Evo 8 ECU In 2g Works!!

[Hal]

Back in the early days of the "6 bolt swap" there was a documented circuit that could be added to the harness to aid with the CAS inversion.

A little digging found this:

RRE Instructions

 

[the_mork]

I saw a snippet of discussion in which a 3g Eclipse was having the same issues as we have had with USDM roms and put what he called an "evo reductor wheel" (IIRC) in his CAS and it fixed his problems.

Is this the same sort of thing that you are working on mike96? I don't know if you want this to be your achievement but if you don't care some details might let someone else pick up some of your work.

I'm still working on SD but I can't get anyone to do my welding in a timely manner.

 

[_Madman_]

as far as the CAS is conserned... not exactly. i am running an evo8 ecu in a 96 so yes it will work, you just need to switch the pins at the ecu for the injectors and coils, not a big deal. i am working on something to try to get the CAS correct on the eclipse... right now it looks like the CAS pickup wheel is a bit different on the evo. i have seen two different pictures of an EVO cas and both were different (may be a difference dependent on year???) but i am working on it i just don't have the time right now to press forward with it, that and my car is down for the count right now. i did some wire harness re routing/shortening and apparently i cant solder for s#!t....5 bad solder joints can cause a lot of issues.

Do you know which pins do you have to swap? I have this problem with swapping H8 ECU in my '96 DSM. And I will have the same problems when and if I get the Evo8/9 ECU.

I tried different configs, the one shown in DSM link - I swapped injectors the same way as you have to swap plug wires, then I tried swapping the same way as RRE instructions say. None of these configs seemed to work, basically they all vibrated like crazy at 2500 RPMs, and some of the configs were knocking like crazy as well.

The coils are simple, as you have both spark plugs firing at the same time, but injectors are tough. There are 24 ways you can swap them! Are you sure about cylinders 3 and 4 being swapped?

If someone has an IDA Pro and knows which port is for CAS it should be pretty simple to make a software fix, but the IDA Pro is costly

 

[GSXDNA]

:

may be try changing it to 0 or 1, only problem is that you will go into closed loop almost instentantly... not that you aren't going into closed loop fast enough already... could be a CTS going bad or a pour connection at the CTS. have you ever gotten a CEL for CTS p0115 i think?

I doubt it is anything with my CTS. If I switch to my JDM ROM, the code goes away. The problem is, I just got this ROM tuned by Scot Grey, and his maps are hidden, so I can't even transfer the maps to another ROM.

 

[mike96]

I saw a snippet of discussion in which a 3g Eclipse was having the same issues as we have had with USDM roms and put what he called an "evo reductor wheel" (IIRC) in his CAS and it fixed his problems.

Is this the same sort of thing that you are working on mike96? I don't know if you want this to be your achievement but if you don't care some details might let someone else pick up some of your work.

I'm still working on SD but I can't get anyone to do my welding in a timely manner.

yea the newer member i have been talking to has a 98 or 99 (i don't remember which) that he just put an evo8 ecu in. and he told me that he already has an EVO8 wheel that he is going to try in his car. i have a picture of it and it looks similar to the 2G wheel only the smaller solid section is a lot smaller then the 2G. i was going to look into putting in an older 1G or 2G CAS on my car and mess around with changing around the wheels/plates and see if i could get a combination that works with the 9417 roms but my car has been down for the count and i haven't been able to find a decent CAS at the junk yard for the right price yet.

Do you know which pins do you have to swap? I have this problem with swapping H8 ECU in my '96 DSM. And I will have the same problems when and if I get the Evo8/9 ECU.

I tried different configs, the one shown in DSM link - I swapped injectors the same way as you have to swap plug wires, then I tried swapping the same way as RRE instructions say. None of these configs seemed to work, basically they all vibrated like crazy at 2500 RPMs, and some of the configs were knocking like crazy as well.

The coils are simple, as you have both spark plugs firing at the same time, but injectors are tough. There are 24 ways you can swap them! Are you sure about cylinders 3 and 4 being swapped?

If someone has an IDA Pro and knows which port is for CAS it should be pretty simple to make a software fix, but the IDA Pro is costly

all i did on my evo8 swap was swap pins 1 and 14, 2 and 15, 10 and 23 nothing more. i assume you have the factory cam sensor under the intake cam gear? it does make me wounder if it is possible to log individual injector firing sequence and see if the firing order is off sequence. that might be mood point if the evo is batch fire and not sequential.

I doubt it is anything with my CTS. If I switch to my JDM ROM, the code goes away. The problem is, I just got this ROM tuned by Scot Grey, and his maps are hidden, so I can't even transfer the maps to another ROM.

sounds to me like he may have done something to the tune that may have caused the problem. i would talk to Scot and see if he is willing to try fixing the issue. it may have goofed something up when he did whatever it was to "hide" the tables.

 

[_Madman_]

all i did on my evo8 swap was swap pins 1 and 14, 2 and 15, 10 and 23 nothing more. i assume you have the factory cam sensor under the intake cam gear? it does make me wounder if it is possible to log individual injector firing sequence and see if the firing order is off sequence. that might be mood point if the evo is batch fire and not sequential.

Like this http://www.dsmtuners.com/forums/tun...07561-new-ecuflash-1-38-2g.html#post151631922 - ?

Yes, I have the sensor under the gear. But this swap yielded excessive knocking in 2500 range ( http://www.dsmtuners.com/forums/tun...-excessive-knock-vibration.html#post151830713 ) are you completely sure this setup works correctly with Evo ECU?

I can't find where to buy an Evo ECU, so I can't verify this yet.

 

[mike96]

yea thats exactly what i did only i swapped the wires at the ECU pins 10 and 23 instead of swapping wires at the coils that way if/when i put the stock ECU in it is an easier swap. i just unplug the factory harness from the jumper harness where all the pin swapping got done.

i got my evo ECU from another member on EvolutionM for $100. it had no sticker on it so i took a chance that it was an evo ECU, but it was. i was told that another car has the same ECU as the evo, i just cant remember which one right now. i will try to look it up.

 

[delta448]

I tried different configs, the one shown in DSM link - I swapped injectors the same way as you have to swap plug wires, then I tried swapping the same way as RRE instructions say. None of these configs seemed to work, basically they all vibrated like crazy at 2500 RPMs, and some of the configs were knocking like crazy as well.

You must have done something wrong somewhere. The shudder/vibration is most likely an incorrect injector firing sequence.

The coils are simple, as you have both spark plugs firing at the same time, but injectors are tough. There are 24 ways you can swap them! Are you sure about cylinders 3 and 4 being swapped?

Actually there can't really be that many ways. Remembering that 1 and 4 are TDC at the same time and 2 and 3 are TDC at the same time (hence the wasted spark ignition), the -insert either injector or spark- firing order has to be either 1243 or 1342. Cylinders on the same coil must fire their injectors a revolution apart, or in other words; the second coil must fire a power stroke before the first coil can fire another power stroke, hence the firing order logic I listed in the last sentence. Also in support of this is the fact that the injector should open at the right moment for all the fuel to completely reach the cylinder, and not just wash down the walls of the intake port at low engine speeds / airflow.

So, boiled down, that basically means that if you swap the coils by changing plug wires 1&4 to 2&3, it creates a 180 degree difference in ignition timing, which will offset the CAS being turned in the opposite direction since thankfully in the case of our 4G63s there is a wasted spark ignition and you only have to guess the correct coil.

If you then need to also swap injector harnesses, they have to be swapped between cylinders originally on the opposite coil too, since the ecu still thinks it's firing off one of the cylinders now on the other coil. So since we know the firing sequence of the 4G63 is 1342, the injector harnesses should correlate to a firing order change of 3124 (could also write it 2431, which is exactly backwards from the original firing sequence, but is technically listing the wrong coil first). In order to maintain the correct firing sequence, when looking left to right in the engine bay from passenger side to driver's side you'd put the injector harnesses in the order 3412. Just for additional clarity, injector plug 3 is on cylinder 4's injector, plug 4 on cyl. 3's, plug 1 on cyl. 2's, and injector plug 2 is now on cyl. 1's injector.

it does make me wounder if it is possible to log individual injector firing sequence and see if the firing order is off sequence. that might be moot point if the evo is batch fire and not sequential.

Surely a batch fire design would only use a single ecu output / controlling actuator. Evoscan allows you to test each injector's actuator, so I'm going to assume Evo's still use proper sequential injection. Batch fire is for wet manifolds, nitrous fueling and meth/water injection, not the best route for main fuel supply IMHO.

 

[GSXDNA]

sounds to me like he may have done something to the tune that may have caused the problem. i would talk to Scot and see if he is willing to try fixing the issue. it may have goofed something up when he did whatever it was to "hide" the tables.

No, I was throwing that code before he ever touched it. I've been emailing him, but can't get a response

 

[the_mork]

FWIW SD is now available on 9653.

 

[mike96]

FWIW SD is now available on 9653.

now i just need to get my car running again and emissions tested so i can make the switch.

 

[the_mork]

I'm stepping out of my area of knowledge here but is it possible that our injector orders are backwards because of the reversed engines?

 

[delta448]

It only depends on what the ecu and the wiring are originally set up for. The 95-96 CAS is the only issue I'm aware of.

Cylinder #1 is the cylinder next to the timing belt on both engines.
And still appears the same to the ecu, no matter if it's on the left or the right side of the engine bay. The ecu opens injector #1 when fueling cylinder #1 whether it's on the left or the right.

 

[the_mork]

I guess wouldn't the easiest way of checking be just to unplug all of the injectors but one and then use Evoscan to activate each injector individually, then when it clicks you know which one you landed on?

 

[clockworktoy]

FWIW SD is now available on 9653.

Link? Been looking for it but haven't found it yet.

 

[clockworktoy]

No, he has a point, CAS issues are still troubling for 95/96 cars. So you can only say it will definitely work on 97-99 cars now.

Not necessarily, i don't know if anyone else who is posting/reading the thread has a 6bolt swap but i know that my 1g cam sensor seems to be working fine with my evo ecu.

 

[the_mork]

Link? Been looking for it but haven't found it yet.

You have to PM mrfred. It is still in testing.

After a long ordeal with a local exhaust shop my sensor bung will be arriving from summit this week (what I should have done to start with) and if they can't get me in quickly I'm going to pitch a fit, I've been fighting with them for two weeks.

 

[jrohner]

I think that it's definitely necessary to do some changing to the knock sensing, it's way too sensitive on my car. On my car, it shows knock even on the stock Evo8 timing map and I'm on E85. I can rev it up in neutral and it shows knock! It's not bad under full load, but at like 12psi of boost it always shows knock and it almost always does somewhere around 3-3500 rpms no matter what the load (anywhere above 0 psi). I ran at least twice as much timing on my EPROM ECU as I can run on this one. I could run whatever timing I wanted on the EPROM ECU, but on this thing I've only been able to add timing mostly at higher rpms.

At 6500+ I can get to 20+ degrees, so it makes my peak hp at 6500 when it was at 5500 on the dyno (and much higher hp). According to the excel spreadsheet from EvolutionM, I'm only at 325hp or so, and quite a bit less than that for torque (but the torque curve is really flat & the hp is like a straight diagonal line about like a supercharged vehicle). On my cousin's 2G, the spreadsheet makes almost identical curves & numbers to his real dyno chart; his shows more power/torque than mine everywhere below 5500, where on the real dyno (with my other ECU) I had more everywhere, even long before full boost.

 

[_Madman_]

You could try to borrow the knock sensor maps from 98/99 DSM ECU. We just have to figure out which ones are the knock maps.

 

[the_mork]

There is a small thread on evom started by evo kid on the knock issue and that might be the appropriate place to develop this discussion.

The major suggestions at this point are to use det cans to verify phantom knock and to switch to an evo knock sensor since they are a bit different. I actually don't know what a det can is though.

It is worth checking but at this point I doubt that the knock is real.

 

[delta448]

Det cans are pretty rudimentary. A long thin piece of metal (like a collapsable FM radio antenna) is attached to a tin can by a length of wire, preferrably less strands and moderate gauge.

To use: place the thin metal probe on the engine block near the cylinders and hold the can to your ear. Listen for pinging.

Here's a good DIY article on making a slightly more sophisticated detonation detector.

 

[the_mork]

I just saw something I thought I should share, evo fuel pressure is 38psi which we need to compensate for. I know I've been running a 513 injector scaling and wondering if I had a big boost leak because of my fuel trims. I would add the link but I'm posting from my BB so it is a PITA.


Anyone know what my scaling should be?

 

[delta448]

Something around 549, whatever value you can set that gets you closest to 548.87cc.

Try this calculator.

Sorry for the edit -anyone who saw my original post, I'm not thinking clearly today.

*Note to my dumbass self: The number is what the ecu sees as flow rate, a larger set value means smaller resulting pulsewidths... and obviously changing to a higher pressure makes the ecu need to open the injectors a shorter time.

 

[the_mork]

I made the same mistake you did after you pointed the mistake out. I'm an idiot sometimes.

The only thing is it came out to 580.5 for me.

43         x
-     =    -
38        513

x=580.5

This forum fails at handling white space so I needed to use code tags


If I use the calculator for -^2(38/43) x 513 = 545.71

I think the calculator is right but after a full day of work I can't for the life of me understand why the simpler method isn't right.

 

[delta448]

I used 43.5 as the new pressure, you used 43; that's why you got a different value. You'll still have to fine tune certain cells anyway for best performance, so long as you're close on the global you'll be fine with that to start from. Some tuners actually prefer to use a global that is extra conservative and just cut out more fuel during tuning for the added margin of safety for initial changes.

Just remember that an AFR around 11.1:1 burns the fastest so use less timing to start with if you're leaning out the upper ranges, and an AFR of 12.2 produces the most power for lower load where extra fuel is not needed to cool combustion. Timing makes more power difference than AFR, but AFR affects how much timing is optimal.