Caps advice

Bypassing

I'm actually doing the same thing with my Lite DAC-AH but I'm bypassing some Russian 2.2uF K75-10 (PLIO) caps with some .22uf Russion T3 (teflon) caps.

1) actually bypass caps are usually of a higher grade, not lower. the purpose fo a bypass is to have a smaller value faster cap, often a lower quality larger cap will be used with a higher grade bypass because high grade caps in large values are physically large and expensive. so this allows you to get some of that better performance at the extremes without paying the penalty. what works best in any given situation is often more like cooking, trying different combinations and seeing what works best.

2) no, your cascade values are wrong, the bypass caps need to be set up in a way so the harmonics work well together, generally a factor of 10 or 100 is used, so for a 2.2uf cap the recommended bypass would be 0.22uf, for audio signal path I wouldnt worry about the 0.022uf (22nf) as the 0.22 is already operating at the very extremes of the audio band. smaller bypasses like the 0.022 are IMO more useful in power supply bypassing and other filter purposes in the digital realm.

3) I would just connect it in parallel. you wont have both connected at the same time and the small amount of extra wire will not add a meaningful amount of capacitance. also the bypass itself is in parallel, what makes the connectors any different if you really want to you could install a simple 2P2T switch, leaving the grounds connected. make sure to ground the case if you can (provided its metal) it will provide some shielding.

hope that helps

I agree with gusp on almost everything he said.

#1 is particularly important - "like cooking" is a nice way of saying that bypassing caps in the audio path is often unpredictable. Combinations that you may think work well can turn out sounding awful. Others that you may not suspect will work, can turn out sounding great.

Ditto on #2 as well. I've found that there's little benefit in even going much beyond 0.2uf for bypassing in any cap arrangement. 0.1uf works well and 0.47uf can, too. Anything much above that or below that is useless for most loads around here. That maybe why many of us settle on ~0.22uf.

More often than not, you may have a great sounding (and perhaps expensive) primary cap that may have a glorious midrange and deep bass, but is often not fast enough to handle the higher frequencies needed for detail (often referred to as "sparkle"). That's where bypassing is useful. A good, fast, film cap can do the trick. In many cases this need can be met with a good polypropylene box cap such as a Wima or Roederstein MKP - even MKT's can work well under certain circumstances. In the Millett MiniMAX/MAX world, we also use VitaminQ's for bypassing to good effect.

These two links may prove useful to you:
ecp.cc - Notes on Output Coupling Capacitors
Humble Homemade Hifi - Capacitor Test

The Great Capacitor Shoot-Out
Humble Homemade Hifi
ecp.cc <<< There's more links at the bottom of this page, thanks to dsavitsk.

There's lots more info on the web, but I have found that tinkering around meself is by far the most rewarding. It could be a hobby in and of itself.

hehe that would be a first as my experience in building anything beyond cables becomes more solid, I would think that our paths will converge somewhat. with my preferences probably still remaining slightly more esoteric when appropriate. my username bolded for future reference

yep, much to my (and probably your) dismay, even the best critical thinking and planning never quite predicts the reality in this area, since we are pretty much leaving the organic chemistry and physics do its thing in a passive combination of components. this way even our most sophisticated modelling can and will often produce chaotic results. there is nothing actively directing electron flow through a bypass arrangement, we are just letting the combination of materials interact; willful little things electrons

so you wouldnt agree with putting say a 100pf silver mica cap across the input/output pins and/or power supply of an opamp? I have found good results with opamps that are very sensitive to noise or 'cranky' in other ways, I will often combine this with a 1uf. though I havent tried this with a scope and is totally subjective; but yes as stated i'm pretty much with you on that.

for sure, I suppose in the case mentioned above I wouldnt always go for a smaller value mundorf SIO, a 0.22uf teflon would do just as well and these are not always expensive. I also have found great effect from wima and vishay films (mostly 0.1uf) in fact I just bought a crapload of silver mica, wima and vishay MKP for bypassing high quality electrolytic caps in power supply and signal duties.

wima MKP are a bit pricey though (at least at farnell australia they are) compared to the equivalent vishay MKP, I havent actually used them that extensively and usually use vishay films for this. have you found the wima to be worth the premium? I may grab a few more to try out

of course I doubt we'll ever agree on cables but hey we dont have to

Actually, most of what you mention in this paragraph wouldn't be classified as "audio capacitor bypassing." In a power circuit, real and measureable results are obtained by bypassing large electrolytics with smaller film caps. Likewise, bypassing certain opamps pins is a provable electronic improvement for opamp performance - usually by bypassing certain pins to ground to maintain stability and prevent oscillation.

I'm not quite sure about the case across the input/output pins.

But yeah, we agree on the rest of it. Moreover, you mentioned cables - I'm not sure what you are referring to that I might disagree with, but I've always thought that cables made a difference with headphones. I have a set of Cardas on a pair of HD600's myself - they definitely make a difference.

oh, I wasnt sure you were narrowing it to only audio band stuff as I mentioned digital circuits as well, but ok for sure its mainly just local bypassing of the power supply.

me neither that was badly worded, i've been playing around with some new opamp adapters that have multiple pads for different purposes, mainly supply bypassing of the supply, but also the option for the input and output to ground, thus my qualifier that I really didnt know whether there was improvement or not. I was actually meaning across input to ground and output to ground, not providing a highway for any HF noise from the input to the output hehe

aha, right I guess second hand info is exactly that, I had made an assumption (dangerous) based on the lack of any 'audiophile grade' parts in any of the BOMs or even recommendations that you might be a non-believer as many engineers are. agreed headphone cables IMO make the most profound difference of all. plus someone mentioned that you had mentioned you didnt think upgrading hookup wire etc in the bet would make much difference. I suppose in your position you must also maintain a bipartisan outlook to a degree

You may not even hear a difference with bypassing. Your Mundorf's may be doing an excellent job on their own.