When you look at the circuit, you have to realize that this
is a mild form of negative feedback. It points to a
possible equivalence between low amounts of negative
feedback and the effect of plate-cathode voltage on the
transconductance figure of a triode.
When you work a load line so as to cancel these two
distortions, the result is lowered 2nd harmonic but the
appearance of 3rd and other harmonics, similar to
Baxandall's analysis of negative feedback.
is a mild form of negative feedback. It points to a
possible equivalence between low amounts of negative
feedback and the effect of plate-cathode voltage on the
transconductance figure of a triode.
When you work a load line so as to cancel these two
distortions, the result is lowered 2nd harmonic but the
appearance of 3rd and other harmonics, similar to
Baxandall's analysis of negative feedback.
I was browsing through Horowitz and Hill (a little light after-
work reading) and noticed the "Fet linearizing" circuit which
is used to straighten out the characteristics of fets used as
variable resistors. It is another example of this technique,
but it's used to straighten out the curves at low voltages.
work reading) and noticed the "Fet linearizing" circuit which
is used to straighten out the characteristics of fets used as
variable resistors. It is another example of this technique,
but it's used to straighten out the curves at low voltages.
Reread Linsley-Hoods LINIAC article from -71 the other day and saw some similarities to what have been delt with here.
If we combine the Schaded 2SK170 with a "voltagecontrolling" gyrator and a SF we are there.
One could of course take the NFB from the output but does a SF really need more NFB?
Using a P-channel output device there is no need for insulation against the heatsink as drain is to ground.
RDC is used to achieve symmetrical clipping.
Unfortunately the gyrator needs additonal voltage to work. If instead a choke was used the 24V would be enough. The gyrator FET doesn´t need to be depletion, anything with high Gm will do.
If we combine the Schaded 2SK170 with a "voltagecontrolling" gyrator and a SF we are there.
One could of course take the NFB from the output but does a SF really need more NFB?
Using a P-channel output device there is no need for insulation against the heatsink as drain is to ground.
RDC is used to achieve symmetrical clipping.
Unfortunately the gyrator needs additonal voltage to work. If instead a choke was used the 24V would be enough. The gyrator FET doesn´t need to be depletion, anything with high Gm will do.
Attachments
Last edited:
Are there a schematic available in the meantime?
Measuring results are to find here:
EAR/Yoshino M100A monoblock power amplifier Measurements | Stereophile.com
and the listening test results here:
EAR/Yoshino M100A monoblock power amplifier | Stereophile.com
Hello revintage. A bjt has pentode-like characteristic curves. Will your methodology of this post also apply to a bjt; meaning a bjt sounding like a triode?So what happens if we adapt what Schade taught us to a JFET as buffer?
Works like wonders
Zout below 200ohms, together with low THD and acceptable bandwith (-3dB at well over 200kHz). Zin chosen to 10kohm. With 22kohm/91k bandwith is 100kHz.
But don´t try to a volume-pot at the input! Must be fed from a lowZ source.
Brgds.
Just for fun, I traced some curves for the NTE451 JFET that show the effect of parallel feedback. It indeed works as advertized (on the curve tracer).
Link to video of curves using a potentiometer to vary the amount of feedback
Link to video of curves using a potentiometer to vary the amount of feedback
Attachments
Thank you avincenty for your interesting results. Did you convert a run of the mill bjt to a new and linear solid state triode? What is the relevance of "extremely linear" in application? I hope it is good if it sounds triodish?
Thanks avincenty for the interesting video. Creative, instructional and showed that one gets a tweakble solid state triode performance! Was feedback from drain to gate increasing as a function of time? Do you have another video like it for bjt to understand its behavior?. Best regards.
avincenty: Schade feedback concerns itself with a current source amp; also called a transconductance amp [TCA]. Please take a look at the F1 schematic in its service manual at www.firstwatt.com. It is a differential TCA. Mr. Pass used feedback from the drain to the gate of each of the two amplifying MOSFETs. Clearly Schade feedback [inadvertent or not]; which may have sweetened the performance of this amplifier in favor of a triodish sound. Best regards.
It is a voltage divider bias. Please note that it also has the signature of Schade feedback of at least the following 3 requirements:
- MOSFET Q1 [for example] is a current source amplifier.
- The Gate of Q1 is a summing junction of 2 signals which are out of phase.
- A feedback resistor is used between the Drain and the Gate of Q1. It generates the out of phase signal which sums with the in-phase input signal at the Gate.
I am a newbee at this fascinating application and learning it . I hope that I am not out of line in my observations. I appreciate your feedback and that of others like wrenchone and Zen Mod. I understand that SIT is a voltage and not a current source amplifier. Thus, it it unclear to me whether Schade Feedback is useful or needed in it to make it behave "more or less triodish" than it already is. Best regards.