Golborne Vintage Radio

Well it keeps me occupied trying to do this and suspect this will keep me occupied for some weeks.

I previously asked a question as to popular/common IF's used in TV, So now with this snippet of knowledge I am trying to build just the vision IF section, using my favourite valve the EF91.

I have a Band 1 & 3 tuner making its way to me, so the idea is to use this and have two EF91's followed by an EB91 (CV140 or CV4007) for the vision detector and if possible AGC, but that can wait.

I have started as I normally do with copper clad board with holes for three valves and also included positions initially for three small screening cans.  I use the copper clad rather than try and build an aluminium chassis, being quicker, and cheaper.  It also allows my typical dead bug building style where I will solder further small sections of copper clad on one side to act as component islands or HT rails etc.  Also being copper clad soldering down to the chassis is not an issue.
[attachment=20151][attachment=20150]
Hear is it at present, input coil going to grid of EF91, tuned anode coil and second winding to grid of 2nd EF91, at present output to test set is via a 0.01 uF  from the anode, so not matching but gives me an indication.  gain presently showing as 26dB, but selectivity is to wide and when trying to add a notch filter for sound it had to much of an effect n resonant frequencies.

The coil formers are these from a Pye Westminster stage, 5mm dia and about 16mm coil space, a bit small for my fingers but managing to wind something on them.
[attachment=20149]

I progress mainly by trial and error, I can not do theoretical designs that include coupling/mutual inductance etc, simple stuff I can get bye, but as I said I think this will be weeks.

Adrian

Great idea for valve circuit prototyping. Easy to work and a solderable earth plane.

Great project!

(17-11-2020, 07:08 AM)ppppenguin Wrote: [ -> ]Great idea for valve circuit prototyping. Easy to work and a solderable earth plane.

It's Manhatten Style prototyping - pretty standard technique for RF (and even AF).

I know and use it for solid state work, though often with colander ground plane boards rather than ordinary solid copper. Hadn't seen it used for valve work before.

(17-11-2020, 07:08 AM)ppppenguin Wrote: [ -> ]Great idea for valve circuit prototyping. Easy to work and a solderable earth plane.

+1

For anything higher than about 2 MHz a decent ground plane and rail decoupling is essential if you want things to be predictable.

Peter

The lower grade alternative I often used was pinboard (like veroboard without copper strips) with a heavy earth bus round the edges and sometimes built up as a grid of wires. Worked well in the 10MHz video bandwidths I often used. The stray capacitances to ground are lower using this technique than with Manhattan or indeed a proper PCB with ground plane. I've selectively removed ground plane around sensitive nodes on occasion. The non-inverting input of a high speed op-amp, notably current feedback types, is very sensitive to stray C.

None of this should be worrying Adrian who has chosen an intrinsically good and predictable way to build his prototypes.

I use the Manhatten method on many projects.
Always works well for me.
The minimalist superhet radio was made this way. It still performs well ten years later.

I wish you great success with the project.

It is a style I have used for many years.  Why someone has to give it a name I have no idea, makes someone feel good I guess.

Anyway the main issue I am having is over-coupling between stages.  I am winding coils as in the old Lynx/Argos designs etc.  With interstage coupling done with coils on the same former.  As the formers are relatively small/short there is not much distance to be had between windings so winding one winding of say 12 turns (anode) and then another at the top of 12 turns (grid) gives way to much coupling and the passband was something like 15 MHz, reducing the grid winding to 4 turns and placing this one over the anode coil at the earthy end has made things better, but with a loss in gain.    The next attempt will be a 12 turn Bifilar winding, in the hope I get the gain back, without the over coupling.

Some will know that I have done work on a sync separator, line and frame timebases.  I was only driving my HP scope to give a picture as a means to spur myself on, but in the end I had reasonable results: -
[attachment=20159]

I will at the end need to find coil formers and screening cans I can buy, the idea is still to use KiCad and make myself a PCB's for the project.

Adrian.

Some success today.

I made two Bifilar wound coils and it seemed to be a lot better, the first one is shown here, in great close up, and what a mess it is!
[attachment=20161]
But it works, the second one was a bit better to look at.  The board now has the input coil, with a 6K8 resistor across the tuned secondary, this feeds the grid of the first EF91, it also has a tuned anode coil again with a 6K8 resistor across the coil.  The secondary of this feeds the grid of the 2nd EF91.  Anode again tuned coil, but at present no resistor across it, the secondary of this going to a CV4007 diode to rectify the output.

My test set can tune filters, but has no means of showing a DC voltage  response like you could with a Wobbulator/Polyskop so before I connected up the rectifier valve I did a quick test to see how it was looking, unfortunately using a 10:1 scope probe for the input to the test set which has a 50 Ohm resistance, so levels mean nothing.
[attachment=20162]

During the day I measured the output from the Hedghog II ad measured them at -45 dBm into 50 Ohm!

So as a last test after connecting up the diode, detector I fed the IF strip with a -45 dBm, 90% A.M. modulated signal using a 20KHz tone and was able to recover a 3.5 Volt peak to peak waveform, off the detector.  That means if the tuner has gain then things should improve and I will have a large enough signal to feed the sync and video stages.

The next item to do on the list with this is add two more coils acting as traps for the sound carrier and see how effective and sharp I can make them.

If successful I will have to be on the lookout for more Westminster coil formers as I am struggling to find a shop that still sells anything new that would be suitable.

I am going to try for an absolute max of 2Mhz bandwidth on video, I do not thing even that would be required for a small DG7-5 tube, but would appreciate comments.

Adrian

Here's 2.5 Mhz on an MW 6-2. I think you'll have to go the extra half a meg. 

Peter

[attachment=20163]