Previous Thread
Next Thread
Print Thread
Page 2 of 3 1 2 3
Re: LED Panel Support on System 12 games #11133 02/21/05 12:30 AM
Joined: Feb 2005
Posts: 16
T
Toonces Offline OP
Junior Member
OP Offline
Junior Member
T
Joined: Feb 2005
Posts: 16
Sorry about that. Links provided.

Re: LED Panel Support on System 12 games #11134 02/25/05 02:15 PM
Joined: Feb 2005
Posts: 16
T
Toonces Offline OP
Junior Member
OP Offline
Junior Member
T
Joined: Feb 2005
Posts: 16
I have been looking into how to get the LED programmed with anything other than the default messages displayed above. After some research, the protocol for programming the LED is still unknown. What I do know is this. The Lower I/O panel which is virtually the same as the upper I/O panel minus the LCd controller is connected via a ps2 type connector and a 4 wire cable. The input feeds into what appears to me to be a custom chip with the markings C77 3202A02F on pins 21 & 22 and the other 2 wires end up at a resistor pack and it is difficult to see where it goes from there. The custom C77 acts as an 8 bit CPU to the LCD controller which is a sed1351. The controller can interface to either an 8 bit or 16 bit cpu and from how it is configured tells me it is interfaced as an 8 bit CPU. The custom chip is also interfaced to a 256Kbit RAM chip and an Atmel AT29C020 eeprom. The Atmel chip looks like it is connected in parallel to the RAM chip and both feed the sed1351 DB0-DB7 input data lines. The sed1351 also requires addressing which also comes from the C77 chip. I am not sure what purpose the RAM chip would have if it is in parallel with the EEPROM. Some sort of scratch RAM? Since I don't have the correct tools to dump the EEPROM I do not know the contents but I am fairly certain it contains the graphics that would be fed to the LED panel. With a 2Mbit chip, you could store quite a few screens of data.

On the output side of things, the LED panel operates in a 4bit grayscale mode with a 64Kbit RAM chip of it's own giving you the possibility of 4 grayscales and up to 256 X 128 display. The LED panel itself is only 96 X 16.

The biggest stumbling block to interfacing the sed1351 directly to the PC's parallel port is that the outputs, instead of going directly to the LED panel go to an Altera Max 7000 chip (EPM7064LC68-15) which makes it very difficult to trace which pins on the output connector perform which function. There are 11 lines of a 20 pin connector used and 4 of those should be data (UD0-UD3). the sed1351 also outputs 5 control signals (LCDENBL, XSCL (clock), LP, YD & WF. That probably leaves GND and +5V for the other 2. It's pretty difficult to tell what is actually used by the LED array though. It looks like Pin A1 goes through a filter and goes back out on pin B1. There's also this same loopback at pins A3 & B3 and A9 & B9 so that would only leave 5 signals that look like they actually go anywhere and that is back to the Max 7000.

As far as game interaction, I have not yet been able to get a hold of a System 12 based game (I was hoping to pick up Soul Calibur or Tekken 3 since I figured they would have the best chance of interaction from what I could find) but have not yet been able to get one. My guess is that the 2Mbit EEPROM contains stock images that react to player control movements. The player 1 & 2 controls would normally be connected to the lower I/O board and they feed the C77 processor on that board. I beleive that based on movement of the joystick or pushing a button would cause the LED display to show to LED characters fighting but it would necessarily match what is on the screen. The lower I/O board most likely feeds the inputs from the controls to the upper I/O board which in turn pulls a graphic out the the EEPROM and displays it on the LED. As you move, it changes the displayed picture and you get the two little guys fighting on the LED. There could possibly be a character set in the EEPROM as well since the sed1351 is strictly a graphical controller and does not contain it's own. I supposed that system 12 games and maybe system 246 games or even Naomi games might also have some way of outputting the name of the game which in turn would put it up on the LED. With the lower I/O board connected to the NTSC board via USB it probably isn't a difficult thing to do.

Anyways, just wanted to give an update on what I found. It does not appear it truly interacts with the games but rather reacts to them. I have a spare set of boards and LED panel so maybe I'll try actually hacking one up and see if I can get it connected to a PC and download graphics directly to the sed1351.

Toonces

Re: LED Panel Support on System 12 games #11135 02/25/05 05:04 PM
Joined: Nov 2003
Posts: 804
S
smf Offline
Senior Member
Offline
Senior Member
S
Joined: Nov 2003
Posts: 804
Quote:
Originally posted by Toonces:
The input feeds into what appears to me to be a custom chip with the markings C77 3202A02F on pins 21 & 22 and the other 2 wires end up at a resistor pack and it is difficult to see where it goes from there. The custom C77 acts as an 8 bit CPU to the LCD controller which is a sed1351. The controller can interface to either an 8 bit or 16 bit cpu and from how it is configured tells me it is interfaced as an 8 bit CPU. The custom chip is also interfaced to a 256Kbit RAM chip and an Atmel AT29C020 eeprom. The Atmel chip looks like it is connected in parallel to the RAM chip and both feed the sed1351 DB0-DB7 input data lines. The sed1351 also requires addressing which also comes from the C77 chip. I am not sure what purpose the RAM chip would have if it is in parallel with the EEPROM. Some sort of scratch RAM?
The c77 could be similar to the c76 used in namco system 11, it has an internal program which we can't dump. The chips being in parallel is pretty common, the data & address lines have to be connected to a common bus. You don't have to lay them out like that but the simpler the better. The chip select lines will be wired seperately though (I hope).

smf

Re: LED Panel Support on System 12 games #11136 02/26/05 03:12 AM
Joined: Mar 2001
Posts: 16,471
R
R. Belmont Online Content
Very Senior Member
Online Content
Very Senior Member
R
Joined: Mar 2001
Posts: 16,471
IIRC the C74 (S22), 75 (FL & NB-2), and 76 (S11) are all known M37702s so it wouldn't be surprising if C77 was also.

Re: LED Panel Support on System 12 games #11137 02/26/05 03:33 AM
Joined: Oct 2002
Posts: 1,017
M
MooglyGuy Offline
Senior Member
Offline
Senior Member
M
Joined: Oct 2002
Posts: 1,017
If it has an undumpable internal ROM, though, what's with the EEPROM chip?

Re: LED Panel Support on System 12 games #11138 02/26/05 04:11 AM
Joined: Mar 2001
Posts: 16,471
R
R. Belmont Online Content
Very Senior Member
Online Content
Very Senior Member
R
Joined: Mar 2001
Posts: 16,471
The (E)EPROM likely contains a font or other data - external ROM is much cheaper than internal, which is why the C7x MCUs keep all their data (and some of their code) outside.

Re: LED Panel Support on System 12 games #11139 02/26/05 06:37 AM
Joined: Feb 2005
Posts: 16
T
Toonces Offline OP
Junior Member
OP Offline
Junior Member
T
Joined: Feb 2005
Posts: 16
That is some very good info. However, the C77 chip is a 64pin chip and from what I could find, the M37702 is an 80pin device. What I did find is pretty useful though. I did some searches on the M37702 and found that it has 2 built in serial ports that can be used in rs-232, rs-422 or rs-485 modes. I also found a few datasheets for serial controllers based on the M37702 which makes a good case for the C77 being used as a serial com device. Furthering that, when the M37702 is used for serial comms, a non-standard clock rate version of the chip is used which operates at 14.746Mhz. Using that speed of clock allows the chip to generate standard baud rates. This happens to be the speed of osc1 which is connected to C77. So it must definately be serial comms happening between the I/O boards. It should just be a matter of what kind now and figuring out the baud rates, etc. with any luck it will be standard rs-232.

Re: LED Panel Support on System 12 games #11140 02/26/05 10:51 AM
Joined: Feb 2005
Posts: 16
T
Toonces Offline OP
Junior Member
OP Offline
Junior Member
T
Joined: Feb 2005
Posts: 16
Doing some additional searches for a smaller 8 bit cousin of the M37702 turns up the 38002 series chips which are essentially 8 bit versions of the M37702. It has a single built in UART, ROM & RAM as the 16 bit chip and comes in a 64pin QFP. From the datasheet, it appears that this could be the chip but in the QFP package the serial pins don't match up with what is being used. They are on the same 8 bit port but wrong pins.

The search continues.......

Re: LED Panel Support on System 12 games #11141 02/26/05 12:34 PM
Joined: Oct 2002
Posts: 1,017
M
MooglyGuy Offline
Senior Member
Offline
Senior Member
M
Joined: Oct 2002
Posts: 1,017
If everything else lines up, you can rest assured that it's the same part. The C7x series being repackaged versions of other processors doesn't necessarily imply that the pinouts will be exactly the same.

Re: LED Panel Support on System 12 games #11142 02/26/05 01:51 PM
Joined: Feb 2005
Posts: 16
T
Toonces Offline OP
Junior Member
OP Offline
Junior Member
T
Joined: Feb 2005
Posts: 16
I think I am on the right track with it being an 8 bit version of an M37702 type of chip but the pinouts just don't seem to line up anywhere. From a functional standpoint it stands to reason that the type of microcontroller would be what is used though.

Growing tired of looking at the numerous microcontrollers that are out there and realizing I am not familiar enough with them to make a more intelligent guess as to what the chip is I decided to get a little more basic and pulled the cable out of my machine and did some continuity checks to get the pinout of the cable. There are 3 wires connected. from the cable, it appears to be the minimum required for a serial connection. Pin 2 on one end goes to pin 3 on the other and visa versa. Pin 6 is connected straight through and is ground on both ends. Pin 2 & 3 of the connector go to pins 20 & 21 of C77. This leads me to beleive that it's probably standard serial at a baud rate of less than 9600 since there are no flow control lines. Interestingly, pins 4 & 5 of the connector go to pins 16 & 4 of C77 but are unused in the cable that connects the 2 i/o boards. perhaps a different cable is used to program the flash rom and the other 2 pins are used when doing so.

I guess my next step should be to see what kind of voltages are on pins 2 & 3 of the connector when the board is powered up and see if they are compatible with what is coming out of the serial port. I think the serial port uses 10V or something like that and I wouldn't want to plug a hacked cable into it if it's expecting TTL levels.

Page 2 of 3 1 2 3

Moderated by  R. Belmont 

Who's Online Now
3 registered members (Alegend45, MacLover490, Dorando), 163 guests, and 3 spiders.
Key: Admin, Global Mod, Mod
ShoutChat Box
Comment Guidelines: Do post respectful and insightful comments. Don't flame, hate, spam.
Forum Statistics
Forums9
Topics8,763
Posts115,160
Members4,889
Most Online890
Jan 17th, 2020
Powered by UBB.threads™ PHP Forum Software 7.7.3