Function of Bypass film caps

"Vast" is a better description of my girth than my knowledge, but I think the whole bypass thing arises from a belief that electrolytics become inductive and have a high impedance at higher audio frequencies. Only one is true, they do become inductive. The impedance, however, remains very low, usually milliohms until the frequency gets well out of the audio band. Since the impedance of a small film cap isn't as low, putting it in parallel doesn't buy much. Example- the impedance of a 0.47 uF Mylar cap will be about 17 ohms at 20 kHz.
 
Reference to Board F2760 on a Sui AU 217 which was recapped (Electros only) and fuse resistors replaced in Aug 2014.
Based on all the comments and advise on this subject, I cautiously replaced Electros C01 and C02 (1uF 50V) with WIMA films (1uF 100V) but left the original Mylar bypass caps (C03 & C04) in. Connected an iPad playing Spotify and noticed there was a slight improvement in clarity. When I removed the bypass Mylars, there was a further improvement in clarity AND gain! The 217 is now playing without bypass caps C03 & C04 & I'm pleasantly surprised.
I think I understand the "clarity" part but why did the gain improve?
Robert.
 
Not a clue! I notice they also parallel at C11/C13, a place where I'd go with a higher voltage part, given the often inferior linearity of 6.3 V parts. Ditto C29.
 
Don't use bypass caps period. They only add harm, if anything. They are a waste of space, parts and money. 100% nothing positive can come out of using them.
 
ZOOM said:
Don't use bypass caps period. They only add harm, if anything. They are a waste of space, parts and money. 100% nothing positive can come out of using them.
Click to expand...

Are you sure?. :idea:.... I mean, not even a little one. :naughty:... now and again. :rolleyes:.... maybe...... :bye:
 
Hi Conrad,
It's C11 & C12 which are paralled. I could not replace Electros C11 & C12 with films as they are high capacitance at 47uF each ... So I retained the bypass Mylars but replaced them with WIMAs of equal value.
I replaced all 1uF (Polar & Bi polar) and 3.3uF Electros with Film & these are not bypassed now. The audio improvement was a pleasant surprise. The other Electros (replaced in Aug 2014) and any bypass caps per the original schematic are retained.
 
I would try a couple of really good quality 47µF replacements - like Nichicon FG, Nichicon KZ, or even Elnas... and lose the bypass?
 
Resurrecting an old thread.

So is a 1-5uf Film cap that spans the terminals of coupling and main filter caps, also considered voodoo/non improvement? Or were people only talking of smaller electrolytics? I wasn't sure from the back and forth.

upload_2021-8-30_1-26-24.png
 
My own experience have been that on the large filter caps, it makes zero difference, but in signal coupling applications it only sometimes makes an audible difference - this difference has been most obvious when an electrolytic feedback cap was bypassed.

The term by-pass is used with some confusion. Bypasses covers both de-coupling applications in which you smooth out DC ripples on a DC rail. The other meaning is when you have by-passes in which you are making up for electrolytic capacitors short comings in the transfer of an AC signal. In the decoupling scenario, you are connecting a cap between a DC rail and ground. This mean the cap holds charge equal to the rail. The instant the rail voltage drops from a fluctuation, the de-coupling cap discharges and thus brings the rail back up. The fluctuations or ripple you are trying to smooth, dictates the size of the de-coupling cap - all the way from the large filter caps to smaller film caps dealing with high frequency ripple on the same DC rail. If you are to de-couple the rail from HF, then do so closest to when the source of consumption.

When bypassing means, transferring an AC signal across an electro, the theory goes that the electro has inherent shortcomings at higher frequencies - even if these distortions takes place at frequencies above the audible threshold, they still inflict distortion downward to audible frequencies. When it comes to the feedback cap (shunt leg) by-passing - in the cases where sansui had an existing polar electrolytic - I've tried a) to replace that for a non-polar b) keeping the electrolytic as is, but by-passed with a PP film cap. The bypassing has sounded the best.

- I'll pass on the word to Bob Cordell from his 608 pages behemoth of a book titled: "Designing Power Amplifies" especially read from - "The biggest offender...."

upload_2021-8-30_10-23-16.jpeg


upload_2021-8-30_10-23-54.jpeg
 
stereofun said:
My own experience have been that on the large filter caps, it makes zero difference, but in signal coupling applications it only sometimes makes an audible difference - this difference has been most obvious when an electrolytic feedback cap was bypassed.

The term by-pass is used with some confusion. Bypasses covers both de-coupling applications in which you smooth out DC ripples on a DC rail. The other meaning is when you have by-passes in which you are making up for electrolytic capacitors short comings in the transfer of an AC signal. In the decoupling scenario, you are connecting a cap between a DC rail and ground. This mean the cap holds charge equal to the rail. The instant the rail voltage drops from a fluctuation, the de-coupling cap discharges and thus brings the rail back up. The fluctuations or ripple you are trying to smooth, dictates the size of the de-coupling cap - all the way from the large filter caps to smaller film caps dealing with high frequency ripple on the same DC rail. If you are to de-couple the rail from HF, then do so closest to when the source of consumption.

When bypassing means, transferring an AC signal across an electro, the theory goes that the electro has inherent shortcomings at higher frequencies - even if these distortions takes place at frequencies above the audible threshold, they still inflict distortion downward to audible frequencies. When it comes to the feedback cap (shunt leg) by-passing - in the cases where sansui had an existing polar electrolytic - I've tried a) to replace that for a non-polar b) keeping the electrolytic as is, but by-passed with a PP film cap. The bypassing has sounded the best.

- I'll pass on the word to Bob Cordell from his 608 pages behemoth of a book titled: "Designing Power Amplifies" especially read from - "The biggest offender...."

View attachment 2329401


View attachment 2329403[/QUOTE!
.
Click to expand...
. Thank you!
 
Hyperion said:
Bypassing signal path electrolytics with a film capacitor seems like a daft thing to do, as the larger bypass capacitors simply add somewhere for noise to be picked up. If you are going to do anything, providing the capacitor is less than about 4.7µF - change the electrolytic for a film, one capacitor, but above all it must be a similar size to the original EC, this is now achievable with modern components.

I still bypass PSU electrolytics with a film capacitor (1-2.2µF), but having also absorbed some of Conrad's vast knowledge on the subject I am considering discontinuing this practise.

BTW: I only get rid of Tantalum capacitors in my audio equipment, I consider them the perfect choice for filtering power lines in digital circuits for example.
Click to expand...

The classic digital power supply bypassing-100uF aluminum elec, 1uF tantalum, 0.1uF ceramic.
 
bubhardy said:
The classic digital power supply bypassing-100uF aluminum elec, 1uF tantalum, 0.1uF ceramic.
Click to expand...

Here is a highly interesting article from Eric Bogatin in the Signal Integrity journal about the "3 capacitor myth" coming from leaded through-hole components that a lot of people take over to the SMD-age:
https://www.signalintegrityjournal.com/articles/print/1589-the-myth-of-three-capacitor-values

On my SMD digital circuits (non-audio circuits I have to add) I use 0.1uF standalone or 10uF||0.1uF (with the 0.1uF closest to pin/pad of the IC).
 
Interesting thread. I particularly like bypassing electrolytics in some locations, and believe the improvements are very apparent. To further, I never got much from bypassing signal path electrolytics, so long as they are good electrolytics. However, in supply decoupling, I notice big gains in doing so, as in locally, where the power comes in to each circuit. Polypropylene is the way to go here too, and I like to use .68-1uF. I find that one bypass each, for the +/- rails of each circuit does the job. Also, bypassing large main filter caps has been a mixed bag for me. On some amps, it helps and sometimes helps allot (one example, bypassing all the large main filter caps on a Pioneer SX-1250), other times it seems to do nothing or even make the sound unpleasant.

Electrolytics in the signal path, I particularly like Elna Silmic II, and perhaps even more important than input coupling electrolytics in amplifiers, the feedback electrolytic, I like to use Elna Silmic II for this duty as well when I can.
 
Last edited:
Bypassing main filter caps to GND to provide a path for RF, no "improved SQ" except the remove/reduction of RF.
Bypassing audio path caps??? Leaning towards snake oil, maybe someone can explain the math rather than "blind testing" or...
 
Signal path
mbz said:
Bypassing audio path caps??? Leaning towards snake oil, maybe someone can explain the math rather than "blind testing" or...
Click to expand...
You can see it as an original configuration in the very best Sansui amplifiers and pre-amplifiers, the addition of, for example a 1000pF (correction not 1000pF but 10,000pF) polystyrene capacitor placed across a signal path electrolytic. It is supposed to improve the passing of high frequencies and tame the inherent tendency of un-bypassed electrolytics to introduce a small amount of distortion. For this reason electrolytics in the signal path are a necessary evil, needed for decoupling, but with that drawback. For a long time audio amplifier designers have been trying to eliminate them from the signal paths of their circuits. Ever heard the term 'Direct Coupled' in relation to audio amplifiers? this has been the goal for a long time so, absolutely not snake oil. :smoke:

Use in Power Supplies
There is continuing debate on here about the use of low value film capacitors to decouple main smoothing/filter capacitors in PSU's. The reason for doing it? - historically to counteract the natural resonance (or inductance), of these large electrolytics. And to give some filtering of HF noise whether this be RF pickup, or rectifier noise, or to counteract the presumed deficiencies of the large electrolytics for this job.

And also discussion about the best location for the PSU film bypass capacitors. Was it just 'convenient' to mount them on the lugs of the large electrolytics? or was this placing the solution closest to the perceived problem? Would it be better to place the bypass capacitors closer to the audio circuitry? on the driver board for example? - the jury is still out on that one...
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom