I think this thing has gone far afield of helping anyone build a
clean PA. It seems Ian, Peter and I agree on the causes of IMD, and
some good ideas to prevent IMD (regulating the screen and grid with
low shunt impedance regulators). But the grid idea does make the
control grid more susceptible to overdrive damage.
On the other hand Rich has a suggestion that would greatly increase
IMD if the tube is ever (even slightly) overdriven, but it does offer
good control grid protection.
Either system would require a warning device for grid current. In the
stiff supply case, it would be mostly necessary to prevent tube
damage. In Rich's case, it would be mostly necessary to prevent
splatter.
So why not use a stiff shunt regulated grid supply (to prevent
splatter) with a warning device that disables the PA if excessive
grid current occurs? After all, we commonly do that with triodes
where the control grid is easily damaged (like the 8877 or
3CX800A7).
That way the bias is stable, occasional peaks won't hurt the grid or
cause needlessly bad splatter, and if you "over do it" your mess is
removed from the airwaves.
It seems the best approach to me, because no matter which system you
use you should have a grid voltage (or grid current) detector anyway.
With the stiff supply it could be a simple op-amp comparator that
trips the T/R relays off in a few mS if grid current becomes more
than a safe value. (With Rich's idea it could be a voltage detector
that disables the PA if the peak RF voltage approaches the bias
level, although I like the stiff grid supply much better).
By the way, Peter is correct. (I think it was Peter that said it)
Adding an additional opposite type of distortion can actually improve
IMD performance. Eimac intentionally adds non-linearity at some
areas of a tubes transfer characteristics to improve IMD performance.
That's why a three tone test is important to measure IMD, so you get
an idea of two tone performance as power level is varied at a slower
rate.
It's quite possible to build a PA that looks excellent in a two tone
test, but generates all kinds of trash with speech modulation. A
three tone test shows up those problems, if done properly. Poor
power supply regulation (including screen and control grid
regulation) will normally ONLY show up with a three tone test.
Hence the G2DAF system or unregulated screen supplies look quite
good, until you talk into them.
73, Tom W8JI
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|