It stands to reason that if you can build a monaural transmitter with one tube, then by logical extension, it should be possible to build a stereo transmitter with two tubes. So...

I had to do something with all those 6CQ8 equivalents that Tom sent me

Here's my sloppy breadboard prototype:


V1a is a crystal oscillator to generate the RF carrier. V1b is a reactance tube phase shift modulator which is modulated by the R-L signal. It's output is a constant amplitude varying only in phase. This is then fed into grid 1 of the modulator V2b. The L+R audio is fed into V2a where is is amplified to a level suitable to screen modulate V2b. At the moment there's nothing in the circuit to generate the required 25Hz pilot, so I've been using a function generator, and feed in that signal to the Pilot In jack. I'll probably have to add another tube at some point to generate the pilot. Darn, that'll make it three tubes.

This circuit isn't necessarily final, but as it stands the audio quality is very good, and stereo separation is pretty reasonable. Using the phase modulator rather than a full blown quadrature modulator doesn't seem introduce any audible distortion, but it does appear to limit the channel separation a bit. Still, for the simplicity of the circuit I can't complain.

All those AM stereo receivers they made in the 1980's are now a quarter century old, and could already be considered antiques. You'll be needing something to transmit music to them.

How does it work with a 400 Hz tone fed to the left side and a 1000 Hz fed to the
right? Can you recover the original sin waves, with reasonable purity at the tuner
end?

That's surely interesting, glad you could put some of those tubes to use... Unfortunately I don't own a AM-Stereo anything, time to check out the thrift stores...

Tom

That's clever Bob. An AM stereo receiver never came my way, but there's always tomorrow.

I have a Realistic TM152 AM-AM stereo tuner on my bench. However finding an AM stereo
/FM stereo + cassette car radio is probably more likely. They show up at thrift shops for
about a buck, and I have been told that on the web you can find a list of these to look for.

I can make the stereo light come on by modulating my signal generator with 25 Hz.

The tuner is handy to find radio stations still using C-QUAM.

I remember there being an AM stereo radio in my dad's mid 80's K-car. I think Chrysler put them in just about everything for a few years. So, there should be a lot of them around.

For my tests I'm using a Sony ST-JX450A, which I've had since the early 1990's.


At the moment, my only function generator is being commandeered to generate the 25Hz pilot tone. I'll have to cobble something together to do this. Maybe I can dig up one of those old stereo test records.

So far, my best test has been to broadcast a bit of modern electronic dance music that has sounds panning left and right, which it reproduces fairly well. I'll try to post an audio clip, but first need to fix some hum problems caused, no doubt, by sloppy construction.

I've uploaded a couple of audio clips. They are exactly as received on my AM stereo receiver.
I found that the hum problem was actually interference from another source. After I recorded the first clip, I managed to adjust the loop antenna on the receiver to minimize the noise. So, the second clip is cleaner.

The first is a big band audio clip that has pretty good stereo separation:
Johnny Favourite
I deleted a bunch of the middle part.
Towards the end, I unplugged the Right audio input to the transmitter for a few seconds, then plugged it back in and unplugged the Left audio input. You can hear some noticeably distorted audio on the unplugged channel. This doesn't surprise me; I expected the worst distortion would occur when there was only one channel present.

The second one is some electronica with some interesting sounds that pan left and right:
Leftfield

I've noticed that the right channel is a bit louder than the left. Adjusting the antenna matching network has an effect on this, but doesn't fix the problem completely. I'll have to figure out whether this is due to the phase modulator, or somewhere in the audio chain.

PS: Best to listen with headphones.

As a broadcast engineer who played with AM stereo on a licensed station in the '80's, I judge the results you are getting with your set-up as SUPERB!

Dave

Split L&R channels to 2 transmitters using separate freqs to 2 old radios on a shelf. Stereo! I did it with a 6BM8 and 6888 transmitters way back when. Stupid experiment, but I can say I did it.

With this you could use one of the 60's era pre-multiplex receivers that had two AM tuners,
like The Sansui SM-80

That's actually a very good idea. It gives you the opportunity to use your two favorite antique radios, and get stereo to boot.


Wow. Well, I'll certainly take that as compliment. Thanks Dave.

Still, I need to sort out the difference in levels between left and right. Having a second look at the phase modulator, it appears that I calculated the capacitor values incorrectly, and they should be smaller, which will probably affect other components. Oh well, back to the drawing board.

Chrysler had AM stereo as standard on most of their radios from 95-91, and their premium sound systems to probably 2000 or so. Delco had a lot of them in the late 80s. Not so sure about Ford though I know their Mach 460 had it through most of the 90s if not somewhat beyond. A number of the new digital radios can decode CQuAM as well as IBOC. Would love to see this circuit in its final form. I only wish I had more time to tinker as I'd like to build the whole thing with quadrature modulator. As it is I have too many projects and too little time and $$.

Scott Todd

The two radios to listen to stereo setup reminds me of something I had read about the early days of stereo, where a station would broadcast one channel on AM and the other channel on FM, so a person with an AM and an FM radio could listen to stereo. Seems pretty cumbersome, but at least it proved the concept worked...

Yeah in my early days at Sears, I serviced a few console HiFi/Stereo built before FM MPX,... Those had a AM/FM setting where one channel played AM & the other FM...

Tom

Considering this rig is using phase modulation, it's essentially a Philips/Magnavox system. There was quite a bit of compatibility between each system except for Kahn ISB.

Scott Todd

Yes. In fact, by the time the Motorola system was finished monkeying around with the QAM signal it was essentially a phase modulated signal. So any difference between the Motorola and the Magnavox system would have been very minor. Maybe some difference in the modulation levels.

BTW, I'm in the process of building this thing into a proper chassis. I've now got the pilot generator included in the circuit, so it's now a 3-tube transmitter. I'll post the final schematic once it's built. It's mostly finalized now, but there may be some minor changes during the build.

Okay, I guess the circuit is finalized, seeing as I've got it into a cabinet, and everything is all closed up now.



Prior to closing up:



Final schematic:


For the 25Hz pilot oscillator I built an LC resonant circuit using a small power transformer as the inductive element. The tube is a 6HA5 triode which I happened to have on hand.

The meter has a 3-position selector switch, and is used for:
A - Setting bias on phase modulator
B - Monitoring audio input level
C - Monitoring RF output, and peaking the antenna matching network

WOW Bob that's nice, superb construction...

What are you using to couple the detector diodes to the antenna output???


Tom

Thanks Tom.

The pickup for the detector diodes is just a short insulated wire wrapped one turn (loosely) around the RF output wire, right at the antenna jack solder lug.

I'd guess he's using a gimmick. And I agree- superb looking construction. If he ever gets a multimode receiver, he could adjust the pilot for the Magnavox pilot frequency and get it to play in that mode as well.

Regarding the output tuning- does the second arrangement put out more signal than the first iteration? Seems to me the first one would work better, but then I'm not the one experimenting with it.

Scott Todd