The AM Forum

I put this together this morning from parts that I had in my junk box. The thought crossed my mind if I could build a PW RF deck from just stuff that I had laying around, and this is what I've come up with.

A 5763 tube oscillator driving a 7984 for 16 watts output. The loading cap is fixed at 1500pf and plate voltage is 275vdc. If you work backwards, the plate loading impedance is 1675 ohms.

The plate choke and the grid tank inductor are almost the same except for the fact that I ran out of wire winding the plate choke, but it should work ok. I'm not going to neutralize the 7984. At this power level and low frequency, it'll be fine. No metering, but I'll but an LED in the grid circuit to tune for grid current.

The only parts I need to buy are incidental small parts, like hardware and a few caps/resistors. I might put a PNP transistor in there for a transmit switch.

Aside from being a novelty piece for the shelf, a person could hi-fi modulate this and drive a linear.


Jon

That's a fun project Jon. Like it. I made something similar, photos attached. It has an EL84 modulator and an EL84 final. AM modulation by a balanced output transformer for EL95, one side to the PA, other side to the modulator anode and center tap to the +HV. This to prevent iron saturation, so I could use a much smaller iron.
It has also a VFO
Unfortunately the customs in Puerto Limon, Costa Rica did open my container and stole it together with many, many other materials, thousands of dollars lost and things I never can recover. All my VHF stuf stolen, 2 metal storage cases with 120 small drawers and thousands of integrated circuits, my chain saw, a 1790 flintstone rifle inherited from my father, a box with watches of my wife, my antennas, the transmitter and many things more

Very nice indeed! It's fun to make stuff out of copper. It looks all nice and shiny, at least for a little while.

I did some work on mine today. Here's a few pictures.

For a while, I was asking everybody at work to give me the toilet paper rolls rather than throw them out. They're the nice thick cardboard. I finally put them to use! Just in time too, because after I had collected a bunch the manufacturer switched to a cardboard-less center. My guess is to be more eco-friendly.

Jon

Looks nice!! But don't you expect coupling and oscillation if the grid coil and the tank coil are that close? Or do you screen one of those?
Please keep us informed about the progress

I put the copper shield in between the plate tank and the grid. The coil in behind the grid coil is the plate choke. I’ll see what happens.

Jon

The plate choke will carry some RF current. It may give some problems. In order to prevent RF current through a choke, I normally test by resonating an LC combination and touch the hot part with one side of the choke and holding the other side. The RF resonance circuit should not de-tune for an ifeal choke at that frequency. You can use the LC circuit of an existing transmitter to test, e.g. the drive circuit. The grid current should not change when touching the grid with the choke under test

Jon...The parts from your junquebox are so nice...no gobs of solder on them...no snark or dust...The TX looks very cool..good luck and keep us informed..
Steve

A little trick I've used when unable to find a cardboard tube is to take brown kraft paper or similar, coat it with varnish and roll it to desired size.  After a day or so it dries to a very hard surface tube suitable for winding on.

Thanks!

The incidentals are turning out to be more than I expected. My Digikey order is now $50!

I'm going to use an LED to indicate plate current (actually, cathode current) as well.

Also, the 6-32 nuts were too large for the 5763 tube socket so I had to grind down the edges.

Jon

I decided not to put a PNP transistor in for keying. I've put a switch on the front to key the rig with a separate set of contacts going to an RCA jack on the back to key up auxiliaries.

I put adhesive heat shrink one the coils to hide the electrical tape and made the output coil more stout with beefier wire.

I've been pondering modulators and everything costs money. But I do have an Elmac AF67 sitting unused. Elmac put pins on the back connected to the plate modulation transformer secondary (someone throw some flowers on that engineer's grave). I'll run the same 275vdc to the 6L6's plate and screens so I won't need an additional power supply. The data sheet says to run 16 volts bias. Even with the low voltage, it still should be plenty for 16 watts carrier.

Jon

This project keeps evolving. I did wind up putting a shield in behind the grid coil (thank you for the suggestion). It also adds a little aesthetically. 

I put the transmitter on a piece of wood for a nostalgic look. I countersunk fishing weights in the wood to give it some stability when adjusting the controls.

The box next to it will be the modulator. All that will be inside is an audio output transformer (8 ohm to a few thousand).  I'll reverse feed it with an audio amp, and I'll incorporate a negative peak limiter too.  The meter will read plate current rather than using the LED idea.

Jon

J,


Putting the device on wood adds to its gravitas.

Yer meter box looks really nice too.


klc

I finished the power supply today (toolbox picture).

The audio amp came in today, and this is how the rig's lineup will look. I'm waiting on my Digikey order to finish everything up.

Jon

Here's the inside of the modulator. I haven't finished it yet, but it'll also have a negative peak limiter powered from a 13.8vdc power supply.

Jon

Completed modulator

Jon

Jon

I’m wondering if this particular transformer can accommodate the DC plate current (without saturating the core) flowing through the output (high impedance) winding.

The information that I found on line shows it as designed for “push pull” amplifier applications... which implies that there would be essentially no unbalanced DC in the high impedance winding.

Stu

Right, push-pull... I should’ve bought single ended. I got this one in particular because it fit nicely in the box for the power handling capability. Could I use the center tap to one side of the winding?

I guess alternatively I could use a Heising circuit on the output of the high impedance winding. The current would just flow in one direction through the Heising choke. I don’t want to do this though because I want everything to be compact and fit in the box. For once, I want to build something that isn’t the size of a Sherman tank!

Jon

Just a simple 6L6, KT88 or so single ended instead of the audio amp, one side of the balanced winding to the 6L6/KT88 plate, other side of the balanced winding to the transmitter, centertap to +HV. That cancels the DC magnetization for the most and you have sufficient Iron. A 12AX7 is sufficient to drive and gives you microphone sensitivity. All that goes in the space of a mod choke and less weight

I could do that, and might, but I was still thinking about using my existing setup.

Feeding the B+ into the center tap and the transmitter from one side. The other side would need a resistor to balance the load. My plate impedance is calculated at 1650 ohms. I’d need to put that in with a sufficient size to dissipate the heat. I’m clearly not worried about efficiency here! The whole setup would be about 10% efficient.

Jon

** On second thought, I’m pulling 85mA through the final. I would need a resistor of about 3300 ohms to balance out the other side... at a minimum of 25 watts to handle the heat.

Jon

You need a Heising choke that can accommodate around 100mA of DC, and having an inductance value of 20H.

With 275V on the plate and 75mA of DC plate current, you would have the following:

DC input power at carrier: 21W
RF output power at carrier (assuming 80% efficiency): 16.5W
Modulation resistance:  3667ohms

Set 2 x pi x H x 30Hz = 3667 ohms => H= 19.5 henries

Stu

Another great resource for boxes like your using for the modulator are cigar boxes from an upscale cigar store. If a box gets beat up over time, an exact copy of a well known cigar box will be available. They also make nice jewelry boxes for kids and great places to put spare parts in. A cigar box is man's second best friend! Please make a note of it ;D ;D ;D

Most cigar stores will give the boxes to you for free but I have seen them for sale. Even if they are for sale, a good story will usually net you one or so for free :D :D :D

With a balancing resistor you indeed generate a lot of heat, and require double modulation power and DC power. A power tube and a 12AX7 seems simple to me and you don't have to look for a Heising choke and cap, which is difficult for a lot of people that like to copy your project. I even used a 2 x 115 V small power transformer as modulation iron with good results. At 350VDC they run down to 150Hz . which is not bad at all. At the high side 5 kHz and beyond was no problem. That makes it even more a junkbox project

Yeah, there’s lots of options here. It’s turning into a big ordeal just to make my originally intended 16 watts carrier.

The other question is, am I actually going to use this on the air? I have my big class E rig for 75 meters. I might just do the resistor trick to check the audio. Make sure it works, and then put it on the shelf.

I do like the looks of the whole setup, including the audio amp. If I actually use this as a working rig, I’ll probably use Stu’s idea.

Jon

Jon

Check out the Hammond 193C. I used one to build an external Heising modulator for my Johnson Ranger.

Stu

Yeah. I need to order two more ceramic standoffs anyway. Thanks!

Jon

I finished wiring the oscillator and the grid tank. I'm waiting on DigiKey order #2 to finish the final and the modulator. I'm going to mount the Heising choke on the back of the modulator box so it won't be seen.  Lots of iron for 16-ish watts!

Jon

Jon

I recommend a Heising capacitor value that produces a reactance that is equal to the reactance of the Heising choke ... at the frequency where the reactance of the Heising choke is equal to the modulation resistance (3667 ohms).  I.e.

A) Modulation resistance = 3667 ohms

B) 2 x pi x 30Hz x H = 3667 ohms

H = 20 henries

C) 2 x pi x 30Hz x C = 1/(3667 ohms)

C = 1.5 uF

Stu

Good formulas to note. I’ll copy and stash them.

I took a stab in the dark when I placed my DigiKey order and went with 3uF. It’ll work. I can tweak the bass and treble controls if I need to compensate.

It’s QRP AM anyway, so nobody will hear it.  :)

Jon

Jon

3uF will do.

Remember:

The Heising capacitor wil have a DC voltage across it of 275V.

If the modulation has a lot of low frequency content... e.g. a 30Hz sine wave... and if the modulation index is high ... e.g. 100% negative peaks... then the peak voltage across the capacitor can be as high as 2 x 275V.

I usually use two or more capacitors in series (with a 100k ohm voltage equalizing resistor across each)... to divide the voltage among them.

For example, a pair of 3uF 450WVDC capacitors in series (with voltage equalizing resistors across each of them) will have a value of 1.5uF, and a 900WVDC rating.

You definitely don’t want the Heising capacitor to fail in a way that it becomes a short circuit. If it did, it would possibly damage the modulation transformer and/or the Heising choke. A 0.15A fuse in series with the Heising capacitor would be good insurance.

Stu

The 3uF capacitor has a rating of 400vdc, so I'll get another one. Then I'll be at 1.5uF anyway.

All the iron is now mounted and the weight/watts ratio is pretty high! 


Jon

The RF deck is done.  I'll start testing once I finish the modulator.  I'm just waiting on that 2nd 3uF cap to arrive.

In the meantime, here's a link to pass the time:

https://www.youtube.com/watch?v=CdXesX6mYUE&start_radio=1&list=RDCdXesX6mYUE (https://www.youtube.com/watch?v=CdXesX6mYUE&start_radio=1&list=RDCdXesX6mYUE)

Jon

I have it completed, but it's going to take a little more fine tuning.

I'm getting 5 watts output right now.  The grid tank cap is fully unmeshed and the LED isn't glowing as strongly as I think it should. I need to pull a few turns off the coil to try to bring up grid current.  This won't be easy because I used heat shrink with glue.

Also, the plate tuning cap is fully unmeshed.  I did disconnect the 47pf doorknob cap in parallel, but I'm still fully unmeshed.  There is another fixed cap in parallel, so I'll need to install a smaller value one.

Jon

I kept cutting the inductor coil down for the grid tank until the capacitor came off being fully unmeshed for the highest glow on the LED.  I ended up going too far and had to add a red toroid on top to increase inductance a little bit.  It turned out to be a good thing since the big red toroid adds cosmetically to the thing.

I'm now getting 10 watts carrier.  I just need to get some wire to connect the 8 ohm output of the PA amp to the modulator to see how it sounds.  

Jon

Jon

Looks great!

Stu

Thanks, Stu!

I've been puttering with the modulator and the combination that works is connecting the high impedance secondary's center tap to the heising capacitor, and one end to ground.  The other end is left dangling. By doing this I get a low percentage of modulation and the audio is distorted.  I tried connecting the other end of the secondary to ground via a 2k resistor, but it wasn't happy with that.  I was thinking that I needed to balance the secondary.

...if I had only bought a single ended transformer. I should've bought a lot of other BS over the years, and life would be so much simpler now.  Oh well, no rest for the weary.  Enough complaining...

Again, this is a push-pull secondary transformer feeding a heising circuit.  Any thoughts?

Jon

Ok, my bad. I didn't have it tuned properly. Once I readjusted both caps, the audio cleared right up. Monitoring the rectified RF output through my headphones, the audio sounds very good but weak. The amp I chose doesn't seem to have enough drive, but that's ok!  I'm glad that everything works.

So, another nice shelf radio to add to the collection!   :)

Jon

Make at least a few QSO' with it Jon!!  Fun

I’ll go to Las Vegas, drop it at the entrance to Carsar’s Palace and run like hell!

Jon

Jon

Use the entire secondary of the mod transformer. One end should be connected to ground, and the other end to the negative side of the Heising capacitor. Make sure the pair of capacitors is connected properly in series: at the junction beteen them... + to - . Also make sure that you have a 100kohm 1/2 watt (or larger) voltage equalizing resistor in parallel with each capacitor.

The center tap of the secondary should be floating (end insulated to prevent an accidental short to something).

Stu

I didn’t want to use equalizing resistors across the caps because I didn’t want to place B+ (even at low current) across the secondary. It seems to do ok without it.

I’ll go back to using the full secondary now that I discovered my tuning error.

Jon

Jon

If you measure the DC voltage (no modulation) across each capacitor, you may be surprised at how uneven the split is if you rely on the leakage resistances of the two capacitors to be close to equal.

With a pair of 100kohm balancing resistors, the dc through the transformer secondary winding will only be 275V / 200,000 ohms = 1.375 mA.

Stu

Ok, I’ll make that change.

On the other hand though, I was just testing and the plate choke for the oscillator went “puff!” I only applied power, so a short must’ve developed on the load side of it somewhere. It’s a real bother to get to, so I’m going to shelve this project for a while.

It does look cool though sitting on the shelf. 🙂

Jon

Jon

This could be a biasing problem. If you are depending on the oscillator tube to be self biasing as a result of sufficient average grid current flowing when the oscillator is oscillating ... then, if it isn’t oscillating, and if you don’t have any other bias source (e.g. a cathode resistor.... the plate current can be excessive.

Stu

The tube is cutoff during receive. I have the ground lead of the cathode RF choke lifted. I ground it during transmit. It worked fine for days. There’s a lot of connections very tight under the tube socket. I’m sure something got shorted.

As an aside, I’m looking to get my big 75 meter E rig on the air again. I can now, but I’m going through an MFJ remote coax switch, and I’m guessing that it won’t handle boatloads of power. I’m finishing a homebrew version that will handle unknown megawatts of power.

It is almost winter after all. Time to get on 75.

Jon