John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
In my system I distinguish, whether a single gold contact of OMRON relay (input selector in integrated tube amp) is present in the signal chain, or it is shorted by silver wire.
Therefore, I mentioned above, that 99% of listeners have no idea what the "micro sound picture" could be. These tiny effects of a single contact, as also reported Ed Simon, has around -140 - 160dB level.
I understand and agree with your point of view, VladimirK; but, I would disagree with your estimation of the level of the effect, in my opinion it's closer to 50-60dB. In other words, easily audible if not masked by other factors.

IMO, cable differences show up markedly when low level detail which directly relates to the musical message is clearly discernable -- this is the "transparency thing" -- because they influence the spectrum of the remaining distortion. Many systems never get to this point, so for those all cables sound the same.

Edit: John has just mentioned resistors in that context; for me this is all about parasitic behaviours due to less than optimum manufacture, things like how well the leads are bonded to the resistive materal, the electrical integrity of the join .

Frank
 
Last edited:
OF COURSE, VladimirK! And you actually did a test of the Omron relay, the sort I use in everything BUT the CTC Blowtorch! The CTC Blowtorch used silver on silver rotary contacts, with enough pressure to 'clean' them on each pass.
We knew it, intuitively, through long term listening with the JC-80 etc. that relays are good but not GREAT sounding. I still use them by the dozens on many of my advanced projects, I just can't do remote control cost effectively, without them.
 
diyAudio Member RIP
Joined 2005
sorry to intrude - but wonder if any decent P channel SMT JFETS have become available so we have a nice complement to the BF 852/862 ? Thought it might be nice to make some buffers with them. I could not find anything searching through this thread other than the above ref.... Thanksamillion for any guidance :)

Not as yet. It is said that Linear Integrated Systems has made, experimentally, a few PJFETs along the lines of the 2SJ74, but last checked they aren't selling them yet. And silicon P parts with similar breakdown voltages will always be higher capacitance for comparable-to-N-channel transconductance, so you will never get a good complement to the 862.
 
Friday's test

Again here are the curves for some simple amplifiers showing there THD + Noise output for given input voltages.

So for the Gain Clone which shows 1% distortion at 1 mV of input signal and has a gain of 32 what is the value of the feedback resistor that goes from the inverting input to ground if predicted by noise level only?

The test source is rated for 20 ohms output impedance.
 

Attachments

  • 11 THD Noise Small Amps.JPG
    11 THD Noise Small Amps.JPG
    67.8 KB · Views: 185
So for the Gain Clone which shows 1% distortion at 1 mV of input signal and has a gain of 32 what is the value of the feedback resistor that goes from the inverting input to ground if predicted by noise level only?

.

What is the purpose of that exercise? I didn't elaborate on my disdain for these plots. You combine THD a spot frequency phenomena with broadband noise, I find the result essentially meaningless.
 
Last edited:
Not as yet. It is said that Linear Integrated Systems has made, experimentally, a few PJFETs along the lines of the 2SJ74, but last checked they aren't selling them yet. And silicon P parts with similar breakdown voltages will always be higher capacitance for comparable-to-N-channel transconductance, so you will never get a good complement to the 862.

My answer (essentially the same) was eaten by my WiFi. I would add that many FET's previously unknown to me were uncovered in the discrete op-amp thread.
 
Ed, what's the frequency of the test tone, what's the bandwidth of the THD+N measurement?

1 Khz 31,500 bandwidth (Good question!)

What is the purpose of that exercise? I didn't elaborate on my disdain for these plots. You combine THD a spot frequency phenomena with broadband noise, I find the result essentially meaningless.

Because when the spot frequency is 30 db below the noise level it is effectively masked.

Or in other words when the noise is 30 db greater than the distortion, the lack of reproduction quality is noise limited.

Pavel mentioned that when distortion is below .01% it is other factors that determine the sound quality. Now there is a lot to what is really meant by .01% distortion, but even using that shorthand I suspect Pavel is doing an apples to apples comparison.
 
IMO, cable differences show up markedly when low level detail which directly relates to the musical message is clearly discernable -- this is the "transparency thing" -- because they influence the spectrum of the remaining distortion. Many systems never get to this point, so for those all cables sound the same.

This your explaination just helps me a bit, these effects illustrate that some of them stay outside the "macro sound picture" for a long time already, and there are no reliable quantitative approaches for their investigation.
Therefore, I have said before, that science has almost not touched the micro sound effects. These micro effects classify audio systems, whether one of them could pretend to be hi-end or not.
Those who do not have enough experience with "transparent" systems, do have quite reasonable grounds to call some audiophiles as fools, that speak about nonsenses.
 
My take on the A21 is that it is UNDER-BIASED. In fact, due to my personal tests of an A-21, just yesterday, I asked my boss, WHERE they are being biased. TOO MANY higher order harmonics. I should have followed up earlier, but I only got an A-21 to test, recently. Of course, the schematics go back and forth, but not necessarily the real production item. Changes, I hope, can be made.
For the record, the P-3 is NOT my design.
 
OF COURSE, VladimirK! And you actually did a test of the Omron relay, the sort I use in everything BUT the CTC Blowtorch! The CTC Blowtorch used silver on silver rotary contacts, with enough pressure to 'clean' them on each pass.
We knew it, intuitively, through long term listening with the JC-80 etc. that relays are good but not GREAT sounding. I still use them by the dozens on many of my advanced projects, I just can't do remote control cost effectively, without them.

John, after I have realized that this effect "keeps a tail" of my tube integrated amp, I decided to disconnect input selection option. Just one input acts now.
John, you CAN do real things. Those who can not do, they teach (joke).
 
Last edited:
diyAudio Member RIP
Joined 2005
Thanks to the both of you for taking the time to respond. I take it then, the answer remains: no.

That's too bad but that's what it it is I guess.

The Holy Grail of complementarity is possibly a bit oversold. I recently worked out how to achieve the benefits (such as precisely equivalent positive and negative slewing/settling behavior) while still using two pairs of NFETs as the input devices. But it relies on the use of some bipolars, hence will be off-putting to those who shun them.
 
My take on the A21 is that it is UNDER-BIASED. In fact, due to my personal tests of an A-21, just yesterday, I asked my boss, WHERE they are being biased. TOO MANY higher order harmonics. I should have followed up earlier, but I only got an A-21 to test, recently. Of course, the schematics go back and forth, but not necessarily the real production item. Changes, I hope, can be made.
For the record, the P-3 is NOT my design.

I know this amplifier, quite well. One of my customers also made a direct comparison.
 
Status
Not open for further replies.