bypass caps bad idea?

[sydsfloyd67]

Maybe Gordon can help out on this one. Is there any chance that adding a small 10% bypass or even 'cascade' capacitors inline with a tweeter would actually be a disadvantage? I can't find any consencous on if a small bypass cap is a good idea in the first place, or a good explanation of even the theoretical benefit. (Only possibly reducing something like "hysteresis" diststortion.)

A cascade would be using say a 10uF in a parallel with a 0.47uF and a 0.1UF, each with a higher voltage rating, rather than just a 10uF. Can anyone tell me what difference that would make with the flow of signal to the tweeter.
Thanks, -sf

 

[ferninando]

this will be interesting to see the replys. My questions are, do you want to run a separete signal to tweeter by itself and not already in a system.? And will you also be driving a system as well? Where will the tweeter signal come from? A simple high pass filter to the tweeter in question would work and you would have to define the freqs you want to pass. To my mind, there's more involved that just placing a cap in series to the tweeter. Others Im sure will jump in with more details and questions.

 

[clifselina]

Bypass Cap

A bypass cap is a good idea. Put it in Parallel with your existing cap, not series. You should notice better highs with the bypass. Think about how a cap works. It charges up to a certain voltage level and then starts to discharge when the voltage level changes. The amount of time it takes to charge and discharge is dependent on the resistance of the circuit and the value of the cap. A very small value cap charges and discharges much more quickly than a larger value cap. This is why you can get better sound by placing a small value, high quality capacitor in PARALELL with the cap(s) leading to the tweeter.

If you place the bypass cap in SERIES with the existing crossover, the small cap will block almost all of the higher frequencies.

 

[jcmjrt]

I thought that this was interesting information on bypassing and cascading capacitors:

http://www.northcreekmusic.com/Bypassing.html

 

[sydsfloyd67]

A bypass cap is a good idea. Put it in Parallel with your existing cap, not series. You should notice better highs with the bypass. Think about how a cap works. It charges up to a certain voltage level and then starts to discharge when the voltage level changes. The amount of time it takes to charge and discharge is dependent on the resistance of the circuit and the value of the cap. A very small value cap charges and discharges much more quickly than a larger value cap. This is why you can get better sound by placing a small value, high quality capacitor in PARALELL with the cap(s) leading to the tweeter.If you place the bypass cap in SERIES with the existing crossover, the small cap will block almost all of the higher frequencies.

Not really an issue. The thread asks if there is any disadvantage to the bypass cap idea, and for a description of the theoretical benefit. The context is the simplest possible crossover -just an inline (series) capacitor on the tweeter +.
-sf

 

[sydsfloyd67]

this will be interesting to see the replys. My questions are, do you want to run a separete signal to tweeter by itself and not already in a system.? And will you also be driving a system as well? Where will the tweeter signal come from? A simple high pass filter to the tweeter in question would work and you would have to define the freqs you want to pass. To my mind, there's more involved that just placing a cap in series to the tweeter. Others Im sure will jump in with more details and questions.

I had hoped to discusss the concept of a bypass cap in a simple passive crossover as a low frequency filter.
-sf

 

[dgwojo]

I was recently told by the engineers at Madisound that bypass caps are not a good idea anymore as they provide uneven signal paths, they were an OK idea when audio caps had 20% tolerences, todays 5% Solen Fastcaps don't need by-pass caps. I went without them in my Altec N1209-8A crossovers and they sound perfect in my Altec A7's. I do still believe in Zobel's for the woofer, Dave.
:yes:
http://audiokarma.org/forums/showpost.php?p=315664&postcount=30

 

[sydsfloyd67]

I was recently told by the engineers at Madisound that bypass caps are not a good idea anymore as they provide uneven signal paths, they were an OK idea when audio caps had 20% tolerences, todays 5% Solen Fastcaps don't need by-pass caps. I went without them in my Altec N1209-8A crossovers and they sound perfect in my Altec A7's. I do still believe in Zobel's for the woofer, Dave.:yes:

D., Thanks for passing on the comment. I would like to understand what the bypass cap is supposed to be doing in the first place then. (and how a closer tolerance would make that undesirable?) -sf

 

[dgwojo]

D., Thanks for passing on the comment. I would like to understand what the bypass cap is supposed to be doing in the first place then. (and how a closer tolerance would make that undesirable?) -sf

Hopefully I still have the e-mail at work, I'll dig it out and see what else it had to say, I was always a believer in bypass caps just to improve the treble transient response. I don't know why more precise caps have a make a difference, I think it's probably just that Solen Fastcaps are polyprop's and those are fast by nature. Another thing mentioned in the e-mail about unmatched signal paths was that when getting an exact cap value with combinations, use like values, say you couldn't buy a 28mfd, it suggested four 7mfd, or two 14mfd's. I saw this on the web about caps, you might check it out; http://www.eminentaudio.co.uk/Systems/OddMods/OddMods.htm
Dave. :scratch2:

 

[sydsfloyd67]

Hopefully I still have the e-mail at work, I'll dig it out and see what else it had to say, I was always a believer in bypass caps just to improve the treble transient response. I don't know why more precise caps have a make a difference, I think it's probably just that Solen Fastcaps are polyprop's and those are fast by nature. ...... I saw this on the web about caps, you might check it out; http://www.eminentaudio.co.uk/Systems/OddMods/OddMods.htm
Dave. :scratch2:

D., Interesting. What do you have in mind as far as "fast" and "transient" with regard to the bypass cap idea? That web site is pretty interesting, but raises this to a level of mysticism rather than what I was looking for in terms of a sound theoretical idea behind the use of a bypass cap in a simple crossover for running the tweeter.
-sf

 

[dgwojo]

Regarding fast and transients, I use all Altec Speakers with compression horn drivers, they're only 2 way systems, for me the less crossover points the better. The horns I use can go all the way down to 500hz on the low-end to 20khz on the high end, the compression drivers have to cover a pretty large spread, including all the treble. It's amazing how crisp and clear these horns are even though they're covering the mids and highs concurrently, they do it effortlessly, I believe the caps help.

This link had some info that echo'd what was quoted earlier about cheap caps needing bypass caps, here's the link and the quote, now if I could just find the "advanced" section they mention?? :scratch2:
http://www.members.aol.com/_ht_a/pjay99site/FAQs.htm
http://www.partsexpress.com/webpage.cfm?&DID=7&WebPage_ID=70

Do bypass capacitors work?
Maybe. A bypass cap is a very small value high quality cap people put next to a cheap cap on the crossovers to improve the sound of a speaker. The theory is that good electrons going through the bypass cap are somehow going to clean up the bad electrons going through the cheap cap. Some people swear by them. Our theory is that electrons are pretty stupid and any tweak that requires electrons to have intelligence is faulty to begin with. Try it and see if you can hear the difference. A better bet would be to replace the cheap cap with a good cap. Check the advanced section for more theory on this.

 

[Don C]

Our theory is that electrons are pretty stupid and any tweak that requires electrons to have intelligence is faulty to begin with. Try it and see if you can hear the difference.

I think that this is possibly the stupidest explanation of bypass caps that I have ever read. What the hell is he talking about?
Hysteresis distortion is not hard to understand, capacitors do charge at a non linear rate as the voltage applied crosses 0 volts. Adding a small capacitor in paralell helps to reduce the distortion. It is not audiphool nonsense, it is a real source of distortion with a real cure.

 

[Giskard]

I think that this is possibly the stupidest explanation of bypass caps that I have ever read.

I had to read it twice to make sure I read it right... naturally I bypassed the advanced section. No pun intended.

 

[sydsfloyd67]

...Hysteresis distortion is not hard to understand, capacitors do charge at a non linear rate as the voltage applied crosses 0 volts. Adding a small capacitor in paralell helps to reduce the distortion. It is not audiphool nonsense, it is a real source of distortion with a real cure.

D.C., Thanks indeed for getting this going in the right direction again. As someone with a very primitive idea of a capacitor, can I think of the small (low uF rated) bypass cap as being able to charge and recharge more quickly? Will it perhaps be passing signal for the highest signals only, while the main cap will be passing these and those down to the crossover point?

(If that is by any chance correct, then I am wondering if there are these two pathways for signals above a certain fequency, and the smaller bypass can operate more rapidly, a temporal "smear" would occur?)

-sf

 

[rek50]

PE has some AudioCap .01 that won't break the bank. Your ears will tell you, if you made the right move.

 

[ilimzn]

Caps working in a crossover are difficult to understand from a standpoint of charging and discharging (although that is what they do). It is better to think of them as impedances that change with frequency.

Bypass caps do work in principle, the technique is widely used wherever a capacitor needs to work in a very wide band of frequencies. For crossovers, it remains a question wether you will see any benefit from them in a given installation as this depends on the type of both caps (main and bypass) as well as where exactky they are used.

The idea behind a bypass cap is not to miracoulosuly do a better job because it is smaller. It has to do with the non-ideal nature of capacitors - mostly that they are wound components and hence can manifest inductance. Their electrodes also have finite resistances so there is also a series resistance component present (usually this is less of a problem unless a completely inadequate cap is used). The inductive parasitic component results in the cap having a resonant frequency (rarely more than one), which is normally well above the upper end of the audio band (unless you use truly atrocious caps). Over this frequency, the inductive effects overwhelm the capacitive and the cap usually behaves exactly the oposite than expected - as frequency increases, it's impedance increases, instead of decreasing. A smaller bypass cap is used because the point where it stops behaving like a cap is being caused by mechanical reasons - smaller cap has it's problems at a higher frequency. Having two caps with very different resonant frequencies in paralell results in an equivalent component that better approximates an ideal cap because once the larger one's characteristics degrade over the resonant frequency, the smaller one 'takes over'. Of course I am GROSSLY simplifying here, but the basic principle is there.

The trick here is that today you can much more easily get good quality caps large that have very low parasitic inductance, for which obviously a bypass cap will not do much if anything (for instance, their resonant frequency, after which the cap starts behaving more like an inductor, happens in the MHz range, far above audible frequencies). Also, if you use an inductive cap as a bypass cap, you will probably end up having two problems instead of fixing the one you had before you added the bypass. The advantage of splitting a single cap into two is that you can chose the type of the two caps to get closest to an ideal cap. This measure is also prone to the law of diminishing returns - some caps in a crossover will greatly benefit from this approach, for others it would be wasting money.

 

[sydsfloyd67]

......The trick here is that today you can much more easily get good large quality caps that have very low parasitic inductance, for which obviously a bypass cap will not do much if anything (for instance, their resonant frequency, after which the cap starts behaving more like an inductor, happens in the MHz range, far above audible frequencies). Also, if you use an inductive cap as a bypass cap, you will probably end up having two problems instead of fixing the one you had before you added the bypass. .....

I., Thanks for the insights. The part of your reply above goes to the specific question posed by thread. Without any consencous here (or anywhere else) as to the value of using a bypass cap as proposed, the answer for me was "why not"? However if it might actually be a bad idea, that is something else. With a good modern polyproylene film cap like the Sonicap 10uF ($17) or Dayton ($~5), perhaps the use of a small film and tin foil bypass (0.1 uF) is so unlikely to be of any benefit that it is only a liability? (Not so easy to decide simply by listening at this level- and clearly certain reproduction artifacts and deficits can be precieved as favorable by many listeners.)
-sf

 

[GordonW]

Reading the original message, about adding a 10% bypass cap... I wouldn't add THAT big of an additional cap. Maybe 1% or less of the size of the original cap... it it's a 1uf to 10uf, then a .01uf to .1uf bypass cap would be right. Don't want to upset the original value significantly- changing the value too much would change the crossover point, the slope, the Q of the crossover (it's damping near the crossover point) and such... a 1% change shouldn't cause problems, but 10%? Probably asking for trouble...

The idea is, like it was said, to reduce hysteresis... and to counteract inductive effects. A bypass cap can somewhat "short out" hysteresis charge-discharge storage effects, and it can also "shunt" inductive effects present in most ANY large-value cap (a cap is nothing but a WOUND-UP coil of foil... a COIL, ie, it WILL have a certain INDUCTANCE, which is proportional to physical SIZE). A small cap can indeed improve the high-frequency performance of a large cap...

Regards,
Gordon.

 

[WhiskeyRebel]

What if the smaller value cap is not physically smaller? To make my desired capacitance out of the pieces in stock at Abel Electronics, I paralleled a small-value poly cap with two larger-value nonpolar electrolytics, but the poly is physically larger.

 

[sydsfloyd67]

Reading the original message, about adding a 10% bypass cap... The idea is, like it was said, to reduce hysteresis... and to counteract inductive effects. A bypass cap can somewhat "short out" hysteresis charge-discharge storage effects, and it can also "shunt" inductive effects present in most ANY large-value cap (a cap is nothing but a WOUND-UP coil of foil... a COIL, ie, it WILL have a certain INDUCTANCE, which is proportional to physical SIZE). A small cap can indeed improve the high-frequency performance of a large cap...Regards,Gordon.

G., Was hoping you would join in. Yes, 0.1 uF as a bypass at 1%, that was the idea. Is there any simple idea of what 'hysteresis' is (for a simple guy like me)?

Other than that, I have here what seems like a physcially very big 10uF cap from Sonicap. Likely pretty high quality, but if I understand correctly, the big size and corresponding inductance capability does warrant a 0.1 uF film-foil Auricap as a bypass? Is a second 'cascade' bypass like a 0.47 uF here getting ridiculous? -sf