Is anyone here successfully running MS2 on a TPI car?

I have the following setup:
91 TPI LB9 mostly stock runs pretty good on factory ECU with some tuning.
MS2 v3 with standard B&G firmware. I did the following mods for use with a GM 7 or 8 pin HEI:
OPTOIN <-> TACHSELECT
TSEL <-> OUTOUT
XG1 <-> XG2
HEI Pin "B" to relay 5v only in run not crank as shown in the MSExtra Manual.
I used the trigger wizard to set the MS timing advance to what the timing light shows.
The problems I am having:

I had to make the mixture too fat so it would idle smooth. The MS2 needs 12.5 or so to make it idle where I have it at 13-14 on the factory ecu.

My Autometer tach is erratic at lower rpm and seems jumpy and delayed when you rev the engine. It is perfect with factory ECU.

Everything seems to be pointing to ignition signal into the ecu. If someone out there has this setup working, please let me know how you did the ignition setup and modifications.

 

Update in case anyone else tries this:

I did "The transistor modification" it is #3 on the page I found here:

http://www.megamanual.com/ms2/inputHEI.htm

It helped a lot, hach is usable and timing light is way more steady. Now I need to figure out why it needs 12.0 AFR to idle...

 

Get anywhere with it?

Make sure you configure your injection settings properly based on the injector dead time and size. Your firing strategy, etc.

-- Joe

 

I have MS3 extra on mine. In figuring out idle, to begin make sure all the idle spark advance timing cells are the same. 600-1000 should have the same value for now. Make sure the fuel idle cells are also fairly flat. That should stabilize the idle enough to start to tune it. Also remember to have it where with idle air control valve is fully closed as it should be for now and that when it is the engine sees a slightly cracked throttle blade in order to prevent running off of spark advance alone, iac valve alone or a combo of both. If this doesn't get idle tuned you have a setting that is incorrect. Joe might know more about MS2 but I remember most of its parameters.

 

I had given up on the MS and thought something was wrong with the hardware or my settings. However last night driving home the last 10 min of my 40 min commute the car started acting up. It was basically doing the same thing on the factory ECU that the MS was.

The car started stumbling and bogging. I had rapid change in AFR from over 16 to under 12. I don't think the engine could really change as fast as the gauge was reading, so I think the O2 sensor has gone bad.
Also for several miles it was reading AFR over 16. There is no way that could be right. I think either the O2 is bad or the Innovate controller is bad. Probably the o2 because it is pointed up and not down like it should be. I was waiting to replace the whole exhaust system to fix it. Probably a bad idea....

My diagnosis so far: O2 has been going bad for some time, the AFR reading was fairly steady when I first hooked it up. After time it has become more erratic. I think since the factory ECU is connected to the Innovate 0-1V narrow band emulator. It must be better at compensating for the o2 problem and ran ok for longer. The MS is connected to the wideband 1-5v and was more affected by the sensor going bad. Maybe?????

Stay tuned, I will post here when I figure it out.

 

I think you are spot on.

Also, the MS is faster (loops through sensor input quicker) and will react to changes much more rapidly than the stock ECM.

-- Joe

 

To be sure do you have a mega stim to check to the MegaSquirt? Have you checked all the main grounds? Have you done a free air calibration on the innovate recently?

 

Yes I did stim testing on MS when I was building it.
Grounds are all factory, so I think they are ok.
No I have not done free air calibration since I put in the wideband, but that will be my first step.

 

Update:
I fixed the O2 sensor problem and the factory ECU is working fine now.

The MS2 was still giving me the same problems as before, so I decided to upgrade to the latest firmware, 3.4.0 extra.

Now it will not start at all. I hear it fire while cranking, probably because the ignition module is running spark, but when I let off the starter it dies. I think I have the ignition and fuel settings right, but there is probably an issue.
Not sure what to do, if someone can look at my settings file and see what I did wrong that would be much appreciated.

 

Did you read the release notes on the firmware?

A while back, they flipped some of the settings relating to ignition. You also need to re-calibrate everything when upgrading firmware.

Can you zip up your whole project and email it to me? anesthes@cisdi.com

-- Joe

 

The only thing I saw was this:

"Going High" - was called "Going High (Inverted)"
"Going Low" - was called "Going Low (Normal)"
DOUBLE CHECK YOUR SETTING BEFORE CONNECTING COILS.

But that should be the same as I had it.

I re-calibrated TPS, I think the temperature sensors are ok, but will double check.

 

Go back to B&G code. I have the same issue with my MS2 3100 hybrid. Runs pretty decent on B&G code but the second I put in Extra it won't run at all (as in no start PERIOD). And for some reason, I seem to have to repeatedly recalibrate my TPS for the MS to actually read 0% consistently. I don't know about other users of the MS2 but these things seem to be pretty cruddy for my tastes...
Innovate LC-2 reads consistent .4 lambda LEAN in MS2 compared to Innovate gauge. And yes, for those who may ask, I have done free air calibrations 3 times now. It's the MS2 that's the problem. I can get it dead on by playing with the calibration curve, but the second the engine is revved the reading skews again.
Knock sense is a joke. Still supposedly seeing knock counts with knock sense wire completely disconnected from MS altogether (as in pulled out of ECM connector block and taped off). Oh, and not to mention the Knock reading very closely follows the curves for the TPS and MAP...
IAC control NOT automatic. Requires a restart to get idle to settle to 12-1300 RPM hot idle because IAC uncontrolled from cold start and idle speed ends up about 2500 RPM. Manual control works fine.
Have to set all averages for all sensors (CTS, IAT, TPS, Volts, O2S, MAF, MAP) extremely low to filter out sensor spikes non-existant in stock ECM.

Your ignition settings are just fine. Make sure your spark offset is set properly.

I don't mean to intrude on your post by posting up my issues but I'm hoping to point out what you may be up against while tuning your engine. MS forums are a joke, BTW. Don't expect much help there.

 

Strange that some of you guys have problems with these.

Who assembled your MS2 ?

-- Joe

 

Joe: I assembled the MS2 myself. I did all the testing and it seems like the hardware is fine.

Maverick: I agree about the docs being SO bad (even by open source standards) My next step is to go back to the original code, but it really didn't run that great either.

I am going to take another look at my grounds, maybe I have some kind of feedback loop?

I just don't understand why the car will idle fine at 14.7 AFR on the factory ECU and need 12.5-13 on the MS2. The timing is the same (checked with timing light)

I really want to make this work, but I keep spending time on it and not making any progress, it gets frustrating....

 

Most of the problems people have with MS2 is poor assembly and shielding problems. Yes, you might have a ground issue. This is why I recommend most people use the Microsquirt like I did.

http://members.cisdi.com/~anesthes/p...om/megasquirt/

The B&G code shouldn't be used. MS2extra is where it's at. But you need to make sure your ignition settings are proper, and your injector firing strategy is sane.


-- Joe

 

I was going to put the original firmware back on and realized that I had the bootloader jumpered. My bad, I will try it again see if that makes a difference.

 

My Fuel Settings:

Required Fuel: 21.00
Cont Algorithm: SD
Squirts/cycle: 2
Inj Staging: Alternating
Eng Stroke: 4 stroke
# Cyl: 8
# Injectors: 8
Engine Type: Even fire
Size: 5000cc
Injector Size: 217 cc
Seq Injection: Untimed Injection
Inj Drivers: standard

Ignition Settings:

Spark Mode: basic trigger
Trigger Offset: 5.00
Skip Pulses: 3
Ig Capture: Rising Edge
Spark Output: Going High
# coils: single
Spark A output: D14
cam input: cam input
flip priority on cam: Normal
Fixed Advance: use table
use prediction: 1st deriv prediction
cranking dwell: 6.0 (should not be used, cranking spark done by dist. module)
cranking advance: 10.0 (not used)
normal dwell: 3.5
spark duration: 1.7
Spark latency: 0

 

Do you have a relay so +5v is OFF during crank on your ignition module?



-- Joe

 

Yes I did it just like:
http://www.megamanual.com/ms2/GM_7pinHEI.htm
I have tested it and it works.

 

That mimics how the OEM did it. I don't know why OEM doesn't want cranking advance.

Is it working now that the boot jumper is removed?

-- Joe

 

I hooked it back up to the stim and have confirmed that analog readings are coming into the ms2 correctly.

I checked the grounds: my innovate lc1 and gauge are grounded near the master cylinder. The MS2 is grounded to multiple grounds from the factory harness. I will try grounding the innovate to the MS2 and see what happens.

 

You are not using tunerstudio lite are you?

You need the full version of tunerstudio.

-- Joe

 

I have the full paid version.
I tried a new project and it seems to be fine. It must have been my old project file from B&G firmware did not upgrade to the extra code. I needed to start over anyway.
I will post back when I actually hook it to the car.

 

Ok.. Good luck with it.

If you wanna zip up and email me your project I'll take a look too. You've got a lot of flexibility with the MS, I like to set things up a certain way.

-- Joe

 

Tried again to get the car to start on 3.4 Extra and nothing. It starts to fire when cranking, but when I let off the starter it dies. I am going back to the B&G code.

 

Looked at your tune. I wish you had zipped up the whole project rather than just the MSQ, but anyway.

You have 5* initial advance?

19lbs injectors ?

You don't have an actual wideband configured. (under tools->calibrate AFR table).

Your ignition settings are not correct for HEI. You have "Going high" when you want "going low" for spark output. This is probably why it catches and dies, it's running off the module but the second you're relay supplies +5v to the module it has no trigger and dies.







-- Joe

 

Back to the B&G original code. It runs ok, I changed from 2 to 4 injections per cycle. That seems better. Still I can only get it to idle at 13.5-13.8 AFR. Any leaner and it oscillates.

 

Sry, I do have the wideband set, dropbox didn't sync that before I sent it to you.

5 deg is the trigger offset that should be right we set it with the timing light.

Yes 19lb Bosch D3 injectors

I will try "going low" but all the docs I have read say to use going high. Also that is how it runs on the B&G original code.

Thank you for your help, I will zip and sent both projects and resend them.

 

Check this out:

http://www.msextra.com/forums/viewto...?f=101&t=34105


From the thread:

"
The manual suggests to set spark A output to D14 and spark output to Going HIGH (Inverted) this is incorrect.

Correct settings verified with scope after the MS2-Extra HEI manual settings burned up a new module within 10 miles.

If using D14 for spark A output you MUST set spark output to Going LOW ( Normal ), this would be due to Q6 inverting the signal from the proccessor.

If using JS10 for spark A output then Going HIGH (Inverted) should be correct since there is no transistor inverting the signal. I have not scoped this, but it does agree with megamanual http://www.megamanual.com/ms2/GM_7pinHEI.htm

Also might think about recommending using VR input for GM 7 and 8 pin HEI in our testing on the dyno it has proven to give a much more stable signal to the megasquirt ( MS2 cpu V3 pcb )"

-- Joe

 

I need to do some more work on setting the trigger offset, I had the car running, but it idled it 1500RPM, I believe due to the timing way too advanced.

I will post more results here.

 

Worked on it some more. The trigger offset was way off, it went from +5 to -2 when I changed from going high to going low.. The car still will not run right.
It idles around 1500 RPM.
I had to richen the ve table *8 to get it to idle.
Barely touching the throttle goes dead lean and backfires.
I can't put in enough fuel to make it want to idle.
I tried to start over with a new project, but it copies your existing project, so it is just as screwed up as ever.
Giving up for now, might try again later...

 

Send me your current project.

Zip up the whole directory adn send it over.

-- Joe

 

Ok, so after trashing my ignition module, I figured out that you need "charge going high". It does get inverted somewhere in the circuit.
I tested with an LED the way they say here:

http://www.megamanual.com/ms2/configure.htm

You should set the spark output such that no current flows through the coil when the rpm=0. To do this, replace the coil with an LED tester (see step #6).

Setting the spark output one way should result in the LED being off with no rpm - that's the way you want it. If It is on, the coil is flowing current all the time, and it will overheat, and the ignition module (or VB921 or BIP373) may overheat too.

I could not get the "extra 3.4" to run spark. The engine runs with the spark signal from distributor but not from the ECU. Something in my extra setup is not firing the coil at all. (charge going high or low)

 

Looking at your tune. (extra)


Regarding fuel. with 19lb injectors, I'd use simultaneous and 1 fire.


Why is your trigger offset -5.0 ??

I have my dist set to 10* initial. what is yours?


Also, my ignition input capture is falling edge. Yours is rising edge??

Disable prediction.

HEI external coil should be about 2.5 msec dwell.

You set the hardware latency to 38?? I'd go back to 0 and tinker when that when you've got a running tune.

I'd throw a lot more advance in your ignition table. I idle around 22 degrees.. Then again, I don't know what the deal is with your trigger offset.

You should unplug your est so you can get your initial base timing set to 6 or 10 degrees, then fix the trigger offset to match.

-- Joe

 

The trigger offset I copied from the B&G original setup. I can't set it for Extra until I get it running.

The dist is set for 6* the factory setting. I will do the trigger offset once I get it to run.

I will try more initial timing, maybe that will make it run.

 

The trigger offset IS the initial advance. (base timing). If you know it's 6 degrees, set it to 6.

The trigger tool is only to VERIFY it. I usually check it with a light with EST disconnected, and then with EST connected to make sure my numbers match. (i.e, make sure ECM is controlling timing).

-- Joe

 

I tried the MS again with the Extra 3.4 code.
I set trigger offset to 6*, made sure latency was 0, and 2.5 ms dwell.
The MS ecu will not fire the car. It starts to fire on the starter and then dies as soon as you let off the starter.
I put a scope in the signal wire to the dist (pin 36). With the factory ECU I get a good clean square wave. With the MS I get nothing.
I must have a setting wrong so it is not sending the pulse.
It is getting a pulse from the dist because I can run the car with the dist module running spark. The MS is firing injectors and the computer screen shows RPM.
I confirmed on the simulator that there is no square wave on pin 36 and the IG led never fires.
I will load the original B&G code and confirm that it silll works.

 

I flashed the original B&G code back and hooked up the scope. I got a really crappy sine wave, nothing like the factory ecu. I don't know if it because of the simulator?
I will try it on the car and report back.

 

If you have it set up for B&G code, you need to have JS10 wired to IGN and nothing else connected to IGN. An MS2/Extra tune set up for backwards compatibility will require Spark A Output Pin set to JS10. Both code versions will require setting spark output to Going High.

 

I think he has it wired for spark A output on D14.

-- Joe

 

I THOUGHT I had it wired to D14, but I tried JS10 and it worked, so I guess I was wrong.

For the Extra code Going High is correct, for Original it is going Low.

I tried it on the stimulator and I got a really nice square wave. Now I just need to try it on the car...

Here is the scoop on going high vs. going low:
This is if you are controlling the coil through a GM 8 pin module NOT coil directly connected to MS
You want the output pin to be low when at 0 RPM. This will give you the correct square wave like the factory ECU. It can be confirmed 3 ways:
-The D1 (IG) led on the stimulator will be ON at 0 rpm
-An led with resistor hooked from ground to pin 36 should be OFF at 0 RPM.
-If you hook a scope to pin 36, it should have a baseline near 0 and rise for the signal when running, near 0 when RPM=0.

 

I have the car running on the extra 3.4 code for the first time. I have a strange occurrence:

Every 28 seconds or so the engine tries to die. Pulsewidth goes to 0 and then recovers. I have it dataloged, but I don't really know what to look for. Right before the drop there seems to be nothing to cause it.

Very strange, I never had this issue with the B&G code so I am assuming that I have some setting wrong.

 

One other thing, I switched the homing position for the IAC valve to OPEN, since it seemed to be running backwards. I double checked, the wiring is correct. On the B&G code it seemed to be fine.

Anyone know what it should be set to?

 

Can you email me a log of it idling right up until it dies? That's a new one for me..

As far as the IAC, I'm running a Microsquirt module (like a DIY_PNP), not a MS2 so my options are not the same as yours regarding IAC. May need to google that one.

-- Joe

 

Joe: I sent you the log file

Here is a screen shot:

 

Wow..

I looked at the log pretty closely.

I assumed that you were losing DRP's and the MS was shutting down the injectors, but no, it's still seeing DRPs..


So what the hell is shutting off the injectors.


Matt ??


Edit: actually, I wonder if you are hitting a fuel cut mode that's meant for cruise. The thresholds are tps and map if I recall.

Edit again: yep. Just checked your config.

Your fuel cut is set to:

RPM > 1200
TPS < 5%
MAP < 40
CLT > 90

Return fuel RPM = 900



-- Joe

 

Not sure fuel cut was the issue since it never hit 1200 RPM. Now the 28 sec problem has gone away without me changing anything...

I have discovered the idle air test mode. I think the IAC is working properly on "warmup only." I will try closed loop mode more later.

It seems to want IAC set to 130 to idle at around 700-800 RPM.

I drove it and did some basic tuning. It seems to run OK. I think I need to get the knock sensor working before I do any more.

I also need to check the trigger offset.

I am still running 2 pulses per cycle alternating. Injector pulsewidth seems too high, I may need to try 4.

At least I am making progress...

 

Well, it actually does when it recovers but. I still think the feature is suspect. It might even have a bug. But beyond that, the settings are insane. I'd turn it off. I don't even use that feature.

You only have the idle control set to open-loop (warmup). Basically, you have it set to mimic a choke.

2 pulses alternating is normally what we do with huge injectors (55, 60lbs). With small 19 lb injectors I normally set it to 1 pulse, simultaneous.

-- Joe

 

I ran it again last night and got the idle almost tuned in. It is still a little fat, but better than it ever was with the B&G code. I go the trigger offset set.
I have it on 4 pulses alternating, that seemed to run the smoothest. I tried 1 pulse simultaneous, but it was pretty rough. Listening to the injectors they sound the same as on the factory ecu.

I want to get the closed loop idle working eventually, it is nice when you let out the clutch or are idling up hill. On a truck it would be a really big deal.

I am thinking about getting this knock module:
http://www.viatrack.ca/

it was recommended by dyiautotune

For anyone reading this <b>SKIP the Original B&G code</b> go right to the Extra code. There are far more settings to get right, but it is worth it.

 

I'm using a stock ESC module for an '88 L98 car.



Megasquirt has a million configurable options and is designed to work with just about every ignition trigger available. It is very powerful but very complex. This also leads to frustration because people want to just plug it in and fire it up like a stock ECM. It's not "stock" for any particular vehicle. A "stock" ECM or EBL will run on a specific engine combination. (V8/V6 with a HEI type distributor). The Megasquirt will run a bike, boat, airplane, etc.


It's so far advanced in functionality and adaptability. But because of that, and the need to really understand how it works before using it most members opt for a simple stock ECM or EBL. They don't know what they are missing, but they wouldn't know what to do with it anyway.


I'm glad you are making progress. The more you tinker the better you will get.

-- Joe