Any interest in developing an open source fuel economy gauge/computer?

[s2man]

Originally Posted by dcb

F We don't need to reinvent a protocol to talk to the palm...

That was one of my questions, but I decided to wait until we had data before discussing how we should send it. I'll take a look at the elmscan AT commands.

I've registered at SF.

I've ordered an Arduino and a prototype board.

[dcb]

Cool, s2 I added you, metro I couldn't find you so PM me your case sensative unix name.

re Elm protocol, I'm thinking that shouldn't be part of V1 after reviewing the doc. It would be a nice to have, but it will be WAY more than we need for the first run of just injector on time and off time and vss tics. But we can take the nitty gritty stuff over to sourceforge development forum. http://sourceforge.net/forum/forum.php?forum_id=801347

Oh, and this is cute:

[diesel_john]

Originally Posted by dcb

That's the rub there, hopefully the individual will be a variable as (s)he gets better technique. A fair number of details to sort out also, but we should keep it on the table and if anyone runs with it then cool! Also it could be deduced with the MAP/RPM/Air intake temp sensors and a guess at the VE and displacement of the engine, though might be even more to figure out down that path

Having gotten a closer look at all these cheap and tiny computers, I'm starting to think the "fill a little cup, let it empty into the carburetors, and fill it again, and count how many times you've filled it" approach is really viable, accurate at low flows (just not instantaneous), and possibly cheap, especially if you repurpose (drill out) an injector from the boneyard or other common fuel valve/solenoid.

[QUOTE=dcb;15801]That's the rub there, hopefully the individual will be a variable as (s)he gets better technique.

That's why it is so important to be able to see the actually cal. number so it can be tweaked after each tank full without going through the whole cal. process. just tweak the cal number and watch the calculated MPG change until it matches the odometer divided by the pump gals. MPG.

MAF has air temp inside. A MAF is actually has an electronic circuit inside. Output is pounds of air.

"fill a little cup, let it empty into the carburetors, and fill it again, and count how many times you've filled it"

this the way the rain gages work.
It would be possible to have a fixed jet at constant pressure fill up a float bowl with a micro switch on the float and the bowl outlet goes to the regular fuel pump or gravity to the bike carb. (light goes on above my head) Count the total time the float value is filling the bowl.

[dcb]

Originally Posted by diesel_john

It would be easy to have an injector fill up a float bowl with a micro switch on the float and the bowl outlet goes to the regular fuel pump or gravity to the bike carb.

I like how far your drip counter idea has come

I think this will work, for my bike , the microcontroller just turns the pump on for a specific amount of time when the resivoir looks low, and keeps count. The pump doesn't run so long that it would overflow the resivoir, might even want a restriction on the pump outlet. I'm ready to go get a cheap fuel pump.




Edit: some quick figuring,

Lets say 1/64 of a cup, or 1/1024 of a gallon resivoir, 3.7 cubic centimeters, basically a length of clear tubing

@ 60 mpg it would refill every 309 feet

for a 20 mile drive it would refill 341 times, which would be 0.3 parts in 100 accuracy

still need to bench test the rig to see what the lowest amount of time is to run the pump reliably, and still support peak flow rates.

[Coyote X]

more code samples for everyone to play with

http://www.iwsti.com/forums/interior...more-wiki.html

[dcb]

Got the serial dunio today, slapped it together and ran a few programs on it, nice find there lostcause, best $17 I've spent in a while

Here's where I'm tracking it, a couple tips for the serial duino there:
http://planetchampions.org/diympggauge/freeduino.html

[MetroMPG]

Mine also came today - I got USB though. Which means I suppose it won't work with the Palm. Crud. I ordered mine before that particular direction was decided upon.

Anyway, weekend project, since mine is a baggie full of bits 'n' pieces.

[diesel_john]

Originally Posted by dcb

I like how far your drip counter idea has come

I think this will work, for my bike , the microcontroller just turns the pump on for a specific amount of time when the resivoir looks low, and keeps count. The pump doesn't run so long that it would overflow the resivoir, might even want a restriction on the pump outlet. I'm ready to go get a cheap fuel pump.




Edit: some quick figuring,

Lets say 1/64 of a cup, or 1/1024 of a gallon resivoir, 3.7 cubic centimeters, basically a length of clear tubing

@ 60 mpg it would refill every 309 feet

for a 20 mile drive it would refill 341 times, which would be 0.3 parts in 100 accuracy

still need to bench test the rig to see what the lowest amount of time is to run the pump reliably, and still support peak flow rates.

Do you want the frequency of the pump at full flow or the frequency at full pressure? or both.
I might have some IV bags in the attic, do you have ape hangers on your cycle?

[dcb]

Metro, don't sweat the palm, as long as you can get a program on your board you will be set for the standalone gauge.

I'd just start sorting out the standalone bits as there is plenty to do there:

injector circuit and processing (assuming s2man is on this as coyote needs to get his ride back together)

lcd display utilities (lostcause?)

Functions for cool melodies/tunes for various audio alerts (metro?)

Overall menu flow? Do you guys know what you want the first standalone version to look like, balanced with ease of expanding it? Screen mockups? Theres quite a bit to sort out for the standalone yet.

Does anyone else want to play?

[s2man]

Originally Posted by MetroMPG

Mine also came today - I got USB though. Which means I suppose it won't work with the Palm. Crud. I ordered mine before that particular direction was decided upon.

I ordered USB too, but assumed it will work with the Palm. USB to palm cables are like $6, and the serial.print function works on both RS232 and USB ports. Hope I'm correct. I also figured it would be useful for future expansion: USB to OBDII, USB memory sticks for data logging, etc.

Mine has only made it halfway across the country since Monday

Is everyone using straight system voltage to their boards? (12-14.5v) What I'm reading says to use 7-9v, and higher voltage may overheat the onboard power supply.