[titan2782]

You're tuning the MT86 now?

We are working with Alphaspeed as one of their dealers. I have several "remote tuning kits" that I can send to people anywhere so we can get them tuned from their driveway. I personally do not do any tuning. I have been working through a disassembly of the MT86 ECU dump to see if I can discover anything useful. So far I've uncovered zilch.

Great job!

All I know is that dude with the blue line is awesome 😉

He's alright... :grin:

 

[Ryengoth]

We are working with Alphaspeed as one of their dealers. I have several "remote tuning kits" that I can send to people anywhere so we can get them tuned from their driveway. I personally do not do any tuning. I have been working through a disassembly of the MT86 ECU dump to see if I can discover anything useful. So far I've uncovered zilch.

I pulled a rom and tried to disassemble it using a similar architecture and it gave me garbage. Are you disassembling with an Infineon 32-bit instruction set?

 

[titan2782]

I pulled a rom and tried to disassemble it using a similar architecture and it gave me garbage. Are you disassembling with an Infineon 32-bit instruction set?

tricore 1766 but you need a FULL dump which you only get from bench tools. I got a full dump in intel hex format and had to write a converter to convert it to actual binary so I could disassemble it. I spent weeks on dumps from PCMFlash only to realize finally that it only pulls the calibration data, which is useless for disassembly.

Biggest issues even with full dump are

1. No debug symbols, or anything at all to help you even find a starting point.
2. A bug in IDA Pro that causes disassembly to be incorrect. I've seen this first hand, and I'm waiting to get a copy of IDA Pro 7.3 where they fixed it.

 

[Ryengoth]

tricore 1766 but you need a FULL dump which you only get from bench tools. I got a full dump in intel hex format and had to write a converter to convert it to actual binary so I could disassemble it. I spent weeks on dumps from PCMFlash only to realize finally that it only pulls the calibration data, which is useless for disassembly.

Biggest issues even with full dump are

1. No debug symbols, or anything at all to help you even find a starting point.
2. A bug in IDA Pro that causes disassembly to be incorrect. I've seen this first hand, and I'm waiting to get a copy of IDA Pro 7.3 where they fixed it.

I got a full rom dump via physical boot mode back when I was hacking the ECU for my 2013 project. Yeah, forget getting debug symbols. I successfully read/wrote using the Tricore boot mode with the Magpro v17(sold to Swede racing team) and the FG Galletto 4, which I still have. If you can get it properly disassembled I'd be keen to look over the assembler and data blocks.

 

[Juntak3.8]

New update

336/298 - dynojet STD
330/292 - dynojet SAE

352/312 - Dynomite (see below)

The only changes I made this time around performance wise were my new Stage 3 intake manifolds, fixed some exhaust leaks and switched to AlphaSpeed custom tune still on 91 octane.

The dyno sheet shows my setup (in green), the other two are FBO 3.8s (final results) from the same day for comparison. The blue line still has all 4 cats.

At this point in time, this is the basically the limit of the 3.8 NA. The cams are now the bottleneck.



I have some mild port work on the heads which lets the power keep going until redline instead of falling off like it does which means the heads were the bottleneck, but with that fixed, it just won't flow anymore without FI. Blue line is mine, red line is the fbo 3.8 from above, but without the head work.

Dynomite numbers
Last time I was tuned, we did it on a dynomite dyno. Obviously this time the number are lower because it's a different dyno so with the help of a friend and some research, we determined the dynomite was showing 5%-8% higher than the dynojet. We used the data i had from other dynos from several previous events to confirm. The estimated numbers of 352/312 were adjusted for the 5% and were pretty spot on for what i was expecting (360/310). These numbers are just for fun and comparison purposes to avoid confusion since my last update.

At this point, I'll be working to complete my built motor. I'm working with MT Motorsports to build a high compression 3.8 motor with all the goodies and most likely will be running it on E85. But for now, I have plenty of other work to do to help everyone reach their potential.

Was that super insane 460whp Genny done with your help too?

 

[Ryengoth]

Was that super insane 460whp Genny done with your help too?

We got 508whp for stock 3.8 GDi with 14psi on the WinDyn before I tore the project down. Still not the highest around, I gave up due to OE fuel supply(and R&D funds). Work is being done on the DI jerk pump and fuel rail supply issue.

 

[titan2782]

Was that super insane 460whp Genny done with your help too?

You talking about the white 3.8T? No, that was all done by the owner and he's still on stock block too pushing 11psi.

 

[OKC3*8RSpec]

Has anyone ever dyno'd a car with a canned tune and then had it custom tuned to see how much difference there was?

 

[Ryengoth]

Tune alone is minimal. You need at least high flow cats and exhaust to see any change.