I assume it's a clone of some older design, I have a couple Z8 mpus in my collection that are their own thing
Z8 was a popular MCU back then so it's no surprise

Heres the dump.
Chip its on is a National Semiconductors Corporation, NMC27C64Q.

Edit: Music quest rom does nothing, debug returns all FF.

S95Sedan wrote on 2023-10-26, 20:16:

Edit: Music quest rom does nothing, debug returns all FF.

maybe the wiring between the PAL chip and the Zilog is not the same. also, there are no any identifiable strings in the ROM like "(C)opyright", etc. so, unfortunately, no any clues and answers.

[EDIT] or maybe the upper Address lines of the PROM are not wired to the Zilog Z8, that's why it can read the 8KB ROM, but not 64KB ROM of MQ - that actually makes more sense. - if that's the case it should be easily fixable with few bodge wires, between the ROM socket and the Zilog.

[EDIT2]actually, MQ ROM is 8KB as well, but in the ZIP file the file is 64KB, because is the same 8KB ROM repeated 8 times. that's interesting...

[EDIT3] I guess useful test is burn the original ROM to 27C512 and see if it's working or not.

mattw wrote on 2023-10-25, 12:12:

thanks, but my main goal is still an answer: what MDR-401 is as hardware and as ROM? we already know as ROM it's not MQ.

the answer is that MDR is it's own thing
I just tried the opposite: MDR rom on classic old version PCMIDI: there is no life out of it , no detection, no usage at all

both seem to be their own thing and from my point of view the more the better - options are always welcome

keropi wrote on 2023-10-27, 11:37:

the answer is that MDR is it's own thing

but its own thing invented in 2022 or its own thing reproduced from something that existed in 1990, i think it's the same question from the very beginning and nothing actually has changed - all we proved so far is that there is probably minor connection difference between the two boards (if I have more spare time looking at pictures from the 2 boards probably I will even be able to locate those differences - even currently I think it might be just the extra Address lines on the PROM like "A15" that don't exist on 27C64, but are there on 27C512). we cannot even said what is the ROM origin and if it's not just SoftMPU/HardMPU port to Zilog Z8. So, I really see no any real answers. also, no one pointed out case in which MDR401 is better or worse than the currently existing solutions, i.e. what is its level of compatibility.

mattw wrote on 2023-10-27, 12:09:

that there is probably minor connection difference between the two boards (if I have more spare time looking at pictures from the 2 boards probably I will even be able to locate those differences - even currently I think it might be just the extra Address lines on the PROM like "A15" that don't exist on 27C64, but are there on 27C512)

actually, not huge time required to confirm at least that one:

* on MQ card A15 pin of the PROM is connected to VCC, clearly visible on the pictures from here:

https://wiki.preterhuman.net/Music_Quest_PC_MIDI_Card

more specifically - the 2nd picture of the back side of the card

* on MDR-401 at least based on the pictures I found, I am 99% it looks like A15 is left unconnected - that I guess is the reason 27C512 fails in MDR-401. Also, maybe that is the reason 8KB ROM fails in MQ card and it needs to be repeated 8 times one after another in 64KB 27C512 - I am sure smarter people than me even can explain why that is based on the fact A15 is always high (as hardwired to VCC) on MQ card. (I suspect that makes MQ read last 8KB, i.e. 56KB to 64KB from those 27c512 instead 1st 8KB from 0KB to 8KB, because higher address A15 bit is always '1').

Has somebody actually asked the italian seller about the card?
If anybody should know it... its the seller...

another very obvious difference that doesn't require hours to track it down: the PAL chip counting from right-to-left ("right" - being the bracket):

3rd pin is not connected on MQ card to an ISA gold-finger and it's connected on MDR401 card to an ISA gold-finger. considering the PAL chip is sending the IO Port requests via the ISA bus to the Zilog Z8 - maybe that is potential reason simple firmware ROM swap is not working. however, maybe that is the pin of the PAL that is selecting between IO Port 0x300 and 0x330 and that is done with jumper on the MQ card, which can explain why it's permanently connected to an ISA gold-finger on MDR401, which is fixed to 0x330.

mattw wrote on 2023-10-27, 12:19:

[...]Also, maybe that is the reason 8KB ROM fails in MQ card and it needs to be repeated 8 times one after another in 64KB 27C512[...]

Almost all PCMIDI related info except random internet pictures originate from me so I can answer you this with 101% certainty: the original mq card rom device is an OTP 27C64 , the firmware is 8kb in size.

8xROM is repeated to fill the ENTIRE 27C152 eprom , if you use 27C256 you repeat 4x , if you use 27C128 you repeat 2x.
This is ancient trick when you burn data that is smaller than the target eprom and it ensures the eprom will work no matter how the actual PCB is designed.
So nothing new here or tricks or random voodoo that require repetitions of rom data.
The 64kb image you found in my archive is 64kb because this is what I was using at the time: 27C512 - also stated inside the text that accompanies the archive.

For my testing above I used a 27C256 , mdr rom data repeated 4x times and you can rest assure this is not why things do not work.

edit:
cards are obviously different design, with different pinouts and different connections - there is more than 1 way to make something even if it uses more or less the same components

I am looking again at the MDR pictures and I see that the ISA fingers are not gold plated? can someone confirm?

keropi wrote on 2023-10-27, 12:51:

there is more than 1 way to make something even if it uses more or less the same components

true, but more or less the same components means the difference could be just in the connections between them and someone that is not complete zero as me when it comes to digital electronics design can figure out the differences and explain it. after all Zilog Z8 is Zilog Z8 - it's running the ROM code in both cases, the reason it fails to respond has to be broken communication between the ISA and the Zilog, i.e. how that connection is made on the boards and the main difference has to be there. Anyway, as far as the hardware design is concerned that is away over my head and I don't want to speculate further - as I simply don't really know. plus even that answered, it's not answering if MDR-401 really existed in 1990.

This is still a mystery - but I believe this card existed in the 90s at least in Italy (it's a fact that Italy back then was big on pcb cloning/production)
but that's just my guess.... I also do not believe this is a softmpu port (extremely unlikely IMHO) but ultimatelly all we have to go is the seller's word and guesswork.

As to the vetusware upload I have some new info: I was contacted on another forum by a person that reported that he uploaded the file there after he bought the mdr card from ebay - so it was not the seller that did that as the person is 100% trustworthy and an old collector that frequents relevant forums. Maybe they will confirm themselves here but I will leave that to them.

mattw wrote on 2023-10-27, 12:09:

also, no one pointed out case in which MDR401 is better or worse than the currently existing solutions, i.e. what is its level of compatibility.

I also got an MBR-401 from Italy.

While looking for incompatibilities,

PoP 1 doesn't work.

In a 386 system with an SB 2.0 and a Roland CM-32L,

Sierra games run well. The sound is comparable to other MPU-401s like the MQ or the original.

download/file.php?mode=view&id=233144

download/file.php?mode=view&id=233143