These comments are specific to the RV-12 Rotax/D180 configuration.
There are two main components in the system, a sensor to generate the signal and a detector to measure that signal. A Thermistor is used for the sensor; this is a special type of resistor whose resistance is temperature dependent. A voltage signal is generated by placing a resistor in series with the Thermistor and connecting the assembly across a fixed 5 volt supply, we then measure the voltage drop across the sensor. This generates a sensor voltage from about 1 V at 120?F down to about 0.080V at 300?F. These voltages are measured by the Analog/Digital Converter (A/DC) inside the D180. These voltages are extremely low so even small losses will result in large errors and because the sensor is non-linear, these errors increase dramatically with temperature.
Design issues can prevent the D180?s A/DC from seeing all the voltage from temperature sensor. When this happens the displayed temperature indicates higher then actual temperature. With a signal loss of 100mV, the display may indicate 234?F, well into the yellow arc, while the actual temperature is only 204?F. To understand what?s going on we need to calculate the wiring voltage drops and load them into a big picture. In the drawing below (Vsensor) is the voltage provided by the sensor and (Vinput) in the voltage as seen at the input of the D180.
or the pdf version
Looking at the diagram it?s clear where we need to concentrate our attention. The voltage drops are excessive between the D180 Master Ground pin to the airframe grounding point, but only when the master is ON. These are worse case calculated numbers. In the real world the voltage drops are less and will vary between aircraft because the D180 attach screw will shunt some of the ground load. Relying on the D180 attach screw as a grounding point is a poor alternative to a properly designed ground. I was unable to get the voltage drop below 20mV using the attach screw. This caused a 5?F error mid scale and near the top end at 294?F actual, it was indicating 317?. While this was a vast improvement over my initial 100mV losses, we can do better.
In the above drawing I included voltage drops that appear to be irrelevant, these are included to illustrate the big picture. Often with grounding problems the common response is to blindly ?add another ground? or ?run a wire to the battery.? If you look closely, adding additional engine grounding does nothing as these drops are already near zero. And connecting the engine sensor ground directly to the negative battery post, if that were possible, would only add to the voltage drop.
The maximum wire size a D-sub pin will accept is #20. Using material on hand, I installed D-sub pins in one end of three 22-awg wires 1.5 inches long, spliced to a piece of 12-awg wire 15 inches long. (I would have preferred to use 20-awg, but I didn?t have any mil spec wire that size at the time, I recommend using 20-awg.) The female D-sub pins were inserted into the unused ground pins 5, 16 and 17 of the 37 pin connector on the back of the D180. The 12-awg wire?s ring terminal is attached to an existing AN3 bolt, for a wire harness Adell clamp, on the panel base directly behind the D180. This reduced my voltage drop to 4mV. Now at 294?F actual I am indicating 298?. I expect the error to be less than one 1?F mid-scale. Everything is accessible by removing the D180 from its tray. There is plenty of room to work with no need to open the pitot/static/AOA lines.
It?s a simple procedure to test for ground side losses. Other than a 10 cent resistor, no special tools, meters or skills are needed. The greatest errors are on the high end of the temperature scale, so that?s where we need to test it. A 22 Ω resistor will indicate about 290?F-300?F depending on the tolerance. The actual displayed value doesn?t matter because we aren?t checking calibration, we?re testing for losses.
- 1) Temporarily remove the wire from the sensor, connect it to one side of a 22 Ω resistor, ground the other side of the resistor to the engine crankcase.
- 2) Turn the Master Switch ON.
- 3) Record the value displayed on the D180 for that sensor, call this the ?Master ON Value.?
- 4) Turn the Maser Switch OFF (the display will remain on for 30 seconds powered by its internal battery.)
- 5) Record the value displayed for that sensor, call this the ?Master OFF Value.?
- 6) Ideally the two readings should be equal. The larger the error, the more ground side losses you have.
It?s an easy fix. With the wire pre fabricated, the job takes about 30 minuets. If you don?t have the wire or the D-sub pin crimping tool, I?m sure SteinAir can make one up for less than the cost of the $30 tool.