R3VLimited Forums

R3VLimited Forums (https://www.r3vlimited.com/board/index.php)
-   Alternative Tuning / W.A.R / Megasquirt / etc (https://www.r3vlimited.com/board/forumdisplay.php?f=185)
-   -   The Definnitive M42 Crank and Cam Angle Info Thread (https://www.r3vlimited.com/board/showthread.php?t=374858)

bmwman91 12-18-2015 01:25 AM

The Definnitive M42 Crank and Cam Angle Info Thread
 
So, there is a bit of info out there regarding tooth angles and whatnot for the M42, but no single cohesive place with all of it. If you read around enough, you will probably come to the conclusion that the first tooth pulse after the gap is 114 degrees BTDC, which is correct as far as the physical setup on the engine goes. Since I am planning to rock a MS3Pro ECU at some point in the next year or two, I wanted to get a clear idea about this and document it once and for all. I borrowed an oscilloscope and popped the shell apart on the 88 pin connector that plugs into the Motronic so I could directly measure the VR sensors (and yes, both are VR type). It's been too long that only the M20 guys have had nice, well defined information to work from, and it is time for the M42 to get some nerd-love!

Which wire is which? This causes some confusion, mainly because the cam and crank sensors have the same wire colors, but one's polarity is reversed because BMW decided to swap the wires (for whatever reason). Keep in mind that this is the E30 M42, and I do not know if this changed in the E36.
- The crank sensor uses pins 67 and 68 of the Motronic. Pin 67 takes the black wire from the crank sensor, and this is the POSITIVE sensor output. Pin 68 takes the yellow wire from the crank sensor, and this connection is GROUNDED inside the Motronic. So, pin 67 is the crank sensor signal, which by convention would be called the positive output of the sensor.
- The cam sensor uses pins 16 and 44 of the Motronic. Pin 16 takes the yellow wire from the cam sensor, and this is the NEGATIVE sensor output. Pin 44 takes the black wire from the cam sensor, and this connection is GROUNDED inside the Motronic. So, pin 16 is the cam sensor signal, which by convention would be called the negative output of the sensor.

I am not sure why BMW reversed the signal wires between the two sensors. For those that are going stand-alone, you have the option of getting your cam signal from pin 44 (the positive output) and then whatever conditioner circuit you are feeding that (and the crank signal) into will behave the same way. With a VR sensor, both wires will essentially provide the same signal, but one is "flipped upside-down" relative to the other. Most VR conditioner circuits are designed for the VR positive output, and will give a digital rising edge when the VR output crosses zero from positive to negative.

The positive/negative convention is shown in the diagram below.
http://www.e30tuner.com/assist/megas...vrpolarity.png

Great, so that is a lot of reading about VR basics. What about the M42?

First, here's a scope shot of two full rotations of the crankshaft, with some annotations added. The amplitude changes a lot as the crank turns because my damper wheel is either out-of-round, or just worn out, so some teeth get closer to the sensor than those on the other side of the wheel. I did some measurements at 2000RPM and 3500RPM as well, and at 3500RPM the big voltage spike peaked at around 50V. That means that at redline, the VR sensor is going to be putting out a signal that peaks at over 100V.
http://www.e30tuner.com/assist/megas...pro/idleVR.png

Here it is zoomed in a bit, with the important teeth marked.
http://www.e30tuner.com/assist/megas...dleVR_zoom.png

And here's a fun little graphic showing what is what and the relationship between moving parts with respect to the cam and crank sensors. Note that the cam lobes are not to scale or anything; they are just for fun to illustrate the general idea.
http://www.e30tuner.com/assist/megas...ranktiming.png


Note:
This is all based on photos of mine from over the years and a lot of online image searching. Thankfully it all seems to agree with existing information, as well as the oscilloscope measurements. If you find an error or think that you see an error, please let me know so that I can see if a correction is needed.



I plan to do a pretty in-depth look at ignition coil dwell times too, since that is also a critical parameter. That will involve more oscilloscope action with the goal of determining the factory dwell times for the coils (both for the M42 coils, and the M54 coils that many of us run with our COP conversion). In a year or two (or whenever I get my ass into gear and install an aftermarket ECU) I will try to document all of the factory equipment (thermistors, idle control valve, etc) so that others will not have to spend weeks and weeks scratching their heads.

5Toes 12-22-2015 09:54 PM

I am planning on building a custom engine harness for my m42 and megasquirt 3x/pro whatever this Spring and will definitely be needing this info! thank you bmwman!

I will keep y'all tuned of course!

varg 12-26-2015 06:26 PM

Very nice! For what it's worth, these are my verified (with timing light) Megasquirt trigger and dwell settings for my M42 with COP conversion. I have not put a scope on the sensors or done anything fancy with the dwell, but this works well enough that I am running 18psi on stock internals and haven't blown it up!

http://i.imgur.com/Dy7YVPK.jpg

bmwman91 12-29-2015 11:09 AM

Awesome! Your dwell time is very close to what I have directly measured (and will make a separate thread for once I have a chance to do the same measurements on an M42 with stock coils and an M52 with its stock coils which are what we all use for the COP conversion).

Basically, the factory dwell is 3.6ms on average. It jumps around a lot, from 3.0 to 4.2ms and it seems totally random. Maybe the Motronic is old and slow or something, or maybe it has to do with the timing advance changing during the dwell period or something. Anyway, 3.6ms of dwell with the M52 coils on our cars makes for an 8 Amp primary current and the spark duration ranges from 2.2ms at idle to 0.9ms at redline.

Any chance that you can post up the settings you use for the idle control (assuming you use the stock ICV)? How about the tach and MPG gauges? From what I understand, the tach takes a 50% duty cycle square wave with 2 pulses per revolution. As far as the MPG meter, I need to get in there with a scope.

varg 12-30-2015 07:08 PM

Quote:

Originally Posted by bmwman91 (Post 4550241)
Basically, the factory dwell is 3.6ms on average. It jumps around a lot, from 3.0 to 4.2ms and it seems totally random. Maybe the Motronic is old and slow or something, or maybe it has to do with the timing advance changing during the dwell period or something. Anyway, 3.6ms of dwell with the M52 coils on our cars makes for an 8 Amp primary current and the spark duration ranges from 2.2ms at idle to 0.9ms at redline.

Interesting. With the voltage correction I probably average 3.6ms too. Stays steady with MS though - it's nice to be able to see exactly what's going out by plugging in your laptop.

Quote:

Originally Posted by bmwman91 (Post 4550241)
Any chance that you can post up the settings you use for the idle control (assuming you use the stock ICV)? How about the tach and MPG gauges? From what I understand, the tach takes a 50% duty cycle square wave with 2 pulses per revolution. As far as the MPG meter, I need to get in there with a scope.

Shortly after I started to get it dialed in, my ICV stopped working correctly - it would just creep wide open and stay there, not responding to changes in duty cycle. I haven't bothered to replace it yet because they're so expensive.

The MPG gauge I haven't figured out, and likely won't ever try - fuel economy doesn't really matter to me, if I were to try and figure it out it would be to make my own boost mpg gauge. I suppose it wouldn't be so hard, I don't have a function generator though, or a stock car to probe with a scope. The tachometer is happy with the 5V square wave output straight from the MS board, no mods required, set to normal speed which is one pulse per revolution.

bmwman91 12-30-2015 08:46 PM

Cool. Thanks for the info!

Strange about the ICV. Maybe it was all gummed up with oil vapors from the PCV system. I had to soak mine in rubbing alcohol, shake it around and that sort of thing a few times and after that it worked nicely.

Did you just plug the ICV hose and crack open the throttle plate or something?

varg 12-31-2015 09:53 AM

Quote:

Originally Posted by bmwman91 (Post 4551097)
Strange about the ICV. Maybe it was all gummed up with oil vapors from the PCV system. I had to soak mine in rubbing alcohol, shake it around and that sort of thing a few times and after that it worked nicely.

I put plenty of brake cleaner through it, if it was dirty, it wasn't after that.

Quote:

Originally Posted by bmwman91 (Post 4551097)
Did you just plug the ICV hose and crack open the throttle plate or something?

Pretty much. I turned the AC on and adjusted the throttle plate screw until it idled around 900rpm. I also used the megasquirt idle advance settings, when the TPS signal is low enough it adds timing when idle speed drops and pulls timing when it rises.

ccsdo5 12-31-2015 02:05 PM

This thread has some awesome information! Props to all your work bmwman91! Varg also has some really good settings for his car. He helped me a lot trying to get my car to run correctly.

I just welded a barb to my IC pipe for the ICV the other day I'm going to play around with closed loop idle settings sometime this weekend and see how happy I can get my car to idle. When I have some sorta decent idle I'll post the settings I used here. Nice seeing the M42 get a little love lmao!

bmwman91 12-31-2015 02:51 PM

I am starting to seriously consider VEMS now. There has been a group-buy going for PnP VEMS setups for 88 pin M/S5x engines, and the guy is now going to do an M42 one as well. While VEMS is much older hardware, I sort of doubt that there is much of a functional difference since the M42 isn't that complicated of an engine. Either way, I guess that this info will be of use though!

ccsdo5 01-01-2016 12:50 PM

Okay so here are the settings I'm currently with the stock M42 ICV. I'm they aren't perfect but my car is idling pretty good at ~900 rpm give or take with my AFR's at just around 13.7-14.5:1 It does tend to dip pretty low when I come off throttle to a stop sometimes but I added a little bit more advance in the ignition map and it seems to help with that.

http://i.imgur.com/wWIIm6gl.jpg

http://i.imgur.com/U0ydyXil.jpg

http://i.imgur.com/JsIhMztl.jpg

http://i.imgur.com/8umOeeIl.jpg

http://i.imgur.com/zj40kL2l.jpg

I'm no master of megasquirt or really know what the hell I'm doing. This is just what is working for me. If you guys have some other suggestions I can definitely give it a try and see what happens. Just trying to help give some more info so anyone that wants to run megasquirt in the future has an easier time with it than me.

blueapplesoda 01-07-2016 08:50 AM

2 Attachment(s)
About the econometer: Econometer funtions on PWM. What I've come to conclude is that BMW likes to take the injector pulsewidth and do some sort of conversion with wheel/vehicle speed. The cluster obviously receives its own VSS input signal and requires injector cylinder 1's pwm signal to calculate and display what the MPG is. Check attached "genericpwma.png". Note this is running on an completely stock M50NV running MS3X, but it shouldn't be any different for an m42 or m20 car.

Now I tried to make it a boost gauge, but the gauge contains the VSS input, which is hardwired into the econogauge. This causes a problem because boost has nothing to do with VSS, and making math in tunerstudio to negate a variable that isn't related is a nightmare. I want to know more about this, because I really want to make the econogauge a boost gauge or a A/F gauge, but I'm not going to go any further into it until I start getting into cluster mods. It's on my list of things to do, but its priority is pretty low.

The pdf is the pinout for the e30 m42. It's where I start for building PnP units.

Derek
DedericMS

varg 01-07-2016 09:42 AM

Quote:

Originally Posted by blueapplesoda (Post 4555118)
Now I tried to make it a boost gauge, but the gauge contains the VSS input, which is hardwired into the econogauge. This causes a problem because boost has nothing to do with VSS, and making math in tunerstudio to negate a variable that isn't related is a nightmare. I want to know more about this, because I really want to make the econogauge a boost gauge or a A/F gauge, but I'm not going to go any further into it until I start getting into cluster

Cut off the VSS input from the gauge and see how it responds to just a fake injector signal, if it doesn't respond well, you'll need to spoof a constant VSS signal. Gonna require some custom boards and pulse generator circuitry, lots of fun.

blueapplesoda 01-07-2016 07:19 PM

Without VSS it doesn't move at all. 0 MPH=0 MPG I'm guessing. May have to snip the motor control and control the gauge direct.

pklym 07-16-2016 03:09 PM

Forgive these O-scope newbie questions, we have access to an o-scope to help with troubleshooting that is documented here: http://www.r3vlimited.com/board/showthread.php?t=384330

We are trying to scope both the crank and cam position sensors using a fluke 125.

Questions:
Should we probe both pos & neg wires, or just positive and use a chassis ground?
What should we set the scales too for volts and time?
Is it DC or AC volts?

We are getting some readings but they don't look anything like the pictures up top.

Thanks!

bmwman91 11-03-2016 09:27 PM

Quote:

Originally Posted by pklym (Post 4664033)
Forgive these O-scope newbie questions, we have access to an o-scope to help with troubleshooting that is documented here: http://www.r3vlimited.com/board/showthread.php?t=384330

We are trying to scope both the crank and cam position sensors using a fluke 125.

Questions:
Should we probe both pos & neg wires, or just positive and use a chassis ground?
What should we set the scales too for volts and time?
Is it DC or AC volts?

We are getting some readings but they don't look anything like the pictures up top.

Thanks!

Crap, sorry I didn't see this! It's a bit late, but it looks like you figured it out with the right sensor.

Q&A:
1) Where should you probe?
Obviously for the crank & cam you want the positive probe tip on the black & yellow wires' terminal ends, respectively. For the ground, the bolt where the ECU grounds might be acceptable, but I used the sensor's second wire (the corresponding yellow & black wires). This helps with noise reduction.

2) Time and voltage division settings?
Both depend on RPM. At idle the 5V/div and 10ms/div should work. At 3000RPM you probably want 10V/div and 2ms or 5ms/div.

3) DC or AC coupling?
I had the inputs set to DC coupling.



What did your readings look like? I am guessing that they were super noisy or something. I did some pretty heavy post-processing on the scope captures because there was a lot of noise to filter out to get the proper waveform in a clean format. The ECU does its own filtering and stuff, but if you are probing from the outside you'll probably need to do some of your own.


All times are GMT -7. The time now is 03:57 PM.

Powered by vBulletin® Version 3.8.10
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Copyright 2004 - 2020 R3VLimited LLC