raypsi wrote:
On Nov 28, 12:00 am, Tim Wescott wrote:
On Thu, 27 Nov 2008 13:45:33 -0500, ken scharf wrote:
Anybody here ever build a passive grid linear amp? That's where you
simply terminate the grid of a tetrode with a 50 ohm resistor and just
drive it directly. Svetlana recommended that configuration with their
4cx400, 4xc800 and 4cx1600 tubes, and it should work well with more
commone surplus 4cx250's.
I was thinking of using a 4:1 or 9:1 balum and increasing the grid
resistor value to 200 or 450 ohms to reduce the amount of driving power
required. With the 4cx250 a peak grid drive of about 50 volts is
required in class AB1. With a 50 ohm termination the driving power
would be 50 watts, with a 200 ohm termination 12.5 watts, and with the
450 ohm termination 5.6 watts. That's actually just the power sucked up
in the resistor, but the tube requires less than 1 watt of drive itself
in class AB1. Some power might be lost in the balum, so maybe the
actual driving power might increase by a watt or two. Since I wanted
this to be a final for a QRP rig the larger terminating resistance
looked like a better way to go. The only issue is how high can you go
with the termination resistance and keep the tube stable without
neutralization being required?
This should jump out at you with the proper circuit analysis. Just
analyze the circuit with a parallel tank load on the plate, and look for
negative resistance showing up at the grid. Your grid swamping resistor
would need to take care of the worst-case negative resistance.

Then spend a month tearing your hair out over parasitics that increase
the plate-grid coupling over and above what the data sheet says they are.

--
Tim Wescott
Control systems and communications consultinghttp://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes,http://www.wescottdesign.com/actfes/actfes.html

Hey OM:

I don't see plate grid coupling with a grounded screen and the cathode
at negative screen voltage? eh?

73 OM
n8zu
I think he means that any feedback which would induce instability can be
translated as a negative resistance seen at the grid (all an oscillator
is after all, is a negative resistance generator). And ANY tube
exhibits some degree of plate to grid coupling in the form of stray
capacitance between the elements. Even Pentodes with two grounded (for
rf) elements between the plate and grid show SOME coupling (though in
the case of screen grid tubes the value is reduced to values of a few pf
or less). The grounded grid circuit reduces the feedback even more by
introducing a phase shift as well, but some grounded grid amps still
need some neutralization (especially if you mess things up by putting 4
bottles in parallel.)

On Fri, 28 Nov 2008 10:06:43 -0500, ken scharf wrote:

raypsi wrote:
On Nov 28, 12:00 am, Tim Wescott wrote:
On Thu, 27 Nov 2008 13:45:33 -0500, ken scharf wrote:
Anybody here ever build a passive grid linear amp? That's where you
simply terminate the grid of a tetrode with a 50 ohm resistor and
just drive it directly. Svetlana recommended that configuration with
their 4cx400, 4xc800 and 4cx1600 tubes, and it should work well with
more commone surplus 4cx250's.
I was thinking of using a 4:1 or 9:1 balum and increasing the grid
resistor value to 200 or 450 ohms to reduce the amount of driving
power required. With the 4cx250 a peak grid drive of about 50 volts
is required in class AB1. With a 50 ohm termination the driving
power would be 50 watts, with a 200 ohm termination 12.5 watts, and
with the 450 ohm termination 5.6 watts. That's actually just the
power sucked up in the resistor, but the tube requires less than 1
watt of drive itself in class AB1. Some power might be lost in the
balum, so maybe the actual driving power might increase by a watt or
two. Since I wanted this to be a final for a QRP rig the larger
terminating resistance looked like a better way to go. The only
issue is how high can you go with the termination resistance and keep
the tube stable without neutralization being required?
This should jump out at you with the proper circuit analysis. Just
analyze the circuit with a parallel tank load on the plate, and look
for negative resistance showing up at the grid. Your grid swamping
resistor would need to take care of the worst-case negative
resistance.

Then spend a month tearing your hair out over parasitics that increase
the plate-grid coupling over and above what the data sheet says they
are.

--
Tim Wescott
Control systems and communications
consultinghttp://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes,http://www.wescottdesign.com/actfes/actfes.html

Hey OM:

I don't see plate grid coupling with a grounded screen and the cathode
at negative screen voltage? eh?

73 OM
n8zu
I think he means that any feedback which would induce instability can be
translated as a negative resistance seen at the grid (all an oscillator
is after all, is a negative resistance generator).

Yes, that's what I meant. And if you're looking to figure out the
largest parallel grid resistance that will still effectively swamp out
the instability, I think that's the best approach to take.

And ANY tube
exhibits some degree of plate to grid coupling in the form of stray
capacitance between the elements. Even Pentodes with two grounded (for
rf) elements between the plate and grid show SOME coupling (though in
the case of screen grid tubes the value is reduced to values of a few pf
or less). The grounded grid circuit reduces the feedback even more by
introducing a phase shift as well, but some grounded grid amps still
need some neutralization (especially if you mess things up by putting 4
bottles in parallel.)

In theory, any tube is going to have _some_ coupling straight through the
grids. In practice its probably vanishingly small.

But for many tubes you'll have to spend a long time optimizing the
physical layout of the circuit before you have more coupling through the
tube than you have around it.

--
Tim Wescott
Control systems and communications consulting
http://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html