[HBR] Re: HBR Digest, Vol 54, Issue 5

[Ian Wilson]

I've spent many (happy??) hours over the past month or so
looking at a wideband amplifier that would be capable of the
15v peak that's needed for the 6JH8, etc.

My particular requirement uses high-side injection and a 10.7MHz
IF, so I need about 40MHz bandwidth. If I could manage this
3dB bandwith, I could use series peaking to have a reasonably
flat response out to 40MHz.

The best circuit I could come up with from a consideration of
gm, tube capacitances, and bandwidth, was a 6DJ8 cascode
feeding a 6AH6 cathode follower. I was aiming for a voltage
gain of 5. Looking only at tube capacitances, this setup should
provide adequate bandwidth. However, I could not reduce the
capacitance at the critical gain node (plate of upper triode of
6DJ8 / grid of 6AH6) enough. Each tube socket connection
has a capacitance of about 3pF (measured, tube out of socket).
This was the best I could do, using a low-profile socket (no
screening) and bending the pins away from their neighbors.
This circuit manages about 27MHz 3dB bandwidth. It is
single-ended, so you would need to duplicate it, etc, for
balanced drive.

I'm presently looking at solid-state ways of doing this. Old
TV/color monitor video drive ICs look good except I can't find
a source, and I seem to have thrown away one color monitor
too many :(

73, ian K3IMW