• Please review our updated Terms and Rules here

Olivetti M19

The weird thing is that the other keyboard works fine with the same replacement cable.
I will do some measurements with the oscilloscope to find out which theory turns out to be true.
Thanks for the good pointers.
 
I don't know which is the other keyboard, but maybe it has more modern controller chip and better power consumption design, so it needs less current. Remember, U=R*I. In this case U is the voltage drop over the cable.
 
@1ST1

He has a pair of M19 standard mono bundles. Each of his M19 basic unit, has matching DSM 1912 mono monitor and ANK 1924 keyboard.
However you cannot plug in the ANK 1924 that came with 2nd M19 to 1st M19.

The M19s are, as far as I can tell, different motherboard revisions, possibly different BIOS revisions, and one ANK has discrete ground, the other doesn't.

There are no generic XT keyboards involved, only Olivetti M19-standard hardware, so it is really awkward...
 
Located a piezo buzzer in the newer revision of my ANK1924.
1000042669.jpg
Will try to do some measurements tomorrow. Have to drag myself to the gym this evening instead 🙃

P.S. The font on the keys between the two revisions slighly differs as well. But they're both genuine ANK1924s.
 
The 13 should be a miss type, there was no such model.

That guy faild to adapt it to USB as these (Mini-)DIN to USB keyboard adapters expecting AT (PS/2!) keyboard protocol. The M19 keyboard is a pure XT protocol keyboard. So such a PC-keyboard to USB adapter will fail in any case for this keyboard.

But one more thing let me thinking... The edge connector for the mouse module... It has 5 signal pins plus two paired connetors. Those paired ones should be 5V and ground. For a two button mouse the rest is just not enough, xa, xb, ya, yb for the quadrupple ball sensor and only one more for two mouse buttons... I expected 6 signals and then the edge connector to the DB9 adapter for an M24 or Atari ST mouse might be very simple...
 
I started with the XT DIN5 layout:
1000042715.jpg
Tried to measure clock and data on the "good" keyboard.
Data:
1000042714.jpg
Clock I was unable to capture but I noticed this on the NC:
1000042712.jpg
Not entirely sure what this is, probably picking up noise from the clock line. Since the keyboard works fine I took it as a baseline.

When I tried the second ANK1924 on the same PC it just worked. I thought it had to do with the disconnected ground but I could not reproduce it.
That left the repair styles that I used.
If you compare the repair style I used you can see this:

Keyboard 1:
1000042717.jpg

Keyboard 2:
1000042716.jpg

My conclusion is that method one is more solid. Although I should have used longer heat shrink tubing and close the end. Probably it broke or disconnected when I wrapped eveything up. So no difference after all. Will redo 2nd keyboard properly and wrap it up. Then I will ship it.

By the way: I have a new hypothesis what the piezo is for. Since the keyboard is much more silent than the older revision when pressing a key, Olivetti introduced a very mild clicking sound from the piezo. It clicks only once if you hold a key.
Not sure if it will become annoying, sure there are (blunt) ways to disable.
 
Last edited:
Jazzy said:
Is that a still from a vid? For some reasons the picture gives me Sony Mavica vibes, very retro. :)
Click to expand...
Yes, indeed. The 1997 floppy disk camera. 😂
The wrong part being in focus does the trick. The cut out does not help either.
 
Question: does the M24 monitor have the same slow phosphor as the M19? Or is it totally incompatible?
 
Jan de Lange said:
Question: does the M24 monitor have the same slow phosphor as the M19? Or is it totally incompatible?
Click to expand...
The M24 graphics is totally different. While M19 is "lowcost, below CGA" M24 is "highend, beyond CGA, beyond IBM", higher video frequency (allmost like EGA). So you can't use M19 monitor, you can't use CGA monitor. You will need original monitor or a good multisync. We have already a discussion about how to connect a modern VGA monitor it, even VGA TFT, as long it is "multisync capable", like some NEC Multisync TFT monitors (example 1990nxp): https://forum.vcfed.org/index.php?threads/m24sp-db25-to-db9-adapter.66233/post-813365 Here is a list of "multisync" capable monitors, the list was made mainly for Commodore Amiga / Atari ST / ... series and alike computers, but any of these TFT on the list should also display M24 graphics if you make an adapter as proposed in that thread https://forum.vcfed.org/index.php?threads/m24sp-db25-to-db9-adapter.66233/post-813365 , just follow the discussion.
 
Thanks. I can port the monochrome to VGA (using converted MCEBlaster), but the monochrome dithering for greylevels needs a very slow phosphor. On a LED or normal CRT you get flickering as there is no time integration.

So I am now working on digital time integration to combine frames into grey scales, but it's a pain in the b***t. With a slow phosphor monitor I could skip that step.
 
I sugest you to make such a simple adapter as described in the other thread, you will then have full 16 colors. That's much better than monochrome.
 
You must log in or register to reply here.
Back
Top