I too am interested in the $59 on '808 project. I'll dig through my notes when I get home, but I did do a comparison at one stage between the 730 schematics and the 1227165 (same as '808 but has 8192 baud ALDL) and there where a few differences in pins, but most of them where there. From memory hardware mods where required to get the fan output working and a couple of other things, as the pin was on the processor but was N/C. Is it possible to reassign output pins in the software and create, say a 165/808 patched code 59, as I know this would benefit a lot of people. There where only about 3 or so pins that needed to be swapped/reassigned in total to avoid any hardware mods on the board. I'm still not sure if ALDL would work though even on a 165, it would require testing.

The ALDL thing, whether hi vs. lo speed, is primarily a software thing.
anything, really, that the '165 does with ALDL, $59 should be able to as well.

What pins are the ones that don't seem to work, other than fan?
Fan doesn't surprise me, I guess. Temp switches will get around that.

The 165/808 are even more pin-limited than the 730/749, so we might get into situations where we're running out of pins for things.

As you mention, the proc and a lot of the underlying hardware are the same, just not everything is populated.

Dig

QUOTE:
Originally posted by Sammy

ok few pics for yas...

firstly the reluctor setup inside the dizzy ...


when i said timing all over the place that was with the timing light on the timing marks when it was running .... but i'd say part of that was because of how rough it was running...

as for the map sensor .. at the moment this is all i got to show you what it is
other than that i can tell you it has a different plug to the 1 bar it replaced ... ie. the tabs that allow it to only go in one way are in different spots for each pin.

i'll get this freescan and have a whirl and let you know how i go.

Neat setup. Looks odd to me to see the brake booster on the other side.

Anyway, the setup you have should work just the same as our distributor v6, unless there's something really odd with that ignition module.

Timing will move around quite a bit if the idle is unstable.
We really won't know too much more until we get it to read data, though.

Did you guys try Freescan, to see if you can get data read back?

I'm looking at the BUA code now, which was native to the '165. ($32)
I haven't seen anything that jumps out at me, that would lead me to believe that the hardware is lots different.

RAM seems adequate as well.

Dig

Hi Dig,

I had a look through some of my notes (albeit briefly) and the biggest difference between the 165 and 730/749 is the 749 appears to have a few extra IO expanders linked through the SPI bus. the 165 is missing the 5154, which has the txenable pin for aldl. I need to double check these, but at a quick glance looking at things from a $59 perspective:

txenable needs to move from 5154 pin 20 (out0) to main cpu 4980 pin 19 (MUX
tcc needs to move from pin 6 on 4980 (F4out) to 5148 pin 4 (E9out)
ac relay needs to move from 5148 pin 3 (F1) to 5148 pin 52 (F5out)
fan needs to move from 5148 pin 51 (fan) to 4980 pin 17 (F8out).

there are a fair few more to come.

From a 165/808 perspective:

use b9 as p/steer idle up if needed.
D12 is wastegate control
D8 will be the wideband input.

another option for the a/c relay is moving C2 to A3 (moving it in software would be better though )

please note, I really need to verify these suggestions as I was very tired when I made the preliminary notes....

Edit:
I bought a folder today to organise my notes and went through them. Comparing the 808 to 749,we are missing the 5154, 5153 and 1 4984 ic's.

Things that (should) work as is: TPS, IAT, MAP, CTS, O2, SES (CHK ENG), AC request, park neutral switch, knock sensor.

Things i'm not sure about:
IAC. The controller is the same, but the 808 uses an iacenable pin on the 4993, this is a limp mode pin on 749. this iac pin is used as a MUXA pin on the 749's cpu.

The vss input on the 808 is the same as the 749, but the 749 uses a different circuit normally. Could be a software flag?

ESC pin on 749 is on 5153 expansion ic. this goes to 4988 on 808, on the same pin that the 5153 would connect to on a 749. fault1819 on the 5153 is also a problem, this is used in the injector circuit and connects to main cpu on an 808, pin 3/4/5, ued as MUXC, IIIUCS, GLUCS on 749. otherwise the missing 5153 causes no other problem.

Your TCC (F4out) goes to INJA pin on our ecus. So the TCC pins definitely need to be looked at.

I also know this is all going to be very confusing, its much easier to understand with a copy of both the 808 and 749 schematics + generic wiring for each ecu in hand.

dig - save the image to your pc and mirror it, that will sort the booster out [:p]

i haven't tried freescan yet, i will see if i can download it and have a play over the next few days and let you know.

as for all this schematic talk above its made my head spin and im now going to bed [:-crazy][:-yawn]

ok just a quick one, looking at freescan, what model car should i select for this or should i just try one at a time ?

QUOTE:
Originally posted by Sammy

ok just a quick one, looking at freescan, what model car should i select for this or should i just try one at a time ?

GM610B.

QUOTE:
Originally posted by XR Pilot

Hi Dig,

I had a look through some of my notes (albeit briefly) and the biggest difference between the 165 and 730/749 is the 749 appears to have a few extra IO expanders linked through the SPI bus. the 165 is missing the 5154, which has the txenable pin for aldl. I need to double check these, but at a quick glance looking at things from a $59 perspective:

txenable needs to move from 5154 pin 20 (out0) to main cpu 4980 pin 19 (MUXB)

there are a fair few more to come.

Aha... the plot thickens, as they say. Thanks for sharing!

Txenable is why it's not working... the only time that UART can TX, is when something was flipping muxB to do A/D. Bad Yams.

We can fix that in the ALDL/transmit routine, but it brings up a problem with that mux... everything on it as far as a signal is pretty much invalid, since the A/D read routines are expecting to get that from the mux'd A/D. Makes me wonder, too if the other A/Ds were being addressed right in the code.

We can fix that, too, but it's a bit more daunting. The upside is that doing so would eliminate a few of the re-pin issues.

This would explain why it would run, but kinda crappy.

How are you with code? This will need some additional programmers to get done, as I'm spread kinda thin right now. I might be able to hack the ALDL stuff in the short term, though.

Dig