ps...how do you create some sort of listening refence so you can compare your experiments?....the only way I can think is to do it on a sort of "buddy system" where you have two units so any cap change etc can be dierectly compared by ther flick of a switch to the original and surely there must be a reference setup: made so you can hear how far you have come from a basic sound you felt was acceptable ...but this is expensive
It's like everything in this hobby. Somethings you can measure and some you have to listen to and trust your ears. These power supply caps bypassed with film caps may or may not be measurable? I guess I could hook the scope up and look at the noise on the DC signal? I actually didn't do it.
As far as coupling capacitors and other components I built two identical preamp's with a single power supply. I use a cannon plug connection. Move it from one to another. But this is still subjective because I need to wait for the tubes to warm back-up. But I don't know how to start with measuring coupling caps? But I know there is definitely a difference in their sound. If I like it and it sounds good to me that's good enough.
As far as coupling capacitors and other components I built two identical preamp's with a single power supply. I use a cannon plug connection. Move it from one to another. But this is still subjective because I need to wait for the tubes to warm back-up. But I don't know how to start with measuring coupling caps? But I know there is definitely a difference in their sound. If I like it and it sounds good to me that's good enough.
my .02c -
I've gone through the polyprop (Solen - argh!) power supply capacitor "stage" and after some experimentation ended up preferring good Nichicon / Panasonic / Elna Silmics / etc. Of course I don't like the "sound" of most series regulators either, preferring shunt or if possible, no active regulation.
I've gone through the polyprop (Solen - argh!) power supply capacitor "stage" and after some experimentation ended up preferring good Nichicon / Panasonic / Elna Silmics / etc. Of course I don't like the "sound" of most series regulators either, preferring shunt or if possible, no active regulation.
As above,
I think that good switch mode PSU caps are as good as you will get. The problem with bypassing is you can keep trying different combinations and never be satisfied with the result. All components add their "signature sound" and the numbers of combinations can lead you into a never ending trial and error "loop". I don't think there is any component that has no effect. However if experimenting "floats your boat!" then have fun!
I have bypassed in the past with PIO.
Regards
M. Gregg
I think that good switch mode PSU caps are as good as you will get. The problem with bypassing is you can keep trying different combinations and never be satisfied with the result. All components add their "signature sound" and the numbers of combinations can lead you into a never ending trial and error "loop". I don't think there is any component that has no effect. However if experimenting "floats your boat!" then have fun!
I have bypassed in the past with PIO.
Regards
M. Gregg
Last edited:
Thomas gives good advice: take care before using bypass capacitors. He gives other good advice: if you need an electrolytic, use new electrolytics! I have found this to be VITAL. Even the roughest old guitar amp is improved by renewing the 'lytics. The Panasonic AM is a low cost series of audio grade capacitor that's worth a try (up to 100V). At higher voltages, the Nichicon KG, or even the old VR or VX series should be on your shopping list. Try the new electrolytic for 100 hours of operation before attempting any bypassing - the capacitance is formed by chemical means, and needs time to properly establish itself.
Of course, if you're building at a higher quality level, you won't want an electrolytic in the anode-current loop of your new 300B SE. The Ampohm/LCR 30uF audio polypropylene is a good low cost choice for this position - if you build an LCLC power supply: big choke at the input, Nichicon KG 100uF or more for a first cap, then 1or 2H of choke before the audio cap. Or use the ASCs that Thomas has.
The third place I agree with Thomas is regarding potential resonances of parallel capacitor pairs. Most designers active in the 1980s were aware of the hazard of parallel pairs of ceramic and tantalum caps (leaded types in those days).
Using a stacked-film capacitor to bypass an electrolytic still looks like a risk to me. You will be forming a parallel LC tank composed of the bypass capacitance (with maybe 1nH or less of lead inductance) against the RLC of the electrolytic. The L component here is liable to be 20 to 30nH - putting the resonance below 1MHz in some cases. Actually, the ESR of the electrolytic is helpful here in damping oscillations, indicating that a low ESR part is not necessarily a good thing for audio. Remember that rectifier pulses are a severe provocation for power supply capacitors.
I don't like series regulators in audio either. IC types are bad enough, but regulators or opamps with power transistors wedged into their feedback loops are the worst of all worlds. It takes a great deal of time, instrumentation, and skill, to really tame the transient response of such a circuit.
Curiously, even some IC regulators can be made to oscillate by bypassing the output electrolytic. The older MICREL MIC5205 is pretty much guaranteed to oscillate if you use a ceramic output capacitor.
It's worth studying the reason for this instability, since the same problem can occur if you do make your own regulator. The picture tells the tale. A series regulator's effective output resistance always forms a pole with the output capacitor. With a ceramic or stacked-film (where ESR and ESL are vanishingly small) the rolloff continues above 100kHz, where the phase-shift through the regulators loop, and the effective gain may produce the conditions for instability.
With an electrolytic, the ESR forms a zero at a convenient point, and gain remains too high for oscillation.
Using a 1 or 2.2 ohm resistor in series with the output capacitor can improve the response and stability of a regulator.
Of course, if you're building at a higher quality level, you won't want an electrolytic in the anode-current loop of your new 300B SE. The Ampohm/LCR 30uF audio polypropylene is a good low cost choice for this position - if you build an LCLC power supply: big choke at the input, Nichicon KG 100uF or more for a first cap, then 1or 2H of choke before the audio cap. Or use the ASCs that Thomas has.
The third place I agree with Thomas is regarding potential resonances of parallel capacitor pairs. Most designers active in the 1980s were aware of the hazard of parallel pairs of ceramic and tantalum caps (leaded types in those days).
Using a stacked-film capacitor to bypass an electrolytic still looks like a risk to me. You will be forming a parallel LC tank composed of the bypass capacitance (with maybe 1nH or less of lead inductance) against the RLC of the electrolytic. The L component here is liable to be 20 to 30nH - putting the resonance below 1MHz in some cases. Actually, the ESR of the electrolytic is helpful here in damping oscillations, indicating that a low ESR part is not necessarily a good thing for audio. Remember that rectifier pulses are a severe provocation for power supply capacitors.
I don't like series regulators in audio either. IC types are bad enough, but regulators or opamps with power transistors wedged into their feedback loops are the worst of all worlds. It takes a great deal of time, instrumentation, and skill, to really tame the transient response of such a circuit.
Curiously, even some IC regulators can be made to oscillate by bypassing the output electrolytic. The older MICREL MIC5205 is pretty much guaranteed to oscillate if you use a ceramic output capacitor.
It's worth studying the reason for this instability, since the same problem can occur if you do make your own regulator. The picture tells the tale. A series regulator's effective output resistance always forms a pole with the output capacitor. With a ceramic or stacked-film (where ESR and ESL are vanishingly small) the rolloff continues above 100kHz, where the phase-shift through the regulators loop, and the effective gain may produce the conditions for instability.
With an electrolytic, the ESR forms a zero at a convenient point, and gain remains too high for oscillation.
Using a 1 or 2.2 ohm resistor in series with the output capacitor can improve the response and stability of a regulator.
I'm using the Salas HV shunt regulator for the SRPP 6922
It's not clear if to replace 22uF for an ASC 20uF can be a good upgrade? I'm using BG in PSU....or I will follow with BG for 22uF? Or I replace the BG PSU caps for all ASC caps if yes what amount 220uF x 2 caps?
I experienced full Solen MKP in the PSU & I dislike it du to the thin sound without bass & body....
An externally hosted image should be here but it was not working when we last tested it.
It's not clear if to replace 22uF for an ASC 20uF can be a good upgrade? I'm using BG in PSU....or I will follow with BG for 22uF? Or I replace the BG PSU caps for all ASC caps if yes what amount 220uF x 2 caps?
I experienced full Solen MKP in the PSU & I dislike it du to the thin sound without bass & body....
Last edited:
Hi!
Well, what is it you dislike about the design so you want to change it? If you do not like it as is and as intended by the designer then why not try a different circuit? I for example dislike the SRPP scheme all together, it was a bit of a hype in the 90ies but is much less in use nowadays. I'd very much prefer a simple straight forward RC coupled stage over a SRPP any time. Such changes would provide a whole other level of change in sound than just doing minor tweaks. This is all IMHO and IME, I'm sure others experience might be different.
No surprise, I think the Solens are horrible...
best regards
Thomas
Well, what is it you dislike about the design so you want to change it? If you do not like it as is and as intended by the designer then why not try a different circuit? I for example dislike the SRPP scheme all together, it was a bit of a hype in the 90ies but is much less in use nowadays. I'd very much prefer a simple straight forward RC coupled stage over a SRPP any time. Such changes would provide a whole other level of change in sound than just doing minor tweaks. This is all IMHO and IME, I'm sure others experience might be different.
No surprise, I think the Solens are horrible...
best regards
Thomas
No I don't want to change the design, I only ask if the bypassing cap 22uF can be MKP oil filled or will be better lytic BG?
It's the 2nd time the we share the same experience about caps.
Cheers
Felipe
Hi Felipe,
if you want to experiment, I'd start with the two 22uF decoupling caps in the B+ per channel. Next think about the bypass caps on the cathode. Depending on the requirements on output impedance you might be able to leave them out completely. Changing the 220uF in the B+ to oil will become very bulky and expensive.
Best regards
Thomas
if you want to experiment, I'd start with the two 22uF decoupling caps in the B+ per channel. Next think about the bypass caps on the cathode. Depending on the requirements on output impedance you might be able to leave them out completely. Changing the 220uF in the B+ to oil will become very bulky and expensive.
Best regards
Thomas
As written in a different thread, I am experimenting currently with the ASC x386...I first gave my power amp a separate PSU for the driver stage whichwas great and then in a second step changed there from Mundorf tubecap to ASC x386.
First reaction was disappointment. A remarkable loss in air/spaciousness of sound/3D. It sounded musical, yes, but boring.
OK...I have run the amp with them for two days 24h. Something magical is happening. Nothing is boring anymore, a ton of details are presented, but in a warm, musical way. Treble hizzing is completely gone etc. I am looking forward to the result when it is completely burned-in...not sure how long this will take ?
Thomas, I did not know I could have ordered from you (again), just got them from RS. Next time...
One question I have: How do the x386 which are MKP, so metallized polyprops, compare sonically to the vintage true Paper in Oil Caps like from GE, Sprague, the Russians etc which are even bigger ?
More or less transparency ? More or less mellow etc ?
One question to Thomas: You use the Asc quiet often in 30uF and many parallal instead 47uF...is this purely for estical reasons or because they are faster when smaller values are in parallel ?
Best Regards
Frank
First reaction was disappointment. A remarkable loss in air/spaciousness of sound/3D. It sounded musical, yes, but boring.
OK...I have run the amp with them for two days 24h. Something magical is happening. Nothing is boring anymore, a ton of details are presented, but in a warm, musical way. Treble hizzing is completely gone etc. I am looking forward to the result when it is completely burned-in...not sure how long this will take ?
Thomas, I did not know I could have ordered from you (again), just got them from RS. Next time...
One question I have: How do the x386 which are MKP, so metallized polyprops, compare sonically to the vintage true Paper in Oil Caps like from GE, Sprague, the Russians etc which are even bigger ?
More or less transparency ? More or less mellow etc ?
One question to Thomas: You use the Asc quiet often in 30uF and many parallal instead 47uF...is this purely for estical reasons or because they are faster when smaller values are in parallel ?
Best Regards
Frank
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.