I have often used capacitors in parallel with electrolytics in solid state designs. As most solid state driver stages are current sourcing - bypassing makes sense to me.
Tube amp designs are mostly voltage amplifiers except for the pentode/triode stage were current is being pulled through the transformer.
But I rarely see any schematics showing a parallel film cap along with the B+ capacitor, nor in the voltage gain and splitter stage do I see any bypassing of the power supply capacitors.
Why is this? Is there really no benefit? Or do people do this but they just don't speak about it? (like listening to Lady Gaga).
Any advice would be appreciated.
Tube amp designs are mostly voltage amplifiers except for the pentode/triode stage were current is being pulled through the transformer.
But I rarely see any schematics showing a parallel film cap along with the B+ capacitor, nor in the voltage gain and splitter stage do I see any bypassing of the power supply capacitors.
Why is this? Is there really no benefit? Or do people do this but they just don't speak about it? (like listening to Lady Gaga).
Any advice would be appreciated.
As they say in real estate, "location, location, location." If the power supply cap is 15 cm from the circuit it's bypassing, putting a second bypass cap across it may make you feel good, but does little to nothing electrically. If you can place the bypass cap right at the circuit, it may be useful in some circumstances.
Sy - not sure what you are implying. If a small value bypass capacitor is in parallel with a power supply smoothing capacitor, you are decreasing resistance because you have twice as much wire. Also given the lead length of a capacitor - is it that much inductance you are adding? (typically you put the capacitor right across the terminals of the electrolytic). Are "Valve Amplifiers" and "Building Valve Amplifers" books that are still sold?
Thanks for all the comments.
Jerry
Thanks for all the comments.
Jerry
I think the point SY is making, is that there is no point putting the bypass cap across the electrolytic if that whole part of the circuit is 15 cm from the load. If that is the case, you are better off putting the bypass cap right up close to where it is needed - the load, not the reservoir cap. Hence the real estate term 'location location location'. Both caps are 'in parallel', but only one location will do what you want it to. Yes, this, and many other useful tips, are in the two books. 
I thought the idea behind applying a small-value film cap across a large electrolytic was to decrease the ESR (or impedance?) of those large electrolytic caps at higher frequencies.
If you use an electrolytic cap with very low ESR up to well above high audio frequencies, then a film bypass cap would do nothing, correct?
But if you're using an electrolytic cap whose impedance begins to rise near the upper end of the audio band, then a film bypass cap would help. Right?
Solid state amps use larger value electrolytic capacitors (and lower voltage) than do tube amps. A 10,000uF 100V electrolytic might have much higher impedance with rising frequency than a 47uF 450V cap, no?
What about pulse current?
--
If you use an electrolytic cap with very low ESR up to well above high audio frequencies, then a film bypass cap would do nothing, correct?
But if you're using an electrolytic cap whose impedance begins to rise near the upper end of the audio band, then a film bypass cap would help. Right?
Solid state amps use larger value electrolytic capacitors (and lower voltage) than do tube amps. A 10,000uF 100V electrolytic might have much higher impedance with rising frequency than a 47uF 450V cap, no?
What about pulse current?
--
One thing I have not seen addressed anywhere is the fact that different capacitor types have different self resonant frequencies.
Thus, if you put two different caps in parallel then in the range between their resonant frequencies they will behave as a parallel tuned filter, and you could potentially get less good decoupling than you started with! (But only in that range of frequencies)
Thus, if you put two different caps in parallel then in the range between their resonant frequencies they will behave as a parallel tuned filter, and you could potentially get less good decoupling than you started with! (But only in that range of frequencies)
Last edited:
The issue is not the capacitor wire so much as the wiring to get to the capacitor, as SY says. Whether inductance is an issue depends on circuit typical impedances - probably negligible in most valve cases, but more important for SS.
The effect of paralleling two resonators is shown in several places on the web. The effect can be to make things worse, especially if the caps are high quality low loss ones.
The effect of paralleling two resonators is shown in several places on the web. The effect can be to make things worse, especially if the caps are high quality low loss ones.
Last edited:
With the availability of high quality electrolytic caps like Jensen & Mundorf there is not a lot to be gained with film by-pass caps, although I have seen some folk using Mundorf Supreme film caps (not the silver/oil ect) and the film & oil caps like ASC as by-passes in the power supply.
If the voltage and capacitance is low enough, Solen BP series caps are a nice replacement for an electrolytic (they are only 400v) in a power supply.
Personally I'll avoid by-pass caps if at all possible.
If the voltage and capacitance is low enough, Solen BP series caps are a nice replacement for an electrolytic (they are only 400v) in a power supply.
Personally I'll avoid by-pass caps if at all possible.
Thanks Data - obviously this is a mass produced amp without high quality electrolytics. So I'll roll my own and report back!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.