I am with Aksa and Prasi - beautiful layout Thiagomogi! The on board fat caps have their place in designs where we may not want to have a separate PSU board to clutter things up - such as when using a compact SMPS. This way, everything is off the floor and mounted on the side walls where the big heatsink is.
Have you posted Gerbers for this yet?
Thanks!
@Aksa - thanks for the kind words - it's great to learn from the Masters on this forum.
Have you posted Gerbers for this yet?
Thanks!
@Aksa - thanks for the kind words - it's great to learn from the Masters on this forum.
I am with Aksa and Prasi - beautiful layout Thiagomogi! The on board fat caps have their place in designs where we may not want to have a separate PSU board to clutter things up - such as when using a compact SMPS. This way, everything is off the floor and mounted on the side walls where the big heatsink is.
Have you posted Gerbers for this yet?
Thanks!
@Aksa - thanks for the kind words - it's great to learn from the Masters on this forum.
Thanks guys .
All I know I owe to this forum , so I'm happy to be helpful.
Gerber files attached would be happy to see built.
166x75mm size
Best Regards
Total collector current from the inverter transistor might be 8.5mA
Current through resistor might be 6mA
Current through diode might be 2.5mA
Here is what i measure.I see 4.2mA collector current,do you see something wrong?
Attachments
Thimios,
All your voltages and currents look fine; only a polishing is needed!
1. I would suggest you straddle the Vbe multiplier and the zener with C7 (? - I am doing this on mobile phone lying in bed so can't easily check).
2. I estimate the quasi base is drawing 1.5mA so adding your measured 4.2mA we
have 5.7mA drawn through the phase inverter collector. You have 520mV across the Bax resistor and only 0.5mA drawn through the Bax diode.
When the nmos starts to switch on this diode will turn off swiftly and spray artefacts into the output node very close to the crossover point where discontinuities in the waveform are fatigueing. Solution: Increase the Bax resistor to 120 ohms.
Ciao
Hugh
All your voltages and currents look fine; only a polishing is needed!
1. I would suggest you straddle the Vbe multiplier and the zener with C7 (? - I am doing this on mobile phone lying in bed so can't easily check).
2. I estimate the quasi base is drawing 1.5mA so adding your measured 4.2mA we
have 5.7mA drawn through the phase inverter collector. You have 520mV across the Bax resistor and only 0.5mA drawn through the Bax diode.
When the nmos starts to switch on this diode will turn off swiftly and spray artefacts into the output node very close to the crossover point where discontinuities in the waveform are fatigueing. Solution: Increase the Bax resistor to 120 ohms.
Ciao
Hugh
Ok Hugh take your time,i will wait for all your suggestions when you are ready.
Thanks.
Thanks Kevin, I think You just answered my question. Im a happy owner of Your book, BTW(P of P)
Cheers
J.
Yes, but it could easily be a minor issue with ground impedance too. Very critical to THD, but, and note this, YOUR EAR WON'T NOTICE!!
Sure i can't hear but i can see
Finally it doesn't matter of gnd resistance.
Each channel tested alone with exactly the same configuration .
Why this difference?
Amplifier modules are the left channel when on the Right Channel is an input. out.sound card loop.
Sure i can't hear but i can see
Finally it doesn't matter of gnd resistance.
Each channel tested alone with exactly the same configuration .
Why this difference?
Amplifier modules are the left channel when on the Right Channel is an input. out.sound card loop.
Attachments
Last edited:
Thimios,
If all the active parts are the same, and their betas and Vbe setting are the same, and input and output are identical there should be no difference in THD. We are down to lowish global negative feedback because of the nested fb resistor (68k) and therefore the THD will vary with unmatched parts.
However, you should hear no difference! What are your thoughts now you have a stereo system?
Cheers,
Hugh
If all the active parts are the same, and their betas and Vbe setting are the same, and input and output are identical there should be no difference in THD. We are down to lowish global negative feedback because of the nested fb resistor (68k) and therefore the THD will vary with unmatched parts.
However, you should hear no difference! What are your thoughts now you have a stereo system?
Cheers,
Hugh
Wow, I populated the boards yesterday but was having fits trying to get it working. Current was through the roof. This morning, laying in bed I noticed that D7 is a 3.3V Zener. I had a 1n4007 in there. I just checked my parts bin and the closest I have it 4.7V so I guess I'll have to place an order. Bummer.
Andrew,
My apologies for slow response to your questions.
1. The Bax diode should be passing at less 1.5mA at quiescent. Reason is that while it turns on and off this transition should occura couple volts above crossover when the nmos is passing sufficient current to swamp the switching artefacts of the diode.
2. The Bax 22nF cap helps to squelch the switching noise as well.
3. When you examine the nmos current on LTSpice you notice that careful of Bax network, the quiescent current and the base to rail resistor and the nmos choice, you can keep the nmos ON and passing current across the FULL waveform!
The quasi does switch of course, but the nmos, which determines the voltage at its source stays on the full 360 degrees.
There are benefits for the quasi topology; it blurrs the classic crossover we see with the complementary OPS.
The asymmetrical treatment of the signal generates H2, H4 and H6 and most listeners perceive it as 'warm'
Hugh
My apologies for slow response to your questions.
1. The Bax diode should be passing at less 1.5mA at quiescent. Reason is that while it turns on and off this transition should occura couple volts above crossover when the nmos is passing sufficient current to swamp the switching artefacts of the diode.
2. The Bax 22nF cap helps to squelch the switching noise as well.
3. When you examine the nmos current on LTSpice you notice that careful of Bax network, the quiescent current and the base to rail resistor and the nmos choice, you can keep the nmos ON and passing current across the FULL waveform!
The quasi does switch of course, but the nmos, which determines the voltage at its source stays on the full 360 degrees.
There are benefits for the quasi topology; it blurrs the classic crossover we see with the complementary OPS.
The asymmetrical treatment of the signal generates H2, H4 and H6 and most listeners perceive it as 'warm'
Hugh
Hi Hugh,
I installed a single blue LED and it is working perfectly. I had to install 100k for R5 in order to zero the offset with the 200R trimmer. I actually installed a 100k trimmer in place of R5 and centered VR1. I had the 100k maxed and it turned out to be about perfect so I just installed 100k resistors and now all is well. I have it running on +-44V right now because my variac went out last night. I will have to set up an adjustable PSU to do further testing because I don't have a 24-0-25vac transformer available right now. I can probably get by with the +-44V for some quick testing if I run it on 8 ohm speakers.
Stay tuned.
Blessings, Terry
I installed a single blue LED and it is working perfectly. I had to install 100k for R5 in order to zero the offset with the 200R trimmer. I actually installed a 100k trimmer in place of R5 and centered VR1. I had the 100k maxed and it turned out to be about perfect so I just installed 100k resistors and now all is well. I have it running on +-44V right now because my variac went out last night. I will have to set up an adjustable PSU to do further testing because I don't have a 24-0-25vac transformer available right now. I can probably get by with the +-44V for some quick testing if I run it on 8 ohm speakers.
Stay tuned.
Blessings, Terry
Terry, when you have the time look at this.
I believe that is a good solution for your tests.http://www.diyaudio.com/forums/power-supplies/294017-controling-main-transformer.html#post4774973
looking forward to the test
Yes.. You bet, I am waiting too. Now we have two pairs of ears to give glimpse of SQ.