|Topic Review (Newest First)|
|04-13-2012 06:48 AM|
|Voltar||Glad it worked out !|
|04-12-2012 10:27 PM|
I just finished a test ride, after installing the new military precision resistors (for the gear position sensor voltage divider).
The reading are super accurate, and even though I did not ride much, and the temperature did not change too much, gear reading are rock solid!
NO FI codes
I used 2 X 499K Ohm resistors (CMF55499K00BEEB),
I also installed the DS18B20 temperature sensor in the front box (where the LCD is), so now I have two temperature readings: One which is true ambient air temperature, and another which is the Arduino's on board temperature. I like it.
I got two more military spec resistors for the battery voltage divider but am too lazy to install them, since I need to remove the existing resistors - a lot of clean up work.. next week.
@Voltar: If you decide in the future to have voltage dividers, e.g. for measuring the battery level, please remember my advice/lesson - For ADC ALWAYS use high precision, low PPM/K metal film resistors..
Thanks everybody for helping out!!
I will do more test rides in warmer days, and hopefully can keep reporting a stable solution.
|04-09-2012 11:53 PM|
I am sure I exactly understand the circuit of the ECM wire going into your ADC input. Have you post the simple schematic?
That said, Rs in paralleled do yield a smaller R. It is logical because current now has multiple paths to follow instead of one. More paths for current = less resistance.
Then from Ohms law we also know that as R gets smaller, V will also get smaller (given the same amount of current).
Does that help or am I stating the obvious?
|04-08-2012 08:28 PM|
Originally Posted by Voltar View Post
The ECM resistor (i.e. R1 - from GPS to GND) and my resistors (i.e. R2) are in parallel.
This resistor calculator is very handy:
Parallel Resistor Calculator R1 + R2 = equivalent resistor R resistance equiv total resistor finder - sengpielaudio Sengpiel Berlin
It shows clearly that R2=20KOhms with any value of R1 greater than 20KOhm will always result in a value smaller than 20KOhms.
But why does the voltage go down? I think the ECM measures voltage on the R1, and if there is less resistance on R1, R1 is less dominant to R[GPS] and therefore the voltage on R1 goes down.
What do you think? Makes sense?
|04-08-2012 08:19 PM|
- I think the larger resistors are need to reduce the loading in the ECM or else the FI light will come on as to overload the output.
- don't forget that in "N" the ECM has to drive a green neutral light as well.
Hope that helps some.
|04-08-2012 07:01 PM|
Originally Posted by greywolf View Post
These are the resistors (2 are needed) I'm thinking of getting now (900KOhm, 0.1%, 25ppm also military spec):
|04-08-2012 06:48 PM|
Originally Posted by BigMan73 View Post
|04-08-2012 06:33 PM|
Have some insight!
I am not a person that gets discouraged quickly.. for better and worse, mostly the later :mrgreen:
Took some time to clear and arrange my thoughts and but I got back my to a scientific & methodical approach - experiments.
First, I disconnected the Arduino/Stromputer completely in order to try and determine if it is the Stromputer's fault or the voltage divider as a concept.
I performed 3 experiments:
1. I took a breadboard and built on it a voltage divider with 2X500KOhm resistors. No FI code.
2. I then built another voltage divider with the Military spec 2X10KOhm resistors (I got 4 - 2 were used in the Stromputer, I used the 2 other).
FI Code happened! Hmm. Interesting find - so it has nothing to do with the Stromputer.
3. Then I wanted to determine if the Mil Spec resistors are causing issues or it is the resistance value. So I used two 'simple' 10KOhm resistors (the DIY type with color codes and 10% tolerance)
FI Code repeated! So it was the resistance value!!
But why is the ECU so picky about 20KOhms in parallel to to the GPS line??
Interesting note - when I had the FI Code, and I pulled the clutch lever there was a low volume but repeatable buzz from the ECU's area. As if there a small buzzer there?!?
(Between each experiment I turned off the switch, to clear the FI code)
I guess I need to go back and buy 500KOhm (or larger) military spec resistors. Not a big deal, about $4, just a bit annoying. I guess this is the learning price I have to pay.
Will update you all once I fix this issue. GreyWolf - Thanks for the mental support!
|04-08-2012 02:50 PM|
|BigMan73||Another idea - What if I put a diode on the gear wire? It should reduce the voltage by 0.3V-0.7V (depending on the diode type) and therefore reduce the incoming voltage.|
|04-08-2012 12:04 PM|
Originally Posted by greywolf View Post
So I guess my board is dropping the gear voltage somehow, I do see a drop of 0.4V in my voltage tester - just I don't understand why and how this could happen.
|This thread has more than 10 replies. Click here to review the whole thread.|