John Curl's Blowtorch preamplifier part II

[john curl]

Thanks for the added input Mark. What I think important is that there is a 'change' in the power supply behavior with different diodes. Special thanks for testing the 30A diodes. We changed them without measuring the difference in the JC-1 power amp, just to be consistent.

 

[bcarso]

Somebody else can build a test fixture capable of measuring (tb/ta) for those diodes which don't bother to specify it on their datasheet.

And if we use such diodes in a product there will be no guarantee that they will be consistent.

I fully agree that a proper snubber is the rational path. One caveat, which may be mentioned in your article, is that significant acoustic noise can be emitted by the capacitor. The worst offender I've heard was a high voltage ceramic used in a Hagerman phono preamp kit. When I friend said he was experiencing a lot of noise I assumed he meant electrical noise, only to get hold of the unit and realize that it was acoustic. It was quite audible in even a fairly noisy environment.

It's quite possible that the initial Hagerman design didn't suffer from the piezoelectric noise, and that some part substitute wasn't screened for the possibility.

 

[Mark Johnson]

There is also a 'change' in the power supply behavior when the snubber is added and removed. In my tests this was an even bigger change than I got when swapping between the worst diode and the best.

Oh and speaking of 30A diodes: those are the ones I use in my "RingNot" power supply. Besides epic ruggedness they have two other endearing properties: (i) They're in a TO-220-ISO package so they bolt straight onto a greased heatsink with no insulator required; (ii) their forward voltage drop is waaaaaaay lower than the Vfwd of many high_(tb/ta) diodes: 0.85V vs 2.1V. So they run cool.

 

[Destroyer OS]

I have personally heard the difference with my own design, after 10 years of using standard diodes in the same design. However, I can't say that everyone will hear a difference with every design or with every person's attitude toward changes in audio quality.

Given that it can be very easily heard after a capacitor multiplier voltage regulator, I'm not entirely sure what sort of designs wouldn't notice the difference.

 

[billshurv]

Ones with really good PSRR spring to mind!

 

[Mark Johnson]

Ones with really good PSRR spring to mind!

John Curl's earlier posts about his first soft recovery diode (revision 4 of the Vendetta preamp) speculate that RFI was somehow polluting the Ground of the power supply and thereby, polluting all audio stages. Have a few searches here on DIYA and/or Google in general, to find those. If ground != ground all bets are off.

 

[billshurv]

That is certainly possible, and measurable with the test equipment any professional would have available in their lab. No mystery introductions to component whisperers required surely?

 

[DPH]

Shouldn't we see that RFI injection cause IM distortion? As expected by bad ground, we'd should see substantial additional products.

 

[Mark Johnson]

Hop into your time machine, go back to 1994 (Berkeley, Calif, USA), and tell John Curl exactly what experiments he ought to perform! I'm sure 1994-John would be grateful. (link)

 

[billshurv]

If I could go back to 1994 I'd have been giving myself advice!

But back then John was already a 'genius'. Why on earth would he have listened to me?

 

[DPH]

Can I go back and buy stocks? I didn't have much $$ then as I was 10 years old, but I'm sure any amount would do.

 

[john curl]

I listen to just about anybody who has something positive to share about audio differences. I consider that as having an 'open mind'. Sure some things don't work out, but most do, and that makes all the difference.

 

[jan.didden]

It's not "diodes ringing" it's "diode-stimulated transformer ringing"

Mark I translate that to 'the di/dt caused by diode switching causes the circuit inductance, which is mainly on the xformer side, to ring with circuit capacity'.
Is that the reason why soft recovery diodes cause less ringing to begin with. Am I correct in this line of reasoning?

Jan

 

[kamis]

EI core and flatpack transformers have much higher leakage inductance than toroid and R core and are more influenced by diode speed and soft recovery.
John Curl power supplies have CM chokes on the DC side to cut HF noise. I tried this method with very small improvement in sound quality.

 

[jan.didden]

So the xformer type could also influence sound quality through this mechanism.

Jan

 

[31697B]

what was the snubber values placed across transformer?



THx-RNMarsh

 

[nicoch58]

I listen to just about anybody who has something positive to share about audio differences. I consider that as having an 'open mind'. Sure some things don't work out, but most do, and that makes all the difference.

 

[Bonsai]

Hop into your time machine, go back to 1994 (Berkeley, Calif, USA), and tell John Curl exactly what experiments he ought to perform! I'm sure 1994-John would be grateful. (link)

That is outstanding performance. Period.

 

[Max Headroom]

That is outstanding performance. Period.

Footnote 3: John Curl hand-selects FETs for the SCP-2 with a transistor tester which not only measures noise but classifies that noise according to frequency. Thus he can select both on the basis of low noise and of noise with a particular spectral density.—John Atkinson

JC's secret sauce.....smart.

Dan.

 

[billshurv]

How is measuring noise spectrum a secret sauce? People have been doing that since before I was born, at least in the precision analog world.