J. Gerhard likes it, and some in the past. There is a dinky trimmer in the signal path to be added too. Not so Zen as B1 but more robust technically.
Yes, i use varieties of the JLH buffer. One i made uses darlingtons in the output stage. It is based on feedback, so distortion is really low. Still the best sounding buffer i made so far is open loop. I call it the Hawks buffer. It needs a lot of N and P-channel Fets so it is a bit expensive to make. Because it is very hard to find 2SJ74 Fets right now i have moved on to bipolar buffers recently. A sound comparison is in the making.
JG_Buffer based on Hawksford-Design
This is the best buffer I ever heard.
We have used the JG-Linestage and built several buffers in front
After listening test this is the result
1.J G Hawksford-Buffer
2. Kaneda-Buffer
3. JLH-Buffer
4. LME49600
The sound of the JG Hawksford buffer is marvelous!!
You can build it with only one pair of J-fets. it depends on the load you want to drive.
I am very courious for the bipolar design
Kind regards
Sam
This is the best buffer I ever heard.
We have used the JG-Linestage and built several buffers in front
After listening test this is the result
1.J G Hawksford-Buffer
2. Kaneda-Buffer
3. JLH-Buffer
4. LME49600
The sound of the JG Hawksford buffer is marvelous!!
You can build it with only one pair of J-fets. it depends on the load you want to drive.
I am very courious for the bipolar design
Kind regards
Sam
Attachments
Here is the schematic. The Hawksford connection is when the input Fets drive the Leds.
Decouple well, this beast is fast. If you can not make it stable you can bootstrap the input with a 220pF silver mica from input to output.
Decouple well, this beast is fast. If you can not make it stable you can bootstrap the input with a 220pF silver mica from input to output.
Joachim Gerhard
The J G Hawksford-Buffer
Whether it is necessary to select on Idss transistors in quintuples, included in a parallel?
The J G Hawksford-Buffer
Whether it is necessary to select on Idss transistors in quintuples, included in a parallel?
I have been told some years ago that KP903A are used in industrial communication equipment at Vds near 40V (datasheet tells 20V only). I assembled various amps based on KP903A during last three years, also with Vds=40V operation conditions. Everything is stable.
Just today I tested randomly chosen KP903A as a current source, started to increase Vds with observing Id current. Increased Vds up to 50V, the part still keeps stable Id. Decided not to risk more and not destroy a part.
The Bear (member of this forum) tested these parts by curve tracer, and confirms that 40V is not a problem for KP903A.
I would also like to mention, that jFET quality criterion (transconductance / Ciss) is almost order of magnitude better for RF part KP903A rather than for most popular Toshiba's 2SK170 jFET.
Just today I tested randomly chosen KP903A as a current source, started to increase Vds with observing Id current. Increased Vds up to 50V, the part still keeps stable Id. Decided not to risk more and not destroy a part.
The Bear (member of this forum) tested these parts by curve tracer, and confirms that 40V is not a problem for KP903A.
I would also like to mention, that jFET quality criterion (transconductance / Ciss) is almost order of magnitude better for RF part KP903A rather than for most popular Toshiba's 2SK170 jFET.
Last edited:
- Status
- Not open for further replies.