Nice find ! Theoretically a MDA monitor that can do 720x350 pixels could work as the output is 512x256 pixels. A big unknown is the Horizontal sync signal, would assume it be about 15.7Khz but in the manuals I did not see a reference.
Another thing as you mentioned is no Intensity and the page you got the picture from states the video is in a coax, which is odd because TTL levels do not need that but it could be to keep the signals shielded...
Another thing interesting is that the full pin-out of that slot has R, G, B connections too (and in the manual it describes using a color monitor of either 4 or 8 color) which may line up with the early MDA pinouts also as the very early ones did have those signal pins but were almost never used....
some info if others are interested: http://www.z80ne.com/m20/index.php and the FTP with docs/etc: ftp://ftp.groessler.org/pub/chris/olivetti_m20
I leave the decoding the vid signals to others with more grey matter ;p

My M20 CRT controller is HD46505 by Hitachi. I know that it's not the complete video circutry, but I guess it can be a sign that it's somewhat compatible with IBM's MDA standard.

That z80ne.com site states that M20 uses Hsync not compatible with CGA, so it's not 15,7KHz. But sadly they do not state what the HSync is, but since MDA has different HSync that CGA I guess there is still a chance...

Maybe a NEC Multisync might work ?

HSync is 18,43kHz so there is a chance for MDA monitor

VSync is 70Hz and that's a problem

Bernkastel7734 wrote on 2022-10-22, 12:17:

VSync is 70Hz and that's a problem

What is the pixel clock? vsync and hsync matter less than the pixel clock (even though you might get a rolling image it at least won’t explode)

Frankly speaking I have no idea how to check pixel clock. But when connected to CRT it just looks no scroll in both axis, and nothing expolode.

I've found forum in Italian where (if I understand correctly) someone was able to use MDA monitor and adjust H Sync and V Sync:
https://it.comp.retrocomputing.narkive.com/mo … eo-olivetti-m20

I'm thinking that Sync is not the only problem, but that something is broken.
I've replaced potentiometer on VHold setting and as the screen isn't rolling upside down so I bet it's synced.

I'm not sure if it's MDA compatible or not, but I am sure it is not what a healty video output should look like.

rmay635703 wrote on 2022-10-22, 12:50:

What is the pixel clock? vsync and hsync matter less than the pixel clock (even though you might get a rolling image it at least won’t explode)

The pixel clock doesn't matter at all, it just determines how wide pixels are. Of course, the pixel clock combined with the number of pixels per line (including the blanking time) determine the horizontal frequency. This is the most important timing parameter , because the high voltage generation circuit and the horizontal deflection circuit are designed around this frequency. Especially in fixed-frequency monitors, the design is often resonant, and even worse: There are some monitors that might break if the horizontal frequency is too low, because the deflection circuit is overdriven in that case. Incidentally, the original IBM MDA monitor is one of these rare beasts. That's why connecting a MDA monitor to an unknown signal that might be at CGA frequencies is a dangerous suggestion.

vsync mismatch also is harmless, it will just cause a rolling image, just as you describe.

Bernkastel7734 wrote on 2022-10-25, 17:43:

I'm not sure if it's MDA compatible or not, but I am sure it is not what a healty video output should look like.

What signal are you scoping?

mkarcher wrote on 2022-10-25, 18:05:

What signal are you scoping?

B/W video

H Sync is 18,73 kHz and V Sync is 68 Hz.
The monitor I'm using got a fluorescent lamp on the CRT neck connection board and I guess it's some kind of security. Also that monitor got a H Hold potentiometer (sadly not soldered, just a place for it on main PCB) so I guess it's tolerant for some H Sync changes.

mkarcher wrote on 2022-10-25, 18:05:

What signal are you scoping?

For a comparison that's MDA video output.

Bernkastel7734 wrote on 2022-10-25, 17:43:

I'm not sure if it's MDA compatible or not, but I am sure it is not what a healty video output should look like.

that looks more like interference from adjacent wires. especially given the low voltage.

Yeah, I think the same. While both HSync and VSync looks good, the B/W video is garbage. I do not have much experience with that, but maybe it's RAM related issue. In Olivetti M20 hardware manual I have found such Video logic diagram.

Is the bulb clear see through? That's spark gap protectors to limit voltage *from* the CRT pins, to protect electronic components in that circuit if it go past the spark gap rating.

Cheers,