Another Long Wire question
Moskog
Posts: 554
In my current project, the Tractor Elevator Controller I plan to measure the tilt on the elevator fork compared to the tractor itself and also compared to the true horizon. The system is Propeller based with two Memsic 2125 accelometers, one controlling the body of the tractor and one controlling the elevator fork. The second (remote) Memsic will be some 5 meters away from the main PCB. The remote Memsic is powered (5V) through the cable. I tried different cables including twisted pairs but the reading from the distant Memsic appear to be floating, lets say some two to three degrees. The reading is not stable as when the Memsic is close to the PCB.
If I feed the remote Memsic with power from a different source close to it, with common ground, the reading is stable.
Trying to feed 12V through the cable from the main PCB and regulate down to 5V at the remote Memsic does not have any effect, reading still floating.
I was hoping to use only one cable between the two units because it will be exposed to the environment, so using one for the supply and lets say a coax for the reading was not optimal, but could be done.
Does anyone have any suggestions here to help solve this problem or give an idea what makes the problem?
If I feed the remote Memsic with power from a different source close to it, with common ground, the reading is stable.
Trying to feed 12V through the cable from the main PCB and regulate down to 5V at the remote Memsic does not have any effect, reading still floating.
I was hoping to use only one cable between the two units because it will be exposed to the environment, so using one for the supply and lets say a coax for the reading was not optimal, but could be done.
Does anyone have any suggestions here to help solve this problem or give an idea what makes the problem?
Comments
Try a large capacitor (1000uF) or so across the power terminals on the remoter Memsic. I'd be interested to see how you are dealing with tilt versus Acceleration and differentiating between the two. It's good that you have two sensors... one at the base unit and the other remote... at least you can create some sort of differential and determine a relative relationship between the two, but the other question still arises.
So far the base unit is made on a Propeller development board and the remote unit on a small proto board. In case of nice weather next weekend I plan to make a real test on the tractor. If success the next thing will be making a new base unit out of a pcb and a propeller proto board.
Glad that worked!
"..But still I would like to know why just a few meters of cable will make those kinds of problems." - There are several factors...
1) ... could depend on the environment and how much electrical noise you have. 'Any' length of wire becomes an all band antenna
2) ... IR voltage drop ... depending on your current demands "I" and the distance from point A to point B "R", the voltage on your line can drop and fluctuate significantly.
3) ... sure there are others... 1 and 2 are usually the killers.
What the capacitor does is basically smooth out any fluctuations by providing an 'averaged' stable voltage to the accelerometer at the far end. If you are still having issues, you can setup a simple transmission line from the X and Y outputs of the Accelerometer that can also help increase your distance. <- the reason is because they too will act line an antenna.