t vago's MPGuino workspace thread

[Ardent]

No faults this morning with either the 0.86 or 1.86 code.

I tried to reproduce the scenario seen the last two mornings by setting a short sleep timeout and then starting the engine (repeatedly).

The only behavioral difference noted was that from Key-OFF to Key-ON the back-light would illuminate with the 0.86 code, but not 1.86.

[t vago]

Originally Posted by Ardent

The only behavioral difference noted was that from Key-OFF to Key-ON the back-light would illuminate with the 0.86 code, but not 1.86.

It's enough. If I understand you correctly, you're saying that with DCB's code, your backlight came on, and with my code, it did not. That means that DCB's code is responding to your engine starting, while mine is not. That sounds repeatable, which is progress.

If you don't mind, I'd like to experiment on you... heh, on your setup, anyway. I am going to comment out the bits in the 1.86 code, that relate to the two rationality checks, then send the modified code to you for testing. Depending on your feedback, I am then going to progressively un-comment each bit, and then send the code back to you for further testing. It should not take much more than 4 or 5 tries.

You up for that?

[Ardent]

Originally Posted by t vago

It's enough. If I understand you correctly,. . . That means that DCB's code is responding to your engine starting, while mine is not.

Not quite. This was with the key ON; before advancing to the START position. Both were responding this morning when the engine started.

Edit: I'm guessing that there was a small INJ or VSS pulse sent when the ECU powered up.

Originally Posted by t vago

If you don't mind, I'd like to experiment on you... heh, on your setup, anyway. I am going to comment out the bits in the 1.86 code, that relate to the two rationality checks, then send the modified code to you for testing. Depending on your feedback, I am then going to progressively un-comment each bit, and then send the code back to you for further testing. It should not take much more than 4 or 5 tries.

You up for that?

I'm up for testing.

[t vago]

*headdesk*

Okay, if your setup uses a fuel injector relay to control battery power to your fuel injectors, then that behavior is normal. The code is detecting the pulse caused by applying power to your injector circuit, but the rationality checks are rejecting the pulse.

I think instead that I will just revert the injector interrupt code to use both INT0 and INT1 again, like the original 0.86 code did. It might be that one of your AtMega's pins is not as good as detecting signal changes as the other. We'll test that, instead.

[Ardent]

My testing opportunities are of short duration this week, but I'll be putting on around 350 miles on Saturday that should cover the gamut of conditions and allow for a fairly thorough test.

[josemapiro]

Originally Posted by t vago

Okay... Those are simple buffer circuits. They would certainly work, but my opinion is, is that it's over-engineering the circuit.

One would get the same benefit, merely by changing out the 100k resistors for 47k resistors (or even 22k resistors).

Ok, I thought it was better with the use of transistors to protect the board.

I need help configuring MPGuino, what is the function of each of these lines and where do I find these values​​.

Contrast
Metric 1-Yes
InjTrg 0-Dn 1-Up
Tank (L)*1000
Inj Delay (uS)
Revs/Inj Pulse
Min Good RPM
VSS Pulses/km
VSS Delay (ms)
Mass} (kg)
FrArea*1000 m^2
C(d) * 1000
C(rr) * 1000
Microsec/L
Timeout (s)
Scratchpad(odo?)

Thank you
José Rodrigues

[t vago]

Originally Posted by josemapiro

I need help configuring MPGuino, what is the function of each of these lines and where do I find these values??.

Look for this line of code:

const uint32_t params[] PROGMEM = {

It will go to where these parameters are stored in the code. As for what they mean,

• Contrast - LCD contrast - lower numbers show sharper contrast, while larger numbers will show fading contrast. I do not recommend going above 100.
• Metric 1-Yes - Switches between Metric mode (1) and US mode (0)
• InjTrg 0-Dn 1-Up - Tells MPGuino whether to count injector opening on a downward (0, for 1->0) pulse or an upward (1, for 0->1) pulse. Most vehicles use downward pulses
• Tank (L)*1000 - Tank size in liters, multiplied by 1000.
• Inj Delay (uS) - leave at 550 - this accounts for the delay between when a pulse is applied to a fuel injector by the engine computer, and when the injector responds.
• Revs/Inj Pulse - How many times the crankshaft rotates, from when the injector is opened, to when the injector is next opened.
• Min Good RPM - The engine speed, above which MPGuino will process fuel injector openings. If engine speed falls below this value (set at 100 RPM), then MPGuino will ignore fuel injector pulses.
• VSS Pulses/km - The number of VSS pulses per kilometer traveled.
• VSS Delay (ms) - The amount of time that MPGuino waits, after receiving a VSS pulse, before MPGuino will count it. This is mainly for VSS signal conditioning.
• Microsec/L - This number is how many microseconds it would take to push one liter of fuel through the fuel injector. More details may be found here, although it deals only with microseconds/gallon.
• Timeout (s) - This is the length of time, in seconds, after which the MPGuino will enter an inactivity mode, if the MPGuino does not receive an injector signal, a VSS signal, or a keypress.
• Scratchpad(odo?) - This was left over from the original dcb code. It currently isn't being used for anything. You may store any number between 0 and 4294967295 in it.

All of this will eventually be used in the coastdown test code - you can ignore it for now

• Mass} (kg) - Mass of vehicle, and everything in it, in kilograms
• FrArea*1000 m^2 - Frontal area of vehicle, in square meters
• C(d) * 1000 - Drag coefficient of vehicle - this will be calculated and stored by the coastdown test code
• C(rr) * 1000 - Rolling resistance coefficient of vehicle - this will be calculated and stored by the coastdown test code

[josemapiro]

Ok thanks.
It was even which meant, some already knew.
Not have to define the number of injectors car?
If yes, where it is defined, in the setup menu does not have that option.

Thank you
José Rodrigues

[t vago]

Originally Posted by josemapiro

Not have to define the number of injectors car?
If yes, where it is defined, in the setup menu does not have that option.

It looks like that option has not been enabled in your copy of the code. You will need to go to this line, and uncomment the beginning of it.

//#define useCalculatedFuelFactor true	// Ability to calculate that pesky us/gal (or L) factor from easily available published fuel injector data

You may need to also do a minor code edit, because I think the most recent version in this thread has an incorrect label, somewhere near line 2290.

It should be this:

#ifdef useCalculatedFuelFactor
	idxNumerPressure,
	idxNumerPressure,
#endif

Not this:

#ifdef useIsqrt
	idxNumerPressure,
	idxNumerPressure,
#endif

[josemapiro]

Originally Posted by t vago

It looks like that option has not been enabled in your copy of the code. You will need to go to this line, and uncomment the beginning of it.

//#define useCalculatedFuelFactor true	// Ability to calculate that pesky us/gal (or L) factor from easily available published fuel injector data

You may need to also do a minor code edit, because I think the most recent version in this thread has an incorrect label, somewhere near line 2290.

It should be this:

#ifdef useCalculatedFuelFactor
	idxNumerPressure,
	idxNumerPressure,
#endif

Not this:

#ifdef useIsqrt
	idxNumerPressure,
	idxNumerPressure,
#endif

I've tried but it gives the following error:
mpguino_1_86_tav_c: 5607: error: 'idxCorrFactor' was not declared in this scope
mpguino_1_86_tav_c: 5630: error: 'idxCorrFactor' was not declared in this scope

Thank you
José Rodrigues