Winter project.

[Murphyv310]

   
   
       
Hi.
I have now set up the camera on a tripod and have been trying some pictures on manual. The pictures are much sharper and the line structure is very clear on this picture. There is a little distortion on the picture though as the shed is very cold and a little hum is seen, its a slight leak in the line osc valve base and a job to be done, its fine once the set is fully warmed up though.
I've also uploaded a picture of the after glow from the CRT with no signal.

[Murphyv310]

Hi.
Next thing to do is to fire up the Decca 1000 and do some screen shots from it. To be fair it's impossible to recreate the flicker but with careful positioning and manually adjusting focus and shutter speed I should be able to demonstrate the picture on 240 lines, the line TB won't go to below 240 without modification, perhaps just a resistor change but I will leave the set as is.
I'm also going to make up a sync generator for the mechanical standards over the next week or two.

[Murphyv310]

Hi.
Replaced the valve base for the line osc and sync separator today and rewired it for a 30FL1, nice tight sync and more line drive too.
Also did the same to the video output and used another 30FL1 as a pentode amp and triode cathode follower, far more CRT drive now than with the EF91.

[peter scott]

Much nicer photos!

[Murphyv310]

  TV22mod2.pdf (Size: 693.91 KB / Downloads: 13) http://golbornevintageradio.co.uk/forum/...?aid=16847

Hi.
I have accidentally posted on the wrong thread.
OK so now I have drawn up two circuits, one for the Video modification and now one of the line timebase.
I am using a Thorn 900 lopt with the full EHT overwind, this has to be the best lopt for what we are doing, using a 950 or 1400 type will still need a doubler as the resultant EHT would be too low running the timebase on starvation mode. The snag is the red topped 900 lopt's are quite rare. Don't worry though if you want to try this use the doubler.
On the diagram you'll see T1. It's the blocking oscillator transformer, virtually any will work but values may need changing, I'm using one from an early Ferranti it's small and neat so can be fitted into the bush chassis with ease. Smoothing needs to be upped a bit as the 25 hz frame can breath a bit so an extra 100uf on the main HT line is added and a 22uf on the 105 volt line to line output is added. The screen grid feed to the PL36 is fed from the main HT line via a 5K this works the best, there is also a choke fitted to the PY81 cathode and it helps to suppress some of the line stage ringing, it is one of RS's 1 amp chokes but does the trick, no timebase tuning is required as it is running on very slow speeds and low HT so the ramp up time can not really be damped.
I also will later upload the circuit of the frame timebase.

Any errors I may have missed please let me know.

[Murphyv310]

Hi.
Well I have decided to revert to 405 lines only. I'm also going fit a CRM93 instead of the little radar tube. Using a CRM93 will require a good 9kv that the jellypot can supply with ease. As the CRM93 is aluminized the picture should be brighter than a TV22 was with a mullard tube. Having a stabilised 9kv and more CRT drive I'm hoping results could be very good indeed.

[Murphyv310]

Hi.
Well I have now made the set 405 lines only.
I'm having some issues with left hand foldover and I think it is the blocking oscillator that is misbehaving. The flyback is a little too long, I've tried changing the C/R ratio but no real change is noted. I'll experiment more when I have time.    

[Murphyv310]

Hi.
Well the issue is not with the oscillator. Flyback time is as it should be when driving the line output.
The issue is on the output stage. From what I can deduce it is a mismatch between the scan coils and the Jellypot, fine on the lower standards but the faster the line rate the worse the problem.
I introduced a bit of damping across the scan coils, this reduced the width but the foldover disappeared, increasing the HT full width returned and the picture was perfect although the damping resistor was getting really hot.
I then thought of adding flywheel sync and adding in some DC component to the discriminator for phasing, this works too as I cobbled up the Thorn 1500 circuit that uses a transistor as a DC amp, on 405 I can phase the scan to give a full sized picture with no cutoff either side. This is the way to go to keep dissipation at a minimum and have a very good result.
I'll build the discriminator and DC amp on some veroboard.

[Murphyv310]

Hi.
Today I added in a flywheel discriminator, I used the Thorn 1500 circuit with capacitor values compensated for the lower frequency 405 line timebase. The circuit was built on a piece of matrix board and has been tested with flying leads. A small bit of imbalance has been added by altering one of the shunt resistor values across the lower diode. this has given me near enough perfect picture with no fold over on either side.
Some hooking is seen on the top of the picture on the castellations but not on picture info.
Next is the CRM93 that I'll need to collect in a month or so.
I think I'll call this a BushTV22GTI
   
   
   

[Geordie McBoyne]

there's Hi Trevor,
That an absolutely superb result. The BRC 1500 line flywheel discriminator is a neat circuit, no need for such things as a balanced discriminator which employs the complication of a phase splitter, be it a valve or centre tapped transformer winding.
Actually the 1500 line flywheel discriminator circuit was first used in the Invicta model 136 of 1957. The circuit of the Invicta 136 can be found in the 1957/58 R & T servicing book.

I'm considering adding a flywheel line sync circuit to the Marconi VT161. The metal chassis is pre-drilled to accept extra valves and in fact 625 line export versions of this special BRC chassis employed a flywheel line sync circuit.

Geordie McBoyne.