smoking-amp, regarding your first question
My 6DQ5 crazy drive amp does not employ any output stage local feedback. The amplifier is built with a mu-follower front end dc coupled to a differential phase splitter employing a ccs in the tail. The only coupling caps are between the phase splitter and the FET source follower stages which feeds the both g2 and g1 of the 6DQ5 ouptut tubes. The amplifier does have -12 db global feedback. The distortion as measured with my HP 339 distortion analyzer is showing 0.16% at 1 watt rms output.
Regarding your second question
When I look at the modeled distortion on ltspice (which I give limited respect to), at one watt rms output at 1khz fundamental, the 2nd harmonic is down 99.6db, the 3rd is down by 84.1db, the 4th is down by 162.3db and 5th is down 145.5db. Again, I give little respect to the numbers ltspice measures for Thd. This amp shows 1 watt rms at 1khz to be 0.006% Thd on ltspice.
You maybe interested to know the output tubes are running with 455v at 41ma. or about 19 watts dissipation. At this point in time, I have done very little optimization on this amplifier other than building it and bringing it up. It may not be running optimum at this point it time. I will play more with it as time permits. I am very interested in your crazy drive as it looks very promising. Mickeystan
My 6DQ5 crazy drive amp does not employ any output stage local feedback. The amplifier is built with a mu-follower front end dc coupled to a differential phase splitter employing a ccs in the tail. The only coupling caps are between the phase splitter and the FET source follower stages which feeds the both g2 and g1 of the 6DQ5 ouptut tubes. The amplifier does have -12 db global feedback. The distortion as measured with my HP 339 distortion analyzer is showing 0.16% at 1 watt rms output.
Regarding your second question
I cannot see any visual output stage cross-over distortion with my Tektronix 465b oscilloscope.
When I look at the modeled distortion on ltspice (which I give limited respect to), at one watt rms output at 1khz fundamental, the 2nd harmonic is down 99.6db, the 3rd is down by 84.1db, the 4th is down by 162.3db and 5th is down 145.5db. Again, I give little respect to the numbers ltspice measures for Thd. This amp shows 1 watt rms at 1khz to be 0.006% Thd on ltspice.
You maybe interested to know the output tubes are running with 455v at 41ma. or about 19 watts dissipation. At this point in time, I have done very little optimization on this amplifier other than building it and bringing it up. It may not be running optimum at this point it time. I will play more with it as time permits. I am very interested in your crazy drive as it looks very promising. Mickeystan
Thanks, Mickeystan, for the info.
At some point, when you have the analyzer connected, it would be interesting to temporarily remove the Rg2g1 (from g2 to g1) resistors, thus grounding g1 effectively, to see how the distortion figure changes. That would compare Crazy Drive to g2 drive. Although the input level would need adjusting to use comparable output levels. I suppose biasing might need adjusting too for the optimum for each.
Even more interesting if g1 only drive could be compared. That would require a lot more modification though.
Another point of interest would be to see how the distortion varies with the DC bias current in the tubes. G2 drive typically allows low idle current. Not sure how well it compares with g1 drive for crossover distortion when both are optimised. One would expect the Crazy Drive to have some resemblance to the g2 drive case.
When you say you cannot see any crossover distortion on the O'scope, is that looking directly at the output signal or is that the residual signal from the dist. analyzer? Tuning the bias while observing the analyzer residual should be quite illuminating.
Very encouraging to see some good results. I hope I can get a local feedback Crazy Drive version put together before too long. Either type can provide results, versus g1 only drive, for comparison. Even if the two drive approaches come out similar for distortion when considering gain loss (ie, same tubes), the Crazy Drive should allow an extra approx. 3X gain to be put into the drivers (or front end for global) to make up for the output stage gain loss (giving similar loop stability with the g1 case then, but 3X lower dist. then).
At some point, when you have the analyzer connected, it would be interesting to temporarily remove the Rg2g1 (from g2 to g1) resistors, thus grounding g1 effectively, to see how the distortion figure changes. That would compare Crazy Drive to g2 drive. Although the input level would need adjusting to use comparable output levels. I suppose biasing might need adjusting too for the optimum for each.
Even more interesting if g1 only drive could be compared. That would require a lot more modification though.
Another point of interest would be to see how the distortion varies with the DC bias current in the tubes. G2 drive typically allows low idle current. Not sure how well it compares with g1 drive for crossover distortion when both are optimised. One would expect the Crazy Drive to have some resemblance to the g2 drive case.
When you say you cannot see any crossover distortion on the O'scope, is that looking directly at the output signal or is that the residual signal from the dist. analyzer? Tuning the bias while observing the analyzer residual should be quite illuminating.
Very encouraging to see some good results. I hope I can get a local feedback Crazy Drive version put together before too long. Either type can provide results, versus g1 only drive, for comparison. Even if the two drive approaches come out similar for distortion when considering gain loss (ie, same tubes), the Crazy Drive should allow an extra approx. 3X gain to be put into the drivers (or front end for global) to make up for the output stage gain loss (giving similar loop stability with the g1 case then, but 3X lower dist. then).
Last edited:
smoking-amp, cross over distortion was only looked at while viewing a couple of cycles of the sine wave at 1khz on the scope. I like your idea of looking at the residual on the distortion analyzer. I will take a look at this at various idle currents when I get back to tuning on this amplifier. Concerning your other question about comparing distortion of the crazy drive setup vs. only g2 drive, I did look at this on the distortion analyzer and the 1 watt rms numbers looked virtually the same as I am now seeing on crazy drive all though I cannot say I am sure that the idle current was the same. When I was running straight g2 drive, I had the g1 tied directly to the cathodes. The cathodes were then grounded via a 1 ohm resistor for measuring current. The output transformer in this amp is an Edcor 6.6k to 8 ohm secondary. You might say this is a junk box amplifier as it is made from various parts from stock on hand for the most part. I will share any interesting findings after I do the residual distortion versus idle current investigation. Mickeystan
more different between USA and japan tube amp circuits .usa amp circuit just for loudly sound but Japanese Design circuits is best of sounding .
I agree to compare my amps with best Japanese examples. Both for loudness and for best of sounding.
I'd suggest that an ss rectifier diode is a step backwards, not forwards - no matter how 'fast' it is. Perhaps if the SS diode was a synchronous rectifier where a FET on-resistance changed smoothly and significantly for conduction current near zero, then that would be catching up to where we once were 60 years ago.
hi
can explain abut your amp circuit ?and did you hear the sound of Japanese famous amps before?
Not necessary. Tube rectifiers are more forgiving to wrong wiring and layout. But SS rectifiers help to reduce dynamic resistance of power supplies that is good for minimizing dynamic distortions.
Ahh, so the 'help' is really just related to using a large cap as the interface between rectifier and amplifier.
Yes, capacitance per volume, weight and cost today is much higher than in ancient Egypt. It is affordable. However, electricians demand to correct "Power factor", since it is convenient for them to deliver sinewave AC... ;-)
But if you want, you still can use a swinging choke after diodes, before a cap. Still, voltage drop and high dynamic resistance of a rectifying tube would be out of an equation.
Last edited:
so true, i get many of these 470/450v caps for almost nothing....just look for discarded electronics and there you will find them, an esr tester to screen for defectives and there you have it...
i am quite sure that if designers during the ancient egyptian time were here today, they will be jumping for joy....😀
- Status
- Not open for further replies.