I’ve updated the build thread that I had at GRM, but I haven’t copied everything over here. So here’s a mega-update from there:
Okay, so Megasquirt is hard when you’re using a 16 year old PCB…
I was hoping to add in a knock input and outputs for tach, PWM IAC, and fan control, plus an input for launch control/flat shifting. It looks like I get 5 pins I can use, X11, X12, X13, and X14 plus pin 36 on the DB37. 11 and 12 are already in use for ignition control, so I’m going to put knock on X13 and tach on X14. Pin 36 is orphaned, so if I want to use that I just need to wire directly to the pin. Pins 1-19 are all wired directly to ground.
Not ideal, but I think I can make it work. I did swap out Q5 last night for the PWM-capable transistor. Knock input looks pretty simple – if I connect a jumper from X13 to X6 on the 2.2 PCB, that should put knock input on what MS2 calls JS4.
The tach circuit still perplexes me due to a lack of any specific direction. I have a bunch of the 2N2222A transistors (cheaper to buy 20 than to buy 1…) so I believe that I can use JS5 by wiring to X7 on the 2.2 PCB. That should mean that the right lead of the transistor should go to ground, the center lead wires to a 1k resistor and then to X7, and the left lead wires to a 10k resistor and a jumper wire. The 10k resistor goes to a 12V source and the jumper wire goes to X14. I don’t really understand that, but let’s see if it works.
That will mean that on my harness, the 2A wire (green/white) should be knock input from KnockSense and the 2B wire (green/red) should be tach out.
So I did all that wiring stuff, but it doesn’t seem to work. I suspect that I’ve failed in picking up the +12V and the ground properly, but anything is possible.
I wired up the IAC and the tach connection to the megasquirt harness so I could wrap up the wiring. It’s looking a lot cleaner than before and it should be less prone to any kind of breakage or rubbing:
The throttle cable bracket is installed and the vacuum lines are all hooked up as well.
Next, I need to figure out what went wrong with Megasquirt and then see how my mods work. The knock sensor is still not hooked up – I need to run it back into the cabin of the car because the knock sensor controller isn’t weathertight. I’m thinking that I’ll use the wire that is for the knock input pin to run from the sensor into the cabin, then snip the wire near the MS box and use the part wire that goes back to the DB37 as the knocksense output connection. Then I should just need to find switched +12V and a ground inside.
Well, there was a minor setback tonight. When I wired up the new PWM transistor in place of the old FIDLE transistor on the 2.2 board, I think I literally had my wires crossed. Uncrossing them got it working a bit, but in an attempt to troubleshoot, I cut out the tach output circuit and tried reloading the firmware. On the stimulator it acts fine but it always wants to keep the fuel pump relay turned on, even when it’s reading 0 RPM. I’ve ordered up a new Q3 transistor for the fuel pump circuit so I’ll see if that clears the problem up.
If that works, then I’ll go ahead and rebuild the tach output circuit. I’ll also need to find some place under the hood to pick up +12V when the ignition is on, but that shouldn’t be too tough. Initially, I wired the IAC to the fuel pump relay control, but that’s all ground not +12V so I undid that. I’m not doing anything else, though, until I know if I’ve fried the MS box entirely. If I can’t get this squared away in short order, I’ll just have to order up a new kit. And then I’ll have to see how much I want to gamble — do I hope that my MSII daughterboard is okay and just buy the MS1 3.0 kit and use my current proc, or do I spend the extra 60 or 70 bucks and get the MS2 3.0 kit?
And here things were running so well before I just had to make them better….
All that stuff I posted before – just ignore it.
The real problem is that I blew a fuse that feeds MegaSquirt and it was getting backfed from a sensor somewhere so it was acting like it had power but wasn’t working properly. It’s all good now.
Tonight, I went ahead and re-did all the stuff I undid last night. I gave it a test fire and it started right up! Sort of. It won’t stay running because the idle valve is closed and I need to find +12V for that. That’s a job for tomorrow night. For now, it starts! And the tach works! So I will celebrate with hard root beer and Buffalo Creek Bourbon Cream!
Fans are wired up and installed on the radiator with zip ties. I put a new hole in the core support so I could position the radiator forward a bit. They actually seem to blow a ton of air, so I’m cautiously optimistic.
Megasquirt is working okay. The car will start and idle but the IAC is not working so in order to start, the throttle has to be slightly pressed. The tach gets a signal but it is really jumpy. The IAC makes clicking sounds when I hook it straight to 12v, but MS isn’t sending any signal down the FIDLE wire. I did have the TIP 120 transistor miswired at one point so I’m going to try replacing that.
And I’ve cooked yet another TIP120. With the output connected to absolutely nothing and the FIDLE setting set to “Off”. Something is wrong on the input side of the equation.
The next attempt was to remove the TIP120 completely. No smoke, but the dash tach was very jumpy.
Next test was to remove the jumper wire that sent the +12V to the CPU stepper chip. The instructions don’t say that you NEED that wire if you’re not using a stepper, just that you “should”. With that wire removed, the tach isn’t perfect but it’s pretty close.
The coolant is pretty low from pulling the intake manifold and all that, so I need to top that off, but otherwise, I think it’s ready for a test drive!
This whole idle valve thing has got me pretty bummed out. Same with the tach out. It’s gone back to being super jumpy again, so basically worthless, and the IAC is completely non-functional. I suspect that there is something that isn’t wired correctly, but of course the assembly manual for the 2.2 PCB with MS2Extra is no longer in existence, having been purged from the msextra.com site. Sure, they’ve got a link that says it leads to the manual archive with the older 2.2 manuals, but it’s a dirty lie. I get that it’s an old board, but why eliminate all traces of the old manual?
Anyway, the car starts and runs without idle air, but it does require holding the throttle down a little for a few seconds before the idle settles down to around 600 RPM. I’ll take a drive tomorrow and see what it’s like with the new intake.
Oh, and I found this sitting next to the car:
Stock ECU is out of the car! There’s plenty of wiring that needs to be removed and cleaned up as a result, but at least I know that full standalone is functional now.
Holy E36 M3.
It’s raining today and the car is basically undrivable. The slightest blip of the throttle in gears 1-3 breaks the tires free and gets the car bouncing off the rev limiter. Boost comes on much faster and holds at 15-16psi. Cruising in fifth at 65mph rockets to 80+ with no hesitation when I push the accelerator about half-way. This thing is fast.
Cooling seems okay, but it’s hard to tell on a cool and rainy day. Sitting in the garage, idling after running for 30 minutes, I put the fans on and it did drop the temps pretty steadily, so I’m hopeful.