Bypass capacitors - do they work?

**JoshK** · 13 May 2005, 19:03 Friday

I too am curious as to your opinion JM! Or anyone else with direct experience.

As I mention before, I am soon to be receiving my DEQX and since my current speakers use a Aurum Cantus G2si ribbon tweeter, I have heard it is wise to place a 'protective' cap below the xo point. I scribbled out that I would need around 18-19uf for a pole at ~1500hz (@6R) for actively crossing around 3Khz.

Of course I plan to experiment, but the Thetas are spendy, so I was wondering your guys' take on the subject. Practical experience is usually worth more than 'audiophile wisdom'. I am considering bypassing an 18uf Solen/Dayton/other value cap with a 1uf Theta or SoniCap.

**ThomasW** · 13 May 2005, 20:06 Friday

Jon doesn't use the standard tiny value of bypass cap. His concept is to use high-quality cap for approx. 1/2 the total cap value needed.

I was doubtful of this upgrade, but layed out big bucks for 1/2 dozen of the 10mfd Theta Audio-Cap. I sat in a chair and 'supervised' while Jon upgraded the top end of my Arvos. The difference was immediately audible and certainly not subtle.

I can't comment about the benefit of a tiny bypass cap. But hey, they're just not that expensive.

**cjd** · 13 May 2005, 20:22 Friday

FWIW, if you're considering that upgrade, I would also take a gander at Jon's Modula crossover and consider using that. Though, I have no doubts about Darren's work I suspect the differences are going to be more obvious than a bypass cap might be.

I could be wrong.

C

**JonMarsh** · 13 May 2005, 21:45 Friday

We ALL can be wrong at one time or another... yours truly included.

I'm curious about Darren's crossover and the "BBC Notch" it results in... but finding time to get new baffles cut and measurements made for the RS28A in the Modula's just doesn't seem likel to happen this month- a magazine article deadline was pulled in from late to early June, and believe me, I'm in panic mode on the weekends for "work" work.

Sometime soon I'll at least model it using my existing data, and check that against the prelim modula crossover for the RS28A I've done.

PS, I don't find that putting small bypass caps (like 0.1 uf) in parallel buys one much of anything - it's my experience that you need quite a bit of the overall value to be the film and foil cap.

Part of the Evolution (e) updates for the Ayre electronics now includes the switch from metalized foil polypropylene caps to film and foil... for exactly the same sonic reasons.

I think it's likely you'll hear benefits with your Rotel gear- how much it's not easy to say. Should be more focus and presence in the whole tweeter range (remember, I'm crossing the tweeters over in the 1200 to 1400 Hz range, so they handle a lot of the presence and treble), and the overal sound stage image is better focussed and reveals more of the original acoustic ambience.

~Jon

**JoshK** · 14 May 2005, 01:39 Saturday

Thanks, I'll keep that in mind. To date I have only used a handful of caps in different applications and find that I really liked the SoniCaps (compared to say Solens) but haven't tried many of the other spendier (word?) caps. It doesn't hurt to buy a few, since I'll end up using them somewhere sometime.

**dlr** · 14 May 2005, 10:50 Saturday

Originally posted by mickster1972

Anyway, I am always hunting for my next project or tweak...

Jumping in here, not on caps. If you're looking for tweaks, my suggestion is to consider a felt application to the baffle, primarily for the tweeter itself.

I use this as an integral part of designing my systems, but it can be done as an after-the-fact mod. The changes are orders of magnitude larger than caps, is cheaper and is reversible, all without touching the crossover unless that is also desired.

The June issue of audioXpress (finally) has my article on felt, I just received my two complimentary issues. If interested, check it out.

dlr

**Shawn Solar** · 14 May 2005, 19:44 Saturday

Noob question...

To utilize a bypass cap one splits the initial value and uses two smaller caps in its place. One being foil and the other being a regular off the shelf type?... or I'm I completely of base...hmmmm

**Paul H** · 14 May 2005, 21:24 Saturday

Originally posted by Shawn Solar

Noob question...

To utilize a bypass cap one splits the initial value and uses two smaller caps in its place. One being foil and the other being a regular off the shelf type?... or I'm I completely of base...hmmmm

Correct - and the two (or possibly more) are connected in parallel.

Paul

**JonMarsh** · 15 May 2005, 11:13 Sunday

Originally posted by dlr

Jumping in here, not on caps. If you're looking for tweaks, my suggestion is to consider a felt application to the baffle, primarily for the tweeter itself.

I use this as an integral part of designing my systems, but it can be done as an after-the-fact mod. The changes are orders of magnitude larger than caps, is cheaper and is reversible, all without touching the crossover unless that is also desired.

The June issue of audioXpress (finally) has my article on felt, I just received my two complimentary issues. If interested, check it out.

dlr

Dave's quite right- my three part series on the M8 two way in AudioXpress spent a moderate amount of time discussing the felt diffraction control impelmented and the before and after.

My recommendations for cap changes assume that you've already addressed basic issues like that in your project...

~Jon

**mickster1972** · 15 May 2005, 13:50 Sunday

Thanks to erveryone for the input. I really appreciate it. If Dave and Jon could let me know which issues of audioXpress discuss the felt application, or let me know how I can get copies of those artilces, I would appreciate. I don't subscribe to that magazine.

**dlr** · 15 May 2005, 18:21 Sunday

Originally posted by JonMarsh

My recommendations for cap changes assume that you've already addressed basic issues like that in your project...

~Jon

Which issues were those (to save me the time of digging through my unorganized pile

)?

dlr

**dlr** · 15 May 2005, 18:23 Sunday

Mine's easy...

it's just coming out, the June issue.

dlr

**mickster1972** · 15 May 2005, 22:17 Sunday

Great. thanks for the direction to your article. OK, final question from me on this topic. I guess this one is directed towards Jon Marsh or ThomasW: Assuming I address the other "basic" issues alluded to in this thread (felt application being the primary one), and assuming my source, controller, amp and cables are up to grade, what would you say are the minimum values (as a %) for the film/foil caps in bypassing to get acceptable results? Jon, a few posts up (on 5/13) you said "quite a bit of the value" should be film/foil. Can you give me a range? I would guess 50% to be safe, but if I can get away with 30% and see a real difference, I may do that as well (since audiocap thetas ain't cheap). Thanks.

**JonMarsh** · 16 May 2005, 10:46 Monday

First, the issues with the "Unorthodox Eight inch Two Way " design and construction was in the fall of 2003, September, October, and November.

I'm sure Dave understands all the "basic issues" pretty well,

but the article kind of walks through all that stuff in the context of that particular design- choice of drivers and crossover frequencies, (staying in pistonic range of drivers), LF design and alignment with intended room position, crossover and wiring construction with air core coils and film caps, and no ferrous materials in wiring signal path, design integration of the modified Woodstyle cabinet front baffle (using baffle rise board) and grille in conjunction with felt diffraction control to minimize diffraction effects and boundary load issues (addressed in driver layout with BDS), etc.

I've tried crossovers only in three permutations of "bypass" for what I think are the most important parts (tweeter series pass). So, this isn't an "exhaustive" evaluation (there are too many other things I'd rather be doing).

1) Small film and foil bypass cap (5% or less of value, as some advocate for using in parallel with metalized film, or metalized film in parallel wtih electrolytic. I didn't find this particularly efficacious. Tried using AudioCAP Theta and PE Dayton film and foil.

2) 100% film and foil caps. Sounds lovely, costs dearly. Only tried with AudioCAP Theta. The Dayton values are kind of small, and cap cluster would be rather large using them.

3) 50/50 film and foil with metalized film. Sounds about the same as #2, but cuts the cost in half. Seems a good compromise.

Keep in mind, I'm concerned with the integration and focus of the whole upper range in the upper presence region on up. For higher crossover frequencies, I might be inclined to just use Film and foil. OTOH, I'm not inclined to use 3 kHz crossover frequencies.

YMMV. I'll grant most folks would find it hard to justify spending as much or more on crossover caps as on their tweeter (example, with 27TDFC). My experience is that it does pay off. There are some highly regarded commercial speakers, like Avalon, that go down a similar path. I'd rather have a 27TDFC carefully voiced with good crossover caps than a Millenium Excel with standard caps. Since I have both, and have listened to both, my opinion, at least for me, is not based on speculation.

~Jon

**capslock** · 17 May 2005, 06:03 Tuesday

Just think of how many electrolytic (yuck!) caps the signal has passed in the recording gear (and mixing if analog) gear. Also, most hifi gear (event pretty pricey stuff) will use electrolytic coupling caps. High-endish stuff will sometimes use foil, but it is usually MKT (metallized polyester).

Of course, the theory is that as long as the load is high impedance, the quality of the cap does not matter too much.

My favorite was a Dual amp that I bought for parts a couple of years ago. The electronics were presumably designed and manufactured by Rotel. Anyway, there were impressive metallized PP caps coupling the signal from the volume control pot into the power stage. However, the DC blocking cap in the feedback network, which sees about 100 R load impedance (= pretty low!), was a plain polar electrolytic. What is worse, the coupling and feedback blocking caps in the preamp were 2.2 uF polar tantalums - about as bad as it can get!

I could ramble on about the caps used in well reputed CD players, as well as their grounding and supply decoupling scheme (designed by trained monkeys!) and signal processing (many current Philips and Marantz players simply round the 16th w/o any noise shaping!!!)...

**JonMarsh** · 22 May 2005, 15:58 Sunday

All too true. Digital music players, by their complexity, tend to be a sore spot for cutting corners lest the component costs get out of hand.

But then, that gives the relatively high end guys a lot of wiggle room for improving things.

Even a Hawk DAC from the Netherlands, while not particularly "special" as regards the digital chip sets, get's pretty decent sound, IMO, just by avoiding some of the componegt related issues the mass market players (even at high price points) pay little or no attention to.

~Jon

**TacoD** · 23 May 2005, 12:20 Monday

I like 100% film and foil. The Millenium sounds lovely with this combination, and I think it is worth it to spend those euros on caps

. (Mundorf Supreme/ Silver in Oil/ Gold/ Jensen). Also for the Seas W18 it makes a difference to upgrade the series caps.

For some reason I am not excited with the Seas 27TDC, yes it sounds amazing for this price point, but it is not the Millenium.

Bypass capacitors - do they work?

Bypass capacitors - do they work?

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment