Looking for a Multiface

[Bryce]

Yeah, I've done that (both types of connector on the Ribbon cable) on my RomBoard, so that I can use it on my German 6128 and my English 464, but with DKTronics devices you don't have that possibility, that's why I had to make the converter.
I use a vice to slowly close the IDC connectors, it's the most reliable way of closing the connector. It also helps to slightly warm the cable beforehand. I usually leave it on a radiator for 10 minutes before putting it in the vice.

Bryce.

[Ygdrazil]

I think somewhere there exists a French project to remake the multiface II from present day components!!

/Ygdrazil

Yeah, I've done that (both types of connector on the Ribbon cable) on my RomBoard, so that I can use it on my German 6128 and my English 464, but with DKTronics devices you don't have that possibility, that's why I had to make the converter.
I use a vice to slowly close the IDC connectors, it's the most reliable way of closing the connector. It also helps to slightly warm the cable beforehand. I usually leave it on a radiator for 10 minutes before putting it in the vice.

Bryce.

[Grim]

I think somewhere there exists a French project to remake the multiface II from present day components!!

The guy behind the idea stopped the project. But you can still find his stuff about it on this website (Schematics, PCB Routing and JED files). There are 3 PAL (programmable chip) in the MF2. One of them could not be decoded.

[OCT]

AFAIK there's little that "can't be done" for age-old tech like this. If the chip in question is just an early PAL and the pins used for input have been determined, today's hardware should be able to patiently feed it every pattern known to maths&man until the boolean table has been figured out.
Multichannel logic analyzer, anyone?

[Ygdrazil]

So you mean i can be done!!! 

To replicate/rebuild the Multiface II device!!

/Ygdrazil

AFAIK there's little that "can't be done" for age-old tech like this. If the chip in question is just an early PAL and the pins used for input have been determined, today's hardware should be able to patiently feed it every pattern known to maths&man until the boolean table has been figured out.
Multichannel logic analyzer, anyone?

[Ygdrazil]

I found this about the Multiface II repro project (In French)

http://www.phenixinformatique.com/modules/smartsection/item.php?itemid=141

/Ygdrazil

So you mean i can be done!!! 

To replicate/rebuild the Multiface II device!!

/Ygdrazil

[Bryce]

Judging by the schematic, it should be quite easy to replicate with modern components. The only difficult bit is decoding the PALs, but as stated above, that's just a matter of running a binary sequence through it and creating a truth table, but obviously you'd need a working original to do this. The rest of the circuit components are still available or at least an equivalent. Unfortunately I don't have one, but if I happen to see one going for a reasonable price, I'll grab it and make a copy.

Bryce.

[Munchausen]

I'm grave digging again. 


I don't know if more stuff was added since that page was last looked at, but all the data for the PALs is there in JED format - someone has already done the decoding, so it should be straight forward to build one. I've added the info to the wiki, including a zip file with all the files/pictures and other info from that page. I've also generated the missing boolean formula file for one of the PALs and included that. It's all on the wiki here.

[ralferoo]

Judging by the schematic, it should be quite easy to replicate with modern components. The only difficult bit is...

None of the inputs to the 74LS670s are described, the text just has the comment "exotique (reste...)" so presumably they never managed to trace where these pins went to. Presumably, they're likely to just be the databus too as I think they're used to record the colour registers?

The OE/WE on the RAM/EPROM are also not connected, these presumably go to one of the PALs.

[Munchausen]

None of the inputs to the 74LS670s are described, the text just has the comment "exotique (reste...)" so presumably they never managed to trace where these pins went to. Presumably, they're likely to just be the databus too as I think they're used to record the colour registers?

The OE/WE on the RAM/EPROM are also not connected, these presumably go to one of the PALs.

I hadn't noticed that. Maybe it can be figured out from the PCB?

[TFM]

And that shouldn't be a problem at all. It's a pretty simple one.

[Johnny Olsen]

Unfortunately I don't have one, but if I happen to see one going for a reasonable price, I'll grab it and make a copy.

Bryce.

Hi Bryce

I have a multiface you can borrow, if you still want to build one.

Johnny

[Bryce]

I've actually managed to get one for myself since then (that post is three years old). I just don't have the time to look at it at the moment. I have 20 other projects on the go. It'll be a while before I get a chance to look at this.

Bryce.

[beaker]

I have 20 other projects on the go. It'll be a while before I get a chance to look at this.

Any add-ons for other retro hardware???  If so, would it be OK to post something on the other retro section as you release them, if you have any spares, as I do like my add-ons and I tend to only ready CPC Wiki forum and the amibay forum at the moment despite having a growing number of different systems, and you do make some really nice stuff... 

[Bryce]

The ones that are for Retro computers are mainly Amiga stuff at the moment. But they're not really "projects for everyone". They are more custom solutions and mods that people have asked me to do.

Bryce.

[Gryzor]

I tend to only ready CPC Wiki forum 

That's all you need, really...

[Munchausen]

I've done a bit more digging on this.


The PAL which has two versions (one with a .EQN provided, the other without) is a 16R4, meaning it is not just simple logic but also has flip flops in it, which looking at the schematic are connected to ROMDIS and RAMDIS.


I don't know how this was reverse engineered, but it could be a technically difficult thing to do because, due to the flip flops, it's not just a matter of running it through a logic analyzer with every possible input and recording the output.


So unless someone can sit down and figure out what the equations are doing and if this seems to make sense given the circuit (and I don't think I know enough about the CPC to do this), I don't think we'll know if this works until someone actually tries to build one.


EDIT: I really know relatively little about this. Reading around I can't get a consistent opinion on whether the 16R4 could be protected or not. If not then it could be that the fuse settings could just be read using a programmer. Maybe someone else has some useful information?

[Munchausen]

So I contacted Giants about this, and he says that the information of the 16R4 is not complete, and it is protected. After doing some research, it seems pretty hard to decode the 16R4 by brute force, and I don't have a multiface, so this approach wont work for me anyway. The decoding of the other PALs should be OK already.


I've had a quick look at the caprice emulator source, and it looks fairly clear, so what I've decided to do is try to add emulation of the whole circuit to caprice, including the little information that is already available on the 16R4. In this way it should be possible to figure out what functionality is missing, and try adding it until the emulation works, thus finding out (in a round about way) how to finish the multiface implementation. I've still got a lot to learn about the CPC though, so this might take me a while.


I'll post updates of any progress I make (if any    ) here...

[arnoldemu]

I think me and Ulrich shared information about the Multiface 2 when we made our emulations in Caprice and Arnold.
I did try to work out some of the I/O port decodings.
I probably need to re-test them to confirm.

[Munchausen]

Is there already emulation of it in caprice/arnold then? I can't see anything (looking at caprice reloaded source).


I want to emulate the whole hardware of the multiface as fully as possible to verify it. I think this means getting the ROM and RAM in the right place, and getting the logic for the PALs and other gates in place so that when updates are made to perhipheral registers they also get updated in the multiface RAM. If this is done right then the routines in the multiface ROM should work correctly.


Looking at the caprice source again (I only had a brief look last night) this might actually be more difficult than I thought. Maybe it would be easier in Arnold? It doesn't look as though caprice explicitly models many of the signals I'd need, I think it all happens at a higher level, so I'd probably have to add quite a lot of code (perhaps implement an expansion port object and then add the multiface on to that). But I need to look more at the source (use the source luke!)


This does seem like an easier way forward to me, especially since I can't quite get my head around brute forcing the PALs...

[dcdrac]

Multiface Two by Romantic Robot for the AMSTRAD CPC 464/664/6128 | eBay

[ralferoo]

This great bit of kit lets you expand the ram of your machine by 8k help with your programing duties among other great features.

Oh, so that's what the multiface 2 was for... 

[Bryce]

So he never actually used it then? Or he just doesn't have a clue? 

Bryce.

[dcdrac]

Ihad one used to use it for disc dumps, then I discovered Bonzo Super Meddler and Discology

[Munchausen]

£36 and 5 bids already, and 6 days to go. Don't think I'll be able to get this one...