> From: Rich Measures <measures@vc.net>
To: <amps@contesting.com>
> Date: Sun, 3 Aug 97 00:55:21 +0000
> >Strange.
>
> You thought period=1/frequency and frequency=1/period was strange
Only your use of it in defining "coherent", and calling square waves
"incoherent".
> >I just looked at a bone stock tank two tube Alpha conversion and it
> >worked just fine.
> In serious contesting service? .......With the stock tune-C and 2-holer
> conversion, the tank Q is under 6 on 160m. The 160m output is reportedly
> substantively less than 4kW.
I measured it here and it was about 4 kW. It tuned properly and had
-47dB second harmonic and -59 third at full power.
Are there various production runs, or are all the stock tanks the
same?
>Apparently, you believe that a tank circuit
> which is engineered for 2000 ohms in, 50 ohms out, will also work
> properly for 1000 ohms in, 50 ohms out. I don't.
You have the right to believe what you like. I was not there when
they designed the 77 series, so I have no idea what the goals were..
> - In your opinion, Mr. Rauch, do you feel that a Centralab 5kv 1000pF
> Series 58 is adequate for the 77's load C 160m padder?
It depends on tank Q, the impedance at the point where the cap is
located, and how much C is padding it. 1000 pf 58 series caps are not
generally available in low thermal drift, so they would not be my
personal choice for that reason alone.
> >Why do you think the tanks Q has to be higher than it is, if
> >harmonics are OK?
> >
> When tank Q falls below a critical amount, the tank is incapable of
> performing the needed Z transformation. As I recall, you pointed this
> out to someone in a previous post on rec......homebrew NG.
That right, I did. Finding the sqrt of Zin /Zout and adding one to
that value would be conservative. Let's look at the tank, before we
rush to a blanket condemnation.
We know a few things. First, once Q is high enough to match the
impedances, additional Q only increases losses.
Since this is a Pi-L, we need to know the center impedance of the
network. Let's assume 200 ohms and your value of 1000 ohms Rp.
1000 / 200 is 5, sqrt of that is 2.236 plus one is 3.236. Once
ETO has an operating Q just over three in the pi-section, the network
matches the impedance at the point of the L section connection just
fine.
Now for harmonics. Doubling the Q adds less than six dB to the
second harmonic suppression, and less to higher order harmonics.
The L section, as a general rule, adds about six dB suppression. If
you double the pi- section's Q it adds about the same amount of
attenuation at the second harmonic.
We can assume the total harmonic suppression is probably about the
same with a tank Q of six in the pi-L as a tank Q of twelve in a pi.
The ETO I measured here seemed to work that way. At 3900 watts it
had -47 dB second harmonic suppression, which is typical for class AB
amplifiers with pi-net tank Q's in the ten to fourteen range.
Are there some ETO's of this series that used different components?
Does anyone know?
73, Tom W8JI
(My opinions are my own. I am an independent consultant for several
manufacturers and government agencies. I have no vested interest or control
beyond supplying data and making suggestions. I do receive royalties tied to
the sales or use of certain paten
ted RF
medical devices.)
--
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
|