John Curl's Blowtorch preamplifier part II

[jan.didden]

Actually in a way that phase modulation is quite easy to measure for all of us:
Measure the phase shift between input and output with zero Vout.
Then insert a DC offset of a few volts, and again measure it; you'll find it most probably has changed.
One typical table from his paper:

DC offset at the output Phase shift @ 15kHz
-2V -25 degr
-1V -24 degr
0V -23 degr
+1V -22 degr
+2V -21 degr

So there's you varying phase shift with output level.
Fascinating, isn't it?

jan

 

[SY]

Measure the phase shift between input and output with zero Vout.

How do you accomplish that at zero volts out?

 

[Joachim Gerhard]

Nothing is new under the sun. ( King Salomon )

 

[Wavebourn]

Jan;

I can tel you the similar way to catch a mice. When it is under the stool take a saw and quickly cut stool's legs.

1. Cut the 1'st leg 0.5 centimeters
2. Cut the 2'nd leg 0.5 centimeters
...
5. Go to 1 and repeat until the space is too small for the mice to escape.

 

[31697B]

One typical table from his paper:

DC offset at the output Phase shift @ 15kHz
-2V -25 degr
-1V -24 degr
0V -23 degr
+1V -22 degr
+2V -21 degr

So there's your varying phase shift with output level.
Fascinating, isn't it?

jan

Facinating! OMG. Maybe what M.Otala and JC etal has been talking about is true? And, what would that affect do to the clarity of the sound?

I wouldn't be surprised to find that this affect has also been heard and described by careful listeners without knowing what it was, exactly. And, why certain designs sound better and thus contributes to why JC does design the way he does? OMG.

 

[Wavebourn]

What is fascinating?

Did you hear SSB station on AM radio? How does it sound?

Similarly, very small amount of phase intermodulation causes "harsh" highs and fuzzed stereo image. It's one of reasons why measurements of THD and IMD don't tell how good will be the sound.

 

[kevinkr]

How do you accomplish that at zero volts out?

Phase shift is referred to changing levels of DC offset at the output. (Not changing levels of AC signal amplitude) This apparently is pretty typical behavior for op-amps operating on relatively low supply voltages, shifts internal operating points, etc. (Look at GBWP and phase margin at different supply voltages for a hint)

 

[jan.didden]

How do you accomplish that at zero volts out?

Zero V DC. You need to have a small signal riding on the DC offset to be able to measure the phase shift.
Of course.

jan

 

[jan.didden]

Phase shift is referred to changing levels of DC offset at the output. (Not changing levels of AC signal amplitude) This apparently is pretty typical behavior for op-amps operating on relatively low supply voltages, shifts internal operating points, etc. (Look at GBP at different supply voltages)

Except that the opamps all measured essentially the measurement floor. Even the lowly TL074. Low ol bandwidth, very high feedback factor, go figure.

jan

 

[jan.didden]

Facinating! OMG. Maybe what M.Otala and JC etal has been talking about is true? And, what would that affect do to the clarity of the sound?

I wouldn't be surprised to find that this affect has also been heard and described by careful listeners without knowing what it was, exactly. And, why certain designs sound better and thus contributes to why JC does design the way he does? OMG.

Ohh yes of course it is true.
But not where we thought it was.

jan

 

[kevinkr]

Except that the opamps all measured essentially the measurement floor. Even the lowly TL074. Low ol bandwidth, very high feedback factor, go figure.

jan

Measurement artifact?

 

[31697B]

many opa are not symmetrical topologies. especially, the 'classic' opa design.
Your more likely to have these issues with such opa design topologies.

-Thx RNM

 

[SY]

Zero V DC. You need to have a small signal riding on the DC offset to be able to measure the phase shift.
Of course.

jan

Phase shift is referred to changing levels of DC offset at the output. (Not changing levels of AC signal amplitude) This apparently is pretty typical behavior for op-amps operating on relatively low supply voltages, shifts internal operating points, etc. (Look at GBWP and phase margin at different supply voltages for a hint)

Thanks for the clarification and education, guys!

 

[Wavebourn]

many opa are not symmetrical topologies. especially, the 'classic' opa design.
Your more likely to have these issues with such opa design topologies.

Not necessary. Phase modulation twice per period sounds nastier than once per period.

 

[scott wurcer]

Facinating! OMG. Maybe what M.Otala and JC etal has been talking about is true? And, what would that affect do to the clarity of the sound?

I wouldn't be surprised to find that this affect has also been heard and described by careful listeners without knowing what it was, exactly. And, why certain designs sound better and thus contributes to why JC does design the way he does? OMG.

But pease note:

"Except that the opamps all measured essentially the measurement floor. Even the lowly TL074. Low ol bandwidth, very high feedback factor, go figure."

Not exactly what JC believes, i.e. "op-amps, can't use 'em" so I would make exactly the opposite conclusion. Or OTOH go back to the old discussion about how the euphonic colorations make people prefer the less accurate amplifiers.

 

[31697B]

There's always a not necessarily. ??

A lot of opa topologies are quit asymmetrical in their topology designs. Again -- show great with symmetrical waveforms (sine waves). I'm not going back into - an FFT of anything is a bunch of sine waves discussion. Just note that amp circuits dont care about such subjects and work differently with signals which contain a dc component or average which is non-zero. Symmetrical/balanced topologies seem to work the best for audio. My instincts tell me this might be a factor in the authors results. Thx-RNM

 

[john curl]

I use IC's exclusively in the JC-3 phono preamp. It is just not as good as the Vendetta Research phono preamp or the new Constellation phono preamp that are made with discrete components.

 

[jan.didden]

Well Ron used all kinds of test signals, pulses, squarewaves with sines superimposed, etc. You really need to read the paper to appreciate his findings in all detail. But it has a couple of surprises.

jan

 

[31697B]

yep. various waveform shapes -- other than relying on symmetrical sine waves --- is the only way you'll find anything new. I look forward to reading it. And, his future follow-up work in sorting things out further. Then, I can comment in a more specific way. Thx-RNM

 

[31697B]

all of which brings me back to a distortion limiting parameter: common-mode rejection.... rejection of the dc component, as i call it, or the 'average' of a non-sine wave.
After all, that what CMR is about. [Something the FFT throws away (ignors), usually.]

We need to have DIY, SIM'ers and others maximize cmr also and not get tied only to thd and noise. If they want to hear great music at the highest level of performance.

How are opa used that compromise cmr? Might be a good point to start. Then, How does the cmr change with differing waveform shapes and differing topologies?

How much of the cm signal which is not rejected, can be turned into diff-mode and amp'ed? These things are not seen when sine waves of seperate, individual tests are done... they just look like spec numbers. CM to DM conversion probably happens a lot in audio systems.... even with opa that would otherwise measure super cmr on the bench.

Thx - RN