Hi All,
I'm working on this Class A Amplifier concept and am hoping for some feedback.
The main aim is to use a mosfet output stage and obtain rail to rail output without boosted supplies. This is successfully attained using the double bootstrap and folded drivers seen in the schematic. This part of the amplifier works very well in simulation. It can run all the way to the rails with no problem.
The next aim is to use Allison biasing to prevent thermal runaway, but still have a smooth class A to AB transition. I think I have achieved this by beating the poor Allison into submission and making it unable to change the bias particularly quickly. This is done using the 22uF cap across the Allison bases, substantial base stoppers, frequency compensation on the Allison transistors and a huge 220uF cap across the bias spreader. In effect this makes the bias voltage almost static and only able to change very slowly.
In simulation the idea is looking promising. There's still some tweaking to be done but it is looking good so far I think.
If no one can find any obvious issues with my approach I will post some further waveforms and develop things further, etc.
I like the idea of being able to turn it on and have it self adjust to operating conditions.
Any thoughts would be appreciated!
Greg.
I'm working on this Class A Amplifier concept and am hoping for some feedback.
The main aim is to use a mosfet output stage and obtain rail to rail output without boosted supplies. This is successfully attained using the double bootstrap and folded drivers seen in the schematic. This part of the amplifier works very well in simulation. It can run all the way to the rails with no problem.
The next aim is to use Allison biasing to prevent thermal runaway, but still have a smooth class A to AB transition. I think I have achieved this by beating the poor Allison into submission and making it unable to change the bias particularly quickly. This is done using the 22uF cap across the Allison bases, substantial base stoppers, frequency compensation on the Allison transistors and a huge 220uF cap across the bias spreader. In effect this makes the bias voltage almost static and only able to change very slowly.
In simulation the idea is looking promising. There's still some tweaking to be done but it is looking good so far I think.
If no one can find any obvious issues with my approach I will post some further waveforms and develop things further, etc.
I like the idea of being able to turn it on and have it self adjust to operating conditions.
Any thoughts would be appreciated!
Greg.
Attachments
Member
Joined 2009
Paid Member
Excellent set of goals, I like your thinking !
I don't know much about the Allison, Keantoken introduced me to it a couple of years ago where he put together a headphone amp. He may well be along to join this thread.
I do like the double bootstrap for FETs, this is the third thread on the topic of late and I suspect it's something Hugh Dean is familiar with.
No doubt there's some experimenting to be done and some finessing too - I would add some emitter degen to the LTP for starters.
Depending on how this thread goes, I may develop a version based on a symmetrical input stage.
I don't know much about the Allison, Keantoken introduced me to it a couple of years ago where he put together a headphone amp. He may well be along to join this thread.
I do like the double bootstrap for FETs, this is the third thread on the topic of late and I suspect it's something Hugh Dean is familiar with.
No doubt there's some experimenting to be done and some finessing too - I would add some emitter degen to the LTP for starters.
Depending on how this thread goes, I may develop a version based on a symmetrical input stage.
Member
Joined 2009
Paid Member
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.