Porsche 944 Megasquirt conversion

[TireSmoker]

Scott brought the Fickle Bitch Racing Porsche 944 over for a Megasquirt conversion today. The car was upgraded to a 3.0L S2 engine over the winter, which should be good for +60hp over the previous engine. Since I'm one of the drivers for Mid-Ohio in September, I have a very vested interest, and also a tight timeline.

I'm going to install a Megasquirt 3X, a Spartan 2 WBO2 controller, and a set of GM LS coils for ignition.

-Dave

[Basement Paul]

That motor takes up as much space as an LS. 3.0 is a giant four pot.

-BP

[TireSmoker]

The intake manifold is unusually large, I think.

[Basement Paul]

The angle makes it look bigger too.

-BP

[TireSmoker]

I got some progress under my belt here the past couple days. Starting with a cardboard template, I made a mounting plate from some aluminum (courtesy CutNPaste). I mounted the MS3X on some MSD rubber isolation mounts. I got fancy with the fuse panel and relays, using rivet nuts to mount them. No self-tappers here.

MS3X has two main connectors, the main MS and the 3X. I went through each and pulled aside / bundled the wires that won't be needed in this installation. I'll prob leave 24" of anything I don't use, just in case some unknown future case. Properly taped/secured. The bundled stuff you in the pics will all go away. The two bundles going out the firewall get loomed and secured.

I was able to reuse the factory firewall grommet, which is a really nice piece.

This car has (I think) a on/off switch for TPS. There is also no idle-air (IAC) valve. That's gonna be a little bit of a challenge.


Wires not used & trimmed on MS connector:
Green Outputs: INJ1/2 (batch fire)
Blue Outputs: INJ1/2 (batch fire)
Tan: CAN-L, CAN-H, SPR3, SPR4
Brown Output: SpkA


Wires used on 3X:
INJ A/B/C/D
SPK A/B/C/D
Green input: Cam signal
Green output: Tach out
Green output: PWM Idle (for Fan relay control) ** (there is also a "PWM idle" pin on the MS harness, they are two different outputs)
Blue Input: Ext_MAP (Map Sensor, prob GM)
Blue Input: SpareADC (for Coolant pressure sensor

Wires not used & trimmed on 3X connector:
INJ E/F/G/H
SPK E/F/G/H
Gray Inputs: Nitrous In, Datalog In, Launch In, Tableswitch In
Green Outputs: Nitrous 1, Nitrous 2, VVT, Boost Control
Blue Input: EGO2
Purple Output: Logic Output
Orange Input: Flex Fuel

Updated July 2024:
Crankshaft sensor:
Stock Porsche 3.0 S2 crank sensor, gapped at .9mm / .032"
Danst Engineering 36-1 crank trigger wheel and bracket. (not used)

On replacement pigtail, black = orig connector white, blue = original connector tan/brown
On replacement pigtail, black goes to MS VR+ (MS white) and and blue goes to VR- (MS black)
(see picture down thread)

Tooth #1 angle - 114 degrees (factory 60-2)
Tooth #1 angle ~ 230 degrees (36-1)


TPS Sensor:
1994 BMW 525 6cyl
Falk Engineering adapter for sensor
944 Throttle body rotates opposite BMW, so sensor is wired 'backwards'
*) black = +5V
*) orange = TPS signal

[Fred32v]

Impressive job 'smoke. Can't wait to see it run at Nelson Ledges.

[TireSmoker]

It's been a while since I posted an update, but a lot of progress has been made despire losing almost a week to being sick and other assorted ailments. However, a lot has gotten done.

I'm using the S2's factory 60-2 crank sensor. I'm seeing RPM signal when cranking the engine. I need to verify TDC and make a mark on the front of the engine.

Injector and coil harnesses are done and tested.

MAP sensor is a GM 1-bar and is working as expected. It is plumbed to the 5/16 (8mm?) big vacuum port on the intake, and a 5/16<-->1/4 hose barb adapter was used to get it hooked to the sensor.

Coolant sensor showed a sane value. This might still need to be calibrated.

TPS is from an E30, and due to opposite rotation on the Porsche throttle body, the +5v and GND had to be swapped to get increasing values with the throttle opening. It is mounted using an adapter made by Falk.

Wideband O2 is a tried-n-true Spartan 2 from 14point7.com, along with a Genuine Bosch 17025 sensor, with about a 36" pigtail length.

I picked up a 2.75" to 3" adapter, to make a generic CAI.

I ran across one issue -- the stock injectors are low-impedance. I can't use those. 1-year only spec, and we have it.. booo. Stock injectors are 28lb/hr, I have some 30lb unit that are high-impedance. I'll have to swap those in.

[TireSmoker]

So I got the injectors swapped during my lunch hour today. I removed the low-impedance 28lb injectors and installed high-impedance 30lb injectors. Still no start. Not even a cough, which was puzzling. Further troubleshooting revealed that the fuel lines were hooked up backwards. No wonder I wasn't getting any fuel.

Unfortunately, it's not the easiest to swap the lines, but that's what has to happen.

-Dave

[TireSmoker]

So I haven't posted too many updates, because the car **still** doesnt run. The signal from the factory crankshaft position sensor would drop out to 0 while cranking. I tried reversing polarity, adding a 10k resistor, tried adjusting these 'pots' on the mainboard. Nothing was helping, the Megasquirt guys told me the signal was noisy.

Yesterday (Saturday), Tom Smith came over and we mounted a completely new 36-1 trigger wheel and sensor bracket on the front of the engine. I was planning on using a certain Ford sensor, but it doesn't fit the bracket. The 2nd sensor I tried was another Ford one, and while it fit, it wasn't long enough, to get close enough to the teeth (0.8m/.032" gap). The factory Porsche sensor does fit, and it's sorta working -- at least a way better signal than I got from the factory 60-2 trigger wheel. (same sensor, different trigger wheels!)

When the injectors aren't firing, I get a perfect signal during several seconds of cranking. When in the injectors are firing, it sounds like it's almost gonna catch, the RPM signal drops out, and it doesn't catch. I tried moving/isolating ground wires in an effort to reduce noise, to no avail. I may try a different set of injectors that I have in inventory.

It's been super frustrating, and we've basically decided we won't make Mid-Ohio. Scott has until Sept 1 to cancel the house rental. That gives me a few more days. I'm not too optimistic, though.

-Dave

[TireSmoker]

IT RUNS!!

https://files.tiresmoke.org/f/7ce8ef9ec ... 9ba4/?dl=1 (60mb)

[wxo]

So what fixed it??

[TireSmoker]

This setting. Which, ironically, is what I use on my Chevelle and Malibu, but its a different sensor, and the default is to use Rising Edge. Even when my cars were set wrong, they still started up and ran. I mean, the Chevelle ran *great* with the wrong setting until 5100 RPM, when it had a hard cutout.

But, it's fixed and I even took it for the first test drive today. No headlights, so can't do anything after dark. I just went down my street to the other side street, around the cul-de-sac and back to the house.

-Dave

[wxo]

I know how satisfying that is to get that behind you. How did you discover it? Mid-Ohio now??

[TireSmoker]

I discovered it by posting a thread on the msextra.com/forums -- in TunerStudio, you can take a 'tooth log' which shows the trigger pattern seen by the CPU in detail. I don't know how to analzye/interpret those logs, but there are guys on the forum who know and help. So I posted a bunch of logs, made the changes requested, took more logs, both for the factory setup, and the new setup.

These are the two threads where I was asking for help:
https://www.msextra.com/forums/viewtopic.php?t=78447 (factory setup)
https://www.msextra.com/forums/viewtopic.php?p=594844 (36-1 setup)

Yes, we are still on for Mid-Ohio!! We were *really* close to postponing, and trying to run at PittRace 3 weeks later. We have until Sept 1 to cancel our AirBNB/Vrbo rental.

[TireSmoker]

Also, Tom came over again last night and we got the timing belt cover put back, zip tied some wiring and took it for the first test drive -- just down my street and the next street over to the cul-de-sac. I'm sure I looked pretty hillbilly to the woman and her daughter near the cul-de-sac with no hood, engine not running great, etc.