ECUflash First Flashed DSM ECU! (98/99 Flashing FAQ)

[dsm-onster]

I seriously doubt that the boost control helps with creap. Td05h creap is due to another issue. . .

 

[rush2252]

I have a s16g and I took my exhaust off at the track and it would literally run away with itself. Creeping from 17psi upto like 23 it wanted to go more but i let off. What is up with that?

 

[VPrime]

I have a 98 ecu (flashable one, with the same board as yours) in my 95.. I switched the spark plug wires and get misfires..
Is it possible to fix this with openecu as well? Sort of like how dsmlink does it (but opposite)?

 

[Ceddy]

I have a 98 ecu (flashable one, with the same board as yours) in my 95.. I switched the spark plug wires and get misfires..
Is it possible to fix this with openecu as well? Sort of like how dsmlink does it (but opposite)?

Which CAS setup do you have? The 95-96 inverted CAS and Crank Trigger setup will most likely give you issues.

You can turn off the P300 Misfire CELs in the Periphery Bits, but if you are getting real misfires that most likely won't help.

You can install a 1G Black Top CAS, thats what I run without any problems.

 

[VPrime]

Which CAS setup do you have? The 95-96 inverted CAS and Crank Trigger setup will most likely give you issues.

You can turn off the P300 Misfire CELs in the Periphery Bits, but if you are getting real misfires that most likely won't help.

You can install a 1G Black Top CAS, thats what I run without any problems.

I don't have the one on the side of the engine... I have the crank sensor.

 

[roni-wan]

I have a 98 ecu (flashable one, with the same board as yours) in my 95.. I switched the spark plug wires and get misfires..
Is it possible to fix this with openecu as well? Sort of like how dsmlink does it (but opposite)?

tried it about a month ago with no luck, changed firing order and also swapped injector plugs, had a lot of vibrations in the 2k-3k range like misfiring or knocking ended returning to my 95 ecu

 

[VPrime]

tried it about a month ago with no luck, changed firing order and also swapped injector plugs, had a lot of vibrations in the 2k-3k range like misfiring or knocking ended returning to my 95 ecu

Yeah mine currently does the same thing except I have not done the injector plugs yet.. But my eprom ecu needs some repairs so was hoping I could get this working instead

 

[_Madman_]

Well, patching the CAS is theoretically possible, but we need to find CAS code first.

How hard it is to install 1G CAS?

Does swapping plug wires didn't helped?

 

[ed_GST]

My 98 GST 5psd ECU just arrived
it's a MD346675
and the board says JE331B156C
Then yes, it is flashable

 

[_Madman_]

My 98 GST 5psd ECU just arrived
it's a MD346675
and the board says JE331B156C
Then yes, it is flashable

In that case it seems there is zero difference between AWD and FWD ECUs. The board numbers match, the difference must be in software. Can you post FWD ROM so that we can compare it with AWD ROM.

I think Ceddy has the same board, but in AWD ECU.

 

[Ceddy]

75 are FWD
76 are AWD

I'd be interested in checking out your flash image to compare it.

 

[_Madman_]

75 are FWD
76 are AWD

That's the MD number, but the plastic circuit board (PCB) numbers are the same, so I guess MD number changes only because there is different ROM image inside ECU, not because of different hardware.

If this is true that would mean we can buy whichever ECU we want (AWD/FWD) and just flash it to needed ROM.

The only question is if JE331B156D ROM can be written to JE331B156C board. That would be very interesting, because we already have a latest ROM of JE331B156D board posted in one thread.

 

[draculia]

What's going to be different between fwd and awd ROMs though? Speed limiter on fwd ROMs, but that should be scaleable, no? I don't know if this is applicable to the 98/99 boards, but I came across this: Warning!!! This Mod Is Extremely Dangerous!!!

 

[steve]

My 98 GST 5psd ECU just arrived
it's a MD346675 and the board says JE331B156C

Post up some good pictures of the board so we can see what the difference between the C and D PCB versions is.

 

[Ceddy]

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

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

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

Tuner won't let anything over 700x700.
I can host some big pics elsewhere, if need be.

Steve, you have any idea what the other large chip is?
Printed on it ->
(Mitsu Symbol) E328
7BA1Y6J

 

[steve]

Steve, you have any idea what the other large chip is?
Printed on it ->
(Mitsu Symbol) E328
7BA1Y6J

You mean IC2? It's programable logic. I haven't traced out my MD35904 but from a quick look that what I expect.

 

[Ceddy]

Did some testing today with the OP 2.0 cable.

First I tried Quad Speed Logging 62500 vs 15625 baud. I only noticed a 15% increase in speed, not the 400% you would expect.

The OP 2.0 doesn't log super fast because of its pass-thru design. I'm going to try my cheapo ebay cable at 62500 baud to see if I get better results, I think the OP 2.0 cable is the bottleneck.

I tried the set 5 degrees, for base timing, works perfect.

Trying to pull DTCs failed. Got OpenPort not responding error.

Run Open Loop Always, is very nice, makes scaling your MAF and Injectors very easy.


Tomorrow, I'll try to some Logging Speed Comparisons between the OP 2.0 and the ebay VAG 409.1 cable.

 

[ed_GST]

Does running in Open Loop all the time hurt MPG's?
Is there a new XML for the Full Time Open Loop?

 

[the_mork]

Open loop will hurt your gas mileage and your emissions (if you care). It is useful for tuning purposes and if your car is extremely consistent and you can tune you could increase your fuel efficiency by moving from a 14.7 target to 15.3 using the fuel map but much nicer would be if someone could write a patch that allowed the closed loop system to be controlled by a wideband and then set to a user selectable (or just 15.3) crossover point.

 

[Ceddy]

Is there a new XML for the Full Time Open Loop?

"Disable Closed Loop" is in the Periphery Bits.

 

[_Madman_]

For advancd stuff we need the [ECU pin <-> CPU pin] diagram. Does anyone have a fried ECU for sale?!!!

First I tried Quad Speed Logging 62500 vs 15625 baud. I only noticed a 15% increase in speed, not the 400% you would expect.

IIRC, h8/539F spec mentioned something about speed synchronization. That is, if flash speed was faster or slower than something CPU would adjust to needed speed anyways.

I need to re-read the spec again, but there was some sync stuff inside.

 

[logic]

Did some testing today with the OP 2.0 cable.

First I tried Quad Speed Logging 62500 vs 15625 baud. I only noticed a 15% increase in speed, not the 400% you would expect.

Are you doing this with plain old MUT (with an increased bitrate), or did you rework the protocol a bit to accommodate? Over on the Evo side of the house, we discovered pretty quickly that increasing the bitrate only served to get us to the maximum rate that the serial interrupts and MUT requests could be processed, which ended up being a pretty disappointing speed (an improvement, but nothing to write home about).

Going off on a tangent: IMHO, better would be a different protocol than MUT altogether; jcsbanks went down that road with his DMA implementation on the SH ECUs. While I don't believe H8 supports DMA, the idea is basically sound: instead of a request->reply cycle for each value you're interested in like MUT does, transfer blocks of memory either directly (ie. give me 0xFF bytes starting from 0xEE80) or via a table of indirect references (like the MUT table). Using DMA was nice, since you could force the ECU to disable interrupt handling while a block was transferring, but surely you could cook up something on this processor.

Part of that was live RAM updating as well, which is fantastic if you have the memory (which, sadly, H8s seem to be pretty weak on), since you could just move fuel, timing, and other "interesting" maps to RAM. Even the MUT table, which turned out to be a pretty cool side-effect.

If you're at all curious, check the EcuFlash forum on EvoM; there's a thread by jcsbanks titled "How the DMA code works" or something to that effect. I replied in there with a rudimentary "spec" for it's behavior, and the custom MUT commands that it uses as triggers. (It's actually rather nice in that respect: you can test for the existence of the mod, and use it if it's there, or go back to MUT if it isn't.)

 

[ed_GST]

"Disable Closed Loop" is in the Periphery Bits.

Got it!!!

I notice that the ROM u have has the EGR Function disable
This means that i can remove the EGR and wont have a CEL because of it

 

[jrohner]

You can disable most CEL's simply by disabling them in the periphery -- change a 1 to a 0 and that's it.

 

[steve]

Really there's no reason the 96-97 ECU's can't do the same thing.
any obII computer in any car from any manufacturer is capable of doing this if someone actually makes the program to do it.

Sorry, the earlier 2G ECUs hardware doesn't support Electrically Erasable ROM for changing the software. Only the 95's with EPROM support and the 98/99 ECUs with Flash ROM in the CPU allow changing the software.