "100v Non-Polarized" (npe) caps. Known as "NPE" (non-polarized electrolytic) due to the fact that they use an electrolytic "paste" inside, and YES, the paste can dry out over a period of time, causing opens, shorts, and just unpredictable performance. If you had 100 of these that were made in the 1970's, about 25% would NOT be operable TODAY. These average about ONE SIXTH the cost of the metallized polypropylene (mpp) types. Typically they AVERAGE about 5% over their stated value (running between their stated value and +10%, and sometimes even higher), but if you account for their value, they can be perfectly acceptable for use in all areas of passive crossovers.

Since MOST higher powered home audio amps seem to be capable of pushing about 40 volts out, the 100v rating of the npe's isn't really an issue. If you need values above 100uF, I'd DEFINATELY use npe's, not the mpp's (which you'd have to "bundle" to get values that high). If you need values below 1uF, you'll have to go mpp, since the npe's don't go that low. Generally, I LIKE to use the mpp's up to around 25uF, where they start to get pricey.

By application, MOST people feel that if you CAN resolve a change in sound (due to being young, with excellent hearing, and using primo drivers) that you'll be most likely to hear it on caps in series with a tweeter, and not so much with caps in parallel (to ground) or those on mid or woofer circuits (so, you COULD use mpp's on the high pass section of a midrange circuit, but npe's would be OK for caps in the low pass section (which are in parallel) and for Zobels).

Have I ever used an npe in series with a tweeter? Absolutely!

You know what? On your next project, build it with the mpp's (like you have been), and, since the npe's are so cheap, make a version of the XO using npe's and see if you can hear a difference. If you can't, and you don't care if your caps dry out in 20 or 30 or 40 years, try the npe's and use the money saved for upgrading other aspects of your projects.

Chris

[critofur]

Re: NPE Caps

So how to you implement the cap?
-In the beginning of the circuit?
-before or after the NPE?
-Before or after every NPE?

MUST be in PARALLEL with the NPE!!!

...they are 100V instead of 250V. If this is really the only difference why not parallel them to match the volt rating.

I believe that paralleling them - say, 2 x 10uF 100V yeilds: 20uF 100V.

However, 100V should be plenty adequate, anything more is generally overkill (or, as some like to call it: "headroom").

[PWR RYD]

Re: NPE Caps

I'm pretty much with Chris Roemer on this one. If you're buying buyout drivers or drivers under $50 each, then the NPE cap's quality isn't going to be the limiting factor. If you are designing and spending the $$$ on Seas or Scan Speak or (insert high end driver name here) drivers... then of course, why not spend the cash on high end caps. That said, I always buy the Dayton poly caps unless the value is over 40 uF, otherwise the NPE feel just fine to me Flame suit on :D

[Face]

Re: NPE Caps

I'm pretty much with Chris Roemer on this one. If you're buying buyout drivers or drivers under $50 each, then the NPE cap's quality isn't going to be the limiting factor.

In my experience, you're wrong. $5 drivers, maybe.

[ocdSCHACK]

Re: NPE Caps

I generally use NPE's and had the exact same results as Chris Roemer when I measured my caps:
NPE's were 1-10% on the high side. Only a couple were in the 1% range. Most were 5% or more over the stated value.
Dayton's MP caps were much closer. I never measured a single one that was +/-5%. They were all much closer.
I usually use NPE's on my own speakers...especially for testing. For a speaker I'll never see again, it all depends.

And Critofur, thanks for pointing out the obvious. I'm not sure how that didn't register the first time.

[johnnyrichards]

Re: NPE Caps

There are "in-between" caps out there, Erse Audio has a nice selection of Mylar caps and so does Madisound. They generally run quite a bit cheaper than an equivalent value polypropylene, and more than NPE. IMHO they represent biggest upgrade over NPE (if you believe such things). I use them for woofer circuits, and when voicing.

[daryl]

Re: NPE Caps

Capacitor bypassing is only effective for power supply rails and similar applications where clamping is necessary over an extreme bandwidth.

Bypassing NPE's with film caps has no significant effect in crossovers.

As far as quality is concerned there is no driver in existence at any price which can compare to the quality of an NPE.

You could not build a speaker with a film cap crossover and also an NPE cap crossover and then measure a significant difference in the distortion levels of the finished speakers.

NPE's will dry out at some time but the ones I have (old as they are) measure as they should.

I do like the idea of film caps and use them for small values but if someone uses NPE's I couldn't say they are wrong to do so.

I do have three 500uf NPE's (in parallel 1500uf total) in my subs shaping their third order sealed alignments and have no reservations.