I didn't know that builders were supposed to select Q100/Q101 and Q200/Q201 to have the same Idss. So the gates of Q100 and Q200 are driven positive by the signal?
ok, it pays to read the manual. Mr. Pass says it's best to select them to match...
Last edited:
I need to figure out how this circuit works in the first place. I can see that by providing a fixed positive bias the need for a negative power supply is not needed in order to allow equal positive and negative voltage swings on the input gate. Coming from the tube world I'm confused by these solid state devices.
It looks like this B1 buffer is most like the circuit (d) with Rs set to zero on both transistors.
Attachments
It is (d) without the source resistors and no negative rail. That is why you need matched Jfets. Pretend they were two resistors of the same value. You would have an equal Vdrop across each one as Andrew stated. This is also why, if using a symmetrical supply like the DCB1, you can get away with no output cap. If well matched, the offset should be zero because each fet will be dropping an equal amount of voltage. If it were me and i was interested in the B1, I would try the DCb1.
Last edited:
hmm, maybe more like the variations that followed the B1
I believe they were an attempt to use symmetric +/- supply
B1 have +/0 (ground)
Well, the gate resistor is tied to +9V in the B1, which is like having a -9V supply and the gate resistor tied to ground as in (d). I'm still surprised that jfets are ok being driven both positive and negative relative to the Vgs bias point. I'm not as adept at solid state circuits... in the tube world bad things happen when the grid is driven positive, so it's a habit of thinking on my part, sorry.
Very good point. This is something i learned recently as well. Here is the thread discussing it.
http://www.diyaudio.com/forums/parts/204062-jfet-basics.html
Thanks. Interesting thread. It's still not clear to me if distortion and noise would be reduced even further if the bias was lowered in the B1. R3 could be made adjustable to set the bias anywhere that I wanted it.
Well, besides the fact that the source resistors are zero and the device is operated at Idss, I guess not. One of the references I read says what I thought about operating these things like tubes: always use negative bias otherwise the curves and equations don't apply. But this particular jfet seems to be ok with a few hundred mV of positive Vgs.
Last edited:
No-one is going to doubt that either choosing capacitors or eliminating them altogether will affect the sound of the B1 / DCB1.
My comment was regarding PSU noise.
The B1 behaves beautifully with just a reasonable CRC supply.
There is no audible background noise - from my set-up.
Arcam CD36 / B1 / Aleph 4 / B&W 683s
I tried different ps with the B1 (CRC, regulated, shunted), and I found several notorious differences in sound between them, especially at low frequencies, but not in the noise department.
The best sound I got from a ps with the regular B1 is the Sigma11 (Yes, it's a tank and overdesigned for this preamp, but...) The σ11 Regulated Power Supply