24-10-2021, 11:15 AM
Long post
I am got to a major stumbling block in my head and wondering what to do next, generally down to my lack of knowledge, so I hope to put this into words that someone may point me in a way forward, or give suggestions.
The plan was/is to build a NBTV monitor and use valves as much as possible. From what I can gather a popular specification is 32 lines per frame and a frame rate of 12.5 frames per second, (two more lines than the Baird standard). Scanning goes from bottom right to top left with a 3:2 vertical format.
It is a learning exercise in magnetic scanning for me, trying to make the scan coils and understand it all.
With the NBTV specs I have found things can be a bit vague with synch pulses being anywhere between 0.1 and 0.25 mS
So I have worked on the following:-
Total line time = 2.5mS
Line sweep time = 2.3 mS
Line flyback time = 0.2 mS
Total frame time 80 = mS
Frame sweep time = 79.75 mS
Frame flyback time = 0.25 mS
I have initially wound a scan coil of 1200 turns, two 600 coils in series. Using DC through the coil it takes approx 102mA to deflect from the centre of the CRT to 50mm at the edge, therefore the peak to peak current required would be approx 204 mA to get from one side of the screen to the other, giving Ampere-Turns of 244.8.
Making my own scan coils I do not have access to the iron formers so end up with physically larger coils and higher resistance which may be an issue.
I did a quick spreadsheet with the coil I have and a few potential ones I could wind, basically trying thinner wire, I could also go thicker wire with less turns.
scancoilprint.pdf (Size: 21.55 KB / Downloads: 2)
This was just to figure out if the coils acted as resistances or inductances and at frame rates it is basically a pure resistance, the difference is not so defined at line rates (400Hz).
Now to my issue, I can see why the designs I have seen to date have all been transistorized, using split power rails and push-pull transistor scan drivers. You can cross from a negative supply and then crossover to positive supply to get the full deflection. But, on a single say 250 Volt supply rail at these frequencies I am wondering if the flyback period could provide the negative voltages in valve sets to drive the spot to the opposite side of the CRT and how i could do it.
For the line, I am am considering winding/buying a single ended audio transformer capable of say 5 watts and driving this with a sawtooth wave form and seeing if the flyback emf would be enough to send the spot to the other side of the tube. The transformer should be OK with the frequencies of 400 and 4.4KHz, so basically building an audio amp driven by a Miller-Transitron possibly with a bit of negative feedback?
But at frame frequencies, I feel unsure of what to try, believe it would need to be a big transformer with lots of Henries to get flyback to the other side of the screen, one possible saving grace will be the frame width would be less than the line length, if the line is 100mm in length, the frame should be 66mm across, but would also need less current to drive to one side in the first place, so less flyback?
I hope I have got my thoughts down and would appreciate knowing how anyone else would approach this.
Adrian
I am got to a major stumbling block in my head and wondering what to do next, generally down to my lack of knowledge, so I hope to put this into words that someone may point me in a way forward, or give suggestions.
The plan was/is to build a NBTV monitor and use valves as much as possible. From what I can gather a popular specification is 32 lines per frame and a frame rate of 12.5 frames per second, (two more lines than the Baird standard). Scanning goes from bottom right to top left with a 3:2 vertical format.
It is a learning exercise in magnetic scanning for me, trying to make the scan coils and understand it all.
With the NBTV specs I have found things can be a bit vague with synch pulses being anywhere between 0.1 and 0.25 mS
So I have worked on the following:-
Total line time = 2.5mS
Line sweep time = 2.3 mS
Line flyback time = 0.2 mS
Total frame time 80 = mS
Frame sweep time = 79.75 mS
Frame flyback time = 0.25 mS
I have initially wound a scan coil of 1200 turns, two 600 coils in series. Using DC through the coil it takes approx 102mA to deflect from the centre of the CRT to 50mm at the edge, therefore the peak to peak current required would be approx 204 mA to get from one side of the screen to the other, giving Ampere-Turns of 244.8.
Making my own scan coils I do not have access to the iron formers so end up with physically larger coils and higher resistance which may be an issue.
I did a quick spreadsheet with the coil I have and a few potential ones I could wind, basically trying thinner wire, I could also go thicker wire with less turns.
scancoilprint.pdf (Size: 21.55 KB / Downloads: 2)
This was just to figure out if the coils acted as resistances or inductances and at frame rates it is basically a pure resistance, the difference is not so defined at line rates (400Hz).
Now to my issue, I can see why the designs I have seen to date have all been transistorized, using split power rails and push-pull transistor scan drivers. You can cross from a negative supply and then crossover to positive supply to get the full deflection. But, on a single say 250 Volt supply rail at these frequencies I am wondering if the flyback period could provide the negative voltages in valve sets to drive the spot to the opposite side of the CRT and how i could do it.
For the line, I am am considering winding/buying a single ended audio transformer capable of say 5 watts and driving this with a sawtooth wave form and seeing if the flyback emf would be enough to send the spot to the other side of the tube. The transformer should be OK with the frequencies of 400 and 4.4KHz, so basically building an audio amp driven by a Miller-Transitron possibly with a bit of negative feedback?
But at frame frequencies, I feel unsure of what to try, believe it would need to be a big transformer with lots of Henries to get flyback to the other side of the screen, one possible saving grace will be the frame width would be less than the line length, if the line is 100mm in length, the frame should be 66mm across, but would also need less current to drive to one side in the first place, so less flyback?
I hope I have got my thoughts down and would appreciate knowing how anyone else would approach this.
Adrian
Learning as I go!
Youtube EF91 Valve
Youtube EF91 Valve
