"Let's make GD ROM emulation happen" Facebook group.

well then the question becomes if we have total control of the GD bus then why screw with G2

 

for device g2 bus we need fpga for converting signals to sd/hdd , as "bootloader" we need bios-Flash + sd mods for dreamshell and sofd driver. and we can leave original gd-rom in nateve place)))

but we need get good info about how g2 bus works, i mean high speed transfering (dma), cuz it uses PCI-like protocol, i think it's not well-discovered yet. only low speed pio modes

so gd-rom emu is more realable...

 

I've just hooked up the IDE cable to the DC.

It's not working yet because I need to supply CDDA Clock from an osc module or the FPGA to get it to boot (not tried yet, it's getting late here).
You can plug the GD Drive in temporarily to get it to boot to DS, but the stock BIOS freezes because it's looking for the GDD.

Basically, the original syscalls need to be removed or redirected so we can access a CF / HDD instead.

I've written some basic code for accessing an IDE device using DMA.
Hopefully I will get something working in the next few days.

I've been looking for a cheap PCB prototyping company for the IDE adapter too (I still had to solder the IDE cable to the DC)...

With many of the companies which are supposed to be cheap, it seems to cost at least GBP £70 (USD $110) for 25-30 small boards!
I just found a nice company in the UK though - it should be only £120 for 150 boards!

I don't quite have the money for this atm. Would it possible to do some sort of group buy if enough people are interested?
The obvious issue is that no software or firmware exists for these yet, so I doubt many people will want one?

Also, somebody PM'd me for a copy of my Eagle PCB files (which I gladly sent to them).
I've posted them below if anyone wants to carry on with it or use the libraries (GC socket is included too, sorry Sega fans!).

This is completely untested of course, and I can't be 100% sure that the plug / socket layouts are correct etc.
I'm confident that the IDE and DC pinouts are correct, but everything should be double-checked before even thinking about manufacturing!

It needs checking to make sure it will fit the DC too (it looks like the bottom edge and corners will hit the G2 port / fan).
I suppose the board could be extended out towards the DC PSU instead?

Here it is anyway...

http://www.filedropper.com/dcg1ideadapter-eagleozone30-7-12

(please let me know if it works on the Freeware version of Eagle - I used a "naughty" version to make these files) :sneakiness:

I don't really have the patience to route boards manually, so I tend to use Eagle's messy autorouter.
I will try routing it with Electra at some point.

Hope all of that makes sense. :concern:

OzOnE

 

I've said here I would support and you can count me in.

Let's see how many people are willing to jump and we'll split the value.

 

Use freerouting.net and never look back.

 

I would be interested for the support.

As soon as a way of getting software in through some means or a firmware package up and going, I'd be interested in doing some testing and reporting my results. Perhaps some software can be written and loaded up through Dreamshell. If the board supports connecting the GD ROM on top of the IDE breakout board, then maybe a disk with Dreamshell on it can be loaded up to load the software of firm ware from the disk or from the serial SD reader. Then a simple matter of accessing the connected IDE drive from there. Those with a BIOS mod done have the option of burning Dreamshell into the BIOS so it would still load without a GD drive connected. Maybe having some hacked firmware in there somewhere to make the Dreamcast "think" it has the GD drive in there when it really has the IDE device in place. Again, load all this from an SD card with the serial reader to maybe make it simpler. With the right coding, it should be possible to pool the command lines from the SD card for the CDDA clock. Not sure on that. Perhaps it could fake the needed coding to boot past the BIOS freeze up seen without a GD ROM in place. I think having it loaded up on an SD card will make future revisions easier to do by simply dropping the new files in place of old ones versus burning chips or having to JTAG an FPGA.

 

More than happy to chip in for the support I can Pledge £40.00 so we are a 3rd of the way there already!

When do you want the cash? I can paypal it to you today if needed.

 

Pm incoming

 

I'm definitely willing to donate, but can't do so this month (August). If it still helps, I'll donate in september (unless I somehow manage to get some more money this month, but it's unlikely).

 

I get lost with the technical jargon but if ultimately what you're shooting for is loading games from a hdd with good compatibility (unlike dreamshell) then I am more than willing to donate for your cause.

 

Oh dear, what have I done! lol

I'm VERY grateful for the support, but I was thinking more in the way of a group buy for the boards (eg. work out an intro price, then pay for one / two / how ever many boards you want).
I'm hesitant to take direct donations for something which might not even materialize.

Again, I'm worried that there is no software yet (for direct connection), and that is actually the difficult part. I don't want to let anyone down who has donated because I can't make any promises.

As I've said to others via PM, I'm not an experienced programmer (in C or in Verilog), so it may be a very long road.
I can't even say if all of this stuff is possible without some expert coding. Maybe we should ask the DCEmulation guys to help?

I also need some help with routing the adapter board. It looks simple enough on the face of it, but it always takes me ages to route boards.
I'm running it through freerouter atm (thanks @mikeryan!). It looks like a great piece of software. It's a bit slow though. lol

EDIT: Just to be clear, the IDE adapter board / direct connection is a separate project from the original FPGA project!!
The adapter will be very useful for the FPGA though, so it may help in getting it to work.

@sonicdude - Yep, the idea now would be to redirect the GD syscalls in the stock BIOS and make an HDD version for Flash BIOS.
Unfortunately, I don't think CDDA streaming will be possible when using a direct IDE connection. Not without a small FPGA / CPLD at least.


So, I hope nobody will be offended, but I'll try to get HDD access working first then we can think about it from there.
If I can get basic access to work (especially with DMA) then we're most of the way there. I can then ask SWAT if he can add support to DS?

For better compatibility, we need to rip apart the GD syscalls and completely understand them so they can redirected to our own routines.
There is a lot of info on Marcus' site, and in various emulator sources, I just need to get my head around it all...

http://mc.pp.se/dc/syscalls.html#vecBC

The idea is to redirect the syscalls in a way that's completely transparent to the DC. The routines will DMA from HDD / CF into memory and the DC won't know the difference.
The GD status stuff needs to be handled properly, then each command assigned an ID etc.

Anyway, thanks again guys.
I hope to have basic HDD access working by next week. We can still figure out the PCB layout for the adapter in the mean time.

OzOnE.
P.S. I just received a few GD connectors from Molex today! 8 sockets and 8 plugs.
I tried them on the DC / GDD and they fit perfectly, so they're 100% confirmed.

The part numbers are...
0525840579 (GDD side - I call it a "plug", they call it a Receptacle) http://www.molex.com/molex/products/...cts&Lang=en-US

0534080579 (DC side - I call it a "socket", they call it a Plug!) http://www.molex.com/molex/products/...CB_HEADERS.xml

 

No problem mate when you need donations give me a shout.

Keep up the good work!

 

You can count on a donation from me for the boards. $30 US. Just let me know your paypal address.

 

I am also very interested in this project. If you need funding, I can afford $30 US, just PM me.

 

Wow, seems there's money in this project...

 

Wow! That's a great project. If money is needed to help this, count me in! :tears_of_joy:

 

I have to read more about this, but if its going to have an FPGA and some nice DC tricks i might donate as well.
Re purpose fpga's is cool, and premade pcb's are awesome. And if this project is going to work its even beter no wait, its the best.

 

i told about "redirecting the syscalls" but nobody pay attention to my words as i remember)) , i see - i'm not so popular unlike OzOnE...


but dont forger, DC bios code making pass data thruogh special register when copying it from bios to the memory. so alternate bios can failure cheking - and g1 bus will be blocked. i'm t sure, if we nop'ed this part of code - g1 will not be activated. it's importan if you wat make modifid bios with your routines.

but for any reasons or another, swat acidently make modiffied bios that can check with success and can load files from sd-mod w/o bloking G1. that's why i recomend use sd-mod + flash mod with special ver. and then some exec file with routines with redirecting syscalls for support your device.... but i dunno what happen when original gd-rom is apsend? i think even modifed bios ask original gd-rom board?

also, direct-hdd mod will have same minuses like sd-mod - high/low mem load- it need some area for FS (fat32) handling and other.. so it's justa simple way to use hdd or ata devies for launch games or access data... until FPGA-evice still in progress)))

 

Yes cybdyn, I know you know your stuff. Its just that I thought you were focused on PSIO at the moment.

 

PSIO and any other *-IO has same parts. thats why i plan make my apha for both consoles. also i plan use more chips than it actuly need ,
but it helps figure out what we need in beta version.

so soon i try help OzoNe where i can , for example SD module for interfacing with sd card, show him or share code for pc (win32) for debuging his devices , he can see what data in registers and mem..