I'm glad that my information might help.
When I started this project there was also some concern expressed that
amplifier efficiency would be degraded because the Drake engineers had
"obviously" designed the pi network to be optimized with the plate
impedance realized at the lower anode voltage of the stock power supply.
My efficiency measurements have revealed that if this was their goal, those
Drake engineers didn't accomplish it very successfully. The efficiency
that I measure and calculate with the amplifier putting out 1500 watts with
3300 volts on the anodes is 3 to 4 percentage points higher than the
efficiency that resulted when the amp was tuned to deliver 1000 watts out
with the stock power supply. This is the average across all five bands.
I can't be sure of the absolute efficiency numbers, because my only watt
meter is a Bird 43, which can be off by 5% or 10%, but the before and after
measurement were made with the same slug, so the efficiency comparison
should be valid. I can be sure of plate voltage, plate current and grid
current, which were all carefully calibrated. The efficiency I measure is
always a little lower the average on the 80M and 10M bands, and higher on
40, 20 and 15.
John K7KEY
On Thu, May 12, 2016 at 10:54 AM, Michael Holderfield <
mholderfield56@gmail.com> wrote:
> I agree with your thoughts on this subject and thanks for some good info!
>
> I currently have a transformation in progress. Using an AL-82 transformer
> with a populated Far Circuits 4KV power supply PCB, a K4POZ bias board, &
> a Harbach soft key board. Also directly grounding the grids and
> incorporating a plate voltage relay and step start in the power supply. The
> relay will be controlled by the on/off switch which will reduce the current
> load on the switch.
> I am eliminating the 120v bias and ALC voltage as well.
>
> I had not planned to change the plate choke, but probably will now that I
> have read your post. I also have been wondering how to beef up the cooling.
> Your info on the blower/fan is most helpful. I may go in this direction.
>
> Overall I should end up with about $800-900 dollars in the whole project.
> So far it has been a fun project.
>
>
>
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|