[Amps] 8171

[David G4FTC]

Barrie,

I'm the builder of the 6 metre 8171 amp referenced in Steve G8GSQ's posting.

Although the output capacitance of 23pF may be considered high for a tube
operating at VHF, when I built my 8171 amp I didn't find this to be the main
problem - in fact I used a very conventional PI output but with flappers for
both tuning and loading simply as an experiment to see if it was possible to
build a "tubes with handles" amp without needing to resort to vacuum
variables. The flappers really do work fine.

I built my amplifier in a grounded cathode configuration and I didn't fully
appreciate the losses that I would find with a tube having a 120pF input
capacitance. Consequently the drive requirements were over and above what I
had expected that I would need. IMHO it is the input capacitance which will
cause you many more problems than the output capacitance.

The SK300A tube base isn't really ideal at VHF, and the tube manufacturers
recommend the SK360 for frequencies above 30MHz. I understand the SK360
allows better RF grounding for the filament with a lower inherent inductance
at VHF. However, these are expensive bases compared to the SK300A.

An alternative to the SK300A may be the SK300. The SK300A was developed from
the SK300 to provide improved cooling to the tube but this resulted in
cutting out the "skirt" from around the base. This increases the inductance
and makes providing an effective RF ground somewhat more difficult, and
results in less that optimum gain.  However even with the SK300A base which
I used, I'm still getting 16dB of gain at 50MHz which in reality isn't too
bad.

Currently I'm researching building a new amp for 2 metres, and at this
frequency I feel that the 120pF input capacity for a grid driven
configuration would be a real show stopper.  However, the input capacitance
of a 8170 in grounded grid drops to about 56pF - not much more that the 42pF
of the 8877, and I really feel that this is the way forward.  However to get
really low inductance grounds for the screen and grid will require a certain
amount of metalwork where I will remove the collets from a SK300A, discard
the frame, and mount the collects on individual plates. These plates will be
chassis mounted with PTFE insulation to provide RF decoupling and also the
DC feeds for the screen/grid biases. Cooling of the grid and filament seals
in this arrangement will not be simple - I'm still thinking about this -
particularly as the amp will be run at a QTH almost 10000 ft ASL!

For the plate circuit I'm tempted to try the W6PO style of design - this has
the advantage that the tube's output capacitance is effectively split 50-50
on each of the two quarter wave legs.

Also have a look at http://www.qsl.net/g4ftc/amplifiers/8170_2m/home.htm
Although I didn't build this, there are some very useful technical snippets
of one builders approach to running this range of tubes at VHF.

Hope this gives you some ideas.


David G4FTC