MPGuino release one workspace

[cmags]

There shouldn't be an issue with powering the backlight off a separate power source, but make sure it's a regulated 5V. You can pull the 12V from the Accessory circuit (ACC) in your car which switches off with the car (and during an engine crank). Easiset place to pull ACC is behind the radio, but there are many other things in the car that run on ACC.

Edit: Sorry, forgot you had this powered via signal + transistor (don't have a backlit LCD yet). But yea, that would be a software fix. The above still applies if you don't want to use the software option for brightness control though.

[dcb]

This is a software issue. I think I have auto backlight timeout figured out, will update post 1 tonight hopefully.

[top_down]

First trial run in the 93 Metro XFi. I used the "maybe" values listed in your code for the 98 Metro 3cyl. Two things going on that I don't understand:

1. The MI numbers are clicking off in a controlled fashion, but at about 2x what my odometer says. Another forum member had pm'd me that he thought the VSS value should be closer to 4000. Has anyone verfied the 8204 number?

2. The MH figures are constantly increasing when I'm moving, and slowly decreasing when I'm stopped. I assume when they get to zero they'll stop decreasing. Any idea what this is about?

MPG figures are crazy (in the thousands), probably mostly due to item 2 above.

I double checked my VSS input. I am tapped into terminal B15 on the ECU (yellow with green stripe). My injector lead is tapped into the yellow wire right before it enters the connector for the injector resistor.

Any thoughts would be appreciated.

[dcb]

1. Double the vsspulses per mile with this version in the metro. 16408 should be spot on. The metro has a strange vss signal as far as I can tell.

2. It doesn't sound like the right spot, here is a metro fuel injector hookup pic:


and a vss hookup pic in case anybody else is interested.

[drivenByNothing]

Giving the guino a home. For some reason, I'm having trouble with v.64. It's giving me scrambled eggs on the LCD. I'm making a 2hr highway trip today, so I wanted a decent way to keep everything together and visible.

This is really just a rough attempt, but I'll probably buy another enclosure from the shack (they're only ~$2.60) and do it up nice once I get my power circuit built.

Edit: v.64 seems to be working ok now.

[awillard69]

What size enclosure did you get? What board setup are you using for the 'duino? Iduino, Freeduino, other?

[drivenByNothing]

Originally Posted by awillard69

What size enclosure did you get? What board setup are you using for the 'duino? Iduino, Freeduino, other?

It's the petite iduino. The enclosure is 4x2x1. Everything should fit nicely, but it'll take some more time and attention to detail with the dremel.

[jmassengill]

dcb, I just looked at the code for software version 0.66 and it looks great! Knowing a little c++ from college, looking at your code brought back memories. I might have to go get out a few old text books.

I'm in the stage of building the MPGuino for an old 1992 Geo Metro that I've converted to 13 inch tires. Being as I'm on vacation next week, I plan on building it then.

Let me say THANK YOU dcb. Your hard work has made this happen.

jmassengill

[dcb]

Thanks JM, Nice Unit DBN!

Ok, lets try and regroup again.

re: hardware failures, andrewJ has not had a chance to investigate. Have not heard back from Top Down or Cmags. I do not know of any other unexplained hardware difficulties at the moment. Am I missing any? We've had a few successful installations in the field that I can tell as well.

re: strange speed readings on a metro, Top Down and myself have noticed this, anyone else? From my observations it seems to happen right as you are starting out, so for the first couple blocks my mph reading is 1/2 what it should be (v0.66 16408 vssppm). Seems to imply the pin is floating until something attains a bit of a charge, we might need to rethink the value of the vss resistor or??.

I should have a portable scope on monday to get an idea what kind of noise is getting through the power supply and possibly what is happening on the vss signal and a peek at the injector pin for good measure.

Are we ready to start with the database of by vehicle injector/vss/power hookup pictures, and initial parameters to use? Can we use the wiki to organize that? Is anyone willing to own that?

[cmags]

Sorry, haven't had any time to work on this project - my wife had to have emergency surgery and I've been in the hospital w/ her all week. Don't worry, all is OK now but it certainly threw the priority level of this project down the tubes.

I've got v0.64 coded up such that everything on the LCD is duplicated to Serial, and with the addition of the LCD reset it should be a good place for me to test w/o having to disconnect power when I get back to it. I read through the code and only saw the two-button reset - didn't see any automatic checking for the LCD needing re-initialization, though maybe I read the update wrong.

@ DBN, check the LCD thread to see if the "scrambled eggs" you were getting are the same as the garbage I get on my LCD. If we're both seeing the same issue, we can try to combine notes.

I should have the chance to get back on this project this week when we go home from the hospital. I'll have vacation time and will spend half of it on taking care of my wife and the other on projects like this. Of course I'm already two versions behind, but I'll catchup.

Oh yes, I've got the info for the database for the '05 Subaru Impreza 2.5L and can find out how many other Subarus it applies to. I'll post pics of the schematics and the install when I do it.

[dcb]

Glad the Mrs. is doing ok. It soulds like your display issues are an order of operations thing and not a hardware glitch per-se. Like you might need to bounce the power to the whole unit after you program it. Shouldn't affect in-car operation as long as you adjust your setup so the LCD is powered and connected to the duino before or at the same time the duino is powered up.

So about the metro vss thing, Keep me honest here:

I tried an experiment where I do a bunch of analogReads on the vss pin while at the end of the loop and keep track of the min and max reading each half second.

I was getting min=0 and max=~630 which translates into 0 and 3.08 volts respectively, a solid LOW and HIGH in atmega land if I am not mistaken. So this leads me to believe that we have a 9 volt drop on the 100k resistor and are running 90 microamps through the pin. I don't know if it cares about current or voltage or what, only saw voltage where below .2 is LOW and above .7 is HIGH on an input pin on the datacheet. I also saw that they reccomend max of 50k for a pullup resistor. But our 100k is attached to 12-0 volts, not 5, so there ya go.

I have not seen the behavior on the saturn, but I believe the saturn uses a hall efect->ecu->speedo to make a nice signal the speedo (and we) can use. I believe the metro actually uses a physical switch in the speedometer head. So it may be addressable by adding some debouncing to the vss signal.

To confound (or possibly explain) this, when I ran my "do a bunch of analog reads" trick, I never observed the half speed displayed issue. I am wondering if the analog reads add just enough capacitance to the input signal to clean it up or what?

[awillard69]

Nice work, DCB. I like the variable setup screens, that should save people from having to adjust the variables before compiling. Now, it should just be copy, paste, upload and go!

I need to stuff my circuit monster into an enclosure and install into one of my vehicles, most likely the Jeep. I may need to rework my buttons to make better use of any enclosure, but that should be minor.

[cmags]

Originally Posted by dcb

Glad the Mrs. is doing ok. It soulds like your display issues are an order of operations thing and not a hardware glitch per-se. Like you might need to bounce the power to the whole unit after you program it. Shouldn't affect in-car operation as long as you adjust your setup so the LCD is powered and connected to the duino before or at the same time the duino is powered up.

Moved reply to LCD thread: 2x16 LCD Display issues/discussion

[dcb]

Thx Awillard

[top_down]

Good news, I reprogrammed the MPGuino this morning and my LCD started working again! So if your LCD goes black, try a fresh upload before you start desoldering it!

So now I have my unit with the latest v0.66 and am working through the setup screens. Can anyone remind me what the injector microsecond/gallon number is supposed to be on the Metro? I know it was there in the setup/comments area on v0.64 but I didn't get the number written down anywhere. The default that is loaded in when I upload the program is 0304409714 but I'm remembering the Metro was supposed to be something different.

I'm going to use your new suggestion for VSS pulses per mile: 16408. My tires are 155 80 R12 so that should be within a percent or two. If my math is right the difference between 145 80 R12 and 155 80 R12 should be about 320, but I'll tweak and report on that once I get reasonably close odometer correlation.

Triple checked my VSS connection and am sure that is correct. My injector connection was next to the injector resistor on the yellow wire. I will move it as you suggested to the yellow with stripe wire directly adjacent to the injector - but from what I see in the wiring diagram this simply puts the tap on the other side of the injector solenoid. Would that make a difference?

I wish I could help you more with the Metro speed issue, but I don't have a scope or the knowledge to tell you what's going on there even if I did. I am happy to try any changes you might suggest and report my findings.

"...it may be addressable by adding some debouncing to the vss signal." - Let me know what you think is appropriate and I'll give it a try. Please be specific. I'm very good at soldering but am not a EE!


dcb, as always, thanks for the help and the fine work!!!

[dcb]

That's good news!

metro, MicroSec/Gallon = 684968626 maybe
Pulses/2 revs should be 3 (did get an aplha RPM gauge working with that)

[SVOboy]

I feel like I'm missing something fundamental. When I go to upload the .66 code from sourceforge I just get errors, such as "error: redefition of 'byte brightness []'"

Thoughts?

[dcb]

Did you delete everything before pasting in the new code? I know it sounds like a stupid question but I have to ask.

[top_down]

I had the exact same error the first time I tried to upload v0.64. Reset the arduino and keep trying. I can't say I figured out exactly what solved it, but after the fourth try it worked. Make sure you have the right kind of serial cable (if you are using serial). I plugged my arduino right into the serial port to eliminate that possibility.

My unit is back in the car and reporting some reasonable numbers (although I think maybe too optimistic!). I will report back in a week or two after I manage to burn through some gas - that'll take a while as I am not that inclined to drive the car much when it's 111F outside.

I will say that the display is hard to read in bright sunlight and almost impossible to read when it is hot. I have it perched on top of the steering column in the full Arizona sun. I had hoped to mount it up high but may have to rethink that.

[Wonderboy]

Hello all! I've been building the MPGuino, and I think I've succeeded in everything so far except the injector leads...

I have a 1993 Corolla 1.8l (7A-FE). Just as a guide for anyone with a similar situation, I had a hard time finding my ECU, and it ended up being under and behind the console, best accessible by taking out the glove box and looking left and down.

Anyway, I've got everything almost ready but when I take the multimeter to the injector when the car is on (not running of course), BOTH of them read 12v. Which one should I choose??? Can anyone help me with this?

Thanks!