The ARRL handbook amp used the detectors to fix the load line - you have to
tune for maximum plate RF Volts first and then get the ratio of grid to plate
volts correct to get the load line right. Autotune transmitters tended to go
for a phase detector, and tune the plate for a 180 degree phase difference in
voltage with the grid. As simple passive phase detectors give a centre zero
output when the phase difference is 90 degrees, the usual technique was to
sample the current in the grid tuning inductor to get effectively the grid
voltage and compare with the plate RF voltage for driving the servo motor.
These days, a simple 'hill climbing' circuit would do it or a fast phase
detector using ICs.
Tuning for max plate volts and adjusting the loading for the correct ratio was
certainly in use in a British Army tx designed in about 1957 or so.
The 30L1 system is to my mind, a little strange. The grid of the 811A s are
"grounded" through 220pf capacitors. The RF voltage on the grid of one of them
is sampled and rectified, as is the cathode voltage of that stage, and this is
used for adjusting the loading. If you have RF grid volts in a GG triode stage,
it's not very grounded, and my presumption is that the capacitive divider
formed by plate to grid capacity and the 220pF to ground is intended to give a
constant division regardless of frequency of plate RF volts for the detector.
However, the 30L1 has a reputation for being 'touchy' on the HF bands, and the
original manual required a 20.5 foot length of RG58 between exciter and
amplifier - said to be needed to tame it. W8JI does not like letting the grids
float in this manner, and I do agree. The idea that it gives NFB is dubious
when you take into account strays, especially the inductance.
73
Peter G3RZP
========================================
Message Received: Nov 03 2013, 04:05 PM
From: "Jim Barber" <audioguy@q.com>
To: "Colin Lamb" <k7fm@teleport.com>, "Karl-Arne Markström"
<sm0aom@telia.com>, g8on@fsmail.net, amps@contesting.com
Cc:
Subject: Re: [Amps] Measuring IMD
On 11/3/2013 6:53 AM, Colin Lamb wrote:
>
> A similar linearity-detecting scheme was used in a 3-1000Z design from
> the 1970 ARRL Handbook. "The SS-70 2-KW Amplifier", pp 204-208.
>
> It used diode RF detectors in the cathode and plate circuits, a
> "balance" pot and a 50-0-50 uA front-panel meter.
>
> I have no idea how well it worked, but this thread rang a bell and I had
> to go look it up. It would be interesting to try it. If the mechanical
> meter turned out to be inadequate, a connector for an external scope
> could be added.
>
> 73,
> Jim N7CXI
>
> Instead of using a meter, an led array should work fine, since the led has
> much faster response. If you rectified the input and output of an amp and
> fed to companion led arrays, you could do the same thing as the scope. In
> fact, they have solid state scopes that would work well and occupy almost no
> space.
>
> 73, Colin K7FM
>
The old National app note for the LM3914 bar graph driver shows a
20-segment zero-center meter that drives with a +,- 5V signal. Clamping
zeners on the detector outputs and an op-amp for buffering would make a
fun display.
The HAL-8877 ? ;-)
73,
Jim N7CXI
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|