[AMRadio] monitoring modulation help please

Donald Chester k4kyv at charter.net
Fri Nov 15 00:30:09 EST 2013

To monitor modulation, I prefer to use an oscilloscope in envelope pattern
mode. Not only can you tell if you are overmodulating or undermodulating,
you can easily determine if you are flat-topping on positive peaks, or if
there is severe  distortion on your  modulated waveform. 

I use a highly modified Heathkit HO-10  that I picked up for a few bucks at
a hamfest. First of all, I cleaned up the circuitry and got rid of the hum
and cleared up the fuzziness out  of the pattern, problems inherent to its
poor  design. Instead of running the transmitter output  through the two
coax jacks on the scope as Heathkit designed it, I pulled out the vertical
amplifier tube, and feed the  rf sample directly through one of the coax
connectors to the vertical deflection plates. Using trial-and-error, I made
a simple current transformer using a toroidal core, and feed the output of
the transmitter through the current transformer. With coax, simply let the
inner conductor pass through the toroid. I use OWL, so I use two current
transformers, so that  each conductor of the OWL passes through the  core of
one of the transformers. The output(s) of the current transformer(s) goes to
a small link-coupled tuned circuit located next to the scope.  For the OWL,
I wired the two secondaries in series and phased so that the voltages add.
The output of the current transformer (s) feed(s) rf into the link on the
tuned circuit. One end of the tuned circuit is grounded and the other end is
fed directly to the vertical deflection plate. A small variable trimmer
capacitor is used to resonate the tuned circuit, and for the coil, I use one
of those little B&W turrets (another hamfest find), a QRP version designed
to be used in the grid  circuit of a low powered final or the plate tank of
a QRP rig. It is probably rated at 25 watts at most, and is small enough not
to take up a lot of space. 

I switch bands on the scope tuner whenever I change bands, but it's not a
lot of trouble. For vertical gain, I merely adjust the tuning capacitor
close enough to resonance to get just enough rf to deflect two scale
divisions above base line on  the screen, so that 100% modulation is 4
divisions above base line. When I transmit, occasional peaks usually run up
to about 4 1/2 divisions. 

I picked up a General Radio modulation monitor at Dayton, but haven't tried
to do anything with it yet. I still prefer a  scope. To me, transmitting
without a scope is like driving on a country  road at night with the
headlights turned off. The flashing indicators and flicking modulation
percentage meters don't tell you anything about your modulation waveform.

The trick I used with  the 866 was to feed ground the plate and connect the
midtap of the filament winding to the line running from the modulator to the
final. Whenever the transmitter is overmodulated in the negative direction,
the plate of the final momentarily driven negative. When this line goes
negative with respect to ground, the 866A conducts and flashes blue inside.
This makes a good negative peak overmodulation indicator, but tells you
nothing about your positive modulation peaks or whether you are
flat-topping. I mounted the 866 in a small box with a viewing window, and
painted the interior of the box flat black. This makes even the faintest
blue flash clearly visible. The 866 plate can be biased positive to allow it
to flash just before you reach 100% negative. For example, if you  are
running 2000 volts DC on the final, biasing the 866 plate 100 volts positive
will cause the tube to flash whenever the negative modulation percentage
exceeds 95%.  A negative peak flasher is better than nothing at all, but IMO
still doesn't alleviate the need for an oscilloscope.

The trapezoid pattern is useful for determining modulation percentage in
both positive directions, but I find it most useful of all for determining
the modulation linearity of the final amplifier. The sides of the triangle
should be perfectly straight. If they are curved, that indicates
non-linearity. But like the overmodulation flasher, the trapezoid pattern
tells you very little about the waveform of your modulated signal.

Don k4kyv

This email is free from viruses and malware because avast! Antivirus protection is active.

More information about the AMRadio mailing list

This page last updated 24 Nov 2017.