I built up the breadboard version of Tom's 7AK7 Tx awhile ago. It works fantastic. With a short piece of test lead for an antenna the thing transmitted a solid signal form my shop over a hundred yards to the Kenwood receiver in the house.

I had some difficulty getting rf inductors, so made my own. They are wound on sched 40 plastic pipe. I tried winding some toroids which worked but the oscillator waveform was distorted. Someone suggested a 1K between the oscillator coil tap and the cathode would help and it did. The cathode voltage also increased substantially. Maybe someone can tell me why. Scary winding with #35 wire at first, but went well; so now I am looking at the roll of #40 sitting on my shelf for the next coil project.

I also wound a large toroid, 1.8 inch hole, loading coil; wound full with one layer of #27 wire. That probably helps a lot.

Note the two section variable cap. I tried to make the the oscillator and output plate circuit track so one control could be used. It did a fairly good job but abandoned that idea as I really want to build a vintage looking enclosure. I am currently into that part of the project.

When building the final chassis, I noticed the loctal socket mounting pan-head screws interfered with the tube base such that the tube could not be inserted all the way. Flat-head screws will have to be used.

Thanks Tom (35Z5), the circuit performance is great. I just need to keep fighting off the other pesky projects that keep showing up and focus on the AM Tx.

I am currently looking for some nice cabinet wood. That would be typical for the 1920 or '30s. Maple maybe, I've never seen oak used?

Lyn

Glad it worked out for you... No one seems to be using the 7AK7, guess they don't like Locktals...

Tom

I built one a while back and it was the most powerful of any of the one tube transmitters that I've built. Maybe a little too powerful.

Lack of popularity could be because the locktal sockets are harder to find.

I have a question. Can the 7AK7 be adapted in place of a 6GY6? If so will there be any benefits to doing that?

Thats a pretty healthy tube but not very common. OTOH if you like octals the 7C5 will make plenty of RF or an octal 6AG7 is one of the best oscillator tubes built.

Carl

A 7AK7 is the loctal version of 6888. Yes it will work in place of 6GY6. I am sure Tom has made some resistor adjustments when using these tubes.

When driven with the 5v sq wave from a clock osc the output of a 6GY6 & 7AK7(or 6888) will be approx the same, so other than aesthetic value there isn't much difference... Hit the 7AK7 or 6888 with a 10-15v sine wave and the big guys leap ahead in RF output...

Tom

Would it be possible to add a triode to step up the oscillator voltage after using a choke to make it a sinewave? If so I could replace the 6C4 with a dual triode tube

Well I suppose you could add a "amp" stage to the osc but by this point I'd just go for a totally fresh build... For broadcasting around the house, a 6GY6 as designed will produce more than enough signal...

Tom

oh ok

Tom,

Not sure why the hesitation on the 7AK7 Tx. Yesterday I purchased another 7AK7 and two loctal sockets for $10. I also did an internet search on the 7AK7. That tube was used on the early computers and has a lot of neat improvements over traditional tubes. Used within it's limits, it will probably never wear out. The internal metal parts are made with specially prepared alloys that will not out gas over time.

I have an old Philco console, 1942 I think, that is totally populated with loctal tubes. Sometimes when removing a loctal, a little prying with a small flat blade screw driver helps get past the detent. Maybe folks rock the tube in the socket; That will lead to problems.

Anyway, I am more than satisfied with the performance.

Thanks,
Lyn

If you use a plain crystal instead of crystal oscillator, you can get around 30 Vpp of sinewave on the grid, which is probably why I got plenty of output with mine. Here is the basic schematic:

Oh, my. I thought using self excite was considered almost a crime around here

The correct way to remove a loktal is to read this

viewtopic.php?t=112936

Carl

looks like an interesting circuit. may have to try it some time.

Not really a problem with a crystal. One of the advantages of suppressor modulation is that you have a screen grid between the modulation grid and the oscillator grid, which shields things from each other very well.
I've also used a ceramic resonator with minor circuit changes. The important thing is to minimize the loading on the resonator. An LC circuit is even possible with careful design. A good example of a stable self-excited LC transmitter is here:
http://www.radiomuseum.org/forum/hi_fi_ ... itter.html

That is an awesome transmitter contained in that thread.

If I wasn't so happy with my 6GY6 two tube transmitter I'd be inclined to try that one or this 7AK7 transmitter.

Yes, I ran across Sousa's 'Hi-Fi' transmitter when researching my own idea, which turned out to be the same. Well, the plate/screen push pull part was, which I arrived at by first contemplating a 6ME8 and then later observing the plate/screen characteristics of the dual control pentode (or sharp heptode). It came from mostly virgin ignorance. I was listening to a conversation, on a different board, about the difficulty in getting 100% mod and asked why not 'subtract' carrier? Hence a 6ME8 as the 'first thought' with my logic being there are no phase issues since it's the same 'beam'.

Seemed to me you'd end up with 'exaggerated' side bands with DSBSC at 100% mod being the obvious example but I didn't know what the net effect might be, and didn't get any answers at the time, but suppose it's mainly a radiated power matter.

Anyway, with Sousa's being an exception, I almost universally hear that self excite stinks and I've wondered why because there's almost no modulation component on the cathode as G3 essentially shifts current from plate to screen, so cathode current remains constant. I mean, like when I postulated the 6ME8 version the first thing out of everyone's mouth (different board) was to ditch self excite.

The most common issue I hear is FM mod but where does it come from?

Thanks Carl for showing the correct way to remove the loctal tube.

I found the bump on the loctal base. Originally I thought it was only to locate the base key. Another feature that shows that the locatal tube design was well thought out.

This forum is very helpful.

The most common reason is for freq shift is a high impedance osc coupled directly to the control grid, though as mentioned the dual control & hextode types generally have adequate shielding by the screen grid to prevent this issue...

I've had problems with a direct coupled osc when using screen modulation(talking toobs not the solid state osc), a VFO is prone to FMing and a rock bound osc can be loaded to the point by some tubes(beam power) it ceases to function... I've found a low impedance cathode follower solves the problem...

Tom