A Cool CW and AM Hartley SBE Rig Using 24G triodes (3C24's)

[K1JJ]

Anyone remember the little (circa 1933?) 24G triode? (3C24)  They will handle 2KV and have a plate diss of about 25w or so. I built a Hartley AM and CW transmitter back in 1995 that was a bunch of fun. It had 1800V on the final and put out 90 watts, though with 1000 volts it is good for about 25 watts or so.

This is my own design and has been used on the air. It could be scaled up to even an 833A if desired. Tim/HLR now has an SBE rig on the air (Hartley - called a slop-bucket eliminator) that sounds pretty good.

This rig is sensitive to antenna movements in the wind, so expect some whoop-whoop carrier shift on the air. After all, it is a power oscillator and the whole RF transmitter is just ONE tube, VFO and final combined. I used to have openwire spaced 2 feet apart stretched tight for 100' without insulators - the rig had a slow swing to it as a result.

The 3-diode negative peak limiter is mandatory to prevent the power oscillator from dying when over modulated in the negative direction. The 200VDC on the diodes will limit it to 95% negative and the oscillator will remain on and happy with 130%+ positive peak modulation. It can really sound hi-fi assuming the FMing is held to a minimum through good, rigid RF connections.

I used a vacuum variable and 3/8" inductor for the tank circuit for best stability - plus kept it away from the final's heat and air using a barrier.  These values will work for 24-G's and the rig is stable enough for AWA CW. Frequency is controlled by the main tank capacitor - and fined tuned +- 5kc with a 20pf vernier as shown on the schematic below.


I'm thinking of building up an 833A version, so pulled the schematic out. Figured others could use it as a guide for their projects too.

73,
Tom, K1JJ

CIRCUIT BELOW:

[N3DRB The Derb]

hmmmmm......I got 4 24G's new in teh boxes......

[K1JJ]

hmmmmm......I got 4 24G's new in teh boxes......

I can hear it now - The Tron:  "Calling the SBE Reverb Net... no need to zero beat."

[Ed/KB1HYS]

That looks like it would work FB with 811's too, come to think of it...

[N3DRB The Derb]

can you imagine what a cb echo box hooked to that & cranked would sound like in a slopbucket rig? serious DOUCHE action. Now stack 5 or 6 SBE's on top on each other all with echo. No slopbucket could survive that action for long.   

[W3GMS]

Hi Tom,
Bob Raide, 2ZM put a Hartley on years back and Heising modulated it.  I use to work him on 160 all the time.  It was always thought that you could not modulate a self excited oscillator but you can if you keep the percentage of modulation on the low side.  When doing so it was pretty darn stable.  He used a candle stick carbon mic on the beast.  If I remember right he used a 203.  Of coarse if you really wanted to FM it just turn up the mic gain a bit and hit it a bit harder!! 
Joe, W3GMS 

[KM1H]

Thats a nice rig Tom, looks like something to duplicate if I ever get the time. Altho my PP 211's work fine on 80 CW I never could get them to take any worthwhile amounts of audio and stay in the passband. I fire that off in DX contests and a few other times chasing DXCC.

What did you use for a driver xfmr?

The earliest reference I can find for the 24G is 1943 altho its an obvious takeoff on the 35T of roughly 1936-37.

[w1vtp]

<snip>...  It was always thought that you could not modulate a self excited oscillator but you can if you keep the percentage of modulation on the low side.  When doing so it was pretty darn stable. <snip>
Joe, W3GMS 

Timmy HLR has managed to get his SBE to fully modulate without an excessive amount of FM -- meaning it stays within normal AM modulated bandwidth.  He had to play around with the grid leak and the coupling but he did it

Al

[N3DRB The Derb]

his SBE sounds damn good to me. I've heard many a store bought transmitter sound a lot worse.

of course, I had to go look in some early 30's QST's and read about the evils of self excited oscillators and how modern hams would NEVER run such rigs when xtal control was available. 

[K1JJ]

Yes, the self-excited power oscillator can be modulated fully (with big positive peaks) as long as there is heavy grid current and there is a negative peak limiter inline. The three diode circuit in the schematic was needed to keep the rig from shutting off when hitting neg100%. After all, -100% negative ceases to draw ocillator current.

By adjusting the grid leak value and the tank coil tap (grid feedback) a combination can be found to make the oscillator both stable to warble and heavy modulation.  Mine was unusually stable in operation once all the sweet spots were found. (Except for wind sway of the antenna or openwire.... ) Antenna loading (C2) is also important.


Carl, for an audio driver xfmr I just used a common 8 ohm to 2K? audio output type used in any tube audio amp. Yes, 1937 does ring a bell for the 24-G.  It is a rugged little triode and will put out big power for what it is - I was amazed by it.   Though, this circuit will work with 811A's, 833A's and other triodes. The values will have to be juggled somewhat.

I built it on a large piece of plywood that had a layer of PC board copper on top for the groundplane. Once set, it didn't drift much at all. The trick is to use a vacuum variable and heavy copper coil and copper sheet metal connections in all RF connections. For temp and magnetic stability, they must be away from everything including air flow.

I'm considering building up a single 833A and GRID modulating it. I don't want to go to the trouble of plate modulation to start with and will see if it hangs in there as such. Tron says he's not sure if it won't become unglued, but will be an easy way to go if it works right.

T

[w1vtp]

<SNIP>
... (Except for wind sway of the antenna or openwire.... ) Antenna loading (C2) is also important.


<SNIP>
...
...
T

Isn't the swaying effect the whole idea of a SBE (as long as one doesn't go out of his way to make it happen)?  It was memorizing watching Timmy's signal dance back and forth in time with the swaying of his ant.. on the panadapter in a high resolution setting.  I could tell when the wind increased later on in the afternoon.

Al

[K1JJ]

Yes, the gentle swaying of the signal as a result of the antenna moving around is pretty cool - and should be the only major thing that adversly affects the stablity of the rig. Otherwise, if the modulation affects it too much, the detected audio has a gurgling sound like Tron is choking on his own vomit...    (this comment always gets a good laugh outa Tron)     

In addition, the slow drift as a result of thermal expansion of the tank circuit is a pain to keep on top of and should be worked out too.

T

[k4kyv]

Until sometime in the 70's the SBE rig would have been illegal.  There used to be a regulation against "simultaneous amplitude and frequency modulation".  I once got a pink slip under this regulation for using my own version of SBE, which was a motor-driven variable capacitor connected to the VFO circuit, that caused the carrier to shift back and forth across the frequency +/- about a kc or so at a rate of about 12 times a minute.  With normal AM selectivity it was hardly noticeable, but it gave the slopbucketeers who tried to piggy back ride the carrier a fit.

The FCC said  that although the regulation was not intended to cover a case like mine, it was "not so limited".

Sometime in the late 70's as a final R&O for Docket 20777 they rewrote the regulations concerning technical standards for phone signals.  Evidently the person re-writing the regs didn't take AM into consideration, and they deleted the prohibition against simultaneous AM and FM.  Also, they deleted the regulation limiting modualation percentage to a maximum of 100%.

It is ironic that those restrictions on AM operation were inadvertently deleted by the fee-cee in  response to a docket that had originally proposed to outlaw AM on all amateur frequencies below 28 mc/s.