Not sure the phase margin is relevant to an amplifier with no global feedback.........
So what I see is basically you have
Questions:
How do you define gain without the feedback?
Did you do LG analysis?
What are the THD numbers?
With the feedback version, is it better to use LTP to compare input and output signals?
I wonder is there can be a way to adjust “tubeness” of the sound? So how much of a second harmonic is created?
Also, adjusting the extent of the feedback applied?
That would be cool 😎
- input network filters
- class A VAS
- 2EF output stage with three transistors in parallel
Questions:
How do you define gain without the feedback?
Did you do LG analysis?
What are the THD numbers?
With the feedback version, is it better to use LTP to compare input and output signals?
I wonder is there can be a way to adjust “tubeness” of the sound? So how much of a second harmonic is created?
Also, adjusting the extent of the feedback applied?
That would be cool 😎
- Gain without feedback operates in the voltage domain and its the ratio of 20*logbase10(R22/R20 +1), drain resistor / source resistor
- Without feedback upto 2%, we are running that output stage mostly in class A to improve that THD
- Introducing the LTP raises challenges we don't need at this fun level, the Pathos sounded really good didnt it? Why do people still look for those really vintage amps? We are trying to stay true to our theoretical tube
- Its that dominant second harmonic
These are fun ideas you can play around with if you have these parts in your workshop and see whether or not the sound is reasonable for serious consideration
The feedback one has gain dependent on source impedance, which makes ensuring stability with an arbitrary source trickier to guarantee.
Congratulations on rolling yur sleeves up and actually designing.
Bonus points if you actually build it 🙂
That said:
1) on the non NFB original version, you should split and bootstrap R22 (which you did in the second one).
2) an enhancement MosFet gain stage (Vas here) is notoriousy non linear/unsymmetrical by itself.
The NFB version shows quite unsymmetrical waveform at 20kHz, I´d love to see same out of the non NFB one.
Would expect even higher assymetry.
Same at, say, 1kHz and 100Hz.
Intermodulation would also be strong, any numbers available for it?.
Bonus points if you actually build it 🙂
That said:
1) on the non NFB original version, you should split and bootstrap R22 (which you did in the second one).
2) an enhancement MosFet gain stage (Vas here) is notoriousy non linear/unsymmetrical by itself.
The NFB version shows quite unsymmetrical waveform at 20kHz, I´d love to see same out of the non NFB one.
Would expect even higher assymetry.
Same at, say, 1kHz and 100Hz.
Intermodulation would also be strong, any numbers available for it?.
Yes its quite an interesting amplifier luckily most sources should work just fine
Here are the suggested changes, keep the comments flowing 5V/us,0.3%THD. If everyone is satisfied I can start preparing PCBs. Is this the best midrange amplifier possible? We can create a two way or the three way system crossed off at 200Hz ,2kHz and have 1DIFFET https://www.diyaudio.com/community/...fier-for-whatever-speaker.401371/post-7404062 power the woofers
Attachments
Last edited: