Steve
Maas
Long Beach, California, USA
December, 2008
My Bugeye has an electric fan. The previous owner, who installed it, lived in the California desert, so it seemed appropriate simply to keep it on whenever the car was running. Here along the coast, however, we have our share of cool days, and then it can prevent the engine from coming up to temperature.
The simplest solution is to add an ordinary automotive thermostat. This has a couple of problems, however;
My thermostat is pretty simple. It uses a thermistor for sensing the temperature and a comparator to switch a relay. Positive feedback is used around the comparator to provide the necessary hysteresis, that is, the difference between the on and off temperatures.
Click on any illustration to open a larger version.
The circuit of the controller is shown below. An IC voltage regulator drops the 12V auto system voltage to about 5V, while eliminating virtually all of the electrical noise in the process. This voltage is further regulated to 2.5V with a reference diode; the lower voltage is needed for the thermistor, to minimize its power dissipation and thus self heating.
The thermistor and 1K potentiometer form a voltage divider. The voltage is compared to a second divider, consisting of two 10K resistors. When the thermistor is cool, its resistance is relatively high, so the voltage at the (-) terminal of the comparator is higher than at the (+) terminal. This sets the comparator's output low, keeping the fan off. As the thermistor heats up, its resistance drops, and eventually the voltage at the (-) terminal drops lower that at the (+), setting the comparator output high and turning on the fan.
A 50K resistor provides positive feedback for hysteresis. When the comparator output goes high, turning on the fan, the positive feedback increases the voltage slightly at the (+) input. Thus, the fan does not go off until the thermistor becomes somewhat cooler than the value at which it turned on. Without this modest difference, the fan might turn on and off quite rapidly, quickly wearing out both the relay contacts and the driver's nerves.
The output of the comparator cannot handle the full coil current of the fan relay, so a 2N2219 transistor is used to switch the relay. The diode paralleling the relay coil clips voltage spikes that would otherwise occur when the relay is switched off. This is a standard practice in switching inductive loads. The relay coil operates from the unregulated 12V input; there is no need to regulate its voltage. The coil draws approximately 100 mA, and the rest of the electronics about 20 mA. The fan itself requires 6A, so a 10A in-line fuse is used at the 12V dc input.
A couple pictures of the regulator are shown below. It fits nicely on the side of the wheel well, where it has easy access to the engine compartment. The thermistor, which is about the size and shape of a lentil, is soldered to teflon-insulated wire, the connections are covered with heat-shrink tubing, and it is slipped into the upper radiator hose at the engine. The wire is very thin, so tightening the hose clamp is enough to prevent leakage.
Adjusting the controller is a little tricky. First, it is necessary to calibrate the thermistor. This is done with a water bath (a fancy name for a pot of water on the kitchen stove), a thermocouple temperature gauge, and a multimeter. It involves logging the thermistor resistance at a number of temperature points as it slowly warms up.
Having those data, the hysteresis resistance and voltages can be calculated. Initially the hysteresis potentiometer is adjusted for the expected voltage at the (+) input of the comparator with full positive output. This is approximate and will be changed, but a good initial setting helps to speed the adjustment process. For the initial temperature adjustment, a potentiometer is substituted for the thermistor. The resistance is set to the thermistor's resistance at turn-on, and the temperature potentiometer is adjusted until the fan relay switches on. Then, the potentiometer modeling the thermistor is increased in resistance until the fan goes off. The resistance of that potentiometer then gives the turn-off temperature. If there is too much hysteresis (i.e., the turn-off temperature is too low), the hysteresis potentiometer's resistance is increased and the process repeated until the desired turn-on and turn-off temperatures are achieved.
This process is still not perfect, so it is a good idea to do a final check in a water bath. Dunk the thermistor, warm it up slowly, and note the temperature when it turns on. Then, let it cool and note the turn-off temperature.
The controller is adjusted for a turn-on temperature of 185F (85C) and a turn-off of 175F (79C). This seems reasonable to me, but it may be necessary to change it a bit once I've had some experience with it.
I have found that the engine's mechanical cooling fan moves plenty of air. On all but the hottest days, it keeps the coolant temperature in the 160-170F range. The fan comes on only infrequently, usually when I'm sitting in traffic on a warm day and the coolant temperature is pushing 190F on the gauge. (The gauge, I've determined, reads a little high.) Without the fan running, the engine warms up more quickly, especially on cool days. Even though it doesn't come on very often, I like having the fan; it eliminates any worry about overheating, and with the controller, there's no problem with too much cooling. It's a useful improvement.
I don't like saying this, but I suppose it's necessary, now that the lawyers have taken over American society. If you choose to do anything based on this information, but don't know enough about automobiles or electronics to be comfortable with it, get some help. I make no guarantees that this information is correct, so you take full responsibility for the results of anything you do that is based on it. This is just a report on my experience with these engines. It is not intended to be a set of instructions for duplicating my work or a recommendation to do it. You're on your own.