Hedghog Mk II

[ppppenguin]

I've started this thread about Hedghog Mk II. Link back to original HH thread:
https://golbornevintageradio.co.uk/forum...p?tid=6193&pid=81223#pid81223

I've put it here in "Television" because I think it belongs here rather more than in "Homebrew".

When I visited Frank in Ireland a couple of weeks ago I saw the Hedghog MkII in action. The basic performance is near enough the same as the original HH though the 10 bit DAC is a theoretical imporvement. The most important change is a new FPGA. The old Cyclone II series used in the original HH is near obsolete and the build quality of the FPGA module has been variable, often poor.

The Cyclone IV series used on HH2 is still low cost. It's mounted directly on the PCB, not on a sub-module.

http://www.electronics.frankcuffe.ovh/hedghog-ii

[`FRANK.C]

Thanks for splitting.
I have added a link on the Hedghog II web page to this thread.

Frank

[ppppenguin]

This is me doing idle speculation in the fairly certain knowledge that I'm unlikely to do the work to implement it.

If Frank wants to do a really radical update to Hedghog then adding a framestore would be possible. It would make outputs like 240/25p feasible. Would need an FPGA with more pins. My own framestore based experiments used a couple of original SDRAM chips which need relatively complex control logic. It all works well though I've not done 240/25p.

The other radical thing would be a new modulator. The days of the existing modulator chips are surely numbered. Possibly using direct digital synthesis which Frank himself has explored. This needs a higher performance FPGA and a good high speed DAC. Or look at the modulator work Semir has done in Germany. Or consider DDS at a low-ish frequency (easy on the FPGA and DAC) and an upconverter with frequency synthesis.

[`FRANK.C]

Hi Jeffrey
On the PCB that I have done the DDS experiments on I also mounted two AL422 frame stores. I had them leftover from the earlier days of Hedghog.
My main reason for including a frame store was to see what field interpolation was like on a static picture(test card)
I did some experimenting with it. Using the present 6 line single field interpolater as a benchmark.
I first tried using 6 lines from one field and a 3 lines from the other the resultant picture was better that the 3 line on its own but worse than the 6 line.
I then calculated coefficients for a 11 line interpolater. 6 lines from one field 5 from the other.  The result was not impressive, with no discernible difference between the 11 line and 6 line interpolaters.  Since doing the experiments one thought that occurred to me was that I only ever tried one source of video. A free view box with the test card stored on a stick.  It might be worth trying a different source some day when I get back to it.

Frank

[ppppenguin]

From experience I can assure you that interpolating between spatially adjacent lines on odd and even field works fine on static images. And looks horrible as soon as anything moves. The Pineapple converter used framestores. It had a switch to choose frame or field based interpolation. Interpolating between lines on different fields was utterly hopeless on anything that moved.

I've done 625/50 <> 525/60 conversion using 4 fields and 4 lines. Roughly equivalent to the BBC/McMichael ACE.  It can work very well, but nothing like as good as modern motion compensated conversion. Good de-interlacing is also hard. Look at some LCD TVs to see the horrible mess they make of slow movement.

I don't know if the AL422 is still available. It's a good chip and very easy to use - I've used it professionally. But it lacks flexibility in a standards converter. In my experimental 625>405 converter I used SDRAM. A pair of these giving a 32 bit data bus. Using some fairly clever logic it looks like there's a video input and 4 separate video outputs. All of which have more or less free addressing. This allows me to write 625 video and read 4 different lines of 405 for the interpolator. No separate line stores. In fact I multiplexed the 4 reads on to a single bus running at 4x the 405 clock speed. This also allowed me to use only 1 multiplier to make a 4 line interpolator. A big advantage since I was using an ancient Xilinx FPGA (Spartan 2, XC2S200) which didn't have dedicated multipliers.

[Semir_DE]

I don't know if the AL422 is still available.

 yes it is still available from China. I bought a bunch last year in Autumn for roughly 1€ each. There are still quite few suppliers of this chip on AE I just checked. There is also a field store D42280 with the same pinout as the AL422.

[`FRANK.C]

RS still stocks them as well.

Frank

[Semir_DE]

Hi Frank, hi Jeffrey,

this looks like an interesting project! I just thought about the modulator and had this idea:

A - Generate a modulated vision signal in the IF frequency domain i.e. around 39-45MHz directly in the FPGA
B - Pass this through a vestigial sideband filter.
C - Generate a selectable frequency with the FPGA to be mixed with the VSB signal from step B. This would determine the output channel.
D - Generate a channel dependent DC output voltage. This would be used to control a tracking filter that is needed to filter out the mirror image frequency from step C leaving only the desired VSB signal.

As a VSB filter an SAW filter could be used. Even if this were a filter intended for a different TV System it would be possible to use it as long as the vision carrier in the IF domain was placed in a position to allow only 1.25MHz of the undesired Sideband to pass through. 

Just an idea although this would involve some RF circuitry...

Cheers

Semir

[ppppenguin]

VSB modulation can be done directly in digits using a variant of the Weaver or "3rd method" SSB generator. A quick run through the equations suggests it's possible. Conceptually you can think of splitting the video band with a FIR. Then use ordinary AM to modulate the low frequencies and SSB for th highs. Since it's all digits the delays are 100% predictable and it's easy to recombine the 2 paths.

For the AL422 what I don't know is whether it's still current manufacture. Old parts can take a long time to die. I've also use the 42280 in the past. It's earlier than the AL422.

[ppppenguin]

I think Terry proposed the idea of a SAW filter in another thread but it's more elegant to do the job in digits.

For easy filtering the clock frequency in a digital modulator needs to be at least 4x the highest wanted output. Would need to look at DACs etc but I susepct the sweet spot for an IF modulator lies in the 20Mhz to 30MHz range, with a clock around 100MHz. I think Frank has been experimenting with direct modulation fo channel B1 (45MHz vision) and a clock around 180MHz. This is entirely feasible with modern DACs. Pushing up Band I could get difficult and DDS at Band III just looks too hard with a need for 700MHz+ clocks. It's feasible but not something you'd want to DIY.