Next the sound IF stage, plus!

[AdrianPH]

I thought it about time to get on making the sound IF and detector for my home made TV project.

Over the past few days I have slowly, and this time it is slow progress started to build up the board, copper clad again.  I received some IF cans from Ed which was much appreciated, but was wondering how to mount these things so i can use them.  Then a bit of a lightbulb moment and out with the CAD packages onto a CNC drill and I ended up with 30 4 mm holes in the board to start with and then with use of a draw punch (Q Max) took 3 holes to 16mm. 
    .

Today I have started with populating the board, input will be via co-ax so a single turn coupling at the earthy to the grid tuned coil.  Then a 1:1 tuned bifilar winding in the anode with the secondary ready to go to the grid of the next valve.  The response is quite wide, nowhere near the response of Amie's IF can testing, this could be down to the bifilar winding or the fact I am up at 38.15 MHz, who knows?
   

The plan is for 3 x cans, 2 x EF91's and an EB91 detector, very similar to the vision IF but no resistors across the cans.  The idea of bifilar was from the ARGOS article by Dr. Hugo Holden, and I find them easier to do, less coupling to figure out.

So another can to add and two more valves.  The plan at the moment is to use one of my milli-watt audio amps with 6021 valves as the volume control and audio output but that may well change.
If the AF stage works OK I have plans to build another one later for another project, 5 x cans at 30 MHz, 3 x EF91's, 1 x EF92 and EB91 as a high gain video stage with manual gain.

All fun.

Adrian

[Diabolical Artificer]

That's very nice fettling, accurate hole spacing, nice one.

Andy.

[Amie]

Don't know if I've understood correctly, but bifillar winding of primary and secondary together?

I would think if that were the case it would make them over couple.

Amie.

[Kalee20]

Agree - they'd be very tightly coupled, just about as good as it gets for any transformer.

What you'd have effectively is, a SINGLE tuned circuit, with tuning capacitance equal to, whatever's on the primary, in parallel with whatever's on the secondary.

The response would be a single-humped peak. No chance of a broad, almost-flat region in the middle as per Amie's partially-coupled axial inductors in the other thread.

But, for audio bandwidth around a centre frequency of 38MHz, that may well be good enough. It's certainly easier than audio bandwidth around a centre frequency of 0.47MHz.

[AdrianPH]

I doubt I could do it at 470KHz but as mentioned at 38 MHz quite easy when only using 8 turns per winding, the primaries have a 15pF cap across them plus any coupled, valve capacitance etc. Then slug tuned.
So I can not provide any passband tuning in the same can, but as I did on the vision side I could tune the input and output of each stage to provide a bit of stagger tuning when combined with a bit of added resistance to dampen the Q. But as I am only after audio here, everything will be peaked for the centre frequency. As to if it will all work, I suppose it will in a fashion. How good it works and if it suffers with video or not I will wait and see.

Adrian

[Ed_Dinning]

Hi Adrian, neat layout. Don.t forget there are other valves in the EF91 range that are vary-mu if you want a measure of gain control ot AVC

Ed

[AdrianPH]

Hello Ed,

I think that within the TV set I can continue to use the AVC from the vision stage, this feeds back to the tuner to reduce the gain of the rf and if amp. So once I set a suitable level from the AF demod to the amp I hope that side will be taken care of. If all goes well I may add a SE output stage, at a later point but for now the small amps I have will do.

As mentioned in the first post, if it all works I could well later build a 30 MHz straight IF and video detector, this would have an EF92 in it first stage and a manual gain control the rest being standard EF91's. I have a wish to try and build my own simple RADAR A scope, long term project, if you will. If I go back into the 80's I built and still have a PW Exe wideband FM 10 GHz transceiver. From the likes of Winterhill to another hilltop site somewhere in Wales I was able to have conversations I believe over 60 miles fully quieting. So in my head I would like to build a small RADAR set that would do if possible 2 or 3 miles. I have a couple of Klystrons, some uWave diodes, some wave guide and I hope, plenty of time.

Adrian

[AdrianPH]

Another afternoon/evening playing, now have two stages of amp and the diode detector.  I tried 2 x EF91 and in the end changed the second valve to a EF92, this dropped the gain a bit as I run the grids negative wrt cathodes and with the variable mu EF92 it meant a bit less gain.
The rather scruffy underside : -
   
And the better looking topside : -
   
Then out with the radio test set and a few measurements, driving with a -50 dBm output although the display shows as -55 dBm with just a piece of co-ax inserted :-
   
Then with the DUT in place.  On the last coil going to the diode detector, as well as it being a bifilar winding, I also included a 1 turn coil, the idea to allow co-ax to be connected for this test, I guess it would still load the coil, but not as bad as going across the output to the detector.: -
   
This seems to indicate approx 44 dB gain.
   
I changed the Y axis to be 2 dBm per vertical division to give a better indication of bandwidth.
The last picture shows demod, with a -50dBm signal going in 50% A.M. mod, 0.7% distortion: -
   
I am wondering if I will actually have too much gain, time will tell.  I still need to better the PSU before this can be fitted to the TV unit.

Adrian

[Amie]

Hello Adrian.

Good job there!

Are you using an EB91 for detector?

I'm not well up on telly circuits, did the 405 system split the vision and sound of after the tuner? If so, enough gain at those frequencies from two valves is pretty good. I've been trying the EF80's/85's at about 456kHz and it takes off! Probably just need better RF hygiene, its looking like the stage could be run from a fairly low HT supply, loads of gain with its @ 10kHz BW....

Regards
Amie

[AdrianPH]

Yes the vision and audio IF's were generally split after the tuner stage, with the vision IF having traps to remove the sound carrier and the sound being taken from one of the traps.  That is what I will be trying soon.

If you look at most domestic receivers many will only have one stage of gain at IF using double tuned circuits as you are doing.  As the IF frequencies go higher then yes the more gain seems to be required.

When I had 2 x EF81's in there I had something 60 dB plus gain and yes if not careful I could get the stage to oscilate, then I actually got around to connecting the top of the board to the bottom of the board to make a better ground plane and it cured it, but it also showed how much noise was being generated by the switch mode power supply I am using as a temp.  No wonder the HF bands have become unusable.

The detector is a CV4007 which is just a smaller version of the EB91.

Time to play again.

Adrian