I'm looking forward to seeing how turning P3 back and forth affects the shape of that distortion waveform, 6L6!
Yes, indeed, I am also quite interested. 🙂
Also the pots on the output degeneration of the F6. I've got both projects brewing.
Also the pots on the output degeneration of the F6. I've got both projects brewing.
Jim, try also 10R in FE degeneration and null 2nd in FE with symmetrical output stage. Than play with output pots. That what people say, Nelson included 🙂. I am doing so too. You can read about my thoughts on different F6 configurations in F6 thread 🙂.
Attachments
You have to make the probe (it's easy) found here - Electronic Equiptment - DIY soundcard scope probe
I'm reading across a 4ohm dummyload resistor.
FFT is from Electroacoustics toolbox. (An older version) Electroacoustics Toolbox
I'm reading across a 4ohm dummyload resistor.
FFT is from Electroacoustics toolbox. (An older version) Electroacoustics Toolbox
Great thread and nice measurements
The number of output devices and bias per device will move you around on the hfe vs Ic or gm vs Id curves on the data sheet. You want to stay up on the most level part of the curve at the top. Newer transistors have a much better flat area and sound better.
Compare Figure 3 and 4 on http://www.onsemi.com/pub_link/Collateral/MJW3281A-D.PDF
to figure 4 on http://www.onsemi.com/pub_link/Collateral/MJ15022-D.PDF
Ritchie Miller at Harman Kardon had a trick for setting the bias. With the signal and residual on the same screen both with 0V in the middle, he wanted the two to cross where the signal was at 0 volts. You could run the bias up further if you wanted, and it may produce lower measured THD, but it didn't sound any better. He was the guy they invented the phrase 'Golden Ears' for.
Cheers
The number of output devices and bias per device will move you around on the hfe vs Ic or gm vs Id curves on the data sheet. You want to stay up on the most level part of the curve at the top. Newer transistors have a much better flat area and sound better.
Compare Figure 3 and 4 on http://www.onsemi.com/pub_link/Collateral/MJW3281A-D.PDF
to figure 4 on http://www.onsemi.com/pub_link/Collateral/MJ15022-D.PDF
Ritchie Miller at Harman Kardon had a trick for setting the bias. With the signal and residual on the same screen both with 0V in the middle, he wanted the two to cross where the signal was at 0 volts. You could run the bias up further if you wanted, and it may produce lower measured THD, but it didn't sound any better. He was the guy they invented the phrase 'Golden Ears' for.
Cheers
Test Waveform
6L6, I'm wondering what waveform (1kHz testsignal) source did you use? From the PC to Amp to the PC (for the Analyzer) or from CD to Amp to PC?
Thanks,
Mallard
6L6, I'm wondering what waveform (1kHz testsignal) source did you use? From the PC to Amp to the PC (for the Analyzer) or from CD to Amp to PC?
Thanks,
Mallard
The HP 339A has the oscillator. It is connected to the input of the amp under test.
The HP 339A 'monitor output' is connected to the oscilloscope, which shows the distortion residual -- the waveform of the distortion, I.E., the signal at the test tone freq. after the fundamental is notched out.
The computer shows two different things depending on when the post was made;
Where it is a photo of the screen that is an FFT of the distortion residual, looking at the same signal as the oscilloscope.
But where it is a screenshot, that is an FFT of the signal at the 4ohm load resistor on the output of the amp - which is a better way (IMO) of seeing the relationships of the 2nd, 3rd, and higher harmonics.
The HP 339A 'monitor output' is connected to the oscilloscope, which shows the distortion residual -- the waveform of the distortion, I.E., the signal at the test tone freq. after the fundamental is notched out.
The computer shows two different things depending on when the post was made;
Where it is a photo of the screen that is an FFT of the distortion residual, looking at the same signal as the oscilloscope.
But where it is a screenshot, that is an FFT of the signal at the 4ohm load resistor on the output of the amp - which is a better way (IMO) of seeing the relationships of the 2nd, 3rd, and higher harmonics.
6L6,
Nice work.
Can you try this?
Set the bias as desired and monitor the scope as the amp heats up for a few hours. Interesting to see how the circuit responds.
I might try this in the near future as I have a Pioneer SA-9500, SX 950 and a Marantz 2238B here for repairs. The SA-9500 had a thermal runaway issue. Bias ran away after 3.5 hours on.
After any repair, I run the amp for most of the day. This time it payed off.😀
I'll monitor the bias on the DMM while viewing the distortion pattern on the scope. No, I won't stare at the scope for hours.😱
Nice work.
Can you try this?
Set the bias as desired and monitor the scope as the amp heats up for a few hours. Interesting to see how the circuit responds.
I might try this in the near future as I have a Pioneer SA-9500, SX 950 and a Marantz 2238B here for repairs. The SA-9500 had a thermal runaway issue. Bias ran away after 3.5 hours on.
After any repair, I run the amp for most of the day. This time it payed off.😀
I'll monitor the bias on the DMM while viewing the distortion pattern on the scope. No, I won't stare at the scope for hours.😱
Last edited:
Member
Joined 2009
Paid Member
I wonder what the distortion would look like with a real speaker as a load instead of a resistor ?
- Status
- Not open for further replies.