Golborne Vintage Radio

Great stuff there Frank, there's no doubt that will work. This also means that the Aurora SCRF should work if a suitable xtal is fitted.

Frank, another thought. If you decide to work this up into a full 405 to 625 converter you will find that the interpolator has be AFTER the time redistribution store. For down conversion the interoplator is BEFORE time redistribution.

My converter has a framestore so these restrictions don't apply. I store the whole input signal and interpolate after the framestore by accessing multiple lines at the same time.

(06-08-2019, 01:04 PM)Mike Watterson Wrote: [ -> ]I wonder what Frank will use this feature for?

Hi Mike
I am afraid I have no use for it at present anyway. It's just idle curiosity


Hi Jeffrey
The time redistribution would need to take a different form than I am using for down converting. I think something along the lines of a FIFO with the output taped at 1 line intervals would probably work.

But there appears to be a fly in the ointment. I tried to sync it using the Field sync from the TVP5150. It wouldn't sync. It just caused the picture to jump around the place. Occasionally it would appear to sync but just for one field and in the wrong place.

So it appears that the TVP5150 has trouble extracting the field syncs.
Still, a separate sync separator could be used. I think a LM1881 will work on 405 lines.

Frank

Tthe TVP5150 has a separate field sync output on pin 24. Is this conencted to the FPGA? Also a V blanking output on pin 27 and odd/even on pin 23. If there are spare pins on the FPGA these could be hooked up.

I reckon I'll be doing something like this with my SAA7118. At least my sync output pins are already connected to the FPGA though I've never yet used them.

Hi Jeffrey
They are not connected to the FPGA. I will enable them and try scoping them. This should give an indication if they are worth connecting.

Frank

Way back at the start of your work on 625 to 405 conversion you were connecting the sync outputs of your decoder to the logic because you were worried about implementing a digital sync separator. In the end the digital sync sep was no problem.

When I designed the hardware I'm using it was many years(15?, 20?) ago and for a professional client. I planned to use a digital sync separator but I had some spare connectivity available. So I thought it would do no harm if I hooked up a few of the odd sync etc pins on the SAA7118 decoder to the FPGA. You never know, I thought, there may be some function I'll need for a different project on what was designed as very flexible hardware. I have never used them before but now might be the time I need them.

Hi Jeffrey
I have scoped those pins but no joy.
The H sync is nice and clean but the odd/even and V sync are just random.
Pictures below from a DSO odd/even, V sync and H sync.

Frank

Looking at the TVP5150 data sheet a few thoughts:

Make sure you force it into monochorme.
Possibly force TV/VCR mode to either VCR or TV. Try both
Look at VBKO and other controls in register 03h

Fig 2-8, Config shared pins is worth studying

Bit of a long shot, perhaps V sync info can be force to pass by setting register 07h

Address 0Dh. Consider bypassing much processing. Then doing sync sep on digital data in the FPGA. May also need to paly with cropping at 11h etc

Register 0Fh. Try various settings to attempt to get syncs out.

Address 15h may be able to get something useful in the 656 data stream

If trying to get sense out a vertical blnaking then register 18hand 19h becomes important

May need to force a particualr input standard in register 28h. Probably PAL B etc

Reading registers 84h, 85h may give some useful info

If running autoswitch standards then look at register 8Ch to see what has been detected

Might be able to get something useful out of the programmable interrupts.

Correction: A lot of thoughts and plenty to play with. the answer is likely to be somewhere in that lot.

Had a play with some of the registers. No joy yet but more to try.

Frank

Spent some time changing the settings of the TVP5150 but no joy.

I have bodged in a LM1881 and used the odd/even output as a frame sync. The result is the first picture below. The horizontal displacement is not a problem as it can be easily shifted. There is some horizontal jitter. The embedded syncs looked to be a consistent distance from the active video. So I used the SAV signal in conjunction with the odd/even signal from the LM1881 to produce the frame sync. But the jitter is still there.

The second photo is the result of combining the embedded odd/even signal and SAV signal. The photo looks a lot better that it is. There is a lot of vertical jitter as well as the horizontal jitter.

The third photo is of the video clock from the TVP5150 which looks jittery.

As I type this I am begging to realise that I have probably contributed to the horizontal jitter. As the clock on the input side is from a different source to the clock on the output. The video just passes straight across which is not good!
Next I will clock the output from the input clock and see if that helps the jitter.

Frank