[Meridock]

I've seen the initial g-meter mock up - pretty snazzy.

[sti robot]

How much is "finished"? haha

How about finished is accelerometer + timer + performance + a sensor of some type working (probably 2x temp sensors...fullerton?)?

Then I'll call it v1.0?

[Mack]

Yeah temp!

[sti robot]

I should be getting a bunch of sample's from these people: http://www.tekoenclosures.com/ If you go there and see anything you like please tell me.

[sti robot]

I've also been looking at ecu pinouts, because you know, I want to fry an ecu lol.

Seriously though, when I had a greddy informeter it could give me boost values from the obd II which means I might be able to get them directly from the ecu. Remember that I'm not doing OBDII because a) its been done to infinity and back and b) it requires a lot of time/effort/money (on my part at least).

The three pin-outs I'm interested in are IAT sensor signal, Wategate control solenoid valve, and vehicle speed. Using each one would be one less sensor to install, calibrate, etc. However, I'm not sure how comfortable people are with splicing into their ECU harness.
-IAT - this plus intercooler temp (wired up sensor) would solve fullerton's needs, I think - according to the manual this is a perfect 0.3-4.6 V scale, which is cake to hook to the board.
-wastegate solenoid valve - boost - its a fancy term for boost - according to the 04 sti manual it sends 0 or 13-14 volts signal...I'm guessing its a pulse you have to interpret.
-vehicle speed - it says in the manual that this repeated bounces back and forth between 0 and 5 V - this could only be a check signal - or it could be another PWM signal

Question: does anybody have anything they installed that hooks into the ecu harness wires? This is not counting piggyback ECU's and the like.

[Mack]

i'd rather stay away from my ecu.

[sti robot]

For those interested in a gear selector...

this one uses an accelerometer as tilt sensor: http://www.indy-cator.com/download/anleitung.en.pdf
The accuracy of that was probably a PITA to write into code and might be off on big inclines...maybe not though...I've got some ideas

How about using a 7 segment display...this sort of number thing: http://www.clanbbw.com/publicimg/dat...Gv3CLOCK01.JPG

and installing a little sensor to the shaft of the shifter and then powering it...would that work?

It would work like this: The sensor learns what the tilt is at zero and what the difference (x,y,z g forces) is from zero at each gear. Then as you are driving it constantly resets zero (because you could be on an incline and that shouldn't change the display. It would only change when it gets a jolt of tilt/acceleration (acceleration changes with tilt due to the accelerometer reading 1g for gravity and as the angle changes it reads a different level of this 1g) (a certain x, y, z threshold) aka a gear shift. This part is the easy part. The hard part is when you consider that the car is moving through a real life environment. For instance the car will accelerate and brake hard, as well as drive over rough terrain. These are jolts in acceleration/tilt. You could start by placing priority on the z axis (up and down) for most of the gear change sensing, but what if you shift and accelerate up an incline or while cresting a hill? How do you keep these from registering as gear shifts?

-Shifting on level ground with not forward or backward accel is measured by z + y (left right tilt) or z + x (front back tilt)
-Acceleration/braking of the car effects the x component
-Turning effects the y component.

I'm thinking a couple of pots installed in the shift assembly might be easier to do. Like this: http://computer.howstuffworks.com/joystick3.htm It would require more work to install but would be deadly accurate and probably not require recalibration all that often. You'd still have to go through a minute long "learning" thing the first time it was installed to calibrate it.

[sti robot]

I did another autox with the meter in....and the bad news is it did that thing where it stopped working again. Nothing broke...it just stopped working and had to be reflashed. I guess I'll have to get to the bottom of this.

[sti robot]

Ok. I'm think it might be one of 2 causes.

1) heat. The AtiMega168 chip is rate to 185F so it may have reached that. Only temp sensors will tell for sure

2) a short- my wiring in this prototype is messy. Maybe all the jostling from an autox shorted something out on the chip.

[sti robot]

Question for you guys:

How high do you think underhood temps get? Intake Air Temp? Intercooler piping temps?