Atmos ROM

If you want to ask questions about how the machine works, peculiar details, the differences between models, here it is !
How to program the oric hardware (VIA, FDC, ...) is also welcome.
omelette
Officer Cadet
Posts: 60
Joined: Wed Mar 10, 2010 9:54 am

Re: Atmos ROM

Post by omelette »

At the risk of being shouted at for resurrecting dormant threads, I got my Atmos to work with 27c128's eproms without issue. I also used the empty memory socket to house the old v1.0 rom that my Atmos came with, and added a little circuit that allowed me to select which rom I wanted to work with - naturally, it defaults to v1.1. Apart from the little circuit board, I also had to add IC11, absent on my Atmos, and employed the unused d-type flip-flop in IC27 to do the selecting. Push the NMI button 'normally', and it does what you expect, keep it pressed for 0.5sec or so and it toggles which rom is selected. Note that a reboot is not automatically done, this must be done using the Microdisk's reset button. Looking at the circuit diagram above, I think that some trace-cuts were needed, one to the address line anyway. It was done in the 90's so it's a bit hazy now. :)

I actually came across this thread looking for information on eprom-usage problems others might have had. I bought a pile of 2716/32/64/128 eproms from China about a year ago. in the last couple of days, I decided to try to coax back to life 2 Oric-1's in my possession, so naturally needed v1.1 eproms. I had completely forgotten about the pin-27 thing, which coupled with the fact that both Oric-1's had no IC11 installed, made it more trouble than it was worth.

One discovery I did make though was that with my (modified) Atmos, the newly programmed 27c128 eproms (I tried several) will not allow the computer to boot when the Microdisk controller is attached, though it boots fine with the old 27c128 inserted?!? This has me confused. I do know for certain that despite all of the eproms that I bought being labeled as CMOS eproms, NONE of them are, the half-dozen I've tried anyway, all drawing more than twice the current that identical 'genuine' eproms in my possession do. They are almost certainly not produced by STmicroelectronics either!


User avatar
mikeb
Flight Lieutenant
Posts: 282
Joined: Wed Sep 05, 2018 8:03 pm
Location: West Midlands, UK
Contact:

Re: Atmos ROM

Post by mikeb »

omelette wrote: Fri Nov 29, 2019 9:37 pmthe newly programmed 27c128 eproms (I tried several) will not allow the computer to boot when the Microdisk controller is attached, though it boots fine with the old 27c128 inserted?!?
Assuming your EPROMS aren't broken/fake/mislabelled etc :- EPROMs for system ROMS and external devices (the disc interfaces) don't get on.

The disk interface requires the ability to assert ROMDIS (expansion port) to disable any and all Oric internal ROMS to take over the boot process.

The ROMDIS line connects straight to a pin on the masked (original) ROM to shut it off for this purpose.

EPROMs don't have this function, on any pin, so you will have to work it out for yourself (by preventing the OE and/or CS signal reaching one or other ORIC EPROM when ROMDIS is asserted), which will mean more board mods for you.
omelette
Officer Cadet
Posts: 60
Joined: Wed Mar 10, 2010 9:54 am

Re: Atmos ROM

Post by omelette »

mikeb wrote: Sat Nov 30, 2019 7:02 pm
omelette wrote: Fri Nov 29, 2019 9:37 pmthe newly programmed 27c128 eproms (I tried several) will not allow the computer to boot when the Microdisk controller is attached, though it boots fine with the old 27c128 inserted?!?
Assuming your EPROMS aren't broken/fake/mislabelled etc :- EPROMs for system ROMS and external devices (the disc interfaces) don't get on.

The disk interface requires the ability to assert ROMDIS (expansion port) to disable any and all Oric internal ROMS to take over the boot process.

The ROMDIS line connects straight to a pin on the masked (original) ROM to shut it off for this purpose.

EPROMs don't have this function, on any pin, so you will have to work it out for yourself (by preventing the OE and/or CS signal reaching one or other ORIC EPROM when ROMDIS is asserted), which will mean more board mods for you.
Yeah, I understand the theory of why eproms cannot work with the Oric, and before posting this, I figured I knew why the single TMS27c128-20JL that I programmed in the 90's works flawlessly in the Atmos WITH the Microdisk attached - because I modified the Atmos way back then!

Unfortunately, this is not the case!!! Below are 2 pics of my unmodified Oric-1 working perfectly with the same TMS27c128-20JL eprom, while the Microdisk is attached. This should NOT be working, yet it is! For completeness, I've also attached the data-sheet that covers this eprom.

I'm seriously considering doing some (old) TMS-eprom shopping - the data-sheet gives dates of between 1984 - 1993 - just to see if there is some 'undocumented feature' present in this TI-manufactured eprom. Unless someone can explain how ANY eprom could possibly be working on an UNMODIFIED Oric, while the MIcrodisk is attached?


Attachments
TXIIS140788-1.pdf
(141.75 KiB) Downloaded 299 times
User avatar
iss
Wing Commander
Posts: 1641
Joined: Sat Apr 03, 2010 5:43 pm
Location: Bulgaria
Contact:

Re: Atmos ROM

Post by iss »

romdis.png
romdis.png (430.54 KiB) Viewed 9929 times
This is the /ROMDIS signal on an unmodified PCB. Obviously there are no other "hidden" connections and the wiring is exactly as per schematics. So the secret should be in the EPROM chip.
modes.png
modes.png (64.9 KiB) Viewed 9929 times
Looking in the PDF on page 3 in the table with operating modes - the only plausible case for me is that the chip somehow enters "programming mode" and D0..D7 are inputs which allows the external (microdisc) EPROM to be read fine and because we don't have high voltage (12.5/13V) for programming the data is not changed, additionally the /ROMDIS pulse is much shorter than the required /PGM.

This is really interesting because I can confirm that I had the same case with one Oric ...
During the weekend I'll try to "investigate" the mystery. :?
User avatar
mikeb
Flight Lieutenant
Posts: 282
Joined: Wed Sep 05, 2018 8:03 pm
Location: West Midlands, UK
Contact:

Re: Atmos ROM

Post by mikeb »

iss wrote: Thu Dec 05, 2019 10:01 pmthe only plausible case for me is that the chip somehow enters "programming mode" and D0..D7 are inputs which allows the external (microdisc) EPROM to be read fine and because we don't have high voltage (12.5/13V) for programming the data is not changed, additionally the /ROMDIS pulse is much shorter than the required /PGM.
I like your thinking! That might well explain it for certain brands/ranges of chip.

The proper ROMDIS line originally would, of course, cause the D0-D7 to tri-state, so no output. And no input, either, because it's a masked ROM :)

On an EPROM, in programming mode, D0-D7 would become an input, which is next-best thing to being tristate.

I would be wary of using this trick though -- although the EPROM specifies 12.5v for reliable programming on the VPP pin, I bet they don't say that "using 5V will definitely not program anything, ever, and won't degrade the data on the chip". So, you could have a slow, cumulative over-write of the content of EPROM. Whether you ever notice this happening is left as an experiment ...

In the same way they will tell you to erase it you need x mW of UV light at y nm wavelength for z minutes. They won't mention that a standard fluorescent tube will probably give you data errors in 3.5 months, and sunlight might take you out of action in a day :)
User avatar
kenneth
Squad Leader
Posts: 515
Joined: Fri Nov 26, 2010 9:11 pm
Location: France PdD
Contact:

Re: Atmos ROM

Post by kenneth »

It is an unexpected combination by the memory manufacturer. if one day, the 5v is enough to program for new components, the connected drive may corrupt some of the content.
User avatar
iss
Wing Commander
Posts: 1641
Joined: Sat Apr 03, 2010 5:43 pm
Location: Bulgaria
Contact:

Re: Atmos ROM

Post by iss »

Well, just the preparation for the test took lot of weekend time:
for-romdis.jpg
for-romdis.jpg (1.45 MiB) Viewed 9870 times
On the picture: Atmos and 8D "all-on-sockets"; Cumana; Silicebit's, 2x original and one bare PCB Microdisc; 2x 27128 (ST and Hitachi); 3x Gotek (1 HxC and 2 FlashFloppy). So, everything is ready, need some more free time... :(

I thought that I have more 27128 chips, unfortunately I found only these two and whole bunch of 2764 and 27512.
So, stay tuned for the result (I hope very soon).
NightBird
Officer Cadet
Posts: 35
Joined: Sun Nov 24, 2013 7:00 pm

Re: Atmos ROM

Post by NightBird »

Hi

For ROM/EPROM + Microdisc compatibility and Phi2 improvement, please also read CEO-MAG no 288, april 2014, pp. 8-13. :D
User avatar
iss
Wing Commander
Posts: 1641
Joined: Sat Apr 03, 2010 5:43 pm
Location: Bulgaria
Contact:

Re: Atmos ROM

Post by iss »

NightBird wrote: Mon Dec 09, 2019 10:35 pmFor ROM/EPROM + Microdisc compatibility and Phi2 improvement, please also read CEO-MAG no 288, april 2014, pp. 8-13. :D
Thanks for refreshing the memories!
There are also other solutions but here the point is to confirm/bust the myth that with certain models EPROMs it can work "as is".
User avatar
iss
Wing Commander
Posts: 1641
Joined: Sat Apr 03, 2010 5:43 pm
Location: Bulgaria
Contact:

Re: Atmos ROM

Post by iss »

Bit delayed but as promised here are the results from the test:




I used Jasmin instead of Microdisc (because it's on my small working table already) but there should be no differences related to /ROMDIS signal. I'll repeat the test with Microdisc asap just to be 120% sure :).

In the test were tested two UV EPROMS 27128:
20200110_122321.jpg
20200110_122321.jpg (495.25 KiB) Viewed 9461 times
20200110_122512.jpg
20200110_122512.jpg (1.75 MiB) Viewed 9461 times
20200110_122528.jpg
20200110_122528.jpg (1.12 MiB) Viewed 9461 times
ST M27128A from STMicroelectronics - FAILS TOTALLY - no even reset!
HN4827128G-30 from HITACHI - WORKS like charm!

So, only good news:
  • there are obviously models of UV EPROMS 27128 which work flawless with Jasmin/Microdisc!
  • I'm not dreaming and still have reliable memory! :twisted:
I'll continue the tests with 2764 chips because I have lot of them. It will be very useful if someone else find other working EPROMs to give feedback here, so we can collect a list with known working chips (and another list with not working ;)).
User avatar
protek
Pilot Officer
Posts: 102
Joined: Wed May 23, 2018 9:10 pm
Location: Oulu, Finland

Re: Atmos ROM

Post by protek »

Dusting up this thread. I checked out the two EPROMs that I’ve got. They’re:
TMS27C128-2JL
HN27128-AG-20

I don’t have any mass storage peripherals other than my half finished EREBUS, though. Did that also require /ROMDIS to function?
User avatar
kenneth
Squad Leader
Posts: 515
Joined: Fri Nov 26, 2010 9:11 pm
Location: France PdD
Contact:

Re: Atmos ROM

Post by kenneth »

@Protek:
Erebus needs to desactivate the internal rom with the /ROMDIS pin to run with the external one, I suppose that the drive do the same to use its bootstrap program, if Erebus runs correctly, the driver will do the same.
@Iss:
Thanks for the test, its a good new for the Oric computers with a missing Rom memory.
User avatar
protek
Pilot Officer
Posts: 102
Joined: Wed May 23, 2018 9:10 pm
Location: Oulu, Finland

Re: Atmos ROM

Post by protek »

I tried booting with 1.1 ROM burned on the TMS27C128-2JL while Erebus was connected. At least it booted to basic. I couldn’t test accessing the Erebus as I don’t have a keyboard connected to this motherboard.
User avatar
kenneth
Squad Leader
Posts: 515
Joined: Fri Nov 26, 2010 9:11 pm
Location: France PdD
Contact:

Re: Atmos ROM

Post by kenneth »

I never tryed this but normally, if the Rom is deactivated by Erebus, the computer can run without the Oric Rom chip memory.
User avatar
protek
Pilot Officer
Posts: 102
Joined: Wed May 23, 2018 9:10 pm
Location: Oulu, Finland

Re: Atmos ROM

Post by protek »

I tried again my motherboard with the Basic ROM burned on an EPROM. Erebus works all right. Does this mean that the TMS27C128-2JL is a suitable replacement for Atmos Basic ROM?
Post Reply