I while back I posted a amplifier design with only one transistor. I named it SEWA (Seven Watt Amplifier). It seems most people interested in the design actually wanted LESS power (power dissipation I guess). While the SEWA design has evolved into something different (coming later), I have renamed this design OTA - One Transistor Amplifier. Attached is the revised schematic. Enjoy! And please let me/us know if you decide to try it 😉
Mad_K said:
Growing up with OTA meaning Operational Transconductance Amplifier, you fooled me! I also grew up with designs like yours, known as a Source Follower. But what's in a name, as long as you succeed in confusing everyone!😉
Nice schematic, BTW. Why don't you bias the source at the supply midpoint 11.5V to maximise output power?
Jan Didden
Jan, if you do that you get unsymmetrical clipping (negative). The resistor needs enough voltage headroom across it to be able to deliver the required outputcurrent. So when the amp swings - 7,5 volts you have 5,5 volts left over the 5R resistor to deliver 1,1A of current to the load.
Upupa, -yes, that would be nice 😉
Upupa, -yes, that would be nice 😉
Re: Re: OTA - One Transistor Amplifier
For a push-pull or CCS class A, midpoint is most effective.
But
when using a single end with power resistor
it is different.
Say 8 Ohm resistor and 8 Ohm load.
At negative swing, the load will be in series with power resistor.
A theorectical ( not counting drop across transistor )
most effective class A single end
will be if 2/3 (66.66 % ) of voltage supply is across power resistor.
This is if LOAD=Resistor (for example 8 Ohm and 8 Ohm)
The circuit by Mad_K
is actually more effective than if using output at midpoint.
He uses output at 57.77% ( 13 Volt of 22.5 Volt )
I often use output at something like 60%.
Why often used 50%, is to give the transistor some better space to work in.
This will lower efficiency, but possibly reduce distortion.
janneman said:
For a push-pull or CCS class A, midpoint is most effective.
But
when using a single end with power resistor
it is different.
Say 8 Ohm resistor and 8 Ohm load.
At negative swing, the load will be in series with power resistor.
A theorectical ( not counting drop across transistor )
most effective class A single end
will be if 2/3 (66.66 % ) of voltage supply is across power resistor.
This is if LOAD=Resistor (for example 8 Ohm and 8 Ohm)
The circuit by Mad_K
is actually more effective than if using output at midpoint.
He uses output at 57.77% ( 13 Volt of 22.5 Volt )
I often use output at something like 60%.
Why often used 50%, is to give the transistor some better space to work in.
This will lower efficiency, but possibly reduce distortion.
(sorry for my english)
Hi Mad_K,
I have built the " power follower 99" ( Andrea Ciuffoli design) and
have one observation to do.
After a long experimentations ( two years) of listening tests I have found this amp sounds wonderful on the mid and high but sounds also
without any type of punch / slamm on the bass. Big delusion. 🙁
This is not a problem of power, I am sure . I listen at low levels with "easy" speakers ( in mid-field).
I have compared it with a aleph30 and the result is very similar to my ears. Monotriodes also.
Compared with a X-250 the difference is total, another world! One of the best bass I know.
Again I repeat I listen a very low levels.
Is it a sort of "genetic trade/ carachter" of the SE (only output stage
I suppose)?
All the comments are appreciate 🙂
Cheers,
Inertial
Hi Mad_K,
I have built the " power follower 99" ( Andrea Ciuffoli design) and
have one observation to do.
After a long experimentations ( two years) of listening tests I have found this amp sounds wonderful on the mid and high but sounds also
without any type of punch / slamm on the bass. Big delusion. 🙁
This is not a problem of power, I am sure . I listen at low levels with "easy" speakers ( in mid-field).
I have compared it with a aleph30 and the result is very similar to my ears. Monotriodes also.
Compared with a X-250 the difference is total, another world! One of the best bass I know.
Again I repeat I listen a very low levels.
Is it a sort of "genetic trade/ carachter" of the SE (only output stage
I suppose)?
All the comments are appreciate 🙂
Cheers,
Inertial
Inertial, yes your observations are correct. SE excels in mid and treble region, while PP usually has it' strength in the lower regions. The bass caracter of a Common Drain SE amp is way better than a Common Source, so with this design I feel you get the best of both worlds (within the power range). The "power follower 99" is a nice design, and my SEWA design has evolved in that direction.
A good example of this taken to extremes are the Passlabs Rushmore speakers which (from my memory) has A Zen on top, Aleph in the middle and X amp on the bottom.
A good example of this taken to extremes are the Passlabs Rushmore speakers which (from my memory) has A Zen on top, Aleph in the middle and X amp on the bottom.
Oh thank you Mads,
THEN I am not crazy!!! 🙂
Now I understand zero of electronics but what is the physics explanation of this "behaviour"?
Inertial
THEN I am not crazy!!! 🙂
Now I understand zero of electronics but what is the physics explanation of this "behaviour"?
Inertial
I'm shure others can explain it better, but some clues are:
SE: 2. harmonic distortion dominates. often low damping factor designs. limited output current.
PP: 3.harmonic distortion dominates, higher damping factor designs, usually lots of output current avaliable.
SE: 2. harmonic distortion dominates. often low damping factor designs. limited output current.
PP: 3.harmonic distortion dominates, higher damping factor designs, usually lots of output current avaliable.
Hi Mad-K,
Mmmmmh,
SE: power follower 99 have quiete good damping factor ( I believe).
Limited current: If I need ,to say ,2 watt peak, I do not see
current limitations with PF99 (20-30W- 8 ohm) and a good
efficence speaker.
This is the "strange" I do not understand........
Inertial
Mmmmmh,
SE: power follower 99 have quiete good damping factor ( I believe).
Limited current: If I need ,to say ,2 watt peak, I do not see
current limitations with PF99 (20-30W- 8 ohm) and a good
efficence speaker.
This is the "strange" I do not understand........
Inertial
I haven't buildt the PF99, but I have a design which is very similiar, except for the PSU. Basically I have a version of the PF99 with the psu depicted earlier in this thread (With 600VA trafo, Ultrafast rectifier and Evox-Rifa electrolytics) I think the bass is rather good and "slammy". Not as good as a good PP slam, but still. Try another PSU with it and you might be enlightened 😉 I can't remember any significant difference in bass slam going from this (OTA) design to a PF99 design. The differences where more towards greater clarity, less tubey sound (lower distortion). Anyways both designs are great and I believe it is very subjective/system dependent which one you'll prefer. 😉
Thanks for the advices , Mads
Really I have tried various PSU:
Regulated and unregulated
C only and CLC
Massive transformer and capacitors.
NADA, Zero differences ( on the bass). The "carachter" is always the same.
I only constate the evidence. Bass is "anhoressyc" !!
So I have biamplified and now all is OK. An old Threshold 4000
20-400Hz and the PF99 400-20000Hz. Outstanding result!!
Now I am full satisfied.
But the question remain......
Cheers,
Inertial
Really I have tried various PSU:
Regulated and unregulated
C only and CLC
Massive transformer and capacitors.
NADA, Zero differences ( on the bass). The "carachter" is always the same.
I only constate the evidence. Bass is "anhoressyc" !!
So I have biamplified and now all is OK. An old Threshold 4000
20-400Hz and the PF99 400-20000Hz. Outstanding result!!
Now I am full satisfied.
But the question remain......
Cheers,
Inertial
Re: Re: Re: OTA - One Transistor Amplifier
Ahh yes I see. So why noty using a current source instead of the resistor? No, that would need another transistor! Ok, then why not an inductor? Low DC loss, but high AC impedance, so you get less dissipation, more power etc. Still one transistor source follower...
Jan Didden
PS I hate this waste for no particular reason 😉 . It's just bad practise....
lineup said:
Ahh yes I see. So why noty using a current source instead of the resistor? No, that would need another transistor! Ok, then why not an inductor? Low DC loss, but high AC impedance, so you get less dissipation, more power etc. Still one transistor source follower...
Jan Didden
PS I hate this waste for no particular reason 😉 . It's just bad practise....
Obviously, they are Wise.
So, it’s not called funny, dear Upupa, its called wisdom. Wisdom gives (always did and always will) simple and powerful answers to ‘complicated’ questions.
Btw, in my mind you are the best sponsor of all these nice little adaptations of se topology (the more you surf through various threads to remind people that they will get no good sound from simple circuits, the more people you gravitate towards them), and I really admire your work and dedication to these matters;-))
Keep up the good work, Mad_K 🙂
So, it’s not called funny, dear Upupa, its called wisdom. Wisdom gives (always did and always will) simple and powerful answers to ‘complicated’ questions.
Btw, in my mind you are the best sponsor of all these nice little adaptations of se topology (the more you surf through various threads to remind people that they will get no good sound from simple circuits, the more people you gravitate towards them), and I really admire your work and dedication to these matters;-))
Keep up the good work, Mad_K 🙂
- Status
- Not open for further replies.