Anyone got measurements of a real Blowtorch on the usual common or garden tests?
.. and also the new fangled Hirata tests and the PIM test announced at the latest AES?
I'd ask about Blind Listening Tests too but the designer objects ... so pls message me in secret.
Even non-blind will do.
_______________
BJTs only and I dunno about FETs.
.. and also the new fangled Hirata tests and the PIM test announced at the latest AES?
I'd ask about Blind Listening Tests too but the designer objects ... so pls message me in secret.
Even non-blind will do._______________
I've some success in getting my own circuit analysis package to simulate some of the stuff Baxandall reports in the Baxandall papers But this is only linear stuff and I dunno how easy it is to get LTspice to do the same thing.
BJTs only and I dunno about FETs.
The CTC Blowtorch measures relatively poorly, because there is no global feedback loop. It has an open loop bandwidth over 100KHz, however, and that is important (at least to me). Hirata? What do YOU know about the Hirata test? Ever try it? I have. This amp has no mechanism to trigger Hirata distortion.
PFB by current causes negative impedance, plus filter on input causes some boost below Fs. What is different in this article, a language: they mixed electrical correction with mechanical formulas.
Either you were in a hurry to answer or you didn’t read it
It is OK to enjoy what you have done in the past or now but consider the case that there might have been other people too who were technically literate, innovative and productive.
George
Motional Feedback gore -
Modern voltage amplifiers driving moving coil loudspeakers work successfully on the assumption that the Z is roughly constant over the entire freq range. However, this is not the case and although damping factor can minimize the affects, this is one of the causes of coloration.
Placing the speaker in the feedback (as I described), via a sensing resistor that monitors the current, the output of the speaker will be proportional to the amplifier input and independent of variations in the speaker Z. It makes a fairly good approximation of the geometric mean [(in/out)^1/2] The results are quit impressive with even low-medium quality speakers.
Thx- RNMarsh [high-feedback designs are encouraged, if stable]
Try it.
Modern voltage amplifiers driving moving coil loudspeakers work successfully on the assumption that the Z is roughly constant over the entire freq range. However, this is not the case and although damping factor can minimize the affects, this is one of the causes of coloration.
Placing the speaker in the feedback (as I described), via a sensing resistor that monitors the current, the output of the speaker will be proportional to the amplifier input and independent of variations in the speaker Z. It makes a fairly good approximation of the geometric mean [(in/out)^1/2] The results are quit impressive with even low-medium quality speakers.
Thx- RNMarsh [high-feedback designs are encouraged, if stable]
Try it.
Last edited:
At the time we were exploring double moving coil (one front, one back) to study servos, we had to deal deal with several parameters, i'd like to submit to your thoughts.
Increase of frequency range where it need maximal excursion of speakers greatly reduce available power handling by it.
Added retro-action signals decrease it too, amp side, during transients.
At low level, according to Fletcher & Mudson curves, you add a lot of power for low frequencies you don't hear at this level.
We chose to forget about that, and thought it was better and simpler to use a better driver.
The added damping here was of great benefit.
At high level, and not trying to increase basses bandwidth, the system is power consuming, need to multiply the power in amp if you don't want to add distortion. You need too a loudspeaker witch can handle hight peak levels (we destroyed some).
Reasons why we abandoned the idea.
Now, reconsidering all that together, if i should try-it again, i'll do with an overkilling power system, like 1KW amp, and some big diameters hight efficiency speakers able to afford-it.
With a benefit due to the better damping of their heavy cone.
I would not try-it with a single resistance, where direct signal and error signal are mixed, neither double coils in the same magnets for same and other obvious reasons.
I believe the best solution to get the position of the membrane would be an optical system with a mirror on the cone, and a laser or so. (no adding mass, absolute position).
At the end, it is very complicated, and, as traditional loudspeakers can give enough listening pleasure, and because i'm old and lazy...
May-be the market is old and lazy too, reason why servos were never adopted as the standard solution ?
Why do you think so? I even downloaded the PDF and read it.
...and that other people used similar approaches. Before the patent application, during te patent application, and after the patent application. PFB by current was used, EQ was used, both described in articles and used in commercially available gear. This particular patent combines both EQ and PFB by current, and adds the description in mechanical terms.
It is one of valid ways to go, by patenting particular combination of particular solutions, when solutions using each of 2 approaches, and the same combination of 2 approaches, but realized differently, were used before, or after the patent application.
Here is an example: http://www.triodeel.com/heathw6.gif
Actually this is successful cos speaker designers assume the amp is Lo Z.
If they assumed something else, they would design speakers differently as they did in da old days. The original Rice & Kellog is a good example.
Mr. Marsh's -ve resistance trick does good things for modern speakers but the response will not be as the speaker designer intended. Might be worse, might be better.
Last edited:
It does not matter what Z the amp assumes. What does mater, resulting frequency and phase response, and that Z does not have dips that cause excess distortions or damage of the amp. For tube amps it can't go too high to damage an output transformer, but such high Z can't happen practically.
Speaking of power followers ...
I've always liked Andrea's
http://www.audiodesignguide.com/my/Follower_99c.gif
2nd article down:
http://www.audiodesignguide.com/my/index.html
The virtual ground is cool!
I've always liked Andrea's
http://www.audiodesignguide.com/my/Follower_99c.gif
2nd article down:
http://www.audiodesignguide.com/my/index.html
The virtual ground is cool!
Attachments
Last edited:
Yes, it is cool. And cathode follower was even called "Common Anode" stage, but since there were no p-type tubes available, and people wanted to use a single battery for both common anode and common cathode stages, it got common to use "common anode", "common collector", "common drain" stage through "common power supply". What Andrea did, he creatively avoided cliche that was not needed in this particular case: nothing else had been powered from the same power supply. Unlike the majority of "copy-paste" professional designers.
Respect.
Just asking if you've conducted the Hirata test on Blowtorch. I take it your answer is, "No."
Any chance of conducting Ron Quan's test on Blowtorch? I hear Dick Sequerra and some other gurus think this is real important?
Last edited:
Alas, practically no speaker designers follow your rules
Even KEF who made a song & dance about this on their Reference Series at the end of the last millenium.
This often makes life difficult for amp makers.
There's also the small matter of how the speaker reacts with the environment ..when it acts as a microphone. This IS affected by amplifier output Z
The cynical or naive might claim Mr. Marsh's -ve resistance is just improving this bit especially at LF.
Baxandall's paper gives you some clues but you need to swot up on theory of speakers to grasp its full significance.Wave, do you compensate for the bass resonance impedance peak?
Last edited:
How were you able to read my mind ...?
Simple put, very conventional systems can create remarkably pleasurable sound, enough that all this worrying about the n'th, n'th of whatever is quite pointless -- at least for me ...
Frank
Alas, practically no speaker designers follow your rules
Even KEF who made a song & dance about this on their Reference Series at the end of the last millenium.
"And they are the very people who would prevent us to pick our noses"?
Shame on them.
No, I don't. I am too lazy. Audyssey compensates for me, for everything.
I only have a microphone transformer shunted by 0.1 Ohm resistor in series with subwoofer, for PFB by current. Why transformer? because the amp is bridged, class D.
Last edited:
To be complete, when we decided to abandon the research of servoed loudspeakers for industrial production, because we had here or here some nice results in the middle of problems, i had the feeling that we loosed something because i was not able to adress-it correctly.
It is teasing, i understand Mr. marsh appetite. I would love to have a reaction from mr. Kglee's experience vs mine, if I'm not in his ignore list.
BTW: Why don't we use Constantan wires for the loudspeakers moving coils ?
Because that would make the voice coil run much hotter than using copper. Also probably poorer eddy current damping but that might not be an issue in some cases.
- Status
- Not open for further replies.