Explain a bypass cap?

[CT Jim]

Using two caps in parallel, say a 15 + a 17 = 32uf. So is the smaller cap a bypass cap?

Plenty of info about using the correct cap for a speaker, whether a 16 or 32 or 8uf and adding, in parallel a bypass cap of say .1 or .47uf. Using the example of a summed capacitor pair, does adding a .1 or .47uf to it function as a bypass?

What constitutes a proper bypass cap setup, and does it always work???

 

[Yvan_donald]

Hi, mostly by pass cap will be a (good quality) small film cap like .1 uf with an electrolytic cap. That is what I understand about bypass?

 

[CT Jim]

But, does the bypass give a little edge to the woofer, or to the tweeter? In most cases I would think the woofer?

 

[GordonW]

The purpose of a bypass cap, is to "short out" the inductive losses (increase in impedance at high frequencies of a larger cap, due to the unavoidable inductance that exists due to a large "rolled" electrode inside) of a large-value cap. Especially with electrolytic caps- they can have significantly higher impedance at high frequencies. A small film cap should have virtually zero impedance at those higher frequencies- which should make the sum total of the two caps act closer to an "ideal" cap.

As such, it usually has more effect on the tweeter (as they go higher in frequency, naturally), but they can have an effect on woofers with higher-order crossovers (the bypass cap, by improving the behavior of a shunt cap in a lowpass filter, makes the lowpass on the high-end of the woofer crossover act more like an ideally-effective filter)...

Regards
Gordon.

 

[CT Jim]

My examples of a .1 or .47 is because I have some .47 but read using a .1 here on AK.

So, using a bypass everytime we recap isn't a bad thing, but rather a helpful thing, correct, he says after having closed up the NLAs just finished and forgetting the bypass cap.....duh

 

[GordonW]

Well- it CAN be a double edged sword...

IF the original crossover was VOICED around the inductance/HF impedance of a certain cap- adding a bypass cap CAN upset the overall balance.

I've had cases where I had to go BACK to an original design cap- in lieu of completely redesigning the crossover to compensate for the "better" cap.

It's something to always keep in the back of your mind...

Regards,
Gordon.

 

[CT Jim]

OK, maybe when I have time I'll pop one open and add the bypass to the singular 32uf I used.
But, does a bypass function differently when there is a summed capacitor setup of two caps?

 

[GordonW]

It'll have less effect- two or more smaller caps combined in parallel will have markedly less HF problems than one large cap. This is for two reasons:

1) The smaller caps each have less inductance
2) The parallel combination has less inductance than a single cap (since inductors in parallel act as a smaller inductor than either one- in fact two inductors of identical size in parallel, has an inductance of HALF of the original one inductor).

So, the bypass cap has "less to do"...

Regards,
Gordon.

 

[Yvan_donald]

Hi in my X over I have a 60uf electrolytic and change it for Cornell Dubilier film cap and First was not satified. I have in stock some .22uf silver lead janzten film cap, so I decide to use it and install it First on the board and connect the two 30 uf in parallel on the silver lead with it, and the sound was much better. But the way I figure was silver is a better conductor compare to tin copper, better transient. The other way the sound was muffle, don't no if I'm right the way I'M thinking, Yvan.

 

[woodj]

Gordon, your treatise about capacitor induction/capacitance relationships has answered one of my "curiosities".

 

[R Swerdlow]

The purpose of a bypass cap, is to "short out" the inductive losses (increase in impedance at high frequencies of a larger cap, due to the unavoidable inductance that exists due to a large "rolled" electrode inside) of a large-value cap…

Thanks for explanation why bypass caps might be used in a crossover. That’s the first reasonable explanation I’ve seen why bypass caps might make a difference.

I usually try to avoid most threads where various sound qualities of different capacitors are discussed. They tend to quickly get opinionated and controversial. Many of these capacitors, often the high priced ones, are said to give speakers extraordinary sound qualities, but I wonder if any of these have actually been tested. I don’t mean electronic lab bench tests of crossover components. I mean tests of acoustic performance with multiple listeners, under blind conditions.

I once participated in such a test. It involved about 40 DIY speaker builders who were interested in knowing if we could hear differences in speaker crossovers made with cheap non-polar electrolytic (NPE) caps, inexpensive metallized polypropylene (MPP) caps, or expensive caps of various types. The short answer was no one could hear any differences regardless of what type cap was used.

What actually did matter was whether the capacitance values were as specified in the crossover design ±10%. The trouble with the newly purchased cheap NPE caps was that many varied by more than 10% from their printed values. (80-90% of these cheap NPE caps were rejected for this reason when the crossovers for the listening test were being made.) Most, if not all the MPP caps, regardless of price, had capacitance values nearly exactly as printed. This made for the easy choice of inexpensive MPP caps over any other type.

I don't think other capacitor parameters, such as impedance or inductance were tested. I do know that one of the organizers of the test was an experienced EE, who certainly would have known about such issues.

This listening test was done in newly constructed speakers, so it didn’t address the question we often see here about aging speakers with old possibly out-of-spec NPE caps.

Also not tested was whether bypass caps, such as you described, made an audible difference. Maybe some day.

 

[ZOOM]

It's a snake oil theory invented by morons who know just enough of electronics to make dumb theories that sound plausible (to the ill informed). In reality, it does nothing.

 

[R Swerdlow]

It's a snake oil theory invented by morons who know just enough of electronics to make dumb theories that sound plausible (to the ill informed). In reality, it does nothing.

I don't disagree with you, but I'd word it differently. The differences in inductance between a large cap ± a small bypass cap may or may not make a difference in overall inductance in a crossover network, and may or may not make an audible difference to a listener. We wont know until it gets tested.

This reminds me of an old saying we had in the lab:

In theory, there is no difference between theory and practice. In practice, there almost always is.​

 

[sKiZo]

As stated, the smaller cap can improve small signal velocity for improved accuracy. The value of the cap determines the frequency range affected.

Or so the guy said, offering to throw in a free sample of his high end snake oil. ;-}

PS ... would residual VAC on the DC signal be a concern that late in the audio path, assUming good filtering in the amp itself?

 

[toxcrusadr]

I don't disagree with the stated reasoning here but...Why would you put an electrolytic cap in series with a tweeter in the first place? There's no need to minimize size or cost since those caps are typically small anyway. The only place I ever use NPE's is for larger values where the films get large and expensive, which is usually in the woofer section and maybe the shunt across the midrange. If you have a 20 or 30 or 60 uf cap that's shunting bass, what does a 0.1 uf cap have to do in that situation? Very little, it seems to me.

However, in a two-way speaker where the tweeter cap is maybe 12 uf (like an original Advent), maybe it would do something to the high end. This is one place I might try it. But in a part of the crossover that handles lower frequencies, I don't see the point.

 

[Salomon]

"had to go back to the original design"

I have been looking for bypass caps to tweak an already good woofer crossover and never having done this, was inclined to get "the best" I could afford.

As I look I'll bear in mind I may be going back to the original caps.... which don't have a brand name on them. They just say NP 1068-7402 30uf 70V.

Thanks for the insight.

 

[RTally]

"had to go back to the original design"

I have been looking for bypass caps to tweak an already good woofer crossover and never having done this, was inclined to get "the best" I could afford.

As I look I'll bear in mind I may be going back to the original caps.... which don't have a brand name on them. They just say NP 1068-7402 30uf 70V.

Thanks for the insight.

Don't confuse "original design" with "original caps." GordonW said "I've had cases where I had to go BACK to an original design cap- in lieu of completely redesigning the crossover to compensate for the "better" cap." I interpret that to mean he used a single electrolytic cap like was used in the original crossover design, not that he re-used he original cap.

 

[Salomon]

Don't confuse "original design" with "original caps." GordonW said "I've had cases where I had to go BACK to an original design cap- in lieu of completely redesigning the crossover to compensate for the "better" cap." I interpret that to mean he used a single electrolytic cap like was used in the original crossover design, not that he re-used he original cap.

Good point. I trust the existing design and will keep it.

 

[StimpyWan]

"had to go back to the original design"

I have been looking for bypass caps to tweak an already good woofer crossover and never having done this, was inclined to get "the best" I could afford.

As I look I'll bear in mind I may be going back to the original caps.... which don't have a brand name on them. They just say NP 1068-7402 30uf 70V.

Thanks for the insight.

I always use Parts Express for my bypass cap purchases. They sell Dayton Film & Foil capacitors, designed for bypass use. I use 0.1uf or 0.01uf, depending on the size of the cap being bypassed. I've used them on NPE's and poly's.

Parts Connexion sells Vishay film caps, that are for bypass use. Sonic Craft has the SoniCap Gen 2 and Platinum too. Expensive though. The Cornell Dubilier bypass caps are highly rated as well.

Good luck.

 

[Salomon]

Thanks, I'll save those parts sources.
I probably don't understand the terminology correctly. What I need are two 30uf caps for the bass leg of each crossover or 4 each.

If I go with Jentzen Superior Z caps for this as I have in the mid/high legs I'd need 4ea per speaker which puts the cost at about $525 for the bass legs. I might do this just to find out what that sounds like.

I'm looking through what the Humble Home capacitor comparison has to say and found the Jentzen Crosscap is inexpensive and comes in fairly large values but is rated as "average".