if a 2.2k still doesnt work, then try a new chip and check that the opto is triggering. dunna why it would be burned out, but you never know.

it seems that the LED strip is controlled by an analog pin, so it could require some sort of input before doing anything. take your strip, put gnd to gnd on the arduino board. vin to vin on the arduino board header. and the signal wire to A0. then upload the program, unconnect from usb and plug in the arduino via the barrel jack.

For the first part (I want it to be able to charge a battery up to a certain voltage, then when full, it dumps the excess charge into a dump load), use a SPDT relay or a SPCO / SPTT (stable off position for between dummy load and charge). That relay will move the 12v output from the solar panel to the charging circuit and the dump off. So youll also need to make a charge circuit for the solar panel -> battery. The relay can easily be controlled using an arduino. to log data about how much it is charging / dumping, youll need to make voltage and current sensors. here is in example that i made for switching dc to change between rev and fwd for some dc motors: http://1.bp.blogspot.com/-q_O0LU735H4/UYstU-VV1YI/AAAAAAAACfA/_q_CKT7TdWI/s640/motors.png

for the data logging you mentioned later, you just need some of the voltage / current measuring things i mentioned earlier. i've never used a Ethernet shield, but with a few tutorials, you should be ok!

for a voltage sensor, you just need a voltage divider. it essentially "divides" the voltage to a lower range. so instead of 10 to 15v or something measured on teh battery, a 0 to 5v signal is sent to an arduino analog in. go here: http://www.raltron.com/cust/tools/voltage_divider.asp set r1 as 22, and r2 as 10 (ohms or kohms doesnt matter, just the ratio). put 15v as the in voltage and click calculate, you should see the max voltage the arduino will read. now delete the in voltage and replace it with 11v and delete the out voltage number. click calculate and you''ll see the minimum voltage the arduino will see. now, connect the battery(fully charged) to the divider and check the output (to the arduino) with a voltage meter. then find the analog value it corresponds to with the arduino. do the same for a dead battery. then change the analog value back into a voltage number and store it in a byte, int whatever and write it to the ethernet.

this explains the current sensor: http://www.vwlowen.co.uk/arduino/current/current.htm

surely possible!!!

you need to use a 4.7k resistor on the LED side of the opto, not a 10k. neg -> 4.7k -> led -> pos etc.

yes! dont you love it when things go to together like that

i suck at explaining, dont i hah. here is my blog post might make more sense: http://dduino.blogspot.com/2012/02/trigger-buttons-with-arduino.html

specifically this:

turn on the device and measure voltage: one probe on ground, one on the center pad measure that voltage, then measure the outer ring. now, change to continuity test mode on your voltage meter. if the outer pads are connected (~1 ohm resitance, or your meter beeps) then the the center pad needs to be switched by the arduino and have a pull down/up resistor to the outer pad.

the contrast and birghtness of the lcd is controlled by the obduino based on the headlights etc from obd.

i was being lazy haha, so ill put in some ~200 - 300 ohm resistors on the contrast and brightness pins. my contrast is fine, but i dont see how current limiting resistors will fix it. Extra current causes over contrast?

i am using gnd / sig gnd, 12v and the k line directly from the obd connector.

I've built to this schematic with a few modifications:

I made my own "breaboard duino" on some perf board
added a lm7805 for the 12v to 5v conversion in the car.
put switches inline with the TX / RX line between the iso chip and the atmega (wouldn't program over serial otherwise)

my issue is very peculiar so i program the arduino (through the "upload code by programmer" option) with the original obduino32k code from google code and the display works fine while being powered off of the USBASP ISP connector. But, as soon as i use another 5v power supply (connecting the same as the USBASP to the 5v / gnd rails), the lcd contrast is messed up and i cant read it. same happens if i go through the vreg 12v -> 5v.  All voltages were checked to be sure that the sources were ok.

Next, i tried the Hello World LCD example in the Arduino IDE and the LCD contrast works perfect with any power source. i still had to add in analogwrite for the brightness and contrast, of course.

I just did another test: i connected only the gnd and 5v from the USBASP to their respective places on the arduino, and walla the contrast issue is there. somehow the mosi / sck / msio / rst connection are doing something to make it work? makes no sense at all. So, something in the OBDuino code is messing with the contrast output... strange




any thoughts?

thanks.

i dont quite know, haha. i copied that code. bACCx  is probably in a different function. do you want it?

i used another method for level converting 5v to 3.3v seen here: http://2.bp.blogspot.com/-eKW0mVBNZ0k/T1AsgYHyjMI/AAAAAAAABAk/pMAUyS1u0CU/s400/RN42+Scheme.PNG iwth my RN42 module. But, i still recommend that you wire yours according to the documentation of your module seen in the last diagram on page 3 here: http://www.e-gizmo.com/KIT/images/EGBT-04/EGBT-045MS-046S%20Bluetooth%20Module%20Manual%20rev%201r0.pdf

for that scheme i'd use a regular glass schottky diode. you could use one of the common black ones if you cant find one.

good info. id probably be better off doing something more universal like: http://www.radiolocman.com/shem/schematics.html?di=61516
then i can use anything 120v ac

ok, thanks. any tips in general?

im looking for a DC DC boost converter for a car's battery ~10-15v to my laptops charger output (19v, 2.15A when charging). it doesn't make sense to use an inverter to: 12v DC -> 120V AC -> 19v.

im thinking:
Vin: 0-20v
Vout: 10-20V or so.
Aout: <=10A

i was looking at: http://www.linear.com/product/LTC3786 or http://www.ti.com/product/tps55340 or http://www.ti.com/product/tps61175 but all the ones i have seen use transformers. anyway around this, or to get it slimmed downed?

updated original post with new code.

--------------------
the reason at 100:

Code:

if((millis - lastTime) > measureTime) //if 10sec has passed
  {
   ...
  }

there is supposed to be no semicolon after the "measureTime)". simple syntax error.

-----------------
71: was:

Code:

avgTemp += tmp2;

now is:

Code:

avgTemp += (int)tmp2;

beacause tmp2 is a float, but avgtemp is an int variable, so we "convert" the float to an int by adding "(int)" infront of "tmp2" to get (int)tmp2

-------------
an error in the "writeTimeToScratchpad" function was because of a copy/paste error from the code example.


turns out coding later at night results in errors