Sunday, February 22, 2009

Controller Temperature Sensor

I designed the new computer to monitor the temperature of the motor controller but being in a hurry to drive I didn't make and hookup this sensor until now. After each short drive I keep checking the controller and at least in the winter so far heat is no concern at all. I'm only seeing a few degrees increase in temperature.

Either way it'll be nice to see this temperature while driving so I decided to finish that up.
I used a hobby servo extension cable for the wiring. You can depress the little tabs and remove the connectors from the housing.

Note the position of the wires. It's not the same as the original order on the cable. The LM34 temperature IC from left to right is +, signal, -. Use three small pieces of shrink tubing to protect the terminals and keep them from shorting out on each other.

Next I wrapped all three connection in one larger piece of shrink tubing.

Now to make a housing for this sensor. We want to use something that will conduct and transfer heat quickly so that we can see changes to the temperature quickly. I'm using a small aluminum tubing that I cut a 1" long piece from. The bolt and nut are used to reinforce walls on half of the tube while shaping the other half.

After pounding the unprotected half with a hammer you get a nice flat piece that we can drill into for mounting the sensor somewhere.

A little JB Weld to hold the sensor in place. Shove the sensor to the far end so that it contacts the metal to pick up the heat well although the JB Weld does a fair job at heat transfer from what I've noticed.
Here is the final sensor hooked up and ready to be mounted.

I used a self tapping screw and mounting the sensor on to the heat sink of the controller.

This is temporary as I'm still not sure exactly what I want or need mounted in the car. This little LCD screen is showing the RPM from the computer, the temperature of the motor and the temperature of the controller. The RPM are not needed since the dash has that implemented already. Eventually I'll probably use the microprocessor to determine which of the two is hottest relative to its maximum temperature (which I also am outputting here on the screen for reference) and use the dash temperature gauge to represent that temperature. This way it won't matter which of the two is hot, if you see the gauge getting too high, something is too hot and that's all that matters.
After my first test drive I realized I did something wrong. When driving the temperatures tend to bounce around a little and unless you let off the throttle it's hard to tell the real temp. I didn't run shielded wire from the sensors to the computer. I'll need to pull the front batteries out when I do the AC so I'll rerun those wires at that time.

No comments: