|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
73.123.207.47
I'm in the planning stages of a preamp project and plan to use 0.1uF polypropylene caps across the electrolytics in the power supply.
My question is this: in this application, does it make a sonic difference which 0.1uF polypropylene caps are used? More specifically, I am thinking of choosing between Panasonic polypropylenes or Auricap XO, again, to be used across the power supply electrolytics.
What do you think?
Thanks!
Peter
Follow Ups:
I have no practical experience here but I think it depends on what you are trying to achieve. In general bypassing the circuit is better than bypassing the power supply, i.e. apply bypassing close to the circuit.
If your goal is to lower the power supply impedance across a range of frequencies then you can make some reasonable calculations knowing the ESR & SRF of the electrolytic and the bypass caps and an estimate of the wiring inductance in the loop of interest (1nH per mm. is a good estimate).
If you have no objective goal in mind then you will have to sift the answers to find one you like.
Regards
13DoW
Thank you for your response. My understanding is that a somewhat popular use is intended to lower the impedance of the power supply especially at higher frequencies. That is my primary goal. The kinds of calculations you are referring to, however, are a little above my pay grade.
Edits: 07/07/17
Just to recap, the impedance across a capacitor falls as frequency rises until you reach the self-resonant frequency (SRF). Above this frequency the parasitic inductance within the capacitor makes the impedance increase with frequency. The idea of bypassing is to place another capacitor in parallel that has a higher SRF to push out the frequency at which the impedance of the parallel pair starts to rise. Getting a higher SRF means using a smaller bypass capacitance. A smaller capacitor will have a higher impedance than the original capacitor at lower frequencies so you have to choose the right combination of bypass capacitance and bypass SRF for it to be effective. The ratio of the bypass SRF to electrolytic SRF must be greater than the SQRT(ratio of the electrolytic capacitance to the bypass capacitance) for the bypass to be effective. Or, looking at it the other way around, the ratio of electrolytic capacitance to bypass capacitance must not be greater than SQUARE(ratio of bypass SRF to electrolytic SRF).
For example, a 100uF electrolytic with SRF=1MHz and a 0.1uF bypass with a SRF= 31MHz have the same high frequency impedance, so there is little bypassing benefit. If you pick a 0.1uF bypass it must have an SRF > 31MHz to be beneficial.
Hope that helps
13DoW
ps. I was about to answer that the ratio of SRFs should be greater than the ratio of capacitances but I ran a few circuit simulations and realized that SRF varies with frequency squared hence the square/square root relationship.
Thank you for that. What about as the electrolytic ages? They can have a rather limited service life and it may not always be practical or convenient to replace them on a periodic basis...
Edits: 07/07/17
why not just design out the electrolytic caps? Makes a much bigger difference than brands of bypass capacitors, and opens the door to a lot of other related improvements.
------------------------
Favorite Albums of 2016
I believe you mean use larger value polypropylene caps in the PS instead of the electrolytics. Fabulous idea, but I doubt it's possible for this build.
Edits: 07/03/17 07/03/17
Well, that and inductors, and then use shunt regulators with the inductor resistance, standard stuff we do in tube designs. Maybe toss in some zero-feedback lithium references to bring it up to date. Lots you can do in a new design, if no constraints.
------------------------
Favorite Albums of 2016
Small film caps for bypass may induce oscillations, depending on the layout of the wires and traces.
A low value resistor may be needed in series with the bypass, and a high-speed oscilloscope to make sure it is not oscillating. You will also need to check it on music, under various loads.
maybe I should have mentioned it, but it has been rare in my experience. On the other hand, sometimes adding a series resistor to a capacitor bypass will improve the sound in subtle ways, which may be an indicator that there was an undetected oscillation going on and it's been damped by the resistor. Typically, I would use 10 to 1000 ohms in series with .01uF to 0.1uF capacitance.
How might such an oscillation be manifested, sonically?
A touch of edginess to the sound, if the oscillation is very far above the audible range. Can get worse if lower in frequency, such that you wouldn't want to listen. In one extreme case having nothing to do with a bypass capacitor, one of my Beveridge amplifiers was set into oscillation (by the output from a CDP) and promptly blew a fuse.
But the only time I can recall with a bypass capacitor, it was not very audible at all. I added the resistor empirically, whereupon I did hear that a slight tinge of edginess was gone. I don't think one can generalize; it depends upon the circuit, the values, etc. For sure, bypass capacitors do not always cause oscillation.
First time I have heard of this...
It's common.
Back in the day (30 years of hi-fi) I used a 100Khz oscilloscope (ignorance was bliss).
For audio I generally use just a good electrolytic between 22µF~47µF and leave it at that.
(I currently work in L-band RF and chooses bypass caps with great care and use a spectrum analyzer to monitor the results. Where the cap is grounded makes a huge difference, and series resistors on the order of 1Ω~10Ω added as well)
But not as much of a difference as bypassing vs not bypassing. I use Auricap XO in this application very often, because they're small and do outperform similar metallized polypropylene caps I have tried. (For example, I would NOT use Solen Fastcaps.) Auricaps are fine but not exactly a magic bullet.
What else do you like that is small in size for this application?
but they're not so "small". Also, REL RT, which ARE small.
I agree with Lew. Very audible differences between the type (and brand) of cap used. But define "small."
I like the Jensen copper foil/oil caps in this application. Have tried several different types and brands (Kimber poly, Audio 1 PIO, Auricap poly, and a couple others) and prefer the sound of the Jensens. I use a .22 uF (630v) value though. Not what I'd personally define as small....but don't know how much room you have to play with.
"Small," in this case, means I perceive some of the Panasonic polypropylenes to be smaller than many if not most of the boutique class of polypropylenes across a range of values. And, among the boutique polyprops, I seem to think the Auricap XO is smaller than average. Feel free to correct me if I have that wrong, but that's the scale I am thinking of when I say "small." I do have some space limitations.
Edits: 07/03/17
are no longer on my radar. Too many got leaky on me; one got leaky, and that resulted in a blown output tube.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: