97 Subaru Impreza (6/7 Trigger)

[bubba2533]

I now have an MS3/MS3X setup with the same V3.0 board and Jean's Dual VR 2.0 Board.

I have updated my MSQ to the latest running setup and it is doing very well. I am using the base timing map pulled from my stock ECU. It is doing ok considering that it is mapped for MAF based systems.

I have also converted my car to run standalone on the MS3, but can still plug the stock ECU back in anytime in less than a minute.

Latest MSQ attached below. (Closed Loop Idle is still in the initial tuning stages, so I would suggest turning it off if you grab the tune)

I also attached a start up log and a composite log.

OLD POST

Ok, so I haven't gotten that far yet. So I don't want to get ahead of things, but I just got past a huge milestone (to me at least).

As of 4/19/2011 I have an amazing trigger signal from the stock 6/7 setup in my 97 Impreza L. I did this with a MS2 and below you'll see how easy it was.

I got it running without the resistors, but got sync loss at midrange RPM's and I tried the resistors and they worked flawlessly.

I bypassed all of the on board V3.0 signal conditioners and went through Jean's Dual VR 2.0 Board. I tell you what, this thing is magic because my trigger signals look amazing. I did have to swap the trigger to falling edge to get RPM's.I was getting a good looking signal in the composite logger with rising edge, but I got no RPM so I swapped it and that was it.

I will attach a couple logs a took. One with a few revs and one of a start up. I will also tell you that the MS2 is acting as a data-logger right now and is not controlling anything. Therefore it is running in parallel with the stock ECU. This I am also amazed by, because I would have thought the MS would have screwed with the signals and made the stock ECU not run.

Anyway, here are the logs (and an MSQ). Ignore all but RPM as that is all I have wired up at the moment.

P.S. - Expect to see a lot more.

[elturbonitroso]

thats a real clean signal.

[bubba2533]

Update 1:

I have gone through a lot for hardly anything the past few days. Read my story below if you would like to find out why.

It all starts with me drawing up a wire diagram in Photoshop. I took my time to edit the stock wire diagram to suit my application. After finishing it up (or so I though) I printed it out so I could take it to the car with me. Then I realized I had switched the labels for the crank and cam trigger. I thought to myself "I'll fix it later, I want to get something accomplished."

After previously testing in the parking log and having great results with the trigger signals I decided to do some road testing.

Before I did any road testing I decided to hook up the TPS. So I pulled out my patch harness and added a wire for the TPS. Took the harness out with the newly added TPS wire and connected to my car. I remembered that I had to install the crank and cam opposite from my diagram.

Well, my road testing did not go as well. Trigger signals were cutting out and I was loosing sync above 4k RPM. I had assumed it was because my TPS signal was giving me EMI on my trigger signals. So I figured I would shield the trigger wires to negate that. I pulled the harness out again to shield the trigger wires.

This time I installed it according to my printed out diagram. I did not remember that the crank and cam inputs were labeled wrong. I was getting zero sync obviously and i was confusing myself. I then thought I had done something incorrectly with the shielding of the trigger wires. So I pulled off all of that work and I was still getting zero sync.

After I was pulling my hair out for the past couple days I finally remembered what I had done. I have since swapped the labels in the diagram and my wires in the car. So after I swapped the wires I was up and running just like before. Since I pulled off all of the shielding I could not test to see if that would help with EMI from the TPS. It was a lot of work and I wasn't sure it needed to be done.

I then thought I would try what many other people have done with the Subaru trigger signals. I have read that placing a 2.2k Ohm resistor from the (+) to the (-) of the trigger along with a 12k Ohm resistor in the (+) of the trigger that it will help give the VR conditioner a better signal. Well in my bag of resistors I found the 2.2k resistors I needed, but the closest i could find to 12k was 10k.

So after installing those resistors I took my car for ride. Well, I don't get any sync loss anymore. Those resistors worked perfect.

Synopsis
-Drew wire diagram incorrectly
-Had good trigger signal to redline upon free rev
-Connected TPS
-Got bad tirgger signal above 4k RPM under load
-Assumed EMI from TPS was the reason for bad tirgger above 4k RPM
-Shielded Trigger wires
-Installed crank and cam switched (from bad diagram)
-Got no sync
-Assumed from wire shielding (removed shielding)
-Still got no sync
-Remembered diagram mistake
-Still had bad tirgger signal above 4k RPM under load
-Installed 2.2k across (+) (-) and 10k inline (+)
-Great tirgger signal under all conditions

[tpsretard2]

2.2k and in line resistor dos it again
glad you are working now

[bmw5002]

wow this is great news! i have the same sync loss issues with a 6/7 ej20g and ms3. i bought jeans dual vr board but have not installed it yet.

i am curious, you mentioned the resistors - did you need to wire the resistors in right before the dual vr board? i am using 470 ohm resistors now and that is the only thing that actually even allowed the car to run. hopefully the dual vr board fixes the sync loss for me.

[bubba2533]

wow this is great news! i have the same sync loss issues with a 6/7 ej20g and ms3. i bought jeans dual vr board but have not installed it yet.

i am curious, you mentioned the resistors - did you need to wire the resistors in right before the dual vr board? i am using 470 ohm resistors now and that is the only thing that actually even allowed the car to run. hopefully the dual vr board fixes the sync loss for me.

Yes I wired them before the Dual VR Board.

[bubba2533]

Update 2:

Got my LC-1 and XD16 installed and working.

My Analog output 2 (that goes to the MS is not working correctly though). I'm going to try and flash the firmware to see if that does anything. If not I have a brand new LC-1 that I can put in.

I also got My IAT wired up. Works great.

I tied into the stock coolant temp signal, but I need to calibrate the values for my MS so that it correlates to the correct temp.

I have my MAP hooked up via a T in a current vacuum line that goes to the manifold. But I get some oscillations because the line is tapped off of a cylinder runner and not the center plenum. So I'm going to tap the plenum and pull my MAP line off of there.

Next I'm looking to start controlling fuel only.

[tottestad]

so you are using the jbperf dual vr conditioner and the resistors inline and across?

[tottestad]

Did you get the correct values for the subru clt sensor? Woould you be willing to share them?

[tottestad]

And do you have a copy of the wiring diagram you made?

[bubba2533]

so you are using the jbperf dual vr conditioner and the resistors inline and across?

Yes

Did you get the correct values for the subru clt sensor? Woould you be willing to share them?

Not yet. Yes, when I have them and verified them.

And do you have a copy of the wiring diagram you made?

Yes, I will post it as soon as I am happy with it.

My project progress really depends on my free time and the weather because I do not have a garage. So progress will be up and down. I hope to make good progress in the next few weeks but I cannot promise anything.

I am having some really annoying issues with my LC-1 and XD16. I have talked to Innovate Motorsports a couple of times now and both times It seemed like a fairly easy fix. But I just cannot seem to get it to work reliably. I am very close to giving up on Innovate and going with a different wideband. A couple more days of playing with it and if it still angers me I'm going to sell it.

[bubba2533]

Ok, I figured out my issue with the wideband.

After talking to Innovate Motorsports about 5 times about this issue, they finally figured out what I did wrong.

I have the LC-1 with an XD16, which both have a serial in and a serial out. When connecting them together you connect the LC-1 out to the XD16 in. You also plug the LC-1 in with a terminator that they supply. I also put a terminator in the XD16 out (which apparently is wrong). So after trying half a dozen other things, I finally have a working wideband.

[tottestad]

Congrats. The wideband is huge. Ive got my car up and running great, I just have a sync loss every once in a while at high rpm, and my therm values are off a bit. I got them from an older post. Good luck with the build. I have my ej20 in a 914 with ms3x. If there is anyway i can help please dont hesitate.

[flatblack280]

Any updates on the CLT sensor values?

Are you just going through the megamanual for this, or is there a EJ specific writeup for all of the jumpers/etc needed?

The MS1E swap into my Z was way easier

*edit*

Found on legacy central that "The stock coolant temp sensor uses a standard Bosch NTC curve." [From this thread] http://bbs.legacycentral.org/viewtopic.php?f=19&t=32923

And I'm now looking through the Subaru 6/7 Manual: http://www.msextra.com/doc/ms2extra/MS2 ... Subaru.htm

I'm a bit embarrassed I didn't find it sooner. I have gone through the diagrams and converted a list of ECU pins that was originally EJ22 OBD2-->MS DB37 to EJ22 OBD1-->MS DB37 if anyone is interested.

I found the original list here: http://www.rs25.com/forums/f105/2098067-post2.html

Now, to just wrap my brain around the Second VR conditioner circuit...

[tottestad]

Ive gotten rid of all sync loss. I ran the same resistor values(2.2, 12) except i removed the resistor across the cam vr. Also ran my shieled cable ALL the way up to the ms connector. (I used to have about 4-5 inches unshielded to the connector) BAMM! NO SYNC LOSS. And thats throught the entire RPM range. Car runs GREAT!

[racingmini_mtl]

Removed due to the edited message above.

[bubba2533]

Update 3:

Got an Idle valve to test on the bench and had fun playing with it today.

I found through searching that it is a 3 Wire PWM type idle valve.

I think I got the best response from a frequency of 153Hz.

[bubba2533]

Ok so I got it up and running last week with the stock ECU disconnected. I've been on vacation all week, but I hope to get a lot more done in the upcoming couple weeks.

[bubba2533]

Update 4:
Wow, today I was amazed by how great megasquirt works.

I had some troubles the past week with it not running very well, but yesterday I fixed my problem.

So at first I set it up to run the ignition output as going low (normal) because I remembered a thread that stated that is what stock suby coils require to fire. Well, I was wrong in assuming that and it kicked my butt in the end.

I was sending the wrong signal to my ignitor without knowing it and the car just wasn't running to it's full potential. I tried to tune it like that and it helped a little bit. It would fall on its face at 3k rpms and i was just getting tired of it and hooked the stock ECU back up so I could drive it to work the next day.

I started it up the next morning and since I disconnected the stock ECU the CEL was off. Normally it is on because the stock O2 sensor is disconnected and it gives a O2 heater code. So after about 10 seconds of it running the CEL comes on and I think nothing of it because of the O2. Well I back up and start on my way to work and it felt so sluggish that I was kind of worried. But I had to go to work so I kept going. After about a mile into my trip to work my CEL started flashing at me. At that point I turned around and went back to my house. So i parked my car and drove my bike to work even though it was supposed to rain.

I don't know why but when i was at work I decided to go through all my setting for the megasquirt. And when I came upon the spark output setting I didn't remember why I picked going low normal, so I did a little more research and sure enough i should have used going high (inverted) to fire my coils. So I was basically overworked my ignitor and fried it.

Then I got a used ignitor and coil (just in case) from a local and came home that night and installed the ignitor with the stock ECU so that I could go to work the next morning. I started it up to see if that was the problem and it fired up without a hitch and sounded great again.

So after I drove it to work and it drove great (with the stock ECU) I wanted to try the megasquirt with the correct settings and see if that was what my troubles were stemming from.

IT WAS!!! I started it up with the megasquirt and it sounded just like it was on the stock ECU. Then I drove it around for a while dialing in my fuel table and having fun with the amazing throttle response.

I will be trying to drive it to work tomorrow and tweak my fuel table some more.


I am just so happy that the fun has begun and I can start learning how to tune.

Synopsis
-Setup ignition output to Going Low (Normal)
-Car ran but not as good as stock
-Tried tuning it to see if anything changed
-Got fed up and installed stock ECU
-Tried to drive to work and car had no power
-Returned home and drove motorcycle to work
-Researched ignition settings and determined it needed to be set to Going High (Inverted)
-Replaced ignitor and the car ran great on the stock ECU
-Reinstalled the megasquirt with the correct settings and it runs amazing

[carphappy]

ok, a bit lost here. same issue as most, sync loss under load above 5k rpm. jdm subaru ej20 turbo, 6/7 trigger, with mostly stock sensors in porsche 914. using ms/2 extra. the solution seems to be a 2.2k and a 12k resistor and possibly removing an existing resistor. can someone please describe exactly where they installed these resistors and what exactly was removed in anything. also, any suggestions on noise filter settings. thanks.