Hello,

13B-turbo, s4 engine with s5 rotating assembly
LS1 coilpacks
MS2V3.57 with zeal
1000/1600cc injectors
90mm throttlebody and OEM tps from VH45DE



Car will start after some cranking, I'm able to keep it running with the throttle for a minute or two and then it dies.

I noticed that while holding the RPMs steady, I'm getting some spikes of 0 RPM.

I believe I set the potentiometers on the board to the minimum setting per the megamanual during the install but I can't be positive.

Can anyone suggest some troubleshooting steps to fix my rpm signal and get it to idle?

.msl:
http://www.filedropper.com/2015-05-23154605

.msq:
http://www.filedropper.com/2015-05-26184907

Thanks!

-Nathan

 

Everything checks out okay with the megasquirt, pretty sure this is due to noise in the CAS wiring somewhere. I have the CAS wiring split out of the shielding at the firewall bulkhead, and the coils are almost mounted on top of the CAS, so there's two likely sources.

I already gave up and ordered the trigger wheel kit from FFE.

 

Your ignition wires pass directly over the CAS.

You probably know this, but the CAS has two VR sensors inside. Magnets wrapped with many turns of copper wire. Which make wonderful little amplifying antennas for HV cables pulsing right above them. Bet if you scoped the VR lines you'd find big spikes in concert with the ignition firing.

Just a word of warning from tuning a few big throttle bodies: Expect about 25% more idle PWM needed to maintain a steady idle, with a richer mixture. Now that the progressive throttle is gone and all ports are being fed, wall wetting goes all to hell. Same goes for accel enrichment. About 50% more fuel needed.

 

I wasn't sure if the coil mounting would be a problem or not, I couldn't find any info on it so I gave it a shot. Do you think the coils will be okay where they are with better wire routing, or do I need to move them to the fender?

Actually don't have an IAC of any sort at the minute...sounds like I'm in for more surprises. This car is a track toy, so drivability and idle quality aren't a huge priority.

Thanks for the help.

 

I'd move the coils to the strut tower that would completely remove them from the equation.

 

Well, I unbolted the coils and just set them on the inner fender above the brake MC. Car fired right up and happily idled with a little throttle input at 1200rpm until it died out again 20 seconds later. Zero on the lost count gauge...time to start tuning! Thanks guys.

Funny thing is, I just moved the coils OFF the the side of the strut tower to make room for the new charge piping.

 

The CAS is very sensitive to noise.

Glad you got it going!

That throttle body is perfectly suited to constant moderate-full throttle operation. Nice thing about a different throttle body is the new full range TPS an the fact the shaft O-rings don't leak like hell.

 

Okay, a little more insight into the "something else" that is giving me problems...

I was having problems getting the coolant temp sensor to read correctly. I tried recalibrating it several times with different numbers I found online, to no avail, before pulling it and testing the resistance of the sensor to build my own curve. The numbers I measured fit the curve provided by DIYautotune perfectly, meaning the calibration is not the problem.

Meanwhile, the TPS reads perfectly and sweeps cleanly with the engine off, but bounces 1-3% with the engine on. This is enough to kick in TPSdot AE and doesn't allow me to properly set up AE.

With the DB37 plugged into the megasquirt, my multimeter shows continuity between chassis ground and the ground pins on the TPS and CLT. With the DB37 unplugged, the multimeter doesn't beep, but it does show a reading...I believe it was around 1860 ohms.

So...it appears my sensor ground is bleeding over into chassis ground somewhere, or my sensor ground is contaminated with chassis ground through some component. I'm going to print out the grounds section of the manual and pour over the harness with a multimeter...any guesses on likely places I should look first?

 

Chassis ground? Don't you mean engine ground? The MS should not be grounded to the chassis, it should be grounded to the engine.

 

By not reading correctly, what do you mean?

Jittering?

Or offset up or down a number of degrees?

I'm mentally picturing this but they way I see it, the sensor ground with the connector unplugged should not have any reading to chassis ground (or a massively high one in the order of 100s of megaohms) unless you have tied them together somewhere. Now they are tied together on the MS board.

Try unplugging all the sensors and measure again.

 

I was only using the term "chassis ground" to distinguish from "sensor ground". The Megasquirt is grounded to the engine, on the same lug that grounds the engine to the frame.

The other end of this wire terminates on the strut tower and has the coil ground on it...is this a problem? The battery is in the backseat and is grounded to frame only.



The TPS signal was jittery, the CTS was always offset, or static. I have a coolant temp sensor on the dash to check it against.

According to my pinout, it appears DB37 pins 7-9 were tied together inside the harness and run to sensor ground, and pins 15-19 are grounded on the engine. Is this correct?

 

I just went through every ground on the car and cleaned them all up and greased them. None of them were really bad but some were definitely less than ideal.

I started the car and warmed it up until my dash coolant gauge read 130°F.
The Megasquirt CLT bounced between 67 and 68 and then rose to 71, but no higher.

TPS signal is still bouncy, but might be a little better now, I dunno.

I'm gonna make sure I don't have the chassis ground and signal ground wires flipped in my bulkhead connector, but then I'm out of ideas.

 

In all the DIYAutoTune harnesses I've seen, the grounds each lead to their own pin. Then tie together on the board itself. I just went back on a few of my pictures to verify this. That's not to say your harness isn't different so you should pop the DB37 case and check.

Some jitter on the TPS is normal but it should only be 1 or two counts.

Make sure the 5V reference to your sensors is actually 5V.

 

The grounds were clamped together inside the harness to a single thicker wire, 12 gauge or so. I assume I have the sensors grounded correctly, to pins 7-9?

Speaking of DB37...one of the secondary injector wires fell out of the back of the connector. I can see the pin inside where it was once crimped, but it's inside the case and I can't reach it...any hints?

I'll check the voltage at the sensors tomorrow.

 

12.0V at the battery
4.85V at each of the three sensors, whether I check to chassis/engine or sensor ground.

 

Where did this harness come from? The fact an injector wiring fell out (!!!) doesn't bode well. Intermittent connections at the other pins could explain your issues.

I'm not sure how to extract those pins. Google says a paper clip can be used to push out the pin. Never done it myself.

 

It's an older DIYautotune harness, as far as I know. I bought it used.

Anyway, I figured out the problem. I had the chassis ground and the sensor ground pins reversed on all of the LS1 coil plugs. Flipped the wires and all of the sensors read rock solid and the idle is much improved!

Strangely enough, I'm still not seeing zero connectivity between chassis ground and sensor ground unless I have the coils unplugged...I get 8200 ohms now with the coils plugged in. No bother, I found the issue.

I also figured out the other issue. Apparently the IAC pads need jumpers before they work. Ha!

 

IIRC, That's due to how the LSx family coils are setup internally. That 8200 ohm resistance you're reading is the resistance of the primary (IIRC) coil. Coincidentally, unlike old style "dumb" coils, there is no way to read the other coils resistance.

themoreyouknow.jpg

 

Well, that will do it.