Hello again Elmers!

OK .. I'm ready to fabricate a modulator designed
for plate modulation (so that I can compare to my
previous screen grid modulation scheme :-)

I've been lucky enough to procure a Triad M-8A
modulation transformer .. but .. I am having
difficulty locating technical data about this unit.

Can any offer suggestions as to where I might find
specification data/schematic?

I could not locate anything on the Triad home page :-(

Thanks again!

--Tom Nickel KC9KEP

Hello ..

This would work, but it ain't easy :-)

One could cut a circle from anodized aluminum or
maybe grav-o-ply (I think that's what it's called)
do the layout on a computer & have it engraved.

Here's some work that I did with a laser engraver.

The trick is to find a laser engraver :-)

http://www.bignick.net/Morgan_Radio/Radio_6.htm

--Big Nick AKA KC9KEP AKA Tom Nickel

Well, my attempt at studying for my ticket was
when I was about 11 or so .. But I didn't actually
take the exam for many years later.

Now, my wife reminds me that I act like I'm 11 :-(

--Tom Nickel AKA BIg Nick AKA KC9KEP

Hello Elmers,

First of all, I want to thank everyone again for all the help on my recent AM (double sideband) screen grid modulated transmitter project!

(My threads were posted over the last few weeks and had the phrase “screen grid modulation” in the subject line.)

Well, I’ve been recently reading my “Amplitude Modulation Transmitters – Lesson RRT-14, De Forest Training, Inc.” manual.

(I don’t know where this manual came from .. military training?  Tech School?)

Anyway, I was reading the chapters about the various forms of modulation, and here’s what the book had to say about grid modulation:

“Grid current has a tendency to flatten or distort the positive peaks of the a-c modulating voltage and also tends to make the r-f exciting voltage drop off at the modulation peaks.  These actions flatten the positive peaks of modulation and are the principal cause of distortion .. “  (This is actually referring to “grid-modulation”.)

I thought that this was interesting, because I was seeing a lot of “flattening” too ..
So, to some degree, it seems that this may be an expected anomaly of grid modulation schemes?

It is also mentioned about screen grid modulation:  “The r-f amplifier screen voltage must be reduced to between one-third to one-half the value used for amplification of un-modulated r-f , thereby causing a corresponding reduction of r-f output.  Furthermore, about 80% modulation is the accepted maximum .. [therefore] screen grid modulation is seldom employed.

In conclusion, I’ve received a great amount of great analysis from the astute Elmers on this page!  I had just wanted to share the text that I had happened upon, that is describing a behavior similar to what I have been seeing in my project!

I’ll be putting my transmitter on the air soon, but I’m already planning on building a plate-modulator next :-)

Thanks again,

Tom Nickel, KC9KEP

Elmers,

The equipment I'm using is:

a.) Homebrew AM transmitter, output section is a 6146B
    (and pi-network/filter)
b.) A heathkit Cantenna dummy load
c.) A homebrew matcher, basically 2 variable
    capacitors and variable tapped inductor capable
    of multiple L/C configurations
d.) A long-wire antenna & ground.

I am not using an additional amplifier (at least, not
at this time :-)

It seems to me that if one just connects the
transmitter to the antenna matcher without any
reference to where the output pi filter is tuned, or
where the antenna matcher is tuned, that there could
be an infinity of permutations that would yeild
max power/min SWR but I didn't think that this would
be a good way to procede.

My assumption was that once the transmitter is tuned
to a model 50 Ohm, non-inductive dummy load, that
the transmitter would be referenced to an ideal
target load, and then the antenna tuner would be
adjusted to match the antenna impedance to the now
optimized transmitter ..

Is that correct?

--KC9KEP

Hello Elmers!

Here’s a question that I’m certain that many will find simple!

With a dummy load connected, I can tune my tube transmitter’s
pi-network to optimize conditions for a given frequency.
(Fundamentally, locating the grid-dip via PA and loading
capacitors).

Next, I connect my transmitter to my real antenna, and then
optimize my antenna L & C matcher for minimum SWR while
keying the xmitter.  (Of course, I first check to insure that
the frequency is not being used.)

My question is:  Should the transmitter pi-network have already been adjusted
to optimum conditions by using the dummy load .. and ..the
pi-filter should not be re-adjusted when connected to the real
antenna because the pi-network is already set to optimum settings?

Thanks!

--Tom Nickel KC9KEP

Happy Holidays Elmers!

So, you’d think I’d stop working on my homebrew xmitter
On Christmas day?  No way!  This project has become a religion! :-)

Can someone help me out with “how-much” neutralization is enough?

I think I understand the concept; we’re trying to route a small portion
of the output signal back into the grid, out-of-phase, in order to counteract
the small amount of internal grid-plate capacitance that can cause a
self resonance .. which we don’t want.  (or, something similar to that)

In my design, (1965 ARRL handbook), this feedback is done by using
capacitive neutralization consisting of a “doubled length of No. 16 wire ..
to a position alongside the 6146B envelope.”

I guess that the wire couples to the plate adequately, and it saves
the builder the cost of a fancy neutralizing capacitor.

Now, the method of neutralizing the transmitter is to leave the xmitter
in “TUNE” mode (this grounds the screen grid, vastly reducing output
power for the sake of tuning).

Then, at 15 meters setting, adjusting the grid current to about 1.5 mA
by tuning the oscillator tank & drive control.

The PA (200 pF capacitor in the pi output tank) is then rotated while looking
for a grid dip, indicating resonance.

The goal is to adjust the neutralizing capacitance to achieve LESS THAN
0.1mA of grid-current meter deflection at the point of the dip.

I’ve tried numerous variations of the 16 gauge wire, a single wire, doubled,
and tripled.  I’ve tried using two-small plates of brass as a coupling device
as well.  I've tried close & far positioning.

It seems that no matter what I try, I can get, at absolute best, 0.1mA of
meter deflection, and no less.

Question:  Is that good enough?  What if it isn’t?  Will this spoil the transmitter?
What symptoms would the lack of neutralization cause?

Thanks so much guys .. Who else could I ask?

Here's a link to the 1965 ARRL article:

http://www.bignick.net/temp/6146_transmitter.pdf

--Tom Nickel KC9KEP

Yes,

Santa brought me an Icom IC-2200H for my vehicle.

It's cute as a bug's ear .. but .. it ain't got no
tubes.  Sigh ..

--Tom Nickel

Hello Elmers!

Please let me know when y'all have had enough of
my screen grid modulation questions :-)

I've tried a few things today, such as adding a
cathode capacitor to my 6DS5 modulator tube,
increasing control grid current, and flipping the
phase of the modulator transformer.

I've also made some power supply measurements.

I'm not sure that I'm gaining on the problem, but
I am quickly learning about screen & control grids and
easily one can goof up the biasing of these tube
tansmitters!  ;-)

My hope is that eventually, I will have taken enough
data to draw a profile that indicates what's causing
this circuit to behave as it does .. And, I'm really
struggling with the question of "Is that just the
way this circuit was designed to operate?"

Thanks for all the help, suggestions, and time that
everyone has taken to help me out!

Here's my latest data:

http://www.bignick.net/6146/elmers_2.htm

Happy Holidays and celebrate the Winter Solstice!

--Tom Nickel KC9KEP

Hello Elmers,

Thanks again for all the great information that you've
given over the past few days regarding my 6146
screen grid modulation AM DSB transmitter!

What I've tried to do is further examine my audio chain
in my modulation circuit.  I've taken several
photographs and measurements and posted them.

My big discovery is that, apparently, one cannot
simply optimize the output section for minimum plate
current and expect everything else to be in
harmony, modulation wise :-)  (at least with this
screen modulation system.)

It seems to me that I have to do a lot of balancing
between the grid current adjustment, and the PA and
loading capacitor in the pi output filter.

The pi filter really effects the way the modulation
is operating!

I know that several Elmers have been explaining this
to me, but it's just beginning to sink in.

And .. I do *not* know what I can typically expect
in this circuit because it's the fist time that I've
built one of these.

This time, I've placed ALL m photo's etc. into ONE
place, and here it is:

http://www.bignick.net/modulation/Elmers.htm

Please take a look if you would, and let me know if
what I'm seeing is "normal" and if it is just a
matter of me learning how to tune this transmitter.

Thanks a heap!

73

--Tom Nickel KC9KEP

I'll post another image of the unmodulated carrier
later this evening :-)

I can tell you that this transmitter is screen grid
modulated.

The voltage that is normally applied to the grid for
CW operation is reduced when the transmitter is
placed into AM mode.  (I assume to provide head room?)

When in AM mode, it seems that the carrier never
increses in amplitude for the positive peaks.

The modulation seems to be a subractive process ..
i.e., the carrier amplitude becomes reduced for the
valleys, but never increases for the peaks.

It does not behave as I had expected, but then
again, this is my first experience with an AM
transmitter.

Again, I'll post more photo's this evening.

Thanks!

--Tom Nickel KC9KEP

Hello again Elmers!

Yesterday, I was able to fix my transmitter arcing
problem with everyone's help .. thanks again!

There were two brass collars that had disintegrated
in my Cantenna, but I was able to make new ones.

Here's an old collar next to a new one:
http://www.bignick.net/modulation/Cantenna_1.jpg

And, here is a new collar installed .. much better!
http://www.bignick.net/modulation/cantenna_2.jpg

Today, I was looking at my amplitude modulation on
my scope.

I coupled my scope to the transmitter with a few turns
of wire, attached to my scope probe & ground.

I dangled the coil a few inches away from the pi tank.

I applied a 1 kHz test tone.

See if these images look correct to you?  They look
sort of flat-topped to me.  Is this typical for my circuit?

Here is about 20% modulation:
http://www.bignick.net/modulation/20_percent.jpg

About 33% modulation:
http://www.bignick.net/modulation/33_percent.jpg

About 100% modulation.
http://www.bignick.net/modulation/100_percent.jpg

Here is my schematic diagram again:
http://www.bignick.net/temp/6146_xmitter.jpg

Also, is there a better way to show jpg's on this
forum?

The way that I'm doing it doesn't allow them to show up.
(You have to copy & paste the URL into your WEB browser
address line to see them!)

Thanks!

--Tom Nickel KC9KEP

Hello again Elmers!

Say, thanks again for everyone's help!

This is one of the things I love about homebrewing.

Especially when something doesn't work, you start to learn
quickly!

I took another look at my Cantenna. (Recall, it
seemed to me that I did *not* get any arcing when
connected to my longwire antenna & matcher.)

I tried measuring the Cantenna load, and could not
read any impedance at all :-(

So, I opened the can (first time I've ever been inside
of one of these) and the dummy load element's hardware
was both loose *and* just plain broken!

Take a look;
http://www.bignick.net/temp/cantenna.jpg

(I think Bill suggested that I check my load).

I'll need to try a 100 watt bulb next.

But, I'm learning alot that I may still need to use
before this project is over!

Assumming that all my transmitter problems are over
(unlikely) I still need to add T/R switching, and maybe
a transmit monitor for CW ..

But, thanks again everyone!!

--KC9KEP Tom Nickel

Tom,

Thanks for the expedient reply :-)

A couple points of clarification:

My transmitter isn't using plate-modulation .. and,
I got the occasional arcing when I was keying the
transmitter (CW) and adusting the resonance capacitor
of the output section.

I have an identical 200 pF capacitor that I can
use to replace the existing one.  I picked these
up at a hamfest, so who knows why the previous owner
had discarded them :-)

Also, I don't know how to:  "check the tank by looking into it backwards with the anode end terminated in the design loadline of the tube with an SWR analyzer"

Is there a link that talks about this?

And tahnks for the link about amplifiers.  I will read
it next!

Thanks again!

--Tom Nickel KC9KEP

Hello Elmers!

I've completed building my very first AM/CW double sideband
transmitter!  :-)

The construction article was featured in an ARRL 1965
handbook (when the Beatles '65 album ruled!)

It features a 6146B output.  Here is a schematic diagram:

http://www.bignick.net/temp/6146_xmitter.jpg

Everything *seems* to work .. but .. I've been trying
to dial-in the output pi-filter section, using a heath
Cantenna load.  (Actually, I'm not positive that the load is
functioning correctly.)

When I try to dial in the output resonance capacitor,
(C3, 200 pF) it occasionally arcs over, with a sizzle!

I don't seem to have this issue with my homebrew antenna
matcher and longwire antenna.  Of course, I don't
want to dial in my transmitter "live" (on the air.)

Here's a photo of my capacitor.  I think it's beefy
enough?

http://www.bignick.net/temp/200_pf.jpg

I measure about 0.07" between plates.

I think that a load mismatch can create high voltages
in the output section?

Anyway, I'll probably have more questions on how to
operate this sort of transmitter, if that's OK!

The "core" (i.e., ignoring the modulator) appears to be
a fairly common configuration.

If anyone has used a similar circuit, I'd like to hear
form you!

Thanks in advance and happy holidays!

--Tom Nickel KC9KEP