Restoring an Olivetti ETV260 that doesn't start up

[Zare]

Hi @manumuna ,

ETV 260 runs on BA191 mainboard, which is shared between later M19 and typewriters ETV 260, ETV 500 and ETV 3000.

M19's Resident Diagnostics version 3.71 was mostly paired with this motherboard and it has 6 or 7 POST checkpoints (on LPT).

I would expect ETV Diagnostics to behave the same. I still haven't got my hands on ETV BIOS dumps so cannot confirm.

The same mainboard may run these few systems, that does not mean that the video system operates in a same way. What is sure is ETV machines do not have colour selection switch soldered on the board.

What may be happening, and these are my 2 cents because I do not have any ETV gear, is that video system is programmed in a different way than for M19. If I consult that Hantarex MT 3000 repair manual, I can see it operates on CGA frequencies using composite input. This is not the video signal of mono M19 - which is MDA video signal, for MDA monitor DSM 1912.



DSM 1912 is in ~18 kHz range and it will produce hissing if you input 15.5 kHz hsync into it, which is simple to do, just push the M19 switch up to CGA and hear the monitor rant.


I can see this monitor is not compatible with M19's monochromatic video output. It seems to be a composite screen that accepts sync either modulated with video, or discrete sync, separate or h+v modulated. The ETV 260 system probably connects per option 3, separate h/v sync.

@1ST1 have you interchanged M19 and ETV 260 screens, like plugging in M19 screen to ETV 260? I know we discussed it but I can't remember.

Everything about this video topic is quite interesting.
It seems to me the video system onboard these motherboards can be programmed to output different video standards and geometries, with the difference being in BIOS code and PAL logic.
However in order to produce composite video the wiring must be changed, since the composite video lines of Paradise video generator are not connected to mainboard logic in BA191 boards of M19.
I'm not sure whether Olivetti would change the board connections and routing and still keep the same BAxxx designation.

I'm also interested in a detail how manumuna connected the other PC with standard CGA card to this Hantarex screen. I can see it has separate TTL signal functionality but I cannot see 4 separate signal lines, unless these "Brightness" things are it.

But that covers what the monitor can take, not the way ETV 260 sends picture to it.

 

[1ST1]

Strange, because I know I have swapped 400V types.

 

[1ST1]

@1ST1 have you interchanged M19 and ETV 260 screens, like plugging in M19 screen to ETV 260? I know we discussed it but I can't remember.

No this is not possible. At least I am afraid to try it because of the power supply. The machines are too valuable to do such experiments.

With the ETV 3000 I did testing on a CGA monitor, I got a picture, I could see the BIOS post, booting from floppy, but out of sync, but even as I connected a NEC Multisync (1st version with manual sync settings) still no stable picture and this monitor can do from PAL/NTSC up to SVGA. ETV 260 monitor stayed black. I also have the M240 composite video monitor, and also this out of sync. ETV 3000 was sold with ETV 2700 black & white monitor but for those monitors I needed a 2 pin adapter as the composite cable connector is different, and then also out of sync. Maybe because of this reason they choosed different connector and one has to open the monitor to change the sync.

 

[Zare]

The connector on the board of ETV 260 is still DE-9, right?
If so, what pins are present on the screen cable? All of them?

The picture you had linked on other forum is not high resolution but I could see that video section only differs in the switch part - no switch on the ETV. But the port looks the same.

 

[Zare]

@manumuna Please if you're able to provide hi resolution pictures of the entire mainboard, if your system is still open.

 

[manumuna]

The ETV 260 system probably connects per option 3, separate h/v sync.

Yes, this is what I thought. It's quite similar to that monitor I attached the schematic, but it's not the same. The pinout of the monitor connector (a DB9) is this one: 1-2=GND 3=NC 4=GND 5=+12Vdc 6=Intensity (but it's like to be unused for what I understood) 7=Video Signal 8=H Sync 9=V Sync. For what I saw on the board, the ETV screen and the schematic I attached, share the IC for the vertical and horizontal deflection and almost all the ETH part, while the rest is quite similar as working idea but not the same because of the use of various 74xx logic chips. I got the pinout here checking it with the tester and seeing the connection to the monitor board and on this one I read the silkscreen to understand the different function. About the intensity, on the board is written just "N". So I tought it was meaning something like "nuance", similar in meaning to intensity... Just a guess.
To connect it to the CGA card (an isa 8bit card from aliexpress, like this one https://it.aliexpress.com/item/1005006036994388.html) I made a little adapter moving the video signal from one the rgb channel (pin 3,4 and 5 of a normal cga connector) to pin 7 and it worked. The only thing I noted is that it's never completely in sync vertically and it syncs horizontally only when all is "cold" and the horizontal hold to the minimum and the horizontal phase quite to the minimum (or both maximum, I always have a dubt while talking of trimmers...). After a few minutes, just the time for the monitor's board to warm up, and horizontally it gets completely out of sync too. But, one thing is sure, the timing charts of that manual is exactly the same for this monitor, same shapes, peaks and almost timing and in the same points of the circuits. The way the trimmers are soldered on the board are completely different.

Strange, because I know I have swapped 400V types.

About this, I haven't found any electrolytics at that voltage. And I think too this is quite strange for two monitors of the same type to have different components... I really don't know

 

[manumuna]

The connector on the board of ETV 260 is still DE-9, right?
If so, what pins are present on the screen cable? All of them?

The picture you had linked on other forum is not high resolution but I could see that video section only differs in the switch part - no switch on the ETV. But the port looks the same.

Tomorrow I will surely do one: it's still open because it is still not booting. The ETV doesn't have the big switch like I saw on a photo of the M19 board, but connecting to the same spots there is a jumper with a strange silkscreen on one side "LSW" (my guess, but it's really a big one, is that it can mean something like Local Screen hardWare???).

 

[Zare]

From what I've seen online ETV 260 has the same Olivetti font as M24 and M19, which is 8x16, and requires 640x400 native Olivetti resolution for 80x25 display.

The hsync and vsync brackets do correspond to CGA frequencies but not necessarily to CGA picture which has twice less rows. Maybe the monitor has issues displaying 200 lines. Olivetti PCs M19, M24 run monochrome video picture at 640x400 all the time.

Do you have an empty DIL socket right below the video port on the mainboard? On M19 we presume that is for video signal diagnostics, in those sockets M19 traces all the RGBI and VI for both video modes, meaning it connects 5 distinct TTL lines between video port and video system hardware. Poking multimeter at the DIL socket and port could reveal what exact pins does the video system on ETV 260 use.

 

[Zare]

The pinout of the monitor connector (a DB9) is this one: 1-2=GND 3=NC 4=GND 5=+12Vdc 6=Intensity (but it's like to be unused for what I understood) 7=Video Signal 8=H Sync 9=V Sync.

That does correspond to monochrome M19 MDA pinout, except the pin 5.
Figuring out whether Intensity is used could reveal nature of the signal/monitor. If it isn't all bets are on composite video data on pin 7 and separate TTL h/v sync.
That would also imply the BA191 boards are different between ETV 260 and M19, which is very interesting info, potentially.

 

[1ST1]

The connector on the board of ETV 260 is still DE-9, right?

Yes, and ETV 3000 additionally has 2x 2 pin, one for power, one for composite. By defauld a holder is screwed on the DB9 which fixes the two pin cables in their connector.

 

[1ST1]

Intensity (but it's like to be unused for what I understood

It is used, don't worry. The sws wordprocessor software uses it for example to display bold text attribute.

 

[manumuna]

It is used, don't worry. The sws wordprocessor software uses it for example to display bold text attribute.

Ahhh ok!!! That's great, I'm looking forward to seeing it in action and see the difference and its use!!
I attach here the photo of the pc side board.
I also changed back, one by one, the 41256 dram chips. I found the faulty one! just one and I left on a 4164. And I almost cried: the computer finally beeped!!!!!!!!!! Started the video and finally I had an image! It got stuck there. The keyboard respond correctly to caps lock and similar keys.

I ordered 10 new 41256 chips from China... They should arrive at the beginning of March. It show the codes flashing across the parallel port and then revert back to a fixed one.

 

[manumuna]

From what I've seen online ETV 260 has the same Olivetti font as M24 and M19, which is 8x16, and requires 640x400 native Olivetti resolution for 80x25 display.

The hsync and vsync brackets do correspond to CGA frequencies but not necessarily to CGA picture which has twice less rows. Maybe the monitor has issues displaying 200 lines. Olivetti PCs M19, M24 run monochrome video picture at 640x400 all the time.

Do you have an empty DIL socket right below the video port on the mainboard? On M19 we presume that is for video signal diagnostics, in those sockets M19 traces all the RGBI and VI for both video modes, meaning it connects 5 distinct TTL lines between video port and video system hardware. Poking multimeter at the DIL socket and port could reveal what exact pins does the video system on ETV 260 use.

I will check the connection between the DIL and the video DB9.

 

[1ST1]

Ah, good! Broken memory...

 

[Zare]

Thanks for the hi res pic! Awesome, and very informative

- the label down south says BA191, but the sticker up north says BA198.
- FSH1/2 are missing and so is U300 DIP switch - some of their connections are patched in with clips.
- the graphics card thoroughly misses colour mode, no switch, and less lines (count of resistors) near the port
- the JSOUND routing to empty sound hole is gone
- HD68A45P CRTC instead of MC6845P
- PAL logic is designated PL5(A,B,C,D) and not PL1

Which I interpret as : a BA191 motherboard modified for ETV 260 usage - relabled BA198. The phsyical routing and connections look, from the top, the same.
The video system has different PAL logic. CRTC may have differences in behaviour too.
There are four instead of seven resistors running to video port - I assume they are Hsync,Vsync,Intensity and Video since the resistors look the same as M19s.
Thus the colour functionality of the system has been removed electrically.
The FSH row and U300 DIP are used to control DSK 1924 M19 cpu upgrade. The ETV board runs Faraday FE2000A thus this is not needed. So on ETV the components are removed and the "no DSK" setup has been clipped/soldered in.

But - I'm still not certain thats not some sort of composite image because some hi-res screnshots of 'gaming' on ETV 260 I've seen on Reddit, do not look like a digital screen. The text quality is ok, but it is not pristine. It might be due to photo artifacting on that pictures.

You can actually poke oscilloscope or logic analyzer into Intensity signal that goes thru U302 to finally answer this.

Congrats on fixing the system! Do you have means to move data out of it? If you'd be again kind to assist the general knowledge here, I can give you few easy commands to enter into DOS debug tool, to dump BIOS to a file.

 

[manumuna]

Sorry for the late reply but I've been really busy with my job. I still don't have the new memories (they're still on their way from China). I will check in the next days the video signal and post a photo of it. At the moment the system is not really booting (I think quite surely because of that 4164 chips in place of the 41256) so I can't do anything. As soon the system will boot we can try to dump the BIOS file. My only concern is: is it a safe thing? I really don't like the idea to risk to make unusable a machine i just fixed. Sorry for my question but I've never done such a thing.

 

[Zare]

Yes, of course.
It involves using a DOS debug tool to read a part of system memory, a part not mapped to RAM but to BIOS ROM, and save it to a file. So it is totally harmless - unless you have some valuable data in the same filename, as debug will overwrite it

 

[manumuna]

Hi! The memories finally arrived today!!!!!!!! And yes! It booted into DOS!!!! I had to clean carefully the floppy drive from tons of dust and put some oil on the mechanism that move the heads. About the video signal, I had a few shots using two different scopes but giving me the same result. I'm quite sure is not a composite signal, especially because the second one has the capability to sync on the horizontal or vertical sync pulse. Or, at least, a "classical" composite signal.
So, here is the photo of first bank of rams totally on the sockets: only one chip was faulty of all of the rams.

And here the two photos of the signal
I'm sorry but I don't have a digital scope yet... So, this is the max I can obtain.
I attached also a couple of video of the system booting and working.
@Zare , if you like I can help you now with that software. Let me know how to do it.
@1ST1 , may I ask you an img copy of your booting disk? I don't have any of the original ones for this machine: it would be so so great to test it completely. Also, may I ask you a photo of your keyboard? On mine there are a few keys missing and I wanted to rebuild them.
Thanks a lot again for all your help!!! Really really a huge thanks!

 

[1ST1]

Very good! Which language you would need? Currently I have only image of german ETV 260 disk, and it may complain about a foreign daisywheel. The ETV 260 has software driven printer, all done by DOS drivers and the language of the driver must match keyboard and daisywheel. I also have italian ETV 3000 disk, but I don't know if that would work with ETV 260.

Before you can do more with it, you may need to do a cleanup of the printer mechanics. All must move smooth and soft.

The ETV 260 monitor signal is MDA alike, not composite or CGA, EGA, ...

 

[manumuna]

I would like the italian one, but any would be wonderful. I attach the photo of the daisywheel that it's actually in the machine. Maybe we can try with both of them (I mean the 260 and 3000 disks) if it doesn't bother you too much. I will start cleaning all the mechanic soon. What do you recommend as lubrificant?
I attach also the photo of the keyboard. You can see the missing keys. Do you know what those keys are?
Also, on the board that control the typewriter part there is a component is one corner that really seems to me like a battery of some sort... Do you know what that is? Have you changed it on your machines? If you like I can shoot a photo of it.