NA6 CAS -> V2.2 Megasquirt.

[ESDave]

Hi Guys,

My first post so please be gentle.

Project for the summer is to get a V2.2 mega squirt (+ spark 029y4) (its left over from an old old project, essentially a freebie) to run my NA6 MX5.

I like to work in baby baby steps, so firstly i am just using the MS to sample the OEM EMS signals while its running the show, then take over driving the injectors, then the coils etc...

First step is to get the MS to trigger off the CAS, and get an RPM reading, but I am running into some troubles.

Photo of setup: (home brew 1:1 adaptor loom for the factory EMS, at the moment only connections to the MS are power / ground / CMP / CKP)



I've basically been following this guide here:

http://www.diyautotune.com/tech_articles/how_to_megasquirt_your_mazda_miata.htm

The main problem i have is that the MS reads what looks like exactly half the RPM it should be.

I'm using the settings suggested from the above link, to be honest i don't understand what they represent, I started to RTFM but i still don't understand. I don't actually know what the NA6 CAS slots are in number (i think i figured it out from the CRO traces)



At idle (approx 850rpm) Using a dorky single channel USB pen oscilloscope (it has a pretty low sampling rate, if I'm still stuck by Monday i might try and borrow a proper DSO) i get the following pulse train:

V2.2 MS Micro pin 14 / DB37 pin 24 / Main tach input / CKP / White

~1660 pulses / min - if the CAS spins at half crank speed then id say that wheel has 4 slots in it. (But the MS assumes a crank mounted sensor? So the parameter is two base teeth?)



V2.2 MS Micro pin 11 / DB37 pin 25 / Secondary tach input / Top dead centre reference? / CMP / Yellow

I'm guessing the fat pulse is one pair of cylinders TDC reference and the skinny pulse is the other, can the MS distinguish between the two and fire the right coil pack?

The positive edges of these pulses come in at ~ 869 edges / min same as RPM.



Anyway long story short, just to get the RPM correct are my settings for the generic wheel incorrect? If somone could help me understand (or point me to a good explanation with some pictures / examples) what each parameter represents on the CAS wheel?

[ESDave]

Whoops! Looks like it helps when the number of cylinders value is actually set to the number of cylinders in the car

RPM is correct now.

[saboteur]

That will do it!

You're right about the documentation being fairly sparse. This is about as good as it gets for the 2nd trigger configuration:
http://www.msextra.com/doc/ms1extra/MS_ ... m#2senscam

Anything relating to the 4G63 can be applied to the NA CAS, as it works in the same way (if you look at the MX-5 CAS, it is actually a Mitsubishi part).

Good luck and keep us posted with your progress!

[ESDave]

Thanks!

Ive also had a few problems with the MS inputs, interfering / somehow loading the CMP and CKP lines and stopping the car from running.

I'm not really sure why, with one configuration the car wouldn't start with the MS attached but once the car was running i could connect the MS and read rpm.

Perhaps the OEM ecu checks the CMP / CKP lines with its internal pullups off and if it reads a high voltage it assumes there is a short to a high source?

Or maybe my caps on my low pass filter where loading the OEM ecu pullups to much when they where first turned on and it assumed a short to ground?

Anyway i found the following configuration works ok, i can actually start the car with the MS attached with this:



Id love to get my hands on a schematic for a MSPNP or the OEM ECU, are these available anywhere?

[saboteur]

I've only ever used the MegaSquirt standalone, but I know Minh here on the forums is running his in a similar way to yourself. I'm pretty sure he did the fuel and ignition in one step though, so the problem may have not shown up on his setup.

I opened up my MSPNP to do the 0.1uF cap mod to the CAS signal and had a bit of a look at the daughterboard. It didn't look overly complex.

[ESDave]

Almost finished setting up all the sensors, just having a play checking out the OEM ECU in action. (Had to test the rev limiter )



I think my temps are a little on the high side, i used easytherm and these parameters:

90-93: Use Rx7 defaults for the CLT and GM defaults for the AIT.

Piggyback without R7 or R4 (recommended for piggyback sharing the sensors): Use 2625 for CLT and 3475 for AIT for the bias.

Taken from this pretty helpfull thread here:

http://www.miataturbo.net/forum/t13676/

But im not sure if his suggestion for AIT temp are for the DIYAutotune bung sensor or the OEM sensor in the AFM. The manual i have has tables for the CLT and AFM temp sensors and shows them to be almost the same temp / resistance characteristics. But maybe they have different pullup resistors in the OEM ECU?

The temp gauge on my MX5 has always sat at about 3/5ths, on the cold side of half way. Ill use a thermocouple to check the AFM temp sensor later.

Just out of interest I'm sampling the AFM output with a spare A2D input (unscaled 8bit ADC counts on the graph), it starts about 2.5V and gets lower the more the vane moves. At 7000rpm its almost 0V which makes sense i guess, no need for them to scale it beyond what the engine can gulp at the factory rev cut.