Old TV circuits, chapter 43: Sound IF
 

I always get thoughtful and helpful answers here, so I figured I'd try to get some insight into how this circuit works, and a few lead-up questions:

I'm working on a Predicta, using it in Spain. Since all of my other sets are CCIR or PAL, I use a HLLY 2W transmitter in my attic, fed with a PAL baseband signal from a DVB tuner and set to transmit on UHF channel 26, video at 511.25MHz and audio at 516.75MHz. This works great on my older CCIR-B sets. I would like to make the Predicta tune the same channel 26 broadcast, with sound. We'll leave the subject of sweep aside for now.

Given my woefully novice understanding of how the audio gets from the tuner to the FM detector, I'd like some help walking through the Predicta's circuit and hope to determine if what I want to do is both possible and feasible.

To wit:
https://i.imgur.com/7BeMvWB.png

I've been poring over Fink and Gulati to understand what is going on here: what comes from the video detector is, if I understand correctly, a 4.5Mhz audio carrier that somehow made its way through the video IF amplifiers and was then subsequently mixed with the baseband video signal by the video detector diode (somehow again). This much I understand, leaving aside the somehows.

From there, we go to connector L24U in the schematic, where this 4.5MHz carrier is coupled by C16 to the grid of the sound IF amplifier, in parallel with a simple tank circuit which I assume is used to filter the adjacent channel's video carrier? Or something? Fink mentions the purpose of this tank circuit in the section on video IF but glosses over it in the part on audio.

Then our carrier gets amplified by the 3AU6, then coupled to the audio detector by T3, and its associated tank circuits which, again, I am assuming serve to sharpen the dropoff on either side of the 4.5Mhz carrier. And then on we go to make sound.

I'm making a huge assumption that the 5.5MHz audio carrier from my PAL broadcast is going to somehow make it through the three VIF stages just like the 4.5MHz carrier does on an NTSC broadcast. And IF that is the case, would it be possible to tune these above tank circuits to have their hump centered on 5.5MHz?

Regardless of whether it's possible I'm still very interested in understanding how this works.

Thanks everyone for the great answers I'm sure will come.

You have it basically correct. To clarify, the output of the video detector contains an FM sound carrier at 4.5 MHz (for NTSC system M). The video path then usually has a 4.5 MHz trap to prevent the 4.5 MHz carrier dots from appearing in the picture.

Here in the audio section,T5 and C20 (with C16) are a tuned circuit to select the 4.5 MHz carrier and attenuate video components. The amount of tuning obtained from C20 vs. C16 is hard to tell without knowing the output impedance of whatever L24U is connected to. So, at least C20 needs to be reduced and maybe C16 also, to tune this circuit to 5.5 MHz. Similarly, the capacitors inside the T3 can need to be reduced. Part of the problem is that the caps inside the T3 can may not be marked, so there could be a problem knowing what to substitute. C13 needs to be reduced also.

To match the original design parameters, the values of the inductors should be reduced also, but of course there is no information on their current value. Hopefully they can be retuned to lower inductance enough.

The resonant frequency of a tuned circuit is f = 1/(2*pi* (sqrt(L*C)), so to go from 4.5 MHz to 5.5 MHz, you want to reduce both the L and the C by the ratio of 4.5/5. = 0.82. So, C16 becomes 8 pF, C20 becomes 6 pF, and so on.

You might start by ignoring the capacitor changes and try tuning the inductors only, to see how well that works. If it doesn't work well, then you must look at changing the caps.

I don't know what the alignment procedure is in the service manual, but the procedure should work just as well for 5.5 MHz as for 4.5 MHz.

Seems like the first criterion would be whether the Predicta's VIF can pass a 5.5mhz intercarrier.

oops - yes.

Thanks very much for this detailed response. I'm glad to know I'm on the right track.

With respect to old_coot88's point about the VIF passing the 5.5MHz carrier, this is my first "somehow" in the OP. Looking at the circuit, I don't really see any reason why it wouldn't. Fink says the gain of the IF stage is about 8X greater for video than audio, and here I'm only seeing traps that seem to be for increasing selectivity (X3, X7, X8, VC1):

https://i.imgur.com/lHINzZs.png

Along with the interstage transformers T1, T2, T3 which set up the hump around 45.75. Am I correct that the 4.5MHz just sort of goes with the flow through this circuit, on account of being so far removed from the resonant frequency of any of the traps? The gain through the three tubes should be equal at 4.5 or 45.75; Fink's 8X difference in the net gain of the circuit between the two frequencies is due to the frequency response of the interstage transformers? If that is the case then the 5.5MHz carrier should also sail through, no?

The sound alignment procedure is basically to tune a weak station and crank on the coils and transformers until maximum output is obtained. My hope is that there is enough range in those inductors to avoid having to swap capacitors.

I'm wondering about the best method to verify the passing of the 5.5MHz carrier through the VIF section, and then peaking those inductors in the SIF section. I don't have a sweep generator but I wonder if I could just inject an unmodulated 5.5MHz signal at J4, and look for its presence using a scope on the output? If this is viable then I can also presumably then inject a 4.5MHz signal to the SIF and gradually increase the frequency, adjusting the inductors with each increase, so as to walk the center frequency up to 5.5MHz?

Learning is fun. Thanks very much guys.

Nope. 4.5MHz or 5.5MHz does not pass through the video IF as such. Those frequencies are AM modulation on the video IF frequency. It is as if they are in the baseband. Using 4.5 or 5.5MHz after the video detector would work. You would have to be careful you did not load the circuit when you inject the signal. To correct what you said originally, the video detector does not mix a signal.
As far as getting the signal through the video IF, you would have to figure the frequency that the 5.5 MHz sound carrier occupies when it is in the video IF passband.

Correct. The actual aural carrier as received over the air is 5.5mhz lower (instead of 4.5mhz lower) than the picture carrier. That's what generates the 'intercarrier'. In superhet parlance, the intercarrier is the Intermediate Frequency or heterodyne between the two carriers, that difference signal appearing in the output of the V detector.

Oops I read the post "4.5" and in my head saw "41.25," the sound IF fequency of the American system. Assuming the set is still tuned so the PAL video IF frequency is 45.75 (same as American system), the PAL audio will be 40.25 (45.75 - 5.5). You will need IF response at 40.25 to be about the same as it currently is at 41.25.

Error (senior moment)
 

****

The aural carrier over the air is HIGHER than the video carrier. The local oscillator is higher than both, so the IF aural carrier is LOWER than the IF video carrier.

Yep, dumb screwup on my part. Misfiring neurons, time to hang it up.

Doesn't the OP own a B&K!1077PAL generator....if it has a sound output like the American models and if it is the end to 5.5MHz the OP theoretically could inject at the input of the first sound IF, mark the position of the sound IF slugs and try to tune the audio being injected.

The first thing I would do though is see if the UHF tuner on the TV can even tune the transmitter. You should see unsynced video on screen (I have a world standard VCR and once played with its output standards while feeding a NTSC TV and that was what I got)...If that works check if you can retune the sound IF...you may have to or want to retune the video IF too. The video IF does provide some sound carrier amplification and the rolloff to prevent adjacent channel interference in an NTSC set may need to be readjusted to prevent blocking the sound from passing through the IF....Also PAL has more video carrier bandwith so the more of that you can make the IF pass the better tho picture will be.

These types of conversions were common around American overseas bases during the cold war and there is info out there on making the conversion both ways.

Thanks Tom.

I do have a 1077PAL, never got it working reliably and bought a newer solid-state Philips because I needed a known good signal source. So it's on the "to do" pile.

The TV definitely tunes the UHF channel no problem, and the picture will even almost sync up: Horizontal yes, Vertical needs a little more range on the V hold control. I'll do some more searching to see if I can find one of those guides you're talking about...the only one I have at the moment is all in German and designed for Philips sets going to NTSC, and to be honest it's mostly baffling.