• Please review our updated Terms and Rules here

Paradise PVC2 questions on Olivetti M19

Yes it is. I don't have any other 8 MHz.

I found this on vogons

It was an Olivetti M19 PC that had an Intel 8088 CPU running at 8 MHz (I remember it had a turbo switch at the back that allowed you to switch it back to the standard 4.77 MHz but, we never used it since the PC was already slow enough 🤣).
It came with the maximum memory of 640KB installed, had a 12" monochrome (green screen) with two 360KB floppy disk drives which we upgraded 2 years later to accommodate a 20MB hard disk drive in the second floppy disk drive bay.

He's probably misremembering but doesn't cost me anything to try with the system in 'color mode'.
 
I've found a magazine article that is second available internet source for M19. These sources (first is brochure) are used by a developer doing work for Olivetti emulation in 86Box. Autotranslation is too large to fit in a message here : https://pastebin.com/FaMM9Cnk

In 2nd page technical overview table (scheda tecnica) under "Microprocessor" says loud and clear special card with switch is provided with 8MHz upgrade. Maybe this was what vogons person was recalling. Also throughout the text author writes that internal bus is not IBM PC compatible, only Olivetti cards are supported. He is at least partially wrong because it does support various cards, but they need to be small and not have -5V rail because the system does not have one. He goes on to describe expansion box slot as "IBM PC compatible" and can host a variety of adapters, so yeah box space and box PSU can conform to ISA spec 100%.

I have Intel 8088-2 on the way so we'll see.

Who or what is the source for mainboard DIP switch and speed jumpers available online? The speed setting moves them in a pattern. If even jumpers were closed you switch to odd ones, and you switch the DIPS in pairs of 4. I mean this looks overly complicated for a 2-way setting.

Back to 86box, and why I found this material deep in their github issues, is that emulated machine has problems with CPUs faster than 4.77MHz. RTC does not work if you raise the CPU clock. And the CPU clock cannot be raised more than 7.16MHz. A lot of this smells like something extra is needed in hardware to perform the switch.

Would you by any chance have material for other Olivetti XTs and how CPU speed upgrade was performed on them?

Btw, the Olivetti SCSI controller does not work outside of slot 1. I don't remember enough about ISA bus mastering to say can it be a normal 'design' issue or again something specific to M19's bus.
 

Attachments

  • Sperimentare_1986_09_0027.jpg
    Sperimentare_1986_09_0027.jpg
    354.2 KB · Views: 8
  • Sperimentare_1986_09_0028.jpg
    Sperimentare_1986_09_0028.jpg
    443.9 KB · Views: 8
  • Sperimentare_1986_09_0029.jpg
    Sperimentare_1986_09_0029.jpg
    297.5 KB · Views: 7
  • Sperimentare_1986_09_0030.jpg
    Sperimentare_1986_09_0030.jpg
    437.9 KB · Views: 8
Another mag found on archive.org. Extracted and commenting out relevant bits

The RGB socket will drive both monochrome and colour Olivetti monitors or a standard RGB monitor

This should be easily checkable and a major 'discovery'. I don't have an RGB monitor tho.

A go-faster option of 8088 driven at 8MHz will be available by the end of July, at extra cost. Certain chips were socketed rather than soldered directly to the board, presumably for this upgrade.

Oof. Not good. The PLs do a lot and I hacked around that to disconnect the OGC. The official procedure to disconnect the OGC is buy PL bypass chips from Olivetti. Everything about this is undocumented and word of mouth.

The main PCB looked well built and considerably less cluttered than I expected, due to the use of a large custom gate array which encompasses several functions including that of a colour card.

The Faraday chip doesn't do anything with video, maybe, just maybe, it's relevant for horizontal sync rate because the jumpers are close to it but I'll have to check that.

Although the review machine was a late prototype, there was little evidence on the PCB of it being an unfinished product, except for one small area of TTL logic which was covered in patch wires.

So the prototype has patches too, and real-world machines are shown with various amounts of patching. But the promo images don't have them. Probably intentional photo-op. The patches do control the latching and there was no other way to pin up the 'bus configuration' when you lay out the design on board - M19 PCB being laid out very very neatly for this age. Later in the production some latches were changed and/or reordered so less patching was needed.

Olivetti will shortly announce an Australian-made internal modem for the M19; it features 1200/75 and 300/ 300 baud rates and is capable of autodial. Including this modem card, there will initially be five cards available from Olivetti. The other four are: a second RS232C port; a synchronous communications interface; an Olivetti proprietary local area network card (called Olinet) and, best of all, an SCSI hard-disk controller which will allow you to hang a whole host of third-party mass storage devices on the system.

Good too see my card is one of original 5 expansion cards for this PC.

The review machine was supplied with one manual called the M19 Starter Kit which also contained three disks. This was spiral-bound, typeset and made heavy use of diagrams throughout. However, it really isn't enough for the absolute beginner and would be more useful as a simple setting-up guide and sales brochure for Olivetti’s peripherals. The three disks consisted of a Tutorial/Introduction to the M19 — which made up somewhat for the lack of this information in the manual — a system disk and a diagnostics disk.

Curious, I have 4th "Keyboard driver plus utilities" floppy disk to this day (no content tho).
No help from date, it's a label template printing date (1984)
 

Attachments

  • 1986.06 Australian Personal Computer_0089.jpg
    1986.06 Australian Personal Computer_0089.jpg
    144.5 KB · Views: 5
  • 1986.06 Australian Personal Computer_0087.jpg
    1986.06 Australian Personal Computer_0087.jpg
    139.5 KB · Views: 5
  • 1986.06 Australian Personal Computer_0086.jpg
    1986.06 Australian Personal Computer_0086.jpg
    162.8 KB · Views: 7
  • 1986.06 Australian Personal Computer_0084.jpg
    1986.06 Australian Personal Computer_0084.jpg
    93.2 KB · Views: 7
  • 1986.06 Australian Personal Computer_0083.jpg
    1986.06 Australian Personal Computer_0083.jpg
    73.6 KB · Views: 7
  • 1986.06 Australian Personal Computer_0091.jpg
    1986.06 Australian Personal Computer_0091.jpg
    129.7 KB · Views: 6
Is it possible for someone to reach to Olivrea or anyone else to ask for manuals and disks that they have on their site? I sent email to them once, didn't get a response.
@Zare, I apologize. I just checked my inbox today and did not find any messages regarding your inquiry about information for an M19. When and how (mail or contact form) did you try to contact me? Please note that the contact form on my website has been malfunctioning for several weeks. Nevertheless, the only manual I have is the "Installation & Operation Guide," which doesn't provide in-depth technical information but rather instructions on setting up and using the device. As for software, I can share the floppy disk images for "Customer Test" and "Getting to Know the M19." That's all the information I currently possess on the matter. Other then that, i have an M19 with color screen, cpu switch and external case. Some pictures here: https://olivrea.de/olivetti-m19extended/ . If you need further pictures of the internals, please let me know.

I also have no good experience with him. Let's say this way, we could be friends, but we are competition.
@1ST1 I don't understand why such statements have to be made in a public forum. I thought everything between us was clarified!? We reside in a world where freedom of expression is valued, and everyone is entitled to share their opinions. However, our acquaintance is limited to a few email/forum exchanges, and we do not know each other personally. In light of our communication and experiences thus far, I wish to clarify that I am not interested to establish a friendship or partake in any form of competition with you. Let us simply concentrate on independently managing our respective hobbies/collections in a peacefully, straightforward and uncomplicated manner. Thank you!
 
@Zare, I apologize. I just checked my inbox today and did not find any messages regarding your inquiry about information for an M19. When and how (mail or contact form) did you try to contact me? Please note that the contact form on my website has been malfunctioning for several weeks. Nevertheless, the only manual I have is the "Installation & Operation Guide," which doesn't provide in-depth technical information but rather instructions on setting up and using the device. As for software, I can share the floppy disk images for "Customer Test" and "Getting to Know the M19." That's all the information I currently possess on the matter. Other then that, i have an M19 with color screen, cpu switch and external case. Some pictures here: https://olivrea.de/olivetti-m19extended/ . If you need further pictures of the internals, please let me know.

Hello and many thanks for answering back here :)
I sent inquiry for M19CUST.IMD and M19GTK.IMD and whether you could share them. Via site contact form, sometime last year, could be full year ago.

Whatever is in Installation and Operation guide isn't that important, I've had it and as you say it doesn't contain anything more than is already documented on the internet.
The disk images are not available. If you could image them and just attach them here, I would mirror them on archive.org.

I'm also quite interested in the CPU switch. It's an unknown, any picture would worth a ton.
Close-up pictures of expansion box insides would also be very valuable since there are none.
 
A 4.77 MHz M19 might be too slow for the CPU switch, the 8 MHz version was a late production which then turned to ETV 260, 500 and 3000, but these without CPU switch. With the ETV 260 system diskette there is a tool to slow down this machine (and probably the other two ETV as well) to 4.77 MHz, no idea how they do by software. It's not only the M19 CPU which needs to be a 8 MHz type, but maybe also the RAM and probably some TTL logics / PAL of faster type. And as the video timing is coupled to the CPU clock there also should be some logics to keep monitor frequency constant when switching 4.77/8Mhz, or they do it by some waitstates.

As I have enough other faster Olivetti PC XT AT 386... I won't modify anything on my M19, it is lovely like it is.
 
I presumed one of the oscillators is controlling the video frequency.

I'm not modifying anything on mine either. What I have to do about mine is acquire the keyboard. Then it's complete, in original state.
The spare system is here for parts and fun. And maybe to document something important for others along the way.
 
M19CUST.IMD and M19GTK.IMD and whether you could share them

please find the images here:


I'm also quite interested in the CPU switch. It's an unknown, any picture would worth a ton.
Close-up pictures of expansion box insides would also be very valuable since there are none.

i will take some pictures as soon as possible and share them here
 
Last edited:
Excellent thank you very much! It's been...30 years since I last time saw that software running.
Now I'm highly curious what Test disk says about my VGA-modded board.

If you guys that collect Olivetti are interested I can gift the Keyb,drivers,utilities original diskette (overwritten but working). I don't have original sleeve - as I remember they were grayish, stone pattern, I keep mine in latter generic Olivetti white sleeve. I won't probably be able to regenerate the diskettes and manual collection, so I can help someone else that's far closer to that goal than me.
 
@olivetti the customer test IMD is erroneous, it looks like a dump of a blank disk. @jesolo was also kind to image his, in attachment.
 

Attachments

  • M19CUST.zip
    63.6 KB · Views: 6
@olivetti the customer test IMD is erroneous, it looks like a dump of a blank disk.
@Zare Indeed, just tried it myself. Really sorry, should have tested it before sharing.
My original disk is blank :-(

@jesolo was also kind to image his, in attachment.
I attempted to write back that image as well, but it resulted in numerous disk errors. Did it work for you?
 
I attempted to write back that image as well, but it resulted in numerous disk errors. Did it work for you?

Jesolo mentioned "I read the diskette back in a 360KB drive and forced it to 250kbps bit rate in IMD".

I used a conversion tool to make a floppy .img (IBM MFM format) and extract the files. Since I don't have original media to recreate I'll move files to M19 via cable or CF card. In zip below is the content if you want to recreate your disk that way.

By the way, customer test disk is important for emulation projects. It shows that M19 cannot be emulated based on generic XT components and disassembling the test program will give us some info about the discrepancies. The way it's implemented now, stuff like 86box is worthwhile to see how something interacts with Resident Diagnostics BIOS of M19 but it isn't correct at all when it comes to platform emulation. I'll try to establish some contacts and see whether we can move this emulation goal further.
 

Attachments

  • custtest.zip
    63.5 KB · Views: 6
Well those images explain a lot.
2 ISA slots, 3 ISA breakouts at the rear - because third breakout designed for CPU switch control. No botch wiring - it explains why it's there on 4.77MHz model (cpu board has two extra PALs).

So yeah, no CPU upgrade without that board possible.

Extra curiosity is that demo disk mentions two types of expansion boxes - the other one without power supply, just for standard ISA size slot.
 
Even I didn't know that. So these "early" 8MHz M19 had a turbo-card, the mainboard stays at 4.77 MHz, only speedup of the CPU. Later ETV 500, 260 and 3000 have native 8 MHz support without that daughterboard and should be faster than this 8MHz M19 because the full system runs at 8 Mhz.
 
Daughterboard has 21.42Khz oscillator going into a clock selector chip, then on to flipflops and buffer connecting to where CLK pin on 8088 is approx. From CPU up, entire bus side of 8088 is connected to two PLAs which would be unusual since those chips are bus logic (already present on the motherboard), but I can also notice a trace coming from SN74 (selected clock signal) all the way up to PL1W.

This smells like bus is switched to 8MHz too.

There's a big caveat, Olivrea's board doesn't have DIP and jumpers that are documented on internet as CPU speed selection.
I'm also certain I have 3 oscillators on the board, instead 2 on board and one on the daughterboard. And botch wires.

So it's a different revision of the motherboard. This is quite quite curious since the two I have, the one covered in magazine review, and Olivrea's are all different computers. All from 1986.

I would like to check and verify that 8Mhz plug-in upgrade via DIP switch settings can work, 8088-2 should arrive soon, but yet again it's going to be on non-standard configuration because I do tests on the VGA-modded board.
 
@1ST1 I've managed to boot the 32MB CF card created via another means. https://minuszerodegrees.net/xtide/XUB/XUB version R625.zip using xtidecfg did an 'autoconfig' on the ide_xt.bin first, then flashed it. It boots on Resident Diagnostics 3.62 M19 but not on the 3.58 one.

I guess this is something but I want to run the adapter on VGA M19. I'll config the CHS settings directly in and retry.

If you want I'll attach a card dump and XUB BIOS file here.
 
Ok, so translation to normal and user CHS settings and I can boot a 512MB card on the VGA M19. :)
 
I don't think you actually need to.
Just use xtidecfg to update your card to that latest XUB, but first use the menu options to configure it and disable LBA autotranslation and enter CHS settings yourself.
 
Back
Top