[tubesforever]

I like a very fast, detailed and clean sound with lots of detail and dynamics.  You mentioned you have the 1.0 uf DynamiCaps and the AuriCaps.

Which does the above best.  I will be buying a Clarinet in the near future and want to use the best cap.  These are only about 5 dollars difference at many vendors.

Or is there a better option to discuss?

[analog97]

If someone could include a definition of "coupling" and "bypass" capacitors, I would greatly appreciate it.  I have run into this recently in my reading of capacitors and it confuses me.  Thanks.

[hagtech]

definition of "coupling" and "bypass" capacitors

Technically, they are the same, although most folks probably don't realize it.  Some forget the audio signal travels through the power supply caps just like it does the output cap.

However, they do designate different positions within a schematic.  The coupling cap is used to block dc in a signal.  For example C6 in the CORNET schematic.  A bypass cap is for de-coupling (how odd?) a dc power supply rail.  It is added to maintain a dc voltage during load changes.  C3 in the CORNET.

Both positions are very important for sonics, especially in single-ended amplification stages, where the signal current passes through both types.

jh

[tubesforever]

Jim, I agree with you about all caps being important.  I am not sure I have the cash to throw every flavor of the month at my Cornet2 but I am interested in getting a slightly faster and more dynamic sound.

Jim, would beefing up the final B+ electrolytic supply cap also create more dynamics? 

[Bill Epstein]

I put some Kimber Kaps in my crossover yesterday, replacing M-caps in the tweeter attenuation circuit.

Said to be similar to Auricaps, I find them richer sounding and less bright than the M-caps but also dynamic and detailed, moreso than Auricaps.

They're being discontinued and clearanced at several places such as Sonicraft at really good prices.

[PatOMalley]

doesn't 'bypass' mean soldering two caps into the same position to augment the sound of the larger value cap?

[analog97]

doesn't 'bypass' mean soldering two caps into the same position to augment the sound of the larger value cap?

Pat,

That's what confused me.  I am investigating re-doing my speaker x-overs and saw  the original scematic used.  Specifically there was a 12uf mylar and 750 pf silver mica cap in parallel.  The word I got was that was used to improve the sound of the mylar cap.  The opinion of the experts was to just use a good quality film cap (Sonicap, Solens, Auricap, etc) and forget the 750 nf cap.  People called this situation "bypass capacitors" referring to the little 750 pf cap. 

[PatOMalley]

right. 'people' have told me that using pio as a bypass can really really improve things for much cheaper.

[tubesforever]

Regarding bypass caps in a crossover.  Say you are building a high pass filter and the cutoff point you want requires 200 uf as a cap.  If you bypass the 200uf cap with a 1uf cap, then the higher frequencies will just zip along fast and clean while the lower frequencies will get choked by the larger cap.

This is why a speaker crossover should always include a bypass cap for the tweeters.  It really does sound superior.

The bypass caps in a preamp or amp are different.  Like JH said these are bypassing the dc from the power supply so it does not enter the AC signal path.  The final coupling cap just ensures no dc hits the amp and your speakers.

Cheers!

[analog97]

Regarding bypass caps in a crossover.  Say you are building a high pass filter and the cutoff point you want requires 200 uf as a cap.  If you bypass the 200uf cap with a 1uf cap, then the higher frequencies will just zip along fast and clean while the lower frequencies will get choked by the larger cap.

This is why a speaker crossover should always include a bypass cap for the tweeters.  It really does sound superior.

That's what I don't understand (among many things)...If you put a 12 uf and 750pf cap in parallel, why would the high frequencies just pass thru the little "bypass" cap? 

[w8aaz]

I would think you would want to use something like a polystyrene instead of a silver mica cap for bypassing large caps anyway.  Suppose SM would be better in audio circuits than small value ceramic caps where called for, for preventing RF oscillations or the like.  Can't see any need to bypass hi quality film caps with something like that. But if it sounds better to you, go for it. 

[analog97]

Let me rephrase my question.  Consider a speaker x-over where a 12 uf cap is specified.  Look at 3 possibilities:
1. replacing with a 12 uf cap
2. using 2, 6uf caps in parallel
3. using a 12 uf cap and a 750nf cap (a "bypass cap) in parallel

Why are these different? 

[tubesforever]

Once you determine the crossover values you want for a tweeter or midrange the the 1.0uf bypass cap should be a metalized polyprop or better.  The silver mica suggestion doesn't make a great deal of sense to me.  These are rated for high temp and high power applications.    A good metalized polyprop is perfect for a low voltage speaker crossover.

Paper in oil, copper in oil and silver in oil are also interesting options that I have not used in speaker crossovers.  I think these would be best as coupling caps in a high voltage application.

Speaker crossovers are extreme low voltage.  In fact you can measure the output of most amplifiers in miliamps.

[tubesforever]

BTW for a low voltage speaker crossover two 6 uf in parallel is a 12uf cap.  There is no difference if the voltage value is the same.

If you need a 100, 200, 300 uf high pass filter then you use anything totalling 100+1, anything totalling 200+1, and anything measuring 300+1, and so forth.

As long as the voltage rating and uf match your application then it doesn't matter.  On some of my speakers I have needed strange points like 126 uf.  So a 100 a 25 and a 1 are perfect.  The 1 is still considered the bypass cap and it makes up the total value I needed since a 26uf cap is not available.

Hope this helps.

[piedpiper]

On a gross level you are correct, but on a finer level, small caps will pass high frequencies more transparantly (faster) than larger caps. Consequently, two 6s should sound more transparant than a 12, and that is the reason for bypassing. Some designers purposely stay away from large values prefering to stack many small caps in parallel. For bypasses, you ideally want to use a value of no more than 1/100th of the large value so as not to have too much spectral overlap in that the large cap will not be very transparant at the frequencies that the 1/100th value cap is passing. This varies from type to type so, as usual, experimentation is the key as long as you are using burned in caps for your experiments. Also, all else being equal, using caps of voltage ratings no larger than necessary will help minimize the negative effects of excess dialectric, sounding more transparant.

[hagtech]

Also, all else being equal, using caps of voltage ratings no larger than necessary will help minimize the negative effects of excess dialectric, sounding more transparant

I'm not convinced yet this is true.  The thicker dielectric operates at a lower field strength.  Less volts per meter.  Perhaps this makes it more linear?  Also, the higher voltage caps have less leakage, which can be a factor in some high impedance circuits.  On the other hand, for the same uF value, the higher voltage cap will require a much larger surface area.

jh

[analog97]

So, is it considered dogma that using a very small value cap as a "bypass" somehow works by "attracting" high frquencies, i.e. the bypass cap permits high frequencies to "pass thru" more quickly?  This seems reasonable only if there is some scientific explanation for this, e.g. electrons taking the shortest path or maybe a difference in "potential" to shunt the high frequencies to the itsy-bitsy bypass cap? 
BTW, this reminds me of an old college bumper-sticker...."My karma ate my dogma"....

[hagtech]

It's not dogma - just marketspeak based on actual physical reality.  Bascially, the signal takes the path of least resistance.  The construction of capacitors results in imperfection, which is mathematically modeled approximately as a series RLC, ignoring the more exotic and nonlinear effects.  What happens is that a large capacitor generally operates well at lower frequencies, becoming inductive as the frequency increases past self-resonance.  This is where bypassing is beneficial.  The smaller cap has a higher resonant frequency and when combined in parallel with the larger they tend to work together to cover a wider range of frequencies.  So yes, one can simply say the smaller cap attracts the higher frequencies.  It does not, however, alter the velocity of propogation.  That is determined by dielectric.

Similarly, electrolytics often have relatively high ESR, as specified by dissipation factor.  A parallel film cap can lower this.

jh