Lotus Esprit Turbo definition file help please

I've been working on decoding the Lotus Esprit S4 .bin The Lotus Esprit uses a GM ecu with part number: 01228708 It's the same ECU as some GM v6 stuff and 1991 Corvettes.

The catch is that I don't know assembly. I have managed to identify a bunch of tables with the help of using other GM .bin files and masks like $8D. But I fear am I at the end of my knowledge and ability. I am attaching my tunerpro files I have been working on. This is my FIRST ATTEMPT at a definition file so please don't be too harsh on me.

The file starts out just like GM code, but clearly Lotus deviated from the normal rules pretty quickly. For example the number of cylinder at address $09 is AF, that doesn't match any explanation I have found (00-8 C0-6 10-4 18-TBI)

Some people have suggested using the Code59 stuff and trying to run the car on that. That's a reasonably good idea, but I think the boost control code in the Lotus is better (maybe I am wrong) and the Lotus runs two additional injectors (it uses the Quad driver to run them). I have considered setting up the two additional injectors on a separate fuel system (fuel cell, fuel pump, regulator) and running E85 through them. This would give me the long range and easy availability of 93 octane gas for long trips, but the extra octane of E85 when I'm deep into the turbo boost.

Also in a real dream I would like to convert from a 2 bar map sensor to a 3 bar.

Thanks for any help in advance,

Aaron

Hello, and welcome to Gearhead-efi. You've certainly picked a tough project to dive into. Hopefully you can find enough support to stay with it.

I'm surprised folks are still holding onto the definition files that are out there. This application is 25 years old. Is there really that much $$ out there in the custom chip market for these cars?

I'm not sure about your appplication list for the ecm. The 1227808 was used in many cars but the '90 Vette used a 1227727 or a 1228331 in the ZR1. I believe the 8331 was used in a six cylinder Lotus, maybe the Carlton? But it is not the same as the 8708 internally and will not substitute. Additionally, IIRC Lotus used both the 8708 and the 8707 in Esprit models although I do not know enough about those two cars to know the differences.

One of the Lotus turbo cars supposedly had custom code which, after a series of "secret" steps, would engage "Supercar" mode so the car could be imported into the US under the same emissions status as Lamborghini and other high end marques. I don't know which model but I could probably track down the steps that were provided if curious.

Many variables in GM code do not have fixed locations and the number of cylinders is no different. Many four cylinder GM calibrations use a fixed value for the number of cylinders. $58 was actually a bit different because the number could be changed. Lotus did seem to do things differently than GM, and I ran into this fairly early when trying to decipher Quad 4 code, also written for the 1228707.

The values you list for the number of cylinders, "(00-8 C0-6 10-4 18-TBI)," doesn't match the number of cylinders variables in other GM vehicles. The variable for number of cylinders is used to correct the time between reference pulses to the number of cylinders, and the following variables are valid: $00-8, $C0-6, $80-4, $60-4. See: http://www.aces.edu/~parmega/efi/turbo_p4_doc.pdf , p.13. What you've got seems to be the fueling strategy variable which is different.

Do you have any disassembly tools yet? A plan of attack? I don't believe pattern matching with another code mask will be productive.

Thanks so much for your reply.

I do not have any disassembly tools yet, or even a real plan of attack. Do you have any suggestions?

Thanks,

Aaron

Well, this is likely to be a tough path to follow but you need a disassembler. For many years I used a tool called Donald Whisnant's Code Seeking Disassembler. It was free and if you fed it the correct information it would do a reasonable job of separating code and calibration variables. Then I found Ida Pro and realized that it was a big step above the free disassembler I'd been using before.

Aaron,
if you are still interested please contact me. I seem to have a blue print from another project on a source code level rebuild I had been in with Land Rover Lucas 14CUX EFI for my MG RV8.
Mark