[Amps] RE: 4CX250B Screen Supplies

[Will Matney]

Ian,
I've beat this dead horse to death and if anyone doesn't think that IMD 
can be increased by not enough filtering, the best thing to do is just 
try it. Change the filtering on the power supplies for the screen and 
control grids and note the differences as a before and after. I done 
this long ago building smaller amps (RF and Audio) using non regulated 
supplies. I was working around an original design by an engineer down in 
Memphis at the time. He fed the screen off the center tap of a HV 
doubler. The amp splattered like He** and always had that bad name 
branded to it. After looking the circuit over, I added a diode off the 
center tap and another 100 uF @ 450 Vdc filter cap, that cleaned up the 
bad IMD and the hum that the circuit was known for. I then went and 
upped the filtering on the bias supply from 22 uF to 80 uF and it came 
down more. I was actually checking the output the way Richard Measures 
said of using a receiver to tell the difference on a near by frequency. 
I know and seen it was noticeably different. So it had to be the 120 Hz 
ripple causing the IMD, or at least in my way of seeing it. The 
bypassing was already well done enough by the original design using 5000 
pF @ 1Kv ceramic discs at each tubes pin to ground. I did add some 
additional bypassing to the control grid source behind the blocking 
choke at the control grid pin. However, after adding this, I did not 
tell any difference from testing. Also, I used the two tone test and 
audible voice tests using a receiver set for the monitoring and there 
was an increased IMD using the voice compared to the two tone even 
though I cleaned up the filtering but the splatter was less after the 
filter modification. Voice will raise it but it would have been worse 
with 120 Hz hum too because it was! If 120 Hz ripple hum wasn't a 
problem, we wouldn't have to use filtering at all in the power supply. 
All be it that the regulator will drop it more with filtering, but if 
the supply isn't filtered enough, some is bound to get through. This 
being said, why is Rich attempting a resonant choke filter for the 
screen supply but if not to kill out 120 Hz ripple? Last, why does Henry 
go to the extent of using a resonant choke to filter the plate supply?

"I'd like to make a modest proposal here: how about not feeding a 
"dirty" supply to *either* of the grids?".......... Ian G3SEK

Ian, that's exactly the point I'm trying to get across. Get the 
filtering right and I'll bet you'll see a drop, I know I did. I'm not 
saying that a regulator will not clean it up to, it will, but it needs a 
clean supply, that's all I've been preaching here, nothing more. The 
only other thing I mentioned was that voice modulation creates more IMD 
than does the two tone test, and by what I've read, we are all in 
agreement it does.

Will Matney