Re: Project Lex
 

Started on Phase 2 of the MS3 install. The injectors and ignition are all wired up. I had to actually cut these wires, and solder and heat shrink them to the MS pigtail, since MS is taking over full control of them, stock ECU has no control of them anymore.

I need to run a vacuum line to it now for the MAP sensor, then get a base tune from Jordan @ Powersquirt, and another start-up video shall be made :D

http://mitsustyle.com/forums/attachm...1&d=1294367972

Re: Project Lex
 

I'm impressed that you picked up a soldering iron and went to town on your own car like that, very nice. Even tinned the wires and used heat shrink tubing!

Now don't lie, how many wires did you solder and forget to put the heat shrink on first? ;)

Re: Project Lex
 

Haha, only once (the first wire). I'm a quick learner :)

And also, probably no one on here will really care, but as Ive been going, I've been constantly updating my earlier post with exactly how I am hooking up the Megasquirt. But if anyone on here randomly wants to know how to hook up an SC300 to MS3/MS3x the same way I'm doing it, it's all written down in my earlier post. There's still a couple things I have not hooked up yet (wideband, knock) which will come later. But I'm updating it as I go so it'll be complete once I'm all done for others to see.
http://mitsustyle.com/forums/showpos...&postcount=228

Re: Project Lex
 

Most MS cars you see are a hacked up mess, this is a welcome visual treat already. I can't wait to see this thing completely running on the MS unit.

Re: Project Lex
 

Very nice work man, this is way more than I would ever attempt on my own!

Re: Project Lex
 

Been awhile since I've been able to get back to this thing. Been busy lately!

Got the turbo exhaust housing back from Primo Performance from being coated. Looks great thanks guys!

http://mitsustyle.com/forums/attachm...1&d=1295401183
http://mitsustyle.com/forums/attachm...1&d=1295401183
http://mitsustyle.com/forums/attachm...1&d=1295401183
http://mitsustyle.com/forums/attachm...1&d=1295401183

Started it up with the MS3 controlling fuel and ignition for the first time today also. Runs a little rough, but that's to be expected for the first start-up :) Again big thanks to Jordan (Shaodome) for the help!

http://www.youtube.com/watch?v=REOfz7GVyUg&hd=1

Re: Project Lex
 

Nice! So it just needs better base tuning to make it idle better or what? I cant remember but did you originally say that idle would be controled by the stock ecu?

Re: Project Lex
 

Nice work CEO! Keep it up. Maybe this thing can keep up with my bike sometime. ;)

Re: Project Lex
 

Wow! I just looked at this project. Didn't know it was running on a MS system. I am kind of addicted to those standalone.
Good Job!

Re: Project Lex
 

Ya, when I say it retains control over idle, what I mean is the idle components. Like how a dsm has an isc that it constantly adjusts to keep a steady idle. These cars have idle components as well in the TB/IM. And that's what I mean, is the stock ECU retains control over all of that.

The idle I'd say is rough right now because it's a very base tune for now, so the idle and ignition settings probably need a good bit of tweaking.

Re: Project Lex
 

Woot woot!

Re: Project Lex
 

Cool, when I had my DSM running on MS, I didn't even bother with the IAC. Once the engine warmed up (which didn't take too long) it didn't really matter. Idle was smooth without it. However, I didn't have AC or any other stuff that really needed it. The mitsubishi IAC is kind to take control over with MS. I don't know what type IAC Toyota uses but it may be easier to take control over then you think.

Warm up the engine then just tune out the VE table a little. Kind of watch the Air Fuel but mostly what you are going to be watching is the vacuum and RPM and just feeling out the engine.

when tuning Idle you almost have to unplug the IAC because it will throw off your idle tune. ECUs will be fighting against each other. Raise the idle screw to a reasonable RPM (1500). Tune out the VE and most likely it will raise the RPM (to say 1750). Adjust the screw down to lower RPM till it is rough and tune. Repeat till you achieve the Idle that you want.

Also before you start make sure you have your bins set where you want them. I know MS2 has 16x16, I don't know what ms3 has.

Should be really easy to get a steady idle.

Are you using 6 injector channels on the MS3 then? Running sequential injection?

The MS2 had 4 injector channels which works for 4g63.

Re: Project Lex
 

I don't know about half that stuff you just said. My setup is based on another guys setup who's been MS'ing these cars for years, using MS to control ignition and fuel, and leaving the rest for the stock ECU. That's why I chose to do it this way. He's extremely helpful, and extremely knowledgeable on it all, so I'm basically just doing what he tells me to do.

Yes I am running full sequential injection, that was the main reason I chose MS3/3x over MS2.

Re: Project Lex
 

Haha, I was at primo on monday and saw your turbine housing first hand, looked good! I didn't realize it was yours I would have left you a lil note :D

Startup isn't too bad for first shot!

Re: Project Lex
 

That is good that you have someone else guiding you. When I MSed my eclipse I was on my own. Steep learning curve but once you get everything figured out it is really fun.

SO DIVE RIGHT INTO IT!!! except don't blow up your engine...:wink:

Re: Project Lex
 

Very cool! MS sure has come a long way since I last saw/looked at it.

Re: Project Lex
 

First test drive. I keep losing the sync, which deals with the cam signal coming in. I may dig under there next week and resolder that connection, and ensure it's well insulated and see if that helps at all. Anyhow here ya go!


http://www.youtube.com/watch?v=Xha7s9s--VY&hd=1

Re: Project Lex
 

I thought I saw a sync problem in the first video but didn't know for sure if that was the computer or connection to the MS or what.

What kind of signal does the 2jz use? Is it an opto cam or a VR cam signal? are there 2 signals being outputed? cam and crank?

Depending on the type of signal that the ecu is seeing, you may not be able to have both the Stock ECU and the MS system share the same signal. Not a big problem and there are ways around it. You can hook the signal directly up to the MS and have the MS generate another tach output signal for the stock ecu.

Are you tapping the signal from the cam sensor itself or tapping it from where the ECU sees it?

Make sure that the signal cable is shielded with a shielded wire
It should stay away from ignition components. Ignition components have A LOT of noise that could scramble the cam crank signal but if you are using the stock wires I don't think it would be too much of a problem.



What worries me is that the tach on the dash is also seeing the problem and that it isn't just megasquirt seeing the problem. if it was just megasquirt, you know something inside wasn't soldered up or using incorrect pullups to take the signal from the car. Check the signal connection. The lost sync may be present still at higher RPMs but not enough where it makes a difference for the computer to put a check engine light

Are you using someone elses spark and fuel maps for a 2jz or are you using what the ECU came with?

*edit because I really don't know much about the 2jz motor..*

is it a distributor or Coil pack or coil on plug?

You could totally switch to using a coil pack or coil on plug setup on that motor with that MS3 setup.

Re: Project Lex
 

Again, I'm not the first to do this. There is a guy in Florida who has done this on several of these cars. Go on youtube and search SC300 Megasquirt and you'll find a bunch of southern florida guys using it. His name is Jordan, and he calls his product "powersquirt" so you'll see that in the videos a lot.

The 2JZ uses VR sensors within the distributor. Both cam and crank. I am using the cam signal. I think I'm actually bringing both in, but I have USE CAM SIGNAL IF AVAILABLE checked in the program. Look at post 228 for how I wired it. The NE signal is supposidly the CRANK signal, and I'm using G2 for the CAM signal. I am tapping it so both MS and the stock ECU still see it. I tapped the signal on the wire that goes to the ECU (about 1ft prior to the ECU). The MS wire is shielded. I do not have the end where I tapped grounded though. I pulled back and heatshrunk the shielding on that end. The other end is the MS connector, and I'm assuming that end is grounded internally or something. I believe that's how it calls out to hook it up. Don't ground the shield on the "sensor" end, but ground it on the connector end.

Again I don't know much about MS yet, but here's my initial thoughts on why it also affects the cars tach. It is losing sync. Not the TS program, but actual MS is losing sync correct? Well MS is controlling the fuel and ignition. So if it loses sync, the fuel/ignition being output from MS for that split second is "bad", so the car is actually missing for that split second. You can feel it, you can even hear it in the video. It's not just the tach signal, the car is really actually missing when the sync loss happens.

Like I said, next week when I have time I'm going to pull the carpet back and resolder that connection, and ensure no conductors are exposed. Maybe throw some more tape on there. Also I may hook up the stock ignition again. So the stock ECU will still be controlling Ignition, and MS is only controlling fuel. See if that improves anything. I kind of wish I would have done that in the beginning. Just 1 at a time changes, makes troubleshooting so much easier :)

So as for MAPs. I've sent Jordan my MSQ file and he adjusted it and sent it back with a very "base" tune. That's the plan is he's going to help me get it running decent, but then I'll be taking it to Shane to get a real tune done, just want it running fairly decent as I don't want to give Shane an untuned mess that he has to start from scratch on.

Yes I know I can switch to COP with MS3x. But like I said, I'm trying to do 1 at a time here. I plan on keeping the stock dizzy for now. In the future I may look into changing that, but for now, 1 thing at a time...


What system are you running? MS2, 3, etc? I'm working on making myself a custom dashboard in TunerStudio. You got any cool dashboards I could see or use? I like that default Green one that you see in the video that came with TS, but I'm going to make something with histographs so I can watch actual logged AFR vs. Target AFR and what not. And only throw gauges on there that I care about.

Re: Project Lex
 

I know you are not the first to do this. The 2jz and the 7m have both been MS by version 1 2 and MS3 (the MS3 is the most interesting to me though. Lots of options and thought put into that standalone) Myself, I am not familiar with MS ing a 2jz but after building a couple for a couple of different cars, it is really easy to understand how any ECU works on any car. My best description is "It is like hooking up a stereo to your tv, dvd, ps3.... There are inputs and outputs plus you must make sure it is plugged in. " Hopefully I am not sounding rude at all when I make comments. Just trying to help a fellow MSer

After I asked the question I googled it and found out it does use two vr sensors, one is a 12 tooth wheel and the other is a Top Dead Center trigger so ms knows which one to fire on for sequential injecton. You should be able to tap into it with the MS and stock ECU both in place if the stock wire wasn't shielded, it really doesn't matter if the MS is shielded. Rarely is there noise by the stock ECU. it is usually in the engine bay by the coils.

TS is just reporting what MS actually sees and yes the MS is loosing sync, I can even hear it in the video when it misfires.

You don't have the wires for the 12 tooth and TDC switch because if you did, it wouldn't run at all

My comment is that it has to be down stream somewhere because your stock ECU is also loosing sync. Most often for me when it is loosing sync like that, it is because the builder of the MS soldered something inside the case wrong and so MS looses sync. Tach usually reads just fine besides the dip when it misses.

Your case is different. My guess is that your tach on your dash gets a signal from your stock ECU. So both the stock ECU and MS are receiving a bad signal somewhere down stream from the ECUs. It could be that you have to recheck the solder that you made when you tapped in to the signal cable.
It could be that your VR sensor in the distributor is going out when it idles and can't pick the signal up.

Bah it is easier to hook them up at the same time so shane can give you an optimum tune on the car:wink:

It is the spark map that usually needs to get tuned on the dyno. if that is dialed in correctly tuning the VE table is something that can be done on the fly most of the times. I ran someone elses tune on my eclipse for sometime


I use MS2 as that fits my needs for most 4 cylinders (and works pretty well for V8 also). I modified the MS chip so that it would do 4 cylinder sequential injection with the help of an "adder board"
TS is pretty universal with both. I don't have any need dashboard sorry to say. I usually stick with the standard gauges and fit which ones I need to see on the dash. Looking at the green one with bars gives me a headache but Ricers LOVE it. "hey you have a standalone on your car!" now I need to find a warning flasher or something that says my manifold pressure is too great....

GET THE FULL VERSION OF TUNERS STUDIO, there is a tune as you go option that is not on the free ware. Pay the money it isn't that much and that will help you tune as you drive. It helped me dial in my VE table. (it is really hard to tune with only myself in the car...

Re: Project Lex
 

Don't worry about sounding rude. I'm a newb to MS, so I have lots of learning to do as the most complicated system I've ever used before this was DSMLink :)

The way I understood it, the 2jz-ge (non-turbo) uses a 24 tooth nippon style wheel (no missing tooth) with one sensor (NE) for crank position, and a 2nd wheel that has one tooth and two sensors (g1 and g2) for cam position. All within the dizzy. I have the NE signal coming into the tach input on the MS3 card, and the G2 signal coming into the crank input on the MS3x card. Both signals are simply tapped, so both the ECU and the MS are seeing them.


I don't think I ever questioned if I had them backwards. I'm certain they are hooked up correctly.

My understanding is that the 2JZ-GE tach signal does not come from the ECU. It's a distributor setup and that I believe somehow provides the tach signal, not the ECU. And after talking with Jordan, he agrees with my thinking. MS is losing sync. When that happens, my ignition output shuts off. That is why you hear the car miss in the video when it loses sync, and that's why the RPM's drop on both the laptop and on the actual car tach. Because it is actually missing, and the rpm's are actually dropping. The sync loss is what is causing all this, so I need to just get that sync issue fixed and that should take care of most of this.

Yes when I bring it to shane it will have injectors and timing both hooked up. I was saying I may hook the stock ignition back up as part of my sync issue troubleshooting process.

I'm going to have shane do whatever it is he has to, to make the car run as good as new as he can.


Yes I have the full version. I started with the free version, but since I have these sync issues, I decided to get the full version so I can use the trigger/tooth logger function.

Re: Project Lex
 

Yesterday I was able to work on the car for about an hour. All I did was pull the carpet back and add some additional electrical tape to the wires I tapped into to ensure no conductors were exposed. The sync problem still remains however :( My next plan is going to be double checking some settings in TS. If that doesn't fix it, then I am going to double check the ground wire, and maybe see if I can't improve that. Then I will rehook up the stock ignition wire, and see if the sync issue goes away. Then just keep working my way back trying to figure it out...

Re: Project Lex
 

Well today got a little bit done. The original vacuum source I was using wasn't very good (kPa value would freeze soon as the TB closed). So I found a different spot and tapped it from there.
http://mitsustyle.com/forums/attachm...1&d=1296608917


Also decided to work on the sync issue as well. I was told to double check the base ignition timing on the car. So I looked up how to do it and went to it. Should be set to 10*. I found mine was sitting at about 12*. Clocked the dizzy and set it up right on 10. Here's a little how to if anyone ever needs to know how to set base timing on a 2JZGE.

-------------------------------------------------------------------------

SETTING BASE TIMING ON A 2JZGE (1992 SC300)

- Find the diagnostic port in the engine bay, and jumper terminals TE1 and E1 with a paperclip.
http://mitsustyle.com/forums/attachm...1&d=1296608917
http://mitsustyle.com/forums/attachm...1&d=1296608917


- Connect the timing light to the battery, and to plug wire #1
http://mitsustyle.com/forums/attachm...1&d=1296608917
http://mitsustyle.com/forums/attachm...1&d=1296608917
[img][/img]


- The timing marks on the plastic cover will read | 10 | 0 (this is in increments of 5 so it read it as 15 10 5 0 ). With the pins jumpered, you want it set at 10.
http://mitsustyle.com/forums/attachm...1&d=1296608917
http://mitsustyle.com/forums/attachm...1&d=1296608917


- Point the timing light and look for the line on the crank. It should line up at 10. If it does not, there is a single 12mm nut you need to loosen on the distributor. Clock the dizzy until the mark lines up.
http://mitsustyle.com/forums/attachm...1&d=1296608917


- When you're done, tighten the nut back up on the distributor, and check the timing one last time to ensure it didn't budge while tightening the distributor. Remove the jumper. With the jumper removed, you will notice the timing will advance a bit automatically which is normal.

Re: Project Lex
 

Well it's been forever since I've updated this. Mainly because there has been nothing to update as life has been quite hectic since the beginning of the year.

Today I finally got some time to continue troubleshooting the sync problem. Put a scope to the Cam and Crank inputs as noise was suspected as being a possible cause of my sync/tach issue. Here's a nice vid. Looks clean to me. It's a 24 tooth (no missing tooth) crank signal, and I have just a single G2 cam signal coming in. You can see pretty clearly that they look pretty clean. You see the single Cam trigger, then 24 Crank signals. I'm no expert but it doesn't look like there's much noise on there at all.

What I do notice is how close the cam trigger is to the crank trigger. From what I understand, you do not want those to line up perfectly. The cam trigger should come in between two of the crank triggers. Maybe that is my issue?

If anyone has any advice, feel free to share! And big thanks to Shane at DB Performance for lending me the scope. Greatly appreciated!!


http://www.youtube.com/watch?v=gs2vCC5x3KY&hd=1
http://www.youtube.com/watch?v=gs2vCC5x3KY

Re: Project Lex
 

OK did a little bit of playing today per the advice of some gurus. I played with the Zero Crossing pot, and with the different edge triggering options. See pics below.

BEFORE Adjusting Zero Crossing pot log.
http://www.msextra.com/forums/downlo...e.php?id=17743

AFTER Adjusting Zero Crossing pot log.
http://www.msextra.com/forums/downlo...e.php?id=17744

- Zoom to see where the cam trigger falls in comparison to the crank triggers. Ideally it should fall exactly between two of the crank pulses. I played with the different edge options. This pic shows the best combination I could get which is Falling Falling.
http://www.msextra.com/forums/downlo...e.php?id=17745

So the cam signal is much better now. But I'm still having the same problem. The cam pulse appears to be as steady as can be, so the crank signal has to be the cause of all this. Will continue this battle this weekend and hopefully get this thing working!!

Re: Project Lex
 

Wow, seeing you deal with all that makes me feel a little better about the crap I had to go through to get my truck running.

Re: Project Lex
 

Ha, there definately is a learning curve, but everyday I'm figuring it out a little more. Hopefully have this worked out this weekend.

Re: Project Lex
 

Had a couple minutes to look at some more stuff. In this video I'm reading the raw cam signal (G2) coming in, and am reading the processed cam signal on JP2 pin 18 on the silver MS3 board. In a nutshell MS just takes the sine wave signal, and creates a square wave based on where the raw signal crosses the 0V line. Thanks again Shane for the scope, I wouldn't be able to do any of this if I didn't have it.


http://www.youtube.com/watch?v=t3eBiVhBgWo&hd=1

Re: Project Lex
 

All this signal checking crap has finally convinced me to buy my own scope. Big thanks again to Shane for lending me his as it's proving to be a huge help, and helping me learn a ton about how MS works. I'm hoping he'll let me hang onto it for a little bit longer yet until I get the hang of my new scope.

http://www.hobbylab.us/default.aspx
That's what I got. $129 It's a USB based device that you use with your computer (I'll just use it with my little netbook). It's been used by a few of the MS guys and they've all said it works quite well. It's a 2ch Oscope, and a 16ch Logic Analyzer. The software is apparently pretty easy to navigate as well. I heard the probes that come with it leave a little to be desired, so I picked up some better ones off Ebay for $12 shipped.

Someday eventually I'd like to get a true standalone scope w/ 4 channels, but for now this seemed like a great starter unit for a decent price. But in the future, if I could find some old analog Tektronix 4ch sucker for cheap on Craigslist, that'd be cool :)

Re: Project Lex
 

Checked signal at J1 which is the crank square wave. Here's a vid of the crank signal. Was hoping I'd find a smoking gun on this today and be able to work it out with the filtering function in TS, but nothing really caught my eye.


http://www.youtube.com/watch?v=SD4od1097tA&hd=1

After looking at the square waves of both the Crank and Cam, it seems like the correct settings for this in TS would be Rising & Rising.

Starting to run out of ideas here as so far nothing has stood out to me as being an issue thus far. I'm going to take another video comparing the square cam & crank to show how they match up. After that, I really am starting to run out of ideas. I just ordered up a HobbyLab USB scope so I should be able to take some logs with it with the Logic Analyzer function that might be able to tell us more than this old scope meter.

Re: Project Lex
 

Is it an ignition or fuel problem, or a problem at the root of both systems? If it could be one or the other but not both, have you thought about working backwards straight from the ignition or injectors?

Re: Project Lex
 

I have unhooked the ignition, so the ignition is being run by the stock ECU again, and MS is only controlling the fuel. MS is losing sync, which is all driven off the Cam and Crank position sensors. MS needs to see 24 crank pulses, then 1 cam pulse, then 24, then 1, and that's how it stays 'synced' with the motor. When it loses sync, it shuts off the outputs to be safe (so in this case, fuel) which is why you hear the hesitation. The sync error code I receive is 17 which means something like "wrong amount of crank triggers before seeing the cam trigger". Whether or not it's something in the hardware that is the issue, or a setting in the software, I am really unsure of. Support from the guru's are starting to fade little by little as well which doesn't help.

And here is a vid of the crank vs. cam square waves.
http://www.youtube.com/watch?v=kfUace3pZzM

Re: Project Lex
 

OK well just took a wild stab and it seems to be 100 times better, although not completely taken care of. Since earlier some signs were telling me that the crank trigger was the likely cause of this, I decided to enable crank only filtering. I used the Tach Rejection feature, and set it to 30% as that was a value I had seen thrown around in the filtering documentation on here. And what do ya know, the issue seems to have 99% gone away. You'll see in this video that it still lost sync a couple times under WOT, but way better than before. I will play with filtering a little bit more and hopefully be done with this issue.

One thing I did notice was that in the composite log, the crank now shows up as a square wave instead of just a single edge as it did before. No idea what that means however.

http://www.youtube.com/watch?v=EZPwl3fj34I

Re: Project Lex
 

Played with the filtering a bit. I have the global filter time now set to 5uS. That basically means, any pulse that is less than 5uS in length gets filtered out. Anything longer than that is considered good and is used. So basically goes off the assumption that good signals are long, and noise is short.

I then set the rejection percentage to 70%. That means that it predicts when it should see the next pulse. So anything that happens within the first 70% of that time gets ignored. For example say it predicts it'll see the next pulse in 100ms, it'll wait for 70ms until it even starts looking for it again. So any noise that is occurring in the first 70ms is completely ignored.

Car is running like a champ!!! Goint to hookup the ignition again and make sure she's running good still. If so, I may start talking to Shane about getting a quick base tune before the turboing begins!!

Re: Project Lex
 

^Awesome.

Re: Project Lex
 

Bman, the new MS guru.

Re: Project Lex
 

NICE!! Turbo'd by dyno day?

Re: Project Lex
 

Glad to hear you sorted it out...was getting worried there for a while. Looked like a tough little cookie of a problem haha.

Re: Project Lex
 

Glad you figured it out, cause that shit was way over my head haha!

Re: Project Lex
 

I understood a lot of it, and I'm glad it was you doing it, not me. ;)

Good work sir.