John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#9,783
Mr. Curl,

I saw something today. I made an amp with low global feedback. I use spectrum analyzer with computer soundcard. When I put 19khz+20khz IMD test, the peaking at 1khz is quite high. When I put 18khz+20khz IMD test, the peaking at 2khz is quite high.
Is this always the case with low feedback power amps? Or it is just this design only?
 
#9,785
Lumanauw, this is also a "lowfeedback" design.
 

Attachments

  • 13+14_s.gif
    13+14_s.gif
    59 KB · Views: 476
#9,791
Can you elaborate on that?
Click to expand...
this particulary servo senses input and output DC level and arranges for accurate DC amplification.
A servo that only senses the output and controls DC to zero serves basically as highpass with the known effects on amplitude and phase response.
One has to question why not use a good quality capacitor instead. 😕

Such a DC accurate servo is used on the Alexander amplifier, which seems not to have a reputation for good sound, but offers a lot of interesting detail solutions.
regards
 
#9,792
Usually, servos really screw up DC performance when used. Firstly, they remove the DC component, in order to remove the DC offset, and they will NOT pass DC. This is one of the pitfalls with servos. You should NOT use the servo in the audio bandwidth, itself, or else you add the servo to the audio path in a way as to reduce audio quality. Secondly, you will screw up the low frequency transient response, by excessively removing the DC component for low frequency transients. You should retain the dynamic offset as long as possible, in order to preserve the original waveform. This is most probably one of the components of the audio signal, as to why absolute polarity can be important in really good sound systems.
PS Really good audio coupling caps cost BIG MONEY! More than servos, that's for sure.
 
#9,793
"What I learned in 1971, is that IC's did not sound as good as tubes, even if they MEASURED better than typical tube circuits that we had available."

John, that was 1971 . . . . era of the 741 et al. The op-amp world has moved on since then. Hell, in the interim we even discovered TIM, PIM, their root causes and how to fix them. 100db of feedback? 70dB of feedback or even 50dB - who on this forum is advocating that without some very pointed caveats? Why use 1971 as your reference point when mentioning IC based solutions?

😉
 
#9,795
I didn't use the 741 in audio. I first used the 741 for servos at Ampex Research back in 1969. I KNEW that they were next to useless for audio. I used the Radiation Inc, (Harris Assoc.) RA911 with extra leads for added bias, compensation and +/- 24V supplies. It was 'almost' good enough in slew rate (+5/-2.5V/us) and capable of 50 ma output. It would actually drive 600 ohms pretty well. This is what I STARTED with, back in 1970. Things have improved in some ways, you know, fet input, lower noise, higher slew rate, but not enough to overcome every problem. Discrete and tubes still tend to be better for analog circuits.
 
#9,797
is there a guesstimation, what levels of abuse have to accepted and survived by the equipment?
CE calls for ESD protection, but I'm not shure if this level of protection is really nescessary in practice.
At least the tests, that I heard of, seemed a little bit cruel to semiconductors.
regards
 
#9,800
Wavebourn said:

It's a totally different story; no need to invent lamps before sand in order to make it work.
Click to expand...

It's not about inventing the wheel, but about developing the solid state physics. While lamps could be very well developed and understood using the 19th century physics (corpuscular theory) plus the Einstein theory he got the Nobel prize for (thermionic emission), semiconductor devices could not be developed without the solid state physics which is based on the quantum theory that was not complete much before the mid 20th century.
 
Status
Not open for further replies.
Top Bottom