15 Watts - 1 Tube transmitter

[ka1tdq]

Here's a picture of my next work in progress.  It's a single tube (7984) transmitter for 15 watts carrier.  It's plate modulated using Heising modulation.  Audio is fed in through an audio output transformer and into the Heising capacitor and 40 Henry choke. 

1 watt of RF drive can come from any source.

The 20 LED string will indicate RF output power as a bar graph.  I'm using a couple LM3916 IC's for the LED drivers. 

Jon
KA1TDQ

[W4AAB]

That is a good RF tube. GE used them in a lot of their low-band and high-band two-way radios.Never had to change many of them. Always interested in the schematic as you develop it.

[ssbothwell SWL]

very cool project. i look forward to seeing more photos and information. what frequency bands will it operate on?

[ka1tdq]

I picked up over 30 7984 tubes cheap on eBay so I'm going to put them to good use.  It'll be for 75 meters only.  I'll post a schematic when I draw up.  Right now everything is on notes.

Jon
KA1TDQ 

[ka1tdq]

Here's the schematic that I drew up.  I didn't draw out the LED display schematic but something similar is on the LM3914 data sheet. 

The grid tank uses 5 turn around the ground end of the 20uH inductor.  I wound #24 guage wire around a 1/2" wooden dowel 2 inches long for the secondary. 

The audio output transformer is a high fidelity Hammond model 1620 (20 watts) and 20 Henry chokes in series for the Heising circuit.

The RF sample circuit comes from the 1972 ARRL Handbook. 

Jon
KA1TDQ

[W3GMS]

Jon,
Looks like a fun project.  In looking at the schematic very quickly, I noticed that your plate RF choke is on the wrong side of the plate blocking capacitor.  I am sure it was just an oversight. 
Joe, W3GMS

[ka1tdq]

Thanks Joe, good catch...

In one position, you will enjoy countless hours on the air.     In the other, you short out your high voltage supply.   

I drew up the schematic this morning, but I most likely would've caught it during assembly.

Jon
KA1TDQ 

[W3GMS]

Hi Jon,
I try not to "nit pick" others designs and I am sure you would have caught it. 

Have fun with the project and keep everyone posted on your progress.  In my mind nothing is better than doing a real HB project.  I am very much looking forward to creating something myself soon. 

73,
Joe, W3GMS   

[K1JJ]

Hi Jon,

Looks like it should work FB!  The schematic is a good basic design. I've used a similar one for my 50C5 rig.  The outboard audio with the 8 ohm solid state driver always sounds good.

One quick minor point...  when drawing chokes, the two lines next to the coil symbol indicate an iron core. Iron core chokes are used in audio and power supply chokes but usually not in RF CHOKE service. They can be used there, like with slug tuned forms, baluns, etc.,  but I think not in your intended application as 2.5 mh in the grid or plate circuits.  So, eliminate the double lines for RF chokes in your schematic and it will be accurate.

Hope to hear in on soon.

T

[ka1tdq]

Thanks Tom!  Yeah, I borrowed most of this stuff from the AM Window schematics. 

Point taken on the RF chokes...  I should have this going in a couple months.  The drive level is perfect for my homebrew linear (pair of 3-500's). 

Jon
KA1TDQ

[W1RKW]

how does the bias work?

[K1JJ]

how does the bias work?

Looks like straight 20K grid leak bias, Rob.

IE, if he loses drive, (with the cathode resistor keyed) the final may draw some serious current, will probably self destruct, depending on the screen and plate voltage values.. A small fixed bias between the 20K and ground wud solve the problem - just enuff fixed bias to make the tube draw safe current or at cut off - when without RF drive.

T

[ka1tdq]

I might throw in some fixed grid bias for a just-in-case problem.

-or-

I have over 30 of these tubes.  If I destroy one with no drive some day, oh well.   

I do have a new 24 volt transformer though and it'd be easy to make a negative voltage supply.  I might just throw it in there too and make the voltage variable.

[K1JJ]

A good decision, Jon.

Sometimes other stuff can break along with a melted tube too, like the plate RF choke, your Heising reactor, diodes, power supply iron, etc.    Limiting current is a good idea.

You might want to pick a fixed bias voltage and use a zenor regulator there. I have had problems with the fixed bias charging up from RF rectification back into the bias supply. The voltage will slowly creep up higher and higher until the final current drops way off. Unkey and it starts over again.  The fixed voltage with a zenor solves these problems.

Another way is to use an iron core audio choke (8H?) in series with a diode in series with the fixed bias lead to keep the RF charging out.  I've done both techniques in my 4-1000A X 4X1's modulators rig.  Works at any level..

T

[WU2D]

That is a hot tube but the capacitance is low and it is likely very stable without neutralization. But if you want to play...

If you tap up the grid coil a few turns and ground that point instead of the bottom, essentially floating the cap, this will give you a place to wrap around some feedback from the plate (through a piston or very small air trimmer like 1-5 pf) for neutralization. You could also accomplish the same by making a split cap arrangement. This would be done by tacking say a 680 pF to the bottom of the variable, and ground the center point and accomplish the same thing. Again the feedback goes to the bottom of the tank.   

First set the amp up and tune it into a 50 Ohm load with your drive source. Move the 50 Ohm load to the input.

To set neutralization (with fil on but B+ off) attach the drive source temporarily to the output connector. A 50 Ohm pad is nice here. Probe the top of the grid tank with a scope. Peak both controls (grid and plate) for maximum scope deflection. Now adjust the trimmer for minimum. Apply B+. Re-adjust.  Bang you are critically neutralized. This is the reverse method. A little dangerous but it always works and you can see what you are doing.

Mike WU2D

[K1JJ]

First set the amp up and tune it into a 50 Ohm load with your drive source. Move the 50 Ohm load to the input.

To set neutralization (with fil on but B+ off) attach the drive source temporarily to the output connector. A 50 Ohm pad is nice here. Probe the top of the grid tank with a scope. Peak both controls (grid and plate) for maximum scope deflection. Now adjust the trimmer for minimum. Apply B+. Re-adjust.  Bang you are critically neutralized. This is the reverse method. A little dangerous but it always works and you can see what you are doing.

Mike WU2D

Pretty slick, Mike. I've never heard of that procedure of neutralizing "backwards"..     I will try it next time.  I usually put RF into the grid and sample the tank.  

Do you find the isolation better and a cleaner neut dip when using this method?   (My 4X1 is neutralized with an insulated shaft coming out of the front panel.)

T

[WD5JKO]

Jon,

Might try the John Stout circuit with three 7984's:

http://www.amwindow.org/tech/pdf/eckfinal2.pdf

This would be a complete transmitter with modulator, and RF exciter.

I keep wanting to try out a 7984 in one of my Central Electronics 20A's. I have one 20A with a pair of 7591's, and another with a single E-34L.  The 7984 might do well as a class Ab1 linear amplifier..


Jim
JKO

[KB2WIG]

Clicking on the above link yields



" This site has been suspended. "

klc

[WD5JKO]

Clicking on the above link yields
" This site has been suspended. "
klc

   Yea, looks like the AMWINDOW is down, strange.

I am attaching the schematic here.

Jim
JKO

[WU2D]

Since it is essentially a bridge, you can neutralize forward or backwards. I find that backwards is usually easier because I simply strap the gen on the output connector and probe the grid tank with a scope. A ham here in New England showed me this technique a few years ago and I have used it on everything from homebrew stuff to the Apache to and old Hallicrafters SR-150 with sweep tubes. It works.

Mike WU2D

[ka1tdq]

Here are a couple new photos of the work in progress.  The top of the chassis is almost done and the wiring is about 1/2 done.  I'm hoping to smoke test it in a couple weeks. 

Jon
KA1TDQ

[N0WEK]

That's nice clean work!

Looks great!