Audio Virtues of JFET Transconductance Amplifiers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Yes, I also thought the Pioneer SL was very clever, and really good sounding.
That is from someone who builds almost exclusively with FETs. :)

As I was trying to stay with the original Pioneer circuit, I did not want to change it too much.
But I agree that the current mirror can be improved, at the expense of complexity.
Having said that you can see in our measurements how low the distortion, especially H2 is.

And with the BJT input I can trim DC offset at the CCS bias :
http://www.diyaudio.com/forums/head...pioneer-super-linear-circuit.html#post5205757
The main circuit remains completely symmetrical.
And DC is stable to mV's after trimming.

I also have on the drawing board a pure transconductance headphone amp based on the SL.
i.e. totally current drive, no global feedback, (and of course no voltage feedback).
The bug is of course the actual voltage gain depends on and varies with the load impedance.
So only of any use for planar magnetic headphones.
Shall build proto in a few months, just out of curiosity.


Cheers,
Patrick
 
Hi Patrick,

Surprising you mentioned a pure transconductance amplifer, I have tried this with excellent results but you are correct in the fact that you need a load with a flat impedance. With a current multiplying mirror it can be done.

This is what I used for the mirror, there are more complex versions but this worked just fine and you can parallel stages (within reason) to get the required current.


Regards,
Jam
 

Attachments

  • triple.png
    triple.png
    10.4 KB · Views: 187
Yes, I also thought the Pioneer SL was very clever, and really good sounding.
That is from someone who builds almost exclusively with FETs. :)

As I was trying to stay with the original Pioneer circuit, I did not want to change it too much.
But I agree that the current mirror can be improved, at the expense of complexity.
Having said that you can see in our measurements how low the distortion, especially H2 is.

And with the BJT input I can trim DC offset at the CCS bias :
http://www.diyaudio.com/forums/head...pioneer-super-linear-circuit.html#post5205757
The main circuit remains completely symmetrical.
And DC is stable to mV's after trimming.

I also have on the drawing board a pure transconductance headphone amp based on the SL.
i.e. totally current drive, no global feedback, (and of course no voltage feedback).
The bug is of course the actual voltage gain depends on and varies with the load impedance.
So only of any use for planar magnetic headphones.
Shall build proto in a few months, just out of curiosity.


Cheers,
Patrick

Most times an output buffer is used in transconductance ampllifiers.
Otherwise the output voltage is load dependent.
Several transconductance Chips have such buffer transistors onboard one can use.
 
A transconductance amp by definition is voltage in current out.
And output current is proportional to input voltage.
And negative loop feedback is not necessary, as in Jam's example.

The minute one hangs a buffer at the output, it becomes a CFA.
i.e. the output voltage is porportional to input voltage, irrespective of load impedance.


Cheers,
Patrick
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.