More CRT madness

[dominicbeesley]

Hello all,

A quick update, after much help from people in other forums, helping me to understand transformer theory, flybacks and audio amplifiers I've made a little progress.

The EHT generator is made up onto a little PCB of its own with the LOPT and associated parts, mostly robbed from the original TV. Additions are the FET output transistor as this was easier to understand, the grey coil which mimics the original scan coils and the little daughter board which houses a 555 timer to drive it all. I'll probably improve the 555 timer as it drifts about a fair bit.

The main rats nest circuit comprises two 555 timers to make the ramp waveforms for 12.5 and 400Hz, an op-amp buffer and the two FET amplifiers. These drive two transformers, the small yellow one for the line (hand wound on a small DC choke core) and the large frame one (hand wound on a medium sized mains power transformer with air-gap).

Both line and frame are being driven by single-ended FET amplifiers with a bit of distortion feedback to make the scan more linear. This is probably the most difficult and daftest way to do this with solid-state stuff but I thought I'd do it this way as this is most like a valved TV of the 1950s and I more or less understand them!

The picture of it running shows the obligatory test card, this shows about half the height I require but after a bit of fiddling this morning I can get the raster to more or less fill the screen with reasonable linearity. The CRT is being driven with all three cathodes strapped together, hence the pink hue on darker parts of the picture. Video is coming from a Klaas sync seperator and video amplifier. At present there is no frame/line sync these being tweaked by hand...the patterning on the screen is 15kHz pick-up from the EHT generator - this gets into the frame, line and video stages as there are a number of HF spikes. I'll do a bit more taming of this and possibly put the EHT generator in a metal box if all else fails....at least that might stop me constantly brushing my hand against it!

I need to do more work on the frame stage, after spot-wobble is added any non-linearity shows up as gaps between or overlapping of the lines

Dom

[AncientBrit]

Dom,

That's a major achievement.
I know. I've been there with a monochrome monitor conversion.
The hardest part is getting sufficient (linear) scans at NBTV rates.
Well done,

Cheers,

Graham

[dominicbeesley]

Thanks Graham,

it has been a real stuggle to do. But I think a good learning experience, filling gaps in my knowledge about transformers and inductors.

Will be back again soon with spot wobble and then colour!

Dom

[dominicbeesley]

Another few hours of hammering away this evening have resulted in a full-screen raster. The trick was to get a bigger transformer. This one's a pretty heavy 415V/16A autotransformer. I tapped out all the laminations and turned all the E's and I's the same way round. I'm quite happy with how the linearity has worked out.

Hopefully with some better FETs I should get a quicker flyback too! These one's have a low reverse breakdown of 55V but I need something more like afew hundred!

Cheers

Dom

[Harry Dalek]

Hi dom i have been watching and following your project with great interest !
i can't give any advice as you are the MASTER and i am the grass hopper !

Wonderful work .

[AncientBrit]

Dom,

One of the problems with CRT scanned NBTV is that unlike fast scan TV formats the frame blanking/flyback period is not scaled to that of frame rate.

So you are left trying to accomodate frame flyback within line sync period, say 170uS.

It's just possible but you have to allow for a very large voltage pulse in addition to the sawtooth across the frame coils due to the extreme dv/dt.

Keep up the good work,

Regards,

Graham

[dominicbeesley]

Thanks Harry, not sure I'm a master yet!

Hi Graham,

It's not too bad for me I want the frame flyback to happen in 1/64th or preferably 1/128th of a second...in my colour formats that last 1/2 of the last line carries no picture info. So it's not too bad. However even with that I'm expecting a few hundred volts on the spikes.

The current FETS have a reverse breakdown of only 55 volts which stretches out the flybacks. I've just ordered a load of MOSFETs that should be good for 800V and will see if they improve matters.

Cheers

Dom