Hi,
given that my TV does not feature a SCART input, nor an S-Video input (but C-Video and component video), I came up with a design that uses many resistors and two 74HC inverter ICs to generate a YPbPr component video signal.
All required components can currently be sourced in Germany from segor.de. Pretty expensive, but hassle-free for private (i.e. non-business) customers.
Orderd all parts and am going to try to assemble this on a eurocard / matrix board PCB sometime in november.
This is the first uzebox I'm trying to build. Maybe it is too naive to assume that video output will simply work when done like this? Any comments?
Partial/draft schematic attached. Does omit most of the stuff that can be copied over from the original uzebox F5 design.
The DAC resistor network I simulated using the octave script dac.m, attached below. Black is at about 0.3V, 0V, 0V, SYNC at 0V, 0V, 0V. White at about 1V, 0.7V, 0.7V.
Will first try assembling with just the Y output. In this mode it should work similar to the "Uzebox minimal" black/white output.
cheers,
Dave
Uzebox component video (YPbPr) video output
Uzebox component video (YPbPr) video output
Last edited by dave on Sun Aug 21, 2022 8:04 pm, edited 4 times in total.
Re: Uzebox compenent video (YPbPr) video output
Woah!
I don't know much about generating Component video, but I definitely want to see how this turns out.
If it works well, I might have to build one myself. Good luck!
I don't know much about generating Component video, but I definitely want to see how this turns out.
If it works well, I might have to build one myself. Good luck!
-
- Posts: 1480
- Joined: Mon Feb 11, 2013 8:08 am
- Location: Brisbane, Australia
Re: Uzebox component video (YPbPr) video output
Do you want NOT gates or do you want diodes for the "Blue-Luma" and "Red-Luma" channels?
Minus doesn't mean the same as plus-the-inverse does it?
Minus doesn't mean the same as plus-the-inverse does it?
Re: Uzebox component video (YPbPr) video output
Can't currently imagine how one would you use diodes for negation?! The more usual approach is to implement the RGB->YPbPr mixing matrix using Op-Amps, see for example LT6550 datasheet page 14:CunningFellow wrote: ↑Sat Oct 24, 2020 8:38 pm Do you want NOT gates or do you want diodes for the "Blue-Luma" and "Red-Luma" channels?
Minus doesn't mean the same as plus-the-inverse does it?
https://www.analog.com/media/en/technic ... 5501fa.pdf
But video op-amps are difficult to source for me, and maybe also quite expensive. As the uzebox gives us access to binary versions of RGB, a binary inverter can do the inversion for us (INVERT(x) = 2^n-1-x). This results in some offset (2^n-1), but Pb,Pr need to be offset anyway: AFAICS voltage range is 0..0.7V, and not -0.35V when Y is large and Pb,Pr are small. BTW the black&white uzebox-minimal design uses basically the same trick to subtract the SYNC signal from the Y signal (where SYNC is already output inverted by the Atmega IC).
Re: Uzebox component video (YPbPr) video output
Finally found some time to fully assemble the Y/Pb/Pr circuit.
Seems to generate colors almost perfectly. Here is a screenshot taken with an action-cam. It's a fish-eye lens, badly distorted, however action-cam seems to give a better reproduction of colors than photos taken with my tablet:
Seems to generate colors almost perfectly. Here is a screenshot taken with an action-cam. It's a fish-eye lens, badly distorted, however action-cam seems to give a better reproduction of colors than photos taken with my tablet:
Re: Uzebox component video (YPbPr) video output
Looks sharp, this might be a nice alternative for many people
-
- Posts: 1480
- Joined: Mon Feb 11, 2013 8:08 am
- Location: Brisbane, Australia
Re: Uzebox component video (YPbPr) video output
Looks awesome. I might knock one up myself to have a play.
Re: Uzebox component video (YPbPr) video output
Amazing quality, looks as good as S-VIDEO but without the AD725! Awesome work, really love it!