There is a built in ground bridge on the pcm insulator[which you dumped I guess] The contact point is a big circle at the back of the boards, near the centre.
Just put separate grounds on the connetors and you will be good for testing. I think there are some separate grounds on the boards, but you will have to measure it.
I had a look at the OBDII communication section of the Tside and thought it might be worth posting a short description.
OBDII communications are handled by the IC marked:
54545T40
20686
466 {?delco?delphi?other? logo}
I believe this part to be based on/similar to the Harris HIP7030A2. It looks to be based around a 68HC05 with custom interfaces running with an external ceramic resinator of 4MHZ. Pinout is not the same as the Harris part and there are differences. The external input from the data link connector has a capacitor to ground and then a current limit resistor in series for protection. In my notes I have referred to this part as the data link controller. There is a bit of confusion because it shares the TLA with the data link connector, but I will get by...
The Tside reacts to messages targeting both $10 and $18. I believe that messages targeting the Eside are first received by the Tside and then passed across the SPI. It is worth noting that the ALDL is connected through the 30position ribbon cable and connects directly to both T&Eside processors.
My project is progressing very slowly. It is hard to trace circuits through the board. Other things like the conformal coating, components in the way and so on make the job even more difficult. As always I will appreciate any corrections or information you can add to these posts.
GB_CS.jpg
Reverse engineering of these boards is a difficult process and I need all the tools I can get my hands on. One tool I have not used yet is the 68HC11 bootstrap mode. In this mode, you can download a small bit of code into the internal RAM and execute it. Care needs to be taken to ensure the code defaults chip selects and things like that so that other parts on the board won't fight each other but in most cases GM defaulted all these with resistors anyway.
The process for loading is quite simple... The binary code you want to put in ram is pre-fixed with a single $FF character and sent to the serial (aldl). I need to do a bit more checking of the configuration of the serial connections to make sure this can all work out. It is important that when you are bootstrapping one side (E or T) that the other side doesn't try to chat... that would mess up the download for sure.
Please don't attempt this unless you are sure of what your doing. The HC11 reference manual goes into complete detail on the use of this mode. Myself I am trying to decide if to use this or use the built in download and execute through the EL327 clone.
I have two photos to show where the mode lines are. If pulled to low when reset is released, the part will enter bootstrap mode. I have used this technique with success for my Optispark simulator.
It looks to me like this could be used to resurrect a bricked module without de-soldering the FLASH part.
Second picture is full of glare, but you can make out the locations. I will re-take that photo for future work
-Tom
RB_SS.jpg
GB_SS.jpg
I have looked through some of the input signals on the Eside and analyzed the circuit. I have made some assumptions... take care when using the information.
The signals in question are VATS, Traction control and MAF. There is a chip and a good number of resistors and capacitors involved with receiving each of these signals. This is done to control ESD, condition the signal against noise and provide hysteresis.
I have identified the chip to a functional level. It's pinout and function match that of LM339 quad comparitor. Comparitor chip is marked:
014
M66DB
89551
LM339 quad comparitor or a variant that shares the pin configuration.
This chip is used to provide a level switch at 2.5V instead of the standard vil/vih of the receiving part. A network of resistors provide hysteresis to avoid noise problems on the lines.
The outputs of these signals connect to a 68pin PLCC chip that has not been identified and is probably a custom chip for GM. In my notes I will name this part TPU for timer/processor unit.
Input:Output
VATS:TPU pin 20
Traction Contro:TPU pin 21
MAF:TPU pin 22
C1(RED) pin 26:TPU pin 23
The components to connect C1 26 are not installed on my board set. Unsure if/where it was used.
Here are some photos to indicate the components.
MAF_RB_SS.jpg
TC_RB_SS.jpg
VATS_RB_SS.jpg
RB_CS.jpg
-Tom
Ive been looking a little closer at the high and low res signals. Later I will post the ins&outs but I have run into a little snag. There is yet another unidentified part and I could sure use some help. Way back when I had a bunch of contacts in LSI logic / Symbios but with M&A activity I have lost touch. There are a couple of Symbios parts on my board. One is the part that drives the injector transistors. It's the other one that needs investigation. It is probably a small gate array I expect. Here are some details...
Part is marked:
609-3700521
CP08978
144436
9638N
SYMBIOS LOGIC
726
I will attach a 90% hookup wire list for the part. Only 90% because it is SO hard to trace all the copper.
What would help is to know which PCMs have the part. For example if it is on all boards with Opti-spark that is a tell. Also interested if it is used anywhere else (I expect not)
At first I considered if it could be a phase lock loop. This is still one possibility, but I don't think an engine controller needs that kind of resolution. I now believe it is a counter timer chip. It has interconnections with the mysterious 3rd processor TPU chip.
I have not yet connected my optispark simulator but once done I will set the scope on this part and see what sense I can make of it. Some control of this part is handled by the port replacement chip (66285). It would also be nice if anyone knows the pinout for the PRU to send that along.
-Tom
I am adding a picture of the board (red/black connectors...eside) with the component highlighted
DIG1_RB_CS.jpg
Last edited by Tom H; 02-03-2019 at 12:15 AM. Reason: Add picture of where the part is...
I suspect that part might give low res signal to tside. On 94-95 pcm the tside gets an opti signal at IC3I [TIC3H-register]. It doesn`t seem to be connected to eside anyway.
The conversion is also different
tside low res conversion is
2949120/x = rpm
eside conversion is
983040/x = rpm
0.015234*x= msec
Or exactly 3 times lower resolution.
Here are some high res pictures of the 94-95 f-body pcm.
Here are some details regarding the Eside drivers. I wonder if anyone has lifted one of the drivers from the heat sink to get a look at the manufacturer and numbers? If yes please share...
I traced the connections to the two 23 pin power SIPs. I believe they are low side drivers perhaps similar to some of the common available parts. I need to look into the code, but expect to find that the parts detect when a driver is open or shorted to power. The pdf contains the connection info.
-Tom
Posting Permissions
Reply With Quote
Bookmarks