For my work in progress preamp, I've chosen for the Calvin buffer in combination with a Jung Didden superreg. The combo is supplied by a CRCRC quasimodo snubbered dual bobbin super fast soft recovery dual diode bridge, PSU.
The layout is done that way that the sense connections for the superreg are connected as close as possible to the semi power BJT's (2SA1209+2SC2911).
Bias current throught the 2SK389 is 6ma and throught the 2SA1209/2SC2911, 25mA.
With the trimmer it is easily possible to adjust DC offset to below 10µV.
Below the pictures for a one channel setup, temporarily connected to test.
There will be 4 of these PSU in the preamp. 2 for the Calvin buffers, another one for a Muses 72320 volume control and one other for the input selection PCB.
The layout is done that way that the sense connections for the superreg are connected as close as possible to the semi power BJT's (2SA1209+2SC2911).
Bias current throught the 2SK389 is 6ma and throught the 2SA1209/2SC2911, 25mA.
With the trimmer it is easily possible to adjust DC offset to below 10µV.
Below the pictures for a one channel setup, temporarily connected to test.
There will be 4 of these PSU in the preamp. 2 for the Calvin buffers, another one for a Muses 72320 volume control and one other for the input selection PCB.
Attachments
Last edited:
I was just taking a look at the push-pull (modulated ccs) buffer. With 100 ohm load and 10Vp-p output, the 2nd harmonic still dominates at -80dB in respect to the fundamental of 20KHz. It has "second harmonic character", although I consider the distortion quite low. The 5th harmonic is better than -120dB and the 7th is -160dB. As shown, the bias is 8mA in the jfet and 75mA in the pnp. What a beauty.
edit: 2SA1381 is not adequate, it won't sustain hfe (beta) at 100mA. It's ok for simulation purposes. For construction, Q1 and Q2 are TBD
edit: 2SA1381 is not adequate, it won't sustain hfe (beta) at 100mA. It's ok for simulation purposes. For construction, Q1 and Q2 are TBD
Last edited:
Hi,
why not decrease the idle current by increasing R_top and R_bottom?
50mA (45+5mA) would be fully sufficient for 10Vpp into 100R.
Reducing the Id of the SK170BL to say 5mA/45mA would also reduce the heat power loss considerably (which is alot higher than in the cascoded JFETs of the Calvin-Buffer).
If You want run the circuit that hot, You might use a different PNP (plus possibly different compensation caps).
MJE350, 2SA1358, 2SA1837, 2SA1859, 2SA1930(!), BD140, KSA1220, ZTX750, ZTX957 et al.
jauu
Calvin
why not decrease the idle current by increasing R_top and R_bottom?
50mA (45+5mA) would be fully sufficient for 10Vpp into 100R.
Reducing the Id of the SK170BL to say 5mA/45mA would also reduce the heat power loss considerably (which is alot higher than in the cascoded JFETs of the Calvin-Buffer).
If You want run the circuit that hot, You might use a different PNP (plus possibly different compensation caps).
MJE350, 2SA1358, 2SA1837, 2SA1859, 2SA1930(!), BD140, KSA1220, ZTX750, ZTX957 et al.
jauu
Calvin
The Black Forest Buffer was named after its original inventor, namely our most respected Calvin.
https://www.diyaudio.com/community/threads/the-diyaudio-first-watt-m2x.321925/post-6524985
https://www.diyaudio.com/community/threads/the-diyaudio-first-watt-m2x.321925/post-6603447
https://www.diyaudio.com/community/threads/b1-turbo-on-a-chip.140488/post-6847458
Patrick
https://www.diyaudio.com/community/threads/the-diyaudio-first-watt-m2x.321925/post-6524985
https://www.diyaudio.com/community/threads/the-diyaudio-first-watt-m2x.321925/post-6603447
https://www.diyaudio.com/community/threads/b1-turbo-on-a-chip.140488/post-6847458
Patrick
Last edited:
Hi guys,
How do you like the black forest buffer against the Feucht (or cascoded Feucht)?
Patrick, I recall you mentioning in the past some "nervousness" accompanying the sound of CFP topologies? Is it the case that you would prefer the Feucht under light loading, or does the BFB resolve the issues that typically come with CFP?
Cheers
How do you like the black forest buffer against the Feucht (or cascoded Feucht)?
Patrick, I recall you mentioning in the past some "nervousness" accompanying the sound of CFP topologies? Is it the case that you would prefer the Feucht under light loading, or does the BFB resolve the issues that typically come with CFP?
Cheers
