John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Jump to Latest
#46,542
DF96 said:
I assumed the second part of the second sentence was your explanation, not his. Maybe the comma sent me off on the wrong path? However, you appear to agree with him - or are you quoting him out of puzzlement and disagreement?
Click to expand...

I quoted verbatim what he said. He marked and texted the word "delay" on the picture, I merely counted those itty bitty hash marks across the center line, understood the full screen to be 100 microseconds (10 KHz), and put a number to what he labeled "delay".

Cyril has measured and reported a reflection that occurs "4 uSec" delayed with respect to the stimulus.

His explanation is consistent with a short t-line mis-terminated at both ends.

Do you have the article?

jn
 
#46,544
PMA said:
The important thing are reflections at high frequencies (HF interference) and its influence on power amplifier audio band properties. I have done my first measurements of this phenomenon more than 10 years ago. AFAIK NP had published an article on similar theme even earlier.
Click to expand...
Although I have not the equipment to venture into this area, my experiences would suggest this is right on the money ...
 
#46,545
john curl said:
I have found that this 'series' of loudspeakers from the WATT1 to the Sasha works for me to reveal the subtle details that make differences in audio electronics most easily audible.
Another way is direct driven electrostatic microphones, which I also sometimes use.
An excellent counter-example is the Sequerra Met7. I got them cheap! And I have used them 10 times more than the WATT's over the years. They do OK, but they are 'forgiving' and that makes them difficult to use for serious evaluation of other components.
Without acoustic reproduces of similar quality to the WATT's or the STAX headphones, I can understand why many here don't bother much with subtle differences in audio electronics.
Click to expand...

Hope you didn't get the green ones John ........ 😛
 

Attachments

  • Sasha2 Press Release 6.jpg
    Sasha2 Press Release 6.jpg
    218.7 KB · Views: 288
#46,546
Horn...ey

john curl said:
I used my mono AR-1 based system from 1963-1967, in several locations, fairly successfully. Certainly it was better than most other 'student' systems, and I continued to experiment, with phono cartridges and adding a 'tweeter' to the AR-1. I finished college and got my first professional job(s), one in a hi end audio store that specialized in K-horns, triode amps, Mac, and Ortofon, etc., and my first professional job in electronics design (Friden) working under very experienced electronic engineers. I first heard and compared the AR to a K-horn in 1965. Wow! The K-horn just blew it out of the water. I promised myself that in future, I would get a K-horn or two when I could afford them. I also was able to compare the sound of the Ortofon moving coil cartridge with the latest Shure, etc cartridges, and again the Ortofon won hands-down.
So I switched my focus from AR to Paul Klipsch, as far as ultimate speaker quality was concerned, and I even met with and talked to Paul Klipsch, himself on several occasions in that time period. Under many conditions, the K-horns, even today, do pretty well for themselves, especially with large orchestra reproduction. They DO have certain serious problems that effect their stereo imaging, and EVEN the sound of a person's voice due to the rather large path lengths between the drivers. In those days, the ear was presumed to be 'phase deaf' from a mono source, so 2-4 ft path lengths between drivers was considered 'acceptable'. This changed slowly over the next decade by Richard Heyser's papers to the work of Manfred Schroeder that overcame the exasperated complaints by Paul Klipsch that the ear was phase 'deaf' which he was taught from the 1920's and even earlier. (more later)
Click to expand...

I recall building a slew of folded horns from 3/4" ply between 1968 and 1973 , in college, back in Charlottesville, for innumerable fraternity brothers. Great for massive dance (smoke,etc.) parties, highly efficient with the Kenwood, Marantz, Ampzilla,Fisher, Allied etc. amps at the time....

cheap 3 to 6 ft^3 enclosures with EV SP8B's... SP12's JBL LE12's etc. available at the time. Low power... HIGH output!!!

Ohh for the nostalgia.... eh? No worries about healthcare, wives, etc.....

John L.
 
Last edited:
#46,547
I saw this on the Analog Planet web site;

The recording of the National Symphonic Winds conducted by Lowell Graham was minimally miked using Sennheisers and captured to 30IPS ½” analog tape on John Curl’s Ultramaster reel to reel deck. The sound of the enormous bass drum is absolutely sensational as you will hear. Be prepared.

John worked his magic on this reel to reel.
 
#46,548
jneutron said:
I quoted verbatim what he said. He marked and texted the word "delay" on the picture, I merely counted those itty bitty hash marks across the center line, understood the full screen to be 100 microseconds (10 KHz), and put a number to what he labeled "delay".

Cyril has measured and reported a reflection that occurs "4 uSec" delayed with respect to the stimulus.

His explanation is consistent with a short t-line mis-terminated at both ends.

Do you have the article?

jn
Click to expand...

After review Bateman's plots are fatally flawed, he correctly takes the exact expression for Z0 but for some reason takes the magnitude to keep Z0 real and positive, this is just plain wrong. The equation at low frequencies yields a complex impedance and there is no problem with that. Hence his reflextion coefficients are also wrong.

Quick Python script using Bateman's numbers the magnitude lines up exactly with his plots.

from numpy import pi, linspace, log10 , abs, angle, inf, exp, arctan, tan, sin, cos, sqrt


f = 2000
Lseries = .01*(24.82e-6-(2.5444e-6*(log10(f))))
Rseries = (.01118+72.5e-6*sqrt(f))
Cshunt = 1e-6*(439.82e-6-(.1266e-6*(log10(f))))
Gshunt = 67.95e-12*(f)


Z0 = sqrt((Rseries + 2*pi*f*Lseries*1j)/(Gshunt+2*pi*f*Cshunt*1j))

print Z0,abs(Z0)
 
Last edited:
#46,549
Rick Miller said:
I saw this on the Analog Planet web site;

The recording of the National Symphonic Winds conducted by Lowell Graham was minimally miked using Sennheisers and captured to 30IPS ½” analog tape on John Curl’s Ultramaster reel to reel deck. The sound of the enormous bass drum is absolutely sensational as you will hear. Be prepared.

John worked his magic on this reel to reel.
Click to expand...
Is that the 'beast' in question in the background there, John?
I always thougt that yours was based on "VU" model , since they're so popular in US, but it looks like an "RC" to me.

Best,
 

Attachments

  • photo_3.jpg
    photo_3.jpg
    197.5 KB · Views: 280
#46,555
Hedde said:
Hello. I have been following this interesting debate....and thread 🙂

Can this, or part of this, be explained by the asymmetric distribution of current, making the mutual inductance complex in nature (M=Ma+jMb). Supra 2.0 probably have severe proximity issues in the audiorange beeing zip and with the conductors quite closely spaced.

Both Ma and Mb vary with frequency in no easy way changing phase and magnitude with frequency.

I am linking this old article on the topic:
http://www3.alcatel-lucent.com/bstj/vol14-1935/articles/bstj14-2-179.pdf
Click to expand...

Attached the real and imaginary parts of the mutual inductance in wires in severe proximity - from the linked article. The mutual inductance is no longer in phase quadrature, but up to 10-15 degrees away from it. Could this explain the phase/delay? And must the hearability test of short t-line remove proximity from the equation?
🙂
 

Attachments

  • cable crosstalk fig 7.jpg
    cable crosstalk fig 7.jpg
    61.5 KB · Views: 292
#46,556
jneutron said:
Do you have the article?
Click to expand...
No.

scott wurcer said:
After review Bateman's plots are fatally flawed, he correctly takes the exact expression for Z0 but for some reason takes the magnitude to keep Z0 real and positive, this is just plain wrong. The equation at low frequencies yields a complex impedance and there is no problem with that. Hence his reflextion coefficients are also wrong.
Click to expand...
Perhaps I don't need to read it, if it is so wrong.

This business of alleged significant/audible delays on audio cables always seems to come around eventually to either misinterpretation of measurements or miscalculation of theory. Bateman may have done both?
 
#46,559
hitsware said:
Is 'python' sorta the new 'c ' ..... ?
Click to expand...

It's also the language of the "rapture of the nerds" at the singularity. Python is worth it, it's free and has a huge collection of well supported libraries. Google "Hello World in 7 min." They're right. It's become my goto for quick complex calculations with nice plots.

Let me qualify what I said last night. Bateman's results might have very well been fine for his discussion of RF de-stabalization of amps but is no help in separating phase (he throws it away) from real delay. Remember at HF the line has a real positive Z0 and it does not matter. He does not really explore or need to explore the LF implications. I am in the camp that this problem will yield to lumped element analysis, but am willing to do careful experiments to look at the other viewpoint.
 
Status
Not open for further replies.
Top Bottom