John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#6,822
john curl said:

While it is true that I once said that the primary reason for complementary differential fet input was because it gave 2 independent drive sources, it is not the ONLY reason. The other reason is the inherent linearity of the stage during near overdrive. It is better, for the most part, than a single differential stage. This is what PMA found in his measurements put here a few weeks ago. However, this second reason is MINOR for Parasound designs and MAJOR for the 'Blowtorch' design. Can anyone reason why this is so?
Click to expand...


syn08 said:


Really interesting, I don't recall this one. Could someone point me to the message with those measurements?
Click to expand...

And these measurements, please!
 
#6,825
PMA said:


To say that asymmetrical LTP input is same good as complementary-symmetrical is a common wrong point of view of many circuit designers (e.g. Self), who have never analyzed JFET complementary-differential input stages in deep, and who always kept CCS source biasing in their analysis.
Click to expand...


Hi PMA,

Could you please elaborate on why you believe the asymmetrical LTP/VAS I described and use is inferior to the full complementary dual differential pair approach? I believe it is actually superior.

Bear in mind that the approach I am referring to provides full complementary push-pull operation of the VAS (unlike what you may be referring to in referencing Self).

The generic structure I am referring to starts with an N-channel JFET diff pair that is differentially loaded. It feeds a PNP differential pair. One output of that pair drives the top of the bias spreader. The other output drives a current mirror which functions as the NPN part of the VAS, driving the bottom of the bias spreader. The detailed arrangement is documented in my MOSFET power amp paper on my web site at www.cordellaudio.com. That arrangement provides exceptionally low distortion in that amplifier of 0.0006% THD-20 at full power.

Are we talking about the same arrangement?

If we are, could you be more specific about your objections to it?

Thanks,
Bob
 
#6,827
Charles Hansen said:
Edmond,

Your silly post adds nothing to this thread. Your insult to KBK are completely uncalled for.

I am sorry that you have so little imagination that you cannot conceive of the possible existence of directional effects in cables. It's really not hard. Try thinking about two facts:

a) Metal rectifiers. Please refer to:

http://en.wikipedia.org/wiki/Metal_rectifier

and notice that rectifiers can be made from a junction of copper and copper oxide. Now, do you think it is *possible* that there are junctions of copper and copper oxide in a piece of copper wire?

b) Musical signals are not symmetrical. It has been show that in at least some circumstance some listeners (probably not you) can detect the absolute polarity of specific signals.

If you put these two facts together, it is not such a stretch to conceive that maybe, just *maybe* there could be small directional effects in cables that could create audible results.

But instead you prefer to insult people who have better hearing capabilities than you do. Why?
Click to expand...

Hi Charles,

This all makes sense and is plausible. Rectification, if it occurs, is certainly a brazen source of distortion products.

If the rectification effects you describe are there, is it reasonable to believe that they would be measurable? If so, are they measurable? If not, why would they not be measurable?

Thanks,
Bob
 
#6,832
Re: PIM

Edmond Stuart said:


Hi Bob,

Of course you can simulate phase shifts with great precision by means of an AC analysis. However, my point is what does it tell us about the amount of PIM. As far as I know, PIM is defined as the phase deviation/distortion as function of frequency and amplitude (correct me if I'm wrong). Doing an AC analysis, you will get phase shifts as function of frequency (of course) and the DC off-sett and at zero AC amplitude.
So far so good.

In my previous post (about this topic) I was moaning about precision, by which I mean how to translate phase modulation (due to DC off-set variation) obtained by AC analysis, into phase modulation obtained by transient analysis (due to amplitude variation). Not an an easy task, I guess. Do you know how to do that? I don't.

Then we have Pavel's simulation, that shows a shift of the zero-crossings (measured one per cycle!). Of course, a good amp should never show up this kind of behavior, but is it PIM? We don't know, as the same phenomenon might be caused by another kind of disturbance.
As pointed out in one of my previous posts, I suggest to look at the rising and falling ZC as well (i.e. two ZC's per cycle), as that might eliminate the effect (to some extent) of DC off-set, but still the outcome is also affected by the amount of even harmonics.

So, what to do? It depends on how you define PIM. If you define it as function of amplitude, then calculate the real phase by mens of a FFT (or a equivalent of that method). If you define it as function of DC off-set, just do an AC analysis. But in no case a single ZC measurement.

Cheers,
Edmond.
Click to expand...


Hi Edmond,

These are good points and questions.

In all of my discussions, I have defined PIM as Otala did when he coined the term. Here the definition of PIM and how it is measured are synonimous. It is directly analogous to SMPTE IM. 60Hz and 6000Hz are mixed together in a 4:1 ratio and applied to the amplifier under test. The rms phase modulation on the 6000 Hz carrier is measured and expressed in either rms degrees or in rms microseconds.

That is also how my PIM analyzer worked. It used an ultra-low jitter 6 kHz PLL to lock onto the received 6 kHz carrier after the 60 Hz part of the signal was removed. Then the output of the PLL was used to synchronously (coherently) do phase detection on the 6 kHz signal. The quadrature output of the VCO was also used to do synchronous AM detection on the 6 kHz carrier to recover conventional IM in a very sensitive way as a result of the use of coherent detection techniques.

There may be other ways to measure other-defined PIM, of course. In particular, it is very important for people to realize that PIM cannot escape detection by a spectrum analyzer. So, for example, even if you could have an amplifier that was making substantial amount of PIM and was making no THD or amplitude IM, it would still show up as IM sidebands on a 19 + 20 kHz CCIF IM test.

For the reasons you alluded to, it seems relatively unreliable to infer PIM from the time of zero crossings from a transient simulation, as there could be numerous other causes for such behavior.

Cheers,
Bob
 
#6,835
Sigurd Ruschkow said:
Lineup -
I do not want to post in the Pass forum. If I wanted to to do that, I would do it 😉

What I want to know more about is what other designers' experience is with different types of telescopic cascodes.




Sigurd
Click to expand...


Hi Sigurd,

I have not heard the term "telescopic" cascode before, but I have used the term "driven cascode" to mean what I think is the same thing: namely, the bases (or gates) of the cascode are driven with a signal that is a reasonable replica of the common-mode signal applied to the differential pair being cascoded.

I have had very good results in non-inverting power amplifiers by driving the bases of the cascode stage with a replica of the common mode signal derived directly from the output of the amplifier with a divider network that is substantially identical to the feedback network. This yields a very close replica of the actual common mode signal to which the diff pair is subjected without adding and circuitry that might load the input circuits.

The underlying assumption here, of course, is that the differential error input to the LTP being cascoded is small by comparison. In the case of a no-NFB amplifier, this would likely be not true.

Cheers,
Bob
 
#6,837
Re: Re: PIM

Bob Cordell said:

That is also how my PIM analyzer worked. It used an ultra-low jitter 6 kHz PLL to lock onto the received 6 kHz carrier after the 60 Hz part of the signal was removed. Then the output of the PLL was used to synchronously (coherently) do phase detection on the 6 kHz signal. The quadrature output of the VCO was also used to do synchronous AM detection on the 6 kHz carrier to recover conventional IM in a very sensitive way as a result of the use of coherent detection techniques.
Click to expand...

Bob,

You just described a lock-in-amplifier. Never thought about, it will be a nice application for my old PAR lock in.

But then I can't imagine an amp with zero IMD and significant/measurable PIM...
 
#6,838
PMA, thanks again for your efforts.
I find it absolutely astonishing how presumptuous some people can be.
Even in a Pub, after many beers, such a behavior isn't up to grown up people.

It is sad to see John's contributions being flood with tons of useless posts.

Tino
 
Status
Not open for further replies.
Top Bottom