Can this FM detector even work?
 

I'm not getting any audio on my Delco experimental car TV. (The audio output amp works fine with an injected signal.) After tracing out the schematic of the FM detector, I'm wondering if it can even work. It doesn't match circuits I've seen for either the Ratio Detector or the Foster Seeley Discriminator. There is no center tap on the secondary of the transformer. Has anyone seen a detector like this? BTW, pin 7 of the 12AU6 goes to ground.

http://antiqueradios.com/gallery/mai...serialNumber=1

Bob D.

That's a strange circuit, but may be functional as a "slope detector." That means it is simply an AM detector, but the resonance of the tuned circuit is tuned slightly off the IF peak. That way, as the frequency of the carrier is modulated, it moves closer to and farther away from the tuned circuit resonance, causing the amplitude to rise and fall.

Normally you don't need two diodes for a slope detector, and I'm not quite sure what is being accomplished by having two diodes here.

It might be a good idea to use an oscilloscope to check if there is an actual audio IF signal going to the 12AU6. There is no point messing with the detector circuit in an undocumented set if it has no signal coming into it.

Could there be a center tap on the secondary that is tied to ground inside the can?

Good point. There might even be additional components hiding in the can. I'll try injecting a 4.5 MHz FM signal into the grid of the 12AU6. As Tom C. pointed out, the circuit may not be getting a signal to demodulate.

(Just a thought, when did the sound carrier switch from AM to FM?)

Bob D.

FM sound was adopted in 1941.

Looks like an early flux capacitor to me....

At first glance I thought this resembled a (push-pull) voltage doubler more than any FM detector that I'd ever worked with.

I thought I'd compare it with a known pre-WWII TV AM sound detector and looked at the RCA TRK-5 and got a surprise:

http://www.earlytelevision.org/pdf/r...-rider-tv1.pdf

In that set, V-10 the 6B8 pentode, is the 2nd sound IF and its two plates on pins 4 & 5 are the AM detector.

You will notice a circuit similarity (although the TRK-5 detector cathodes are common, and yours are not), and it is drawn quite differently.

It is a known fact that some of these 1939-1940 sets detector transformers were intentionally mis-tuned (for AM) to keep them running in the early days of FM sound when replacement parts were not available.

I'd say that this circuit could probably work as either an AM detector or an FM detector depending on the tuning.

Notice the other components inside the transformer in the TRK-5 circuit.

Also, notice that one of the diodes has both sides connected to ground, therefore it is not used.

In your circuit, I think they were trying to increase gain by capturing both polarities of the slope to increase level.

Jas.

The last slope detectors were used in 1946 and 1947 kit sets. None were used in any production set after that. In addition slope detectors won't work in intercarrier sound designs, which I'm sure is the case with this set. This isn't a slope detector.

There is a slope detector description at the bottom of that TRK-5 file you uploaded,
describing the mis-tuned detector circuit.... Seems to me that actual tuning is more critical because of where on the slope the detection needs to take place.....

Something missing in that hand written schematic...? did you double check it.....?


Still looks like a flux capacitor.....

Skip injecting a 4.5, Sweep it and see what those two sides are tuned to...

Is it a single core ?

Hi Steve,

Is the reason a slope detector wouldn't work with intercarrier sound because there is amplitude modulation present on the 4.5 MHz carrier from the video carrier? Just curious.

Tom, that is correct.

The polarity of the diodes suggests Foster-Seeley. Also, there must be some caps in the can to tune the coil.

Its clearly a discriminator. There's a center tap on the secondary
and a coupling cap inside the transformer, in addition to
tuning caps.

You'll have to open that can up and solve the mystery for us. It sure does look like it ought to be a discriminator with a few more connections that we can't see yet.

The passband on the video IF has a slope to it, where the picture carrier is located. This is the Nyquist slope. This slope introduces some fake FM in intercarrier sets. Only reason we can get away with this in intercarrier sets is the FM capture effect, which helps the FM detector ignore a weaker FM signal in favor of a stronger one. This weak fake FM comes from the video modulation sideband frequencies beating against the FM sound carrier, combined with the Nyquist slope.. Imagine if we tried to do intercarrier if the TV sound was still in AM. Buzz city...

When TV stereo sound was introduced, the sound IF, independent of the video IF, had its own pair of filters at 41-45MHz. One peak at the video carrier, without the above slope. And another peak on the sound carrier, before it's made into a 4.5MHz FM signal. This would also have worked if the sound was AM. And you wouldn't have to do fine tuning just to get the sound.

Well, I have to eat a crow here. The circuit DOES work. I tickled pin 1 of the 12AU6 with FM modulated 4.5 MHz, and a nice pure tone came out of the speaker. It would be a lot of trouble to open up the can, so I won't. (Remember curiosity killed the cat.) Now to find out why I'm not getting any hash sound.

Bob D.

I have some more info. I switched my generator from FM to AM, and this circuit seems to demodulate AM just as well as FM. Also, I found that the IF cable from the tuner was pinched in the sheet metal. After fixing that, I have real audio from the channel 3 output on my cable box. The audio sounds fine as long as there is no digital lettering on the screen. There is a nasty buzz when there is lettering on the screen. Our first TV, some forty years ago, was an old B&W set that did the same thing. I was never able to eliminate the buzz, and finally decided it was a design limitation of the set. Back in those days only the news shows had digital graphics, so it wasn't that big a deal. Is this 'buzz' problem common on old TVs? Is it a design limitation as I assumed?

Bob D.

A couple of thoughts regarding sound buzz on white characters:

1) fine tuning - tune towards lower frequency (softer picture) to move the video carrier up the IF slope. If this helps, it's an indication that the IF frequencies in the lettering are too strong compared to the video carrier, causing overmodulation. However, this may cause a different sound problem by moving the audio IF carrier down the other slope of the IF response.

2) If (1) indicates improvement is possible, an IF alignment adjustment to move the video carrier up the slope of the IF response curve may help, while keeping the audio IF carrier at the correct frequency and therefore at the correct place on the other side IF slope.

Bob, Congratulations on solving the sound issue!

I agree with old tv nut, but remember, alignment is the last thing you do to a TV, not the first.

I'd first check to see if the incoming signal is too strong, and/or the AGC is maladjusted, before thinking about alignment. Remember this set was designed for a auto antenna, not a relatively high level signal generator.

By the way, since your first pictures in another thread did not show a power supply, what are you using for one now?

James.

Ditto! Make sure nothing else (like AGC) is out of whack first.

Further to the internal capacitor not depicted, note the voltage divider on the screen of the preceding stage. The Foster Seeley circuit did not reject AM hence they would always be preceded by a limiter.

The lowered screen voltage would indicate a limiter.

Hi James, there is one large gauge wire going into the chassis that powers the tube filaments. It must be 12V, and originally came from the car battery. (The 6V tubes are wired as pairs in series.) I'm using a large switching power supply, and the TV is drawing 4.5A.

There is a single "B+" input on the TV chassis. (This must have come from a separate power supply that is missing.) There is a 300V electrolytic directly on the raw B+ line. The 6S4-A tube has only one recommended operating voltage on the data sheet, 250V. So 250V seemed like a good choice. (So far no smoke, fingers crossed.)

I just happened to have a pair of Heathkit HV power supplies model IP-17. These are rated at 100mA, 0 to 400V. It became apparent that 100mA was not enough to power the set, so I modified the pair to act a one 200mA supply. Each unit had 2 6L6s in paralled as series pass regulators. So there are now 4 6L6s in parallel.

So 12V @ 4.5A DC (could probably use AC) for the filaments. And 250V DC @200ma DC for B+. Once the set is totally restored, I plan on building a dedicated power supply for both "A" and "B+".

Bob D.