[Amps] Q vs Load vs Tuning: Myths debunked

[Robert Groh]

Excellent email, Bill.  Well reasoned, technically correct and backed up by data - nice job. This one I keep - for the rest, well, as you say, there is always the 'Delete' key (although the Delete key has certainly been getting a work out lately!).

73
Bob Groh, WA2CKY




________________________________
From: "Bill, W6WRT" <dezrat1242 at yahoo.com>
To: Amps reflector <amps at contesting.com>
Sent: Sunday, August 30, 2009 11:07:16 PM
Subject: [Amps] Q vs Load vs Tuning: Myths debunked


For anyone following the thread on plate tank loading and Q vs drive, if your
brain has turned to mush and you are tired of the whole thing, feel free to
delete and move on. But if you are interested in these things I think you will
find what follows fascinating.

A few days ago Joe, W4TV, made the following statement: "  Running an amplifier
in a mis-tuned condition (tune for 1500 W Output and drive to 800 W Out) is
hazardous and inefficient.  If you tune the amplifier for 800 W at the higher
plate voltage the Q of the pi network soars resulting in all kinds of 
problems."

Well maybe, maybe not. Let's see.

I tuned up my homebrew 8877 amp on 40 meters and took some readings. This amp
delivers 1500 watts output according to my Palstar PM2000A when properly tuned
up. The plate voltage under load at that power is 2900 VDC and the plate current
is 880 mA. Efficiency is 59%, rounded off. Strictly routine, nothing out of the
ordinary.

I designed the pi-network for a plate load impedance of 2200 ohms using the
G4AKK Excel spreadsheet, which is derived from the formulas in the ARRL
Handbooks of recent years. I used a "K" factor of 1.5 as explained in the
handbook to come up with 2200 ohms for the plate load impedance. Others may use
a slightly different number, but that is not important to this discussion as
long as they use the same number throughout. There are other spreadsheets based
on the ARRL formulas and they all give the same results. They are available on
the Internet. Use whichever you like.

For this load impedance and a Q of 12, the coil is 5.15 uH. The circulating
current is approximately equal to the tube DC current multiplied by the Q, so
the circulating current is about 10.7 amps. With me so far?

I then reduced the drive so the output was exactly 800 watts. I DID NOT retune
either the tune or load caps. The only thing changed was the drive level. Now
with 800 watts out, the plate current was 605 mA and the efficiency fell to 46%,
about what I expected. 

Now comes the fun part.

I retuned the tune and load caps for max output, while adjusting the drive to
keep the output at exactly 800 watts. Now the plate current is only 477 mA for
the same 800 watts. The efficiency went back up to 58%, only 1% lower than the
full-power, 1500 watt efficiency. So what did change?

The main change is the Q of the tank circuit. The new plate load impedance is
4053 ohms, a far cry from the original 2200 ohms. Since the coil has the same
inductance, the Q has increased dramatically. A few minutes with the G4AKK
spreadsheet will show the new Q is now 21.5. No surprise there except that even
though the Q is that high the efficiency is still high too. Just 1% different..

But here's the real surprise: The circulating current is now just 10.3 amps. It
is actually LOWER than the 1500 watt current, despite the dramatic increase in
Q. The reason of course, is that the  lower power level offsets the higher Q.

So Joe, the ball is in your court. What exactly is "hazardous" about operating
this way? And what "problems" do you see?

We await your volley.

73, Bill W6WRT










_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps