The Sinclair QL Forum

janbredenbeek wrote: Wed Apr 19, 2023 11:03 pm With the upper 16K missing the RAM test should still work as this lies within the first 16K ROM. Given the fact that the test pattern doesn't appear, it looks like there is another hardware failure.

Without knowing what is on the roms, we do not even know there is a RAM test though!

Chr$ wrote: Wed Apr 19, 2023 7:59 pm I think personally my next step would be to check the content of those ROM chips and of course try a known working pair (or single 27512 type), written with a ROM of the full 48kb. Those 27128 chips are both only 16kb (as mentioned before) so it's possible it had an external ROM dongle originally, which is now missing.

Might not be that easy though, as one is labelled 0000 and the other 4000 (and the code from 8000 onwards 'was' external. Normally they are 0000 and 8000 internally. So I think would indicate that addressing might be different. Someone else will be able to offer a better insight.

I agree with Chr$ if you have an eprom programmer than dump the roms, we can then examine them in dissasembler/emulation to see what they actually do!

I can supply a set of JM ROMs for testing, plus the missing LED wires and whatever screws are missing to help get this back to factory condition (assuming the LEDs were ever fitted) - just need the cost of postage as it is such a good find, it would be useful to get it up and running and see what happens.

RWAP wrote: Fri Apr 21, 2023 8:15 am I can supply a set of JM ROMs for testing, plus the missing LED wires and whatever screws are missing to help get this back to factory condition (assuming the LEDs were ever fitted) - just need the cost of postage as it is such a good find, it would be useful to get it up and running and see what happens.

Thanks Rich for offering support with mass production materials.

But as Chr$, myself and others mentioned, this QL should be left/restored as it originally was. No ROMs, just EPROMs. No LEDs as they were not ready when this machine was build.

Chr$ wrote: Sun Apr 16, 2023 3:36 pm Hello Urs. I kept meaning to tell you that your images at:

https://www.sinclairql.net/pic/

don't work, but I suspect you've noticed that (hence the onedrive images)!

Thank you Chr$ for letting me know.

Recently it was cold and rainy and I had/made some free time to get several things fixed on the sinclairql.net website , the QL Picture Gallery is now (again) OneDrive only.

QLvsJAGUAR wrote: Sun Apr 16, 2023 3:25 pm - Most likely this QL came with a ROM port dongle holding at least additional 16 KB of the firmware (in a third 27128 EPROM IC), usually labelled 8000 (at address 8000 Hex). Do you have got a ROM port dongle or any additional EPROM ICs?

I did some investigation, tracking back the seller of this QL on ebay. There was a second auction from the same seller where the
"Sinclair QL EPROM Cartridge Board with an EPROM labelled 13-2-84 8000 0D0C as checksum" was sold. Most likely this would fit to this QL computer with the ZX-83 PCBA. See picture. Key question is who the buyer of this is!?

QLvsJAGUAR wrote: Wed Apr 26, 2023 5:41 pm I did some investigation, tracking back the seller of this QL on ebay. There was a second auction from the same seller where the
"Sinclair QL EPROM Cartridge Board with an EPROM labelled 13-2-84 8000 0D0C as checksum" was sold. Most likely this would fit to this QL computer with the ZX-83 PCBA. See picture.

That 'checksum' needs some investigation... Key question is how it's generated. I've tried byte-sum, word-sum (both big- and little-endian) and word-XOR on FB, PM, AH, and JM ROMs (8000-BFFF) but none of the results looked like 0D0C.
- edit: the '0D0C' should probably be read as 'QDOS', since the other two ROM's stickers had 'BASIC' on it!

janbredenbeek wrote:

QLvsJAGUAR wrote: Wed Apr 26, 2023 5:41 pm 0D0C as checksum

That 'checksum' needs some investigation... Key question is how it's generated. I've tried byte-sum, word-sum (both big- and little-endian) and word-XOR on FB, PM, AH, and JM ROMs (8000-BFFF) but none of the results looked like 0D0C.
- edit: the '0D0C' should probably be read as 'QDOS', since the other two ROM's stickers had 'BASIC' on it!

That‘s a very good read and conclusion, Jan.


Gesendet von iPhone mit Tapatalk

MetroZodiac was kind enough to sell me this ZX83 QL and I've just returned home with it. Credit to him for cleaning it up so well, it's in fabulous condition.

I've read the 13-2-84 dated ROM chip contents and those are attached in the zip file. Someone with better understanding of the ROMs can perhaps find out which version they are... I can see the text 1.00 in one of them with similar looking code after it compared to PM, which is great (or a red herring), but also a bloody shame if the last 3rd has gone to someone else and remains missing! Unless it's FB, for which the last 3rd is available.

I've not powered it up yet. The 'cockroach' upside down chip is I assume a 74LS00 or something very similar, so unless anyone screams no, I assume for testing purposes that I can replace both ROMs with any other version all on a single 27512.

A 27512 EPROM won't work - as the sockets have VCC wired to where A15 should be. Instead - the masked ROM's that Sinclair used are a pair - one activated by a high chip enable - and the otehr by a low chip enable - and these use A15 on what would be the chip enable to decode the 32K and 16K blocks - the sockets themselves are effectively wired in parallel (this is why you can piggy back the ROM chips) and fit them to just one socket.

It is a shame Sinclair didn't wire the socket like a standard JEDEC Eprom - so that a 64K ROM chip could be used.