I have been staring at the addressing decode that is done for chip labeled U3 on page 1 from the Ludis 730 schematic.
According to the schematic it should decode to address space:
0x1800 - 0x1FFFF

Looking through disasseblies of the $8D shows no references to accessing this address space.

I tried shifting the address bits (address space) around in theory to see if maybe the schematic is in error. I still can't get things to line up. Anyone else notice this?

I have to get the bench going to figure this one out.

J

 

Ya, those numbers look right to me. I haven't looked at the hac though.

Maybe there is some kind of table lookup for this?

Is the chip on the 730 board? Maybe its a 749 thing only.

 

It is a direct connect to the external bus so there wouldn't be a look up.

Good point though about checking if the device even exists. I will take a look tomorrow to see if it is even on the PCB.

Thanks,
J

 

I've heard there is room on the 7730/7749 boards for extra RAM chips. Here's a post from Ludis on the diy-efi board (not sure when this was made) :
There was also mention of extra RAM in another ECM (1228708) that was virtually identical to the 7730 in the first post by funstick here: https://www.thirdgen.org/techbb2/sho...hreadid=250915

I assume they're talking about the same type of thing. If nothing else, it might help you figure out what needs to be done to the 7730.

 

u3 is either a rom or ram and can be configured by setting it up the the code or with a pull down or pull up. im not sure how to do it. the thing is that the space is empty on every pcb ive seen to date but could be populated. it is a 24pin slot. id imagine that at some point gm conteplted putting a onboard rom on the board then opted against it.

the pin spacing is perfect for a 2732

 

It was definitely NEVER intended for ROM as the main program storage. I think GM used it for extra RAM since there is a very limited amount in the MCU.

There shouldn't be a need for any resistors if the decode chip is populated.

J

 

remeber 2 things here.

1 the controller is in fact a custom based 68hc11 piece.
2 there are lots of thing on gm controllers that are resistor decoded.
3 i would think GM might store boot loader there for debugging of the development boards. would make lots of sense in reality.

dont assume the hardware works exactly as the schematics show it to be. with all the peripherials that are non stnadard and the custom micro anything is possiable. also where you can put ram you can put rom. onyl requires some minor code changes if you have a micro designed that way. would be the only reasonable explanation for a 24pin dip style layout on the board you know. 27c32 is the only thing ive run across that fits the spacing. there however might be some sort of ram out there for that board location but why in the hell you you make the foot print so big ???

also the memcal itself does handle some of the mcu setup. it also has some code for the i/o imbedded into the mcu.

 

Heh, my old post must have been eaten during the crash.

funstick, that pinout is a standard way of hooking up RAM, look at how the WE (write enable) lines are connected.

Have a look at the 5516 on this page:

http://www.kingswood-consulting.co.u...m/mem-ram.html

 

the 5516 does look like ti will fit. hmm. my personal feeling is that the it was used it for bootloader during the development levels of the board design. ive run this the schematic etc by sevral other ee's they seem to express that it was the most likely use.

however on a good note are those 5516 chips avaiable
?????? that would be interesting.

 

http://rocky.digikey.com/scripts/Pro...=CY7C128A-15PC

Not the 5516 but the same pinout.

Very fast, I don't know what the bus speed is for the 68HC11 but I guarantee that this can handle it.

$6 from digikey.

There's even some NVRAM at digikey, $20 per piece but it doesn't require any battery back up. That might have some neat applications.

 

Back from the dead...

I noticed this while digging for 8d i/o info in the 8F.
Is it what your refering to for 1800?

 

That is exactly the memory location I was referring to. What ECM does the $8F code reside in.

Also, the CY7C128 is not a drop in part. It is a .300 package. The PCB is drilled for a .600 package. You need to get an adapter board(make one) that converts to a .600 inch footprint.
The speed of the part is fine.

The decode chip is a standard LS/HCT/... decoder.

J

 

8F looks to run on a 7730 to me.

Ludis

 

I read PDIP and ASSumed.

 

If anyone is interested, Digikey sells a .300 to .600 adapter for about $10. You can almost fit the entire $8D data tables into a 2K memory. The overflow is about 40 bytes.

It would make it real easy to do ALDL modifications to the data tables in a '730. Install an non-volatile RAM and stuff would not get wiped out at shutdown. Easy to install due to no de-soldering. Cost is cheap........probably $22 + $11 = $33.
Essentially, you could "emulate" via the ALDL on the 730 ecm for about $40 bucks. Good project for those who like to do coding on the 68HC11.

J