[Stimpy]

Bypass capacitors?  What are they, and why use them?  Is there a sonic or electrical benefit, or both, to using bypass capacitors?  And if it's good to use them, is there a formula for determining the correct value of a bypass?  Also, should the voltage of the bypass capacitor be greater/higher than the capacitor it is bypassing?  Any other qualifiers for bypassing?

I know this is a broad range of questions.  But, I read these same questions, almost everyday on various audio forums, and I truly don't know the answers.  For my own benefit, I would love an explanation, if possible?  Any insights, from anyone who would care to contribute, would be greatly appreciated.  Especially if you've used bypass capacitors, and would like to elaborate on their benefit (sound improvement)?

Thanks!   

Edit:

I'd like to share any responses in this thread with the Audio Karma Forum, if OK?  This same discussion is happening there too.  Maybe I'll link this thread, when it get's interesting?

Thanks again.

[GETRDUN]

From what I've read on speaker crossovers, the formula for bypassing is 1% of the total value of the cap you want to bypass. Tony Gee seems to like the .01uF/3000V Cornell Dubilier 940C caps now, even better than the Vishay MKP1837 caps that used to be his favorites. He says to use them everywhere!

Perhaps someone like Jeff G. could give us a better explanation on this! 

[Stimpy]

Thanks GetRDun,

The 1% formula is the only one I've read about too.  But, I seem to just buy either 0.01uF or 0.10uF bypass caps, and be done with it.  Maybe that isn't optimal, but its worked so far.  Now, I was hoping Jeff might comment, though I'm sure he's very busy this time of year.  As such, I'm willing to wait for his input, as I know it'll be worth it.   

Merry Christmas

Oh yea, I've been looking at the CD 940C caps.  I've got the Mouser page open now.  Deciding...

[FullRangeMan]

Bypass capacitors?  What are they, and why use them?  Is there a sonic or electrical benefit, or both, to using bypass capacitors?  And if it's good to use them, is there a formula for determining the correct value of a bypass?  Also, should the voltage of the bypass capacitor be greater/higher than the capacitor it is bypassing?  Any other qualifiers for bypassing?

I know this is a broad range of questions.  But, I read these same questions, almost everyday on various audio forums, and I truly don't know the answers.  For my own benefit, I would love an explanation, if possible?  Any insights, from anyone who would care to contribute, would be greatly appreciated.  Especially if you've used bypass capacitors, and would like to elaborate on their benefit (sound improvement)?

Thanks!   

Edit:

I'd like to share any responses in this thread with the Audio Karma Forum, if OK?  This same discussion is happening there too.  Maybe I'll link this thread, when it get's interesting?

Thanks again.

Capacitors are a way to store electrical energy, its not hi efficiency methoid for audio, but its the usual oldfashioned way. Other way to do power supplies are the inductive power supply as used by the Italian builder Bartholomeo Aloia, there is a AC member that have a Aloia pre/power set.

Inductive power supplies of course use inductors to store energy, I was told on the ear they are much more powerful than regular capacitive PS.

[GETRDUN]

Thanks GetRDun,

The 1% formula is the only one I've read about too.  But, I seem to just buy either 0.01uF or 0.10uF bypass caps, and be done with it.  Maybe that isn't optimal, but its worked so far.  Now, I was hoping Jeff might comment, though I'm sure he's very busy this time of year.  As such, I'm willing to wait for his input, as I know it'll be worth it.   

Merry Christmas

Oh yea, I've been looking at the CD 940C caps.  I've got the Mouser page open now.  Deciding...

You're welcome Stimpy! I'm also looking forward to reading what Jeff has to say! Let me know if you happen to order the CD caps from mouser. I'd like to know what you think.

Merry Christmas to you as well 

[rpf]

Bypass capacitors throw off the timing between the drivers. When I removed them from my Cronus Magnum amp, my speakers gained a great deal of coherency, i.e., the treble info no longer arrived before the mid-range music. (My speakers are time-aligned, phase coherent models).

I would never use them, but rather seek to increase the quality of the main cap.

[Stimpy]

Bypass capacitors throw off the timing between the drivers...I would never use them, but rather seek to increase the quality of the main cap.

That's why I hope Jeff will respond.  Sonic Craft manufacturers and sells SoniCap Gen II and Platinum capacitors.  Both are typically used as bypass caps.  Why market these, if they're a downgrade in audio quality (which I don't think they are).  There's definitely a Love/Hate vibe about bypass caps, which I don't understand?  I suppose because I've never experienced anything negative in their use.  Some have, and have no desire to try them again.

[rpf]

Sonic Craft manufacturers and sells SoniCap Gen II and Platinum capacitors.  Both are typically used as bypass caps.  Why market these, if they're a downgrade in audio quality (which I don't think they are). 

Both of those capacitors are used in other ways than as bypass ones, i.e., it's not that the capacitors in themselves are a downgrade in quality (though of course there are better - and more expensive - ones). Smaller value caps are faster in output than larger value ones and therefore their information reaches the speaker drivers sooner than that of the larger value capacitors. They may indeed perhaps extend the frequency response of the unit they're used in (or just add "air") but at the cost of smearing the timing. People are more or less sensitive to timing issues so, as usual, it becomes a personal preference.

[GETRDUN]

Both of those capacitors are used in other ways than as bypass ones, i.e., it's not that the capacitors in themselves are a downgrade in quality (though of course there are better - and more expensive - ones). Smaller value caps are faster in output than larger value ones and therefore their information reaches the speaker drivers sooner than that of the larger value capacitors. They may indeed perhaps extend the frequency response of the unit they're used in (or just add "air") but at the cost of smearing the timing. People are more or less sensitive to timing issues so, as usual, it becomes a personal preference.

That is interesting! If only there were one simple formula/components to make that perfect sound! I've went down that road some years ago with trying different interconnects, speaker, and USB cables, and noticed changes in quality for better and worse! I also know that from my own personal listening experience, that it comes down to the components that are in the complete system! I don't think that one could hear much difference in a cheaply built system versus a well designed and of higher quality system! "Not saying that less money versus more money is worse" but rather design and quality!!! But usually, you get what you pay for! 

Now a few questions on the bypass caps: Suppose one would take the cheaper Bennic, Solen, or Parts Express film brand caps and use them in the passive crossovers with high quality built drivers, will one notice a difference in sound quality versus using Duelund capacitors?

How about if one likes the sound of bypass caps, should one choose Polypropylene/Teflon, Aluminum film, copper foil, etc.??? And will these different types make enough difference to hear and say that is the magic cap?

How about the Duelund CAST-Cu-Ag Hybrid Loudspeaker Capacitor 100VDC.  Technical specifications (according to manufacturer): "Inside the Duelund CAST hybrid silver-copper capacitor there are two capacitors: a main copper foil capacitor that is bypassed by a small silver capacitor. They are both potted together. The lead-outs are intertwined from both the pure copper foil capacitor & the pure silver bypass capacitor hidden within the CAST potting. Premium VSF-CAST speaker capacitor; pure copper / silver; handmade in Denmark."

BTW, I looked up distributors in the USA for Duelund, and can you believe that they list Parts Connexion, who by the way is IN Canada, a USA distributor?  Now if you are lucky enough to have low valued caps in your tweeter circuit, you may be able to upgrade to these Duelund caps for maybe under $2,000.00 for a 2nd order filter! They better be able to sing me lullabies each night before bed-time! 

[Stimpy]

Both of those capacitors are used in other ways than as bypass ones, i.e., it's not that the capacitors in themselves are a downgrade in quality (though of course there are better - and more expensive - ones). Smaller value caps are faster in output than larger value ones and therefore their information reaches the speaker drivers sooner than that of the larger value capacitors. They may indeed perhaps extend the frequency response of the unit they're used in (or just add "air") but at the cost of smearing the timing. People are more or less sensitive to timing issues so, as usual, it becomes a personal preference.

Danny Richie stated something similar on his board.  He liked using bypass capacitors, on larger values, as he felt it improved the discharge rate of the cap.  I believe Danny felt that it improved the transient response of the speaker; the perceived 'speed' of the speaker.

[HsvHeelFan]

Capacitors store energy.  They charge and discharge.  They're equations for how fast they charge/discharge.

Larger caps charge and discharge slower and at lower frequencies.

Smaller caps charge and discharge faster and at higher frequencies.

In classic power supply design, large caps and small caps are used.  They're primarily used to  eliminate ripple noise on the outgoing voltage.  The important thing is that the voltage be stable under load.

In my experience, you need to know what noise/ripple that you're trying to get rid of before you can pick the correct capacitor value.  An oscilloscope really helps for this.   You can see how much ripple amplitude you have on the line with a voltmeter set to AC voltage, but that won't tell you the frequency.

Here is a webpage on Capacitor basics:

http://www.cmm.gov.mo/eng/exhibition/secondfloor/MoreInfo/2_3_5_ChargingCapacitor.html

Make sure that the cap you use exceeds the maximum voltage expected. Make sure you have some safety factor built-in to the cap voltage rating that you choose.    Also choose fairly tight tolerance capacitors if you want consistency between caps.   The caps are manufactured and then they're evaluated and graded for tolerance once they've been fabricated (resistor tolerance is done the same way). 

HsvHeelFan

[taww]

I have not experimented with bypassing in speaker crossovers much, mostly power supplies and coupling caps in line-level electronics. As a general rule I am not a fan of bypassing in signal applications because you can often hear the effect of having two disparate capacitors with different sonic signatures, and the sound becomes less coherent. This is particularly true if the capacitors are of different types, most extremely electrolytics bypassed by a film, because the ESR/ESL/etc. are so different.

That said, it is a matter of taste, and some things will sound better with a bypass. My Monarchy NM24 tube DAC/line stage came with 3.3uF Wima MPK10's bypassed by .01uF polystyrenes. I took the polystyrenes out because I felt the treble and midrange were a little disjointed - the treble was a bit more prominent and the midrange sounded a step behind. I liked the sound better w/o the bypass, but my wife (an oboist with better HF hearing than me) commented woodwinds lost some of their detail and natural brightness - the polystyrenes definitely improved resolution of highs and harmonics.

And in my Bryston B60 integrated I'm actually breaking my rule big-time by using Mundorf E-cap bi-polar electrolytics bypassed by a Vishay MPK1837 with modified leads (as detailed here: http://audiophile-musings.blogspot.com/2015/01/the-vishay-mkp-1837-review-and.html). I used Cardas litz wire as the lead, very thin (I think 36 gauge) to add a little resistance to the Vishay and get the ESR a little closer to the Mundorf. And I have to say, it sounds really good! The amp sounds fantastic this way... I can't say it's better than a single high-quality film cap (probably not) but it matches the sound of the amp well and there isn't enough room inside the slim chassis for a big film cap anyway.



So yeah, the only rule is there is no hard rule. Experimentation is key but you can do some basic calculations to get a sense of where the effects might fall. The 1% rule is fine but I prefer to factor the circuit load and use the standard 1/2*pi*R*C formula to determine where the corner frequency of the bypass cap will be, and that's where you can expect its effects to kick in. In the case of the Bryston, the preamp drives an effective load of around 8.5k ohms, so the .01uF Vishay kicks in around 1.8kHz. That's not a bad spot, not too far off from where you'd want your crossover frequency to be in a 2-way speaker. I'd avoid bypasses that kick in right in the midrange or lower, it has a higher chance of sounding phasey or discombobulated. Again, YMMV.

[taww]

BTW in my Ayre integrated I was using 4.7uF Russian PETP (Mylar) films bypassed with the .01uF polystyrenes I took out of my DAC, because Mylar for sure needs the help in the upper frequencies. It sounded ok this way but the top end was a tad too fast and crisp relative to the midrange. I replaced this with 2.7uF Jantzen Alumen Z-caps with no bypass and this is overall the best of the few caps I've tried. It is so silky smooth and full yet detailed and dynamic. Yes, the top end could be a little faster and more extended with a polystyrene or Teflon bypass (I'd probably go for 2200pF here given the circuit impedance) but it's bright enough as is and I don't really feel the need to muck around with the sound further right now. At some point I might try a Rel-cap RTE polystyrene, maybe not...

p.s. So far I've found Humble Homemade Hifi's capacitor descriptions to be quite convergent with my own listening, even though he is testing it in loudspeakers and I'm using them in line-level applications.

[Stimpy]

And in my Bryston B60 integrated I'm actually breaking my rule big-time by using Mundorf E-cap bi-polar electrolytics bypassed by a Vishay MPK1837 with modified leads (as detailed here: http://audiophile-musings.blogspot.com/2015/01/the-vishay-mkp-1837-review-and.html). I used Cardas litz wire as the lead, very thin (I think 36 gauge) to add a little resistance to the Vishay and get the ESR a little closer to the Mundorf. And I have to say, it sounds really good! The amp sounds fantastic this way... I can't say it's better than a single high-quality film cap (probably not) but it matches the sound of the amp well and there isn't enough room inside the slim chassis for a big film cap anyway.

I have a friend that was recapping sets of Acoustic Research AR9 and AR90 speakers.  He'd tried several well known and respected poly caps, and didn't like their sound on the upper dome midrange.  Too strident.  He could only listen a few minutes before ear fatigue set in.  So finally, at my urging, he tried the Mundorf E-Caps, and loved them.  He also added a Vishay bypass, and liked that combination even more.

p.s. So far I've found Humble Homemade Hifi's capacitor descriptions to be quite convergent with my own listening, even though he is testing it in loudspeakers and I'm using them in line-level applications.

I haven't tried many of Tony's recommendations yet.  I don't have that many pairs of speakers!  But, I have tried Mundorf EVO Oils and Supremes, in my AR90's.  Not much playing time yet, as I'm still refinishing the cabinets.  But what I have heard, is very promising. 

I also just recapped my AR58S speakers.  This time I used ClarityCap CSA caps on the tweeters, and ESA caps on the midrange domes.  For those caps, the Homemade HiFi comments were 100% spot on to what I was hearing.  Warm, but veiled, and rolled off in the upper mids and treble, to me.  Exactly as Tony said.  I'll give them a few more weeks of burn in before I decide to add bypass caps, since Tony seems to think that helps with the Clarity's.  If the bypass caps still fail to elevate them, I guess it'll be time to order more Mundorf's from Jeff...!   

[GETRDUN]

I have a friend that was recapping sets of Acoustic Research AR9 and AR90 speakers.  He'd tried several well known and respected poly caps, and didn't like their sound on the upper dome midrange.  Too strident.  He could only listen a few minutes before ear fatigue set in.  So finally, at my urging, he tried the Mundorf E-Caps, and loved them.  He also added a Vishay bypass, and liked that combination even more.

I haven't tried many of Tony's recommendations yet.  I don't have that many pairs of speakers!  But, I have tried Mundorf EVO Oils and Supremes, in my AR90's.  Not much playing time yet, as I'm still refinishing the cabinets.  But what I have heard, is very promising. 

I also just recapped my AR58S speakers.  This time I used ClarityCap CSA caps on the tweeters, and ESA caps on the midrange domes.  For those caps, the Homemade HiFi comments were 100% spot on to what I was hearing.  Warm, but veiled, and rolled off in the upper mids and treble, to me.  Exactly as Tony said.  I'll give them a few more weeks of burn in before I decide to add bypass caps, since Tony seems to think that helps with the Clarity's.  If the bypass caps still fail to elevate them, I guess it'll be time to order more Mundorf's from Jeff...!   

I'd be interested to know how you rate the Clarity caps versus the Mundorf's in your high and midrange circuits? What series Mundorf caps will you compare to the Clarity CSA/ESA? I hope that Jeff G will comment soon! I imagine he was busy for the Christmas and New Year holidays.

[Stimpy]

I'd be interested to know how you rate the Clarity caps versus the Mundorf's in your high and midrange circuits? What series Mundorf caps will you compare to the Clarity CSA/ESA? I hope that Jeff G will comment soon! I imagine he was busy for the Christmas and New Year holidays.

It'll be some time before I can do the comparison, but I'll definitely post concerning what differences I hear.  My AR58S speakers are already recapped, but I have to complete the refinish on my AR90 cabinets, before I can listen to those again.

As to the caps, the 58's use a 3.9uF CSA bypassed with a Dayton 0.1uF Film & Foil cap.  The 1.5" dome mid has a 25uF ESA cap.  No bypass there.  The shunt values are all Bennic NPE's, bypassed with 0.1 Dayton Film & Foils.

Now, my AR90's are a bit more complicated than the 58's, with 4 series capacitors, as opposed to the 2 series caps in the 58's.  The 4 series caps are 4uF, 6uF, 24uF and 80uF.  They are made up of Mundorf EVO Oil capacitors.  The 4 is a 3.9 EVO Oil and a 0.1 Auricap.  The 6 is a 3.9 EVO Oil and a 2.2 Mundorf Supreme.  The 24 is a 22 EVO Oil and a 2.2 Supreme Silver/Oil.  The 80 is a 33 and a 47 EVO Oil, with a Supreme 0.1 bypass.  The shunt caps are Axon True Caps, again with Dayton 0.1 Film & Foil caps.

So, both the 58's and 90's have the same tweeters and dome upper mids.  Very similar paper cone drivers for the rest.  When the speakers are finished, the comparisons should be worthwhile.  My system primarily uses a Yamaha RX-A3030 A/V receiver, for preamp and surround duties.  Sources are a XBOX One, a Sony Blu-Ray player, and a Pioneer CD player, with SACD and DVD-A capabilities.  Amps are a rebuilt Hafler DH500, 2 Amber Series 70 amps, and a couple of NAD 2155's, if needed.  Oh yea, a Sonographe SG-3, with a Grado Signature 8MZ cartridge, for LP's.

[GETRDUN]

I'm curious as to why you would use mixed values for the 6, 24, and 80uF caps? I was under the impression that using the same values to get the target value was fine, such as (3) 2uF caps to make the 6uF value or (4) 6uF caps paralleled to make the 24uF value, but using other values outside of this may cause audible effects! I stand to be corrected if this is not the case!

I'm also wondering if substituting the Dayton .1uF caps with say a Cornell Dubilier 940C / 0,01uF / 3000VDC, like Tony Gee suggests, would be an improvement?

FWIW, I have been told that by using the Jupiter copper foil caps in a speaker crossover circuit is not the best, rather the Jupiter VT Round caps would be the way to go! Definitely not the cheapest upgrade..lol

Thanks again for sharing your builds and mods!     

[mikeeastman]

I think you have it backward, the Jupiter copper foil caps are better than the VT caps. I used the Jupiter foil caps to replace the Sonicaps in my speaker crossovers, pricey but  a very nice improvement.

[GETRDUN]

I think you have it backward, the Jupiter copper foil caps are better than the VT caps. I used the Jupiter foil caps to replace the Sonicaps in my speaker crossovers, pricey but  a very nice improvement.

I was actually told from Chris (CEO of Jupiter Condenser) that in his opinion, he preferred the sound better from the VT caps versus the Copper foil caps in speaker crossovers!

[Stimpy]

I'm curious as to why you would use mixed values for the 6, 24, and 80uF caps? I was under the impression that using the same values to get the target value was fine, such as (3) 2uF caps to make the 6uF value or (4) 6uF caps paralleled to make the 24uF value, but using other values outside of this may cause audible effects! I stand to be corrected if this is not the case!

While some brands of capacitors have a wider range of available values, that would have exactly matched my AR speakers, Mundorf didn't.  So, I had to parallel caps, to match up what I needed.  Plus, I was (somewhat) following Jeff's recommendations, and Jeff didn't seem to think 'mixed values' were an issue.  Good enough for me.

I'm also wondering if substituting the Dayton .1uF caps with say a Cornell Dubilier 940C / 0,01uF / 3000VDC, like Tony Gee suggests, would be an improvement?

Yea, I've been considering the CD 940C's.  If I add a bypass to the ESA in my 58's, I'll probably try them.

Thanks again for sharing your builds and mods! 

You are welcome.  My pleasure.