ECUflash EvoScan and EcuFlash tuning

[logic]

That's OpenPort 1.3; has Colby said anything specifically about OP 2.0?

 

[jjgrizz73]

If the flashing doesn't work with 1.3 it wont work with 2.0 either. If it means that much, why not do the evo 8 ecu swap, very easy process and cheap ecus.

 

[the_mork]

But it does mean that there is a possibility here of a workaround. Now that we know what the issue is some hardware should be relatively easy for an enterprising vendor (or user turned vendor) which could really change the face of DSM tuning. Certainly a newly designed cable which used an auxiliary power source would be less expensive than an eprom ecu (and chip).

 

[PieEyedPiper]

If the flashing doesn't work with 1.3 it wont work with 2.0 either. If it means that much, why not do the evo 8 ecu swap, very easy process and cheap ecus.

Sure! So easy that no one has it working flawlessly yet. I'll hold out until the dust has settled on that one. Though I predict that time is coming very near, so that IS still an option. But it also costs an ECU more than simply using what my car came with. You're right though, its a fine alternative.

Do you know if anyone has it working well with stock 450's yet? I know the cars run but when switching to the evo injectors things got much better.

But it does mean that there is a possibility here of a workaround. Now that we know what the issue is some hardware should be relatively easy for an enterprising vendor (or user turned vendor) which could really change the face of DSM tuning. Certainly a newly designed cable which used an auxiliary power source would be less expensive than an eprom ecu (and chip).

Exactly! Is is possible to hook up a 16v external source? Sounds like everything is ready to go and we just need to supply appropriate voltage?

 

[jjgrizz73]

Sure! So easy that no one has it working flawlessly yet. I'll hold out until the dust has settled on that one. Though I predict that time is coming very near, so that IS still an option. But it also costs an ECU more than simply using what my car came with. You're right though, its a fine alternative.

Do you know if anyone has it working well with stock 450's yet? I know the cars run but when switching to the evo injectors things got much better.

Actually a few cars have it working and people are getting simple codes that are usually self inflicted. If proper installation and flashing from someone that in familiar with flashing is done then its easier.

Another thing is, only 98-99 ecus it will possible read/flash correct? If not you need to speed ~100-125$ for an ecu those years. I bought my ecu for 150 shipped but after I recieved it people were offering them for 75$ (go figure, my luck). Thats expensive?

 

[the_mork]

It is $75-150 cheaper.

I got mine for $100 shipped but I'm under the impression that 98-99 ecus are less expensive than evo8 ecus AND if the evo8 ecu becomes a popular mod then likely the ecus will become quite a bit more expensive. When it comes down to it there are a lot more 98-99 ecus out there than evo8 ecus (at least that is what I assume) since something like only 10k evo8s were produced in the US and many are still using the ecus that they came with.

 

[PieEyedPiper]

Actually a few cars have it working and people are getting simple codes that are usually self inflicted. If proper installation and flashing from someone that in familiar with flashing is done then its easier.

Another thing is, only 98-99 ecus it will possible read/flash correct? If not you need to speed ~100-125$ for an ecu those years. I bought my ecu for 150 shipped but after I recieved it people were offering them for 75$ (go figure, my luck). Thats expensive?

$75 to $150 is expensive for something I don't need, since I have a 98. But I'm not just thinking of myself here, its a good point that evo 8 ecus aren't as plentiful, and if we can rekindle the DSM scene a little bit by such a marvelous breakthrough, why not?

 

[Ceddy]

The new version of EvoFlash 1.41, will not allow you to try to flash a 98-99 ecu with a 1.3 cable.
You get a error msg saying use a 2.0 cable.

Problem with the 2.0 cable is none of the Logging software is updated to use it yet.

I was playing with my 1.3U cable today.
The flash connector gets 4.8v normally, and 15.7v when trying to flash.
I might try rigging up simple 16v power source and relay and use a old version of EvoFlash, to see if it will work.


Was wondering if any of the guys with evo ecu swaps, have gone thru emissions yet where they hook up to the OBD2 port?

 

[knochgoon24]

Well guys, I just did a little test drive to wally world and back with Evoscan v2.5. Very, very nice. Much better than that ELM 323 device I was using. Here's what I found out:

I have an early 97 Talon. (4-97) I logged using Mitsubish with the Openport 1.3U cable. Here's what I've found

Throttle Positon: Works
Engine RPM: Works
Battery Level: Works
Airflow / Rev: Doesn't seem to work right. Starts to read much lower as boost kicks in. (Like 1/10th what is was pre boost.)
Coolant Temp: Works
Air Temp: Seems to be my intake temps.
MAF Air Temp Scaled: Not sure why, but it generally climbed as I drove.
Coolant Temp Scaled: This stepped inversely from my coolant temp. Started from 21 and went down.
Timing Advance: Works
Air Flow Hz: Works
Barometer: Works (The pressure actually dropped as I drove for 15 min. A low front is actually moving in.)
Target Idle RPM: It seems to be right where my car wants to idle. When my TPS was mis adjusted, this was reading >1400RPM for some reason.
Fuel Trims: Work? (Don't know much about those yet)
Oxygen Feedback Trim: Didn't try to log it yet
O2 Sensor: Gave it's nice light show
o2 Sensor #2: Didn't get anything from it. Just read 0 volts.
Speed: I get really odd numbers. It doesn't even read zero standing still. It just jumps around between 0 and 325.
Injector Pulse Width: Works
Injector Duty %: Works
Fuel Consumption: Needs speed to work. Calculates MPG.
Gear: Needs speed to work. Goes by speed and RPM
Air volume: Didn't try to log it.
Boost (MDP): Didn't try to log it.
JDM MAP: Didn't try to log it.
External Wideband A/F: Didn't try to log it. Didn't notice it until just now. I have an LC-1 so I'll try it next time I'm out.
Knock Sum: Works! I only had a max of 1 count here and there, but I only had it up to 4000 RPM this run.
ISC Steps: Works
Crank Signal: Didn't try to log it.
Idle Position Switch: Works. Used it to readjust my TPS after replacing my throttle body seals.
Power Steering Switch: Didn't try to log it.
Air Conditioning Switch: Didn't try to log it.
Air Conditioning Relay: Didn't try to log it.
Inhibitor Switch: Didn't try to log it.
AFR Map: Mostly hung around 14.7, but went richer under boost and leaner for ?.
Load Calculated: Not sure what this does. But it logged.
ECU Load: Like "Load Calculated".
Injector Latency: Works
Load MUT 2Byte Mod: Didn't try to log it.
RPM MUT 2Byte Mod: Didn't try to log it.
Airflow MUT 2Byte Mod: Didn't try to log it.
Estimated g/rev: Increased with boost.
Estimated Lbs/min: Increased with boost.
Estimated Horse Power: Didn't work for me at all. If it's true, I have a solid 4.4 hp. I didn't change any of the torque/horsepower settings.
Estimated Torque: This one is reasonable.
E1 RAM 1Byte Mod: Didn't try to log it.
E4 RAM 2Byte Mod: Didn't try to log it.
Knock Voltage: Didn't try to log it.
Octane Level: This one was weird. Any time I had even 1 count of knock, this dropped way down.
ERG Temperature: Seems to display ERG Temps. No clue whether this is even close.

I didn't log anything from here on down the list.



Maybe I'll get it out and log everything as I do a pull tomorrow. I just have too much school work tonight.

I also need to get my coolant temps under control. I hit 220*. I think it's just too much antifreeze to water. I'll test that and fix it tomorrow. It's supposed to be 51* in Erie...in Feburary


EDIT: And for those wanting the info for the HP/Torque Calculations
Gear Ratios
Tranny gear ratios for DSMs

Curb Weights
1000 AAQ - DSM factory specifications
Drag Coefficient: 0.29
Frontal Area: 20.56 ft^2 Drag and Frontal Areas

FYI: Our cars have less drag than someone riding a bicycle. WTF

A 215/50/17 tire is 25.5" diameter.

 

[knochgoon24]

This is some info from a user without the required 50 posts sent me. Hopefully it will help.

I saw your post today and since I can't post I thought I'd send you a bit more info that you can add if you feel necessary.

For fuel trims, there may be a need to update some code in ecuflash before they work. I know that this is the case with Evos.

Load Calculated is a formula that Evo ECUs use for their fuel and timing maps. 0 - 100% load is vacuum, 100% is 0 PSI, 200% is 14.7 PSI.

External wideband does work, it just takes a lot of settings to get right. I am able to log my AEM perfectly fine.

All the MUT 2byte boxes are another, usually more accurate data compared to the parameters that aren't MUT 2byte.

Hopefully that adds a bit more to your post.

 

[_Madman_]

Wow, I have been a little busy for a while, and my car is still siting in a parking lot with broken tranny, so I wasn't logging in for a while, but I'm really glad that you guys are still investigating the issue!

There was a discussion about successful H8/539F flash over at aktivematrix.com forums, with a link to a site that described the process in... well... Japanese LANCER ECU¤Î½ñ´¹¤¨

But the approach used some sort of external wiring.

What I've also heard is that H8/539F boot procedure, well the very start of it, matches the sample boot code in the specification. So I'm still very curios about how the ECU boots in flash mode.

Because if you turn the key, the ECU is booted, therefore you already have a normal operation mode that most likely sets few variables, throws exceptions and disables flashing.

If you don't, then you got to have power applied to the programming pin at boot, so that it doesn't go through the regular initialization, but straight to flashing. Which most likely won't light up CE lights and such. Then again, maybe the EcuFlash restarts the ECU before it tries to start the flashing.

Also, I guess the flashing pin is not the one that is connecting to auxiliary connector, it's the actual pin on CPU. Since it's a winter, I cannot easily check the voltage on that leg of the CPU, but if it's 12v during flashing, then we are good, if not, then that's hopefully only a power issue. The actual voltage on CPU pin should be 12v, not 16v, IIRC from spec.

I will try to update the main post with the gearing and logging info in the latest posts as soon as I get some free time, so that we have a consolidated info in first post. Thank you for input everyone

Concerning Evo ECUs, it's not quite that simple, as far as I understand. You have to disable immobilizer, MIVEC stuff, redo injector scaling, MAF stuff, and if I'm not mistaken then most likely you will have to change the knock sensors as well. Plus there is a lot of pin switching, which sometimes leads to loose connectors, like I did when I swapped my ECUs. Now I sometimes get "Injector x circuitry error".

If there is a way to flash H8, then I bet it's better than to go through all of this.

 

[PieEyedPiper]

If there is a way to flash H8, then I bet it's better than to go through all of this.

No doubt. The guys who have an evo ecu in their cars right now maintain that its easy as pie but I'm still holding out for H8 compatibility.
Thanks for that link! any information is appreciated. What strikes me as odd is that you say 12v is all that should be needed? Then where did this 16v idea originate?

 

[_Madman_]

No doubt. The guys who have an evo ecu in their cars right now maintain that its easy as pie but I'm still holding out for H8 compatibility.
Thanks for that link! any information is appreciated. What strikes me as odd is that you say 12v is all that should be needed? Then where did this 16v idea originate?

I don't know why the +16V thing came up, because the board needs +12V (http://www.dsmtuners.com/forums/tun...-evoscan-ecuflash-tuning-4.html#post151667886), but it might be that the flash pin outside ECU need those +16V as there is some sort of a voltage drop...

But that is kind of strange, as the USB is +5V and OBD2, where you could take the extra juice is only +12V (GM 8192/160 Baud ALDL Interface) I doubt you can get +16V through serial connection as well.

The source I mentioned in previous post is also talking about +12V

 

[the_mork]

It wouldn't be a big deal to ditch the flash port entirely and just run some larger wire to the ecu harness which would help to reduce voltage drop and allow us (hopefully) to use a little less juice so that we can get it from convenient sources.

 

[Jason97GS-T]

So I am looking to get a datalogger for my 97 GST.

What exactly do I need to buy to log with evoscan?

Thanks
Jason

 

[_Madman_]

ECU load, RPM, timing advance, knock count.

I think these are the basic ones. External wideband is also nice if you have one.

 

[mikyrc]

So I am looking to get a datalogger for my 97 GST.

What exactly do I need to buy to log with evoscan?

Thanks
Jason

The evoscan software and a openport USB OBDII cable, the one that he's selling on his site.

 

[the_mork]

Your 97 ecu should connect and work alright with evoscan and all that you will need is the cable and the software. It will not connect with the high transfer rates of the 98/99 ecus but it should be at least as fast as obd2 logging with the additional benefits of a few more loggable parameter the most important being knock.

You should also look at the software that Ceddy is producing here http://www.dsmtuners.com/forums/tun.../326872-mhiscan-new-free-datalogger-dsms.html

Trying it out and reporting back would probably be useful to have a good idea of how well it works on different production date 97 ecus (it has been shown to work well with late production 97 ecus) and it is FREE and produced by a member of this community. I would get both programs and play with them.

 

[Deepnine]

Does someone have a 98 oder 99 GSX rom file? i like to start comparing to the GST roms and start disassembling.

 

[snooopy365]

Does anybody have logged AFRmap on 1G? it returns "Not a number" when I start the car... so is there another way to get the AFR map valve right?
Want to use the "show in map" feature, so that would be nessasary

 

[logic]

Deepnine, I already answered your PM on EvoM, but just in case anyone else is looking for it, here's a stock 1999 USDM GSX Federal MT ROM from my old ECU:

http://esm.logic.net/public/roms/1999-USDM-Eclipse-GSX.hex

 

[knochgoon24]

Your 97 ecu should connect and work alright with evoscan and all that you will need is the cable and the software. It will not connect with the high transfer rates of the 98/99 ecus but it should be at least as fast as obd2 logging with the additional benefits of a few more loggable parameter the most important being knock.

You should also look at the software that Ceddy is producing here http://www.dsmtuners.com/forums/tun.../326872-mhiscan-new-free-datalogger-dsms.html

Trying it out and reporting back would probably be useful to have a good idea of how well it works on different production date 97 ecus (it has been shown to work well with late production 97 ecus) and it is FREE and produced by a member of this community. I would get both programs and play with them.

I noticed that I didn't mention my sample rate speed, about 70 samples / second on my 4-97 Talon. I might be able to get more if I turned off the logging to display and stuff like that.

Compare that to OBDII with an ELM323 chip, 4 sample / second.

 

[logic]

Ugh, my I'm going cross-eyed staring at H8 assembly after spending so much time with SH on the Evo.

It looks like IDA's H8/500 processor option is disassembling this correctly; the init/reset and invalid instruction handlers look basically right to me.

Am I reading Colby's release notes correctly, in that OpenPort 2.0 should be able to supply enough voltage to flash these ECUs? I may not have a totally useless ECU on my hands after all.

 

[_Madman_]

Yea, IDA Pro disassembles our ROMs just fine, you just have to work out the memory layout. Because ROM doesn't have all of it, you have to specify sectors.

Fuel maps are also in some concrete places in memory, as far as I know, and they have signatures depending on their types.

I could create and post a C++ program that creates the layout, but I don't know if it's legal

Also the IDA allows to create signatures that allow to define all maps easily. Again, legality aspect is unknown...

 

[luizdiefenbach]

heres the pictures of my 98 ecu:

You must be logged in to view this image or video.

You must be logged in to view this image or video.

When ecuflash + openport 2.0 works on this it will be great BUT w e already have a way of reflashing this ecu: Kashima way

Translated version of http://kaele.com/~kashima/car/evo/


Notice the Con2 8 pin place to solder the reflashing connector.

If we make this serial programmer we can flash the ecu:
http://kaele.com/~kashima/aki-h8/ISP2.jpg

Here is the schematic: http://kaele.com/~kashima/aki-h8/isp2sch.pdf
and here its the PCB: http://kaele.com/~kashima/aki-h8/isp2pcb.pdf
Here is the asm file to program the PIC: http://kaele.com/~kashima/aki-h8/isp2.asm

The evoflash / evostat and evoread I could not find a link, but with these or other software we can read and reflash the eclipse 98-99 ecu.

With the reflash possible we enconter another problem: editing/ modding the hex file

maybe we could use tunerpro and make definition files for this hex file.

That way we could edit the hex file in a friendly way inside the tunerpro and reflash the ecu with this hex modified file, now with kashima programmer and in the future with the ecuflash / openport 2.0 cable

Maybe someone that actually understand this thing could hop in.

thanks for the space