819line video from a PC without software tools

[Erik-Jan]

Hello,
Let me first introduce myself;
My name is Erik, living in Eindhoven, the city of Philips.
And active in restoring vintage TV`s from the '40 en 50.
I've been a frequent reader of this great forum, and now finally decided to join this comunity.
English is not my native language, so I hope I can be forgiven for some language errors.

The first project i want to discuss is generating 819 line video using a standard windows-PC VGA output.
Of course this is done before, but is always required manual changing of the video settings, using a specialized program.

My idea is to use the VGA EDID data to force the video-card into the desired resolution.
EDID is the handshake a PC en monitor do when connected together,
In it simplest form it tells the name of the monitor and the supported resolution(s) of the monitor.

The goal is to come up with a plug-and-play solution which will work on as many videos adapters as possible.
So i tinkered a simple interface together.
The EDID data is stored in a 24CL02 eeprom, and contains the monitor name (819L to VGA) and only one allowed resolution (1200x776).
(The binary file for the EEProm can be created using CRU 1.5.2)

I tried this setup on three different PCs (coincidentally all with a HP video adapter).
The results are mixed, on all PC's the interface is automatically recognised, but only two PC's start outputting video.

I think a solution in getting this to work the best lies in the combination of choosing the right pixel clock speed and horizontal/vertical pixel count.

This brings me to my first question, what pixel clock to choose which will work on most video cards?

Greetings Erik.

[marceljack]

Hi Erik-Jan,
Welcome to the forum.
Your idea is very interesting however I wonder if it can work with all graphics cards or chips.
A long time ago (in the early 90's) I reprogrammed VGA cards in order to make them work in 640 x 480 on the SCART connector of a 625 lines TV.
Different graphic chips used different ways to be put in interlaced mode, so I had to write different routines for different chips.
I did that under DOS by means of a so-called TSR routine (Terminate and Stay Resident).
I dont think EDID existed at that time.
For your information, I was working at Philips Semiconductors France, so I went relatively often to Eindhoven.
Sorry not to be more helpful !

Greetings

Hervé

[Maarten]

Welcome, fellow Dutchman!

While HP writes some of the specifications, multiple factors such as the video chip, the BIOS and the drivers determine whether the EDID method will work. I think it would be useful to document at least the actual video chip as the other factors are often based on reference designs for this chip.

In your picture, I see an (onboard) Intel HD Graphics 520 videochip.

@marceljack: I think you just missed DDC and later EDID by a few years, not too much in any case. Philips monitors made in the mid 1990's already supported at least DDC.

[Erik-Jan]

Thank you

EDID has indeed known several versions. The interface now works according to the 2.1 standard.
I have now tested with 4 different laptop PCs, all with an onboard HP video card
All PCs recognize the interface, that is positive.
Two PC`s start indeed outputting video in the correct resolution.
One with a Intel HD Graphics 520 chip and one with a HP GMA4500 chip.

The third and fourth PC are not changing the output resolution.

Until now the results are pretty much hit or miss

I think there is some "sweet spot" in the relation between line/frame timing and the chosen pixel-clock.

Greetings Erik