well that too, but the main thing, why are you bothering spending sooo much time on this question of the regulator output caps? they are not only not locally decoupling anything, they are driving another regulator.... any good regulator will render the difference in these caps utterly insignificant to 'sound quality' if it didnt it would be a pretty crap regulator
popular opinion would have you replacing all those evil ceramics on the pins of devices with film caps, even though for local HF decoupling they are orders of magnitude worse than ceramics
it sounds to me like you are taking opinions/recommendations out of context, or whoever is parroting them.
Last edited:
I really suggest that you do some research as to what the roll of capacitors are in this circuit, how that relates to this application of this circuit and audio power supplies in general...
context is everything... not just in my words here, but very specifically for capacitors and each ones roll in the circuits and the specific positions they are placed in.
context is everything... not just in my words here, but very specifically for capacitors and each ones roll in the circuits and the specific positions they are placed in.
Last edited:
The Salas shunt regulator tested very well indeed, but fared poorly in the listening test.
Walt Jung's interview in AX has been archived to his website and has a very good explanation of the reasons why a shunt regulator should perform better. When I can get my panel of listeners together, we'll take a serious listen to the 2012 edition of the Jung Regulator.
Walt Jung's interview in AX has been archived to his website and has a very good explanation of the reasons why a shunt regulator should perform better. When I can get my panel of listeners together, we'll take a serious listen to the 2012 edition of the Jung Regulator.
I wonder what is the advantage of shunt regs. I know shunt regs can source current, but don't see how this is an advantage.
So far I've tried the Salas shunt reg and Twisted Pear shunt reg, and both of them were an improvement over the series regs that they replaced in my DAC and headphone amps. They have a clear advantage over the LM317 chip sonically.
So far I've tried the Salas shunt reg and Twisted Pear shunt reg, and both of them were an improvement over the series regs that they replaced in my DAC and headphone amps. They have a clear advantage over the LM317 chip sonically.
Shunt Regulators have an absolute limit on maximum current.
If the tiny amount of capacitance between the reg and the circuit (typical of most regulators of all types) cannot supply the demanded current then the performance will fail.
This must impinge on sonics.
I think many underestimate the transient current demands of most audio circuits.
Some even think that all ClassA amplifiers/stages draw constant current.
It's fallacies like this that lead unthinking builders down the route of sonic disasters.
If the tiny amount of capacitance between the reg and the circuit (typical of most regulators of all types) cannot supply the demanded current then the performance will fail.
This must impinge on sonics.
I think many underestimate the transient current demands of most audio circuits.
Some even think that all ClassA amplifiers/stages draw constant current.
It's fallacies like this that lead unthinking builders down the route of sonic disasters.
But this is why you over design the shunt (make it waste much more current than its load will ever pull), the load's ground becomes the limiting factor.
I mean it/s not like transients are faster than the current can pull thru the wire right ?
Ok, I'm rephraising that last remark.
So it's your firm belife that the capasitors in this particular regulator
used as a preregulator can not make any difference soundqualitywise
as long as the following regulator is "good" ?
The C2 capasitor should as I see it be there to eliminate inductance
between C1 and the chip.
Wouldn't that count as a decoupling capasitor ?
True.
Interesting.
How this can happen on single-ended circuits?
And for push-pull, which % of headroom would you consider?
On what kind of subjective test load and reg setting? Sounds a bit contradictory. Especially in the presence of many happy subjective reports of various builders.
SSHV2 is much better than SSHV1 for drift, but if you need really very low drift you should use active components in the Vref. Those are noisier than the resistors of course. In the LV ones you can use LM329 alone or in cascade depending on how much voltage you need, or alike very low noise buried Zener refs.
Why don't you search for yourself? A single ended amp draws constant current. The output devices never switch off. Get your own facts straight.
Then we will simply disagree. My understanding is that single ended devices draw constant current and push-pull can be designed to do so as well. In my DIY preamps and headphone amps, the shunt regulators I've used sound more transparent. I was wondering if anyone had any hypothesis as to why. If you do not know, you are not required to respond.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.