I have a circuit that I'm trying to put together to turn the fans on in my car at a lower temperature than what the car's computer is set to turn them on at. (I'm using a lower temp thermostat and trying to prevent large temperature fluxuations while idling).

I would like to use the stock temperature sensor and leave it connected to the stock engine computer (PCM) so that the computer can still turn the fans on if my circuit fails. The stock temp sensor is a thermistor that varies from about 10k ohm at 0 deg F to less than 180 ohms when at 212. The PCM applies a 5 volt reference and monitors this voltage. I would like to use a 741 op amp set up as a comparator so that when the resistance gets below 270 ohms, the relay will trip to ground the fan relay (not shown) turning on the fan.

I built this one on the breadboard and applied a 12 volt power supply. I then applied 3 volts to my reference and 1.5 volts to my input and the relay trips. So everything seems to work as I expect.

My question is, does anyone see any problems with the circuit, or any improvements that they might suggest? Do I need to use a higher rated variable resistor and ground the other leg to provide the correct voltage reference? I've tried to protect the circuit with a zener and a capacitor as a filter on the 12 volt power supply as shown. Any other suggestions?

 

Just a few points: I don't see how the reference voltage is set, the other end of the pot should be connected to 0V to form a voltage-divider. I would probably produce the 5V reference voltage within the circuit rather than take it from elsewhere. Connect the transistor in a common-emitter arrangement so you get more volts across the relay. 741s aren't so good and a dedicated comparator would be a better choice. And perhaps a higher value capacitor across the power rails after the input resistor.

The series resistor and diode on the power input are a good idea.

 

Just a few points: I don't see how the reference voltage is set, the other end of the pot should be connected to 0V to form a voltage-divider. I would probably produce the 5V reference voltage within the circuit rather than take it from elsewhere. Connect the transistor in a common-emitter arrangement so you get more volts across the relay. 741s aren't so good and a dedicated comparator would be a better choice.

The series resistor and diode on the power input are a good idea.

Thanks, connecting the other end of the pot to ground was what I was thinking needed to be done.

I'm sorry, I tried understanding what you mean by common emitter, but after looking at http://www.allaboutcircuits.com/vol_6/chpt_5/12.html, I'm still not getting what changes need to be made. Do you mean I should place the relay coil in series with the collector and the +12 volts, get rid of the 47 ohm resistor and then connect the emmitter to 0 V?

The 5 volt reference will be off of the PCM on the car (there is a dedicated 5 volt reference according to my wire diagram of the vehicle). Is there any reason you would produce a voltage reference rather than use the vehicle's 5 v reference? (I'm trying to keep things as simple as I can).

I would use a comparator, but this is just what I have on hand. What makes the 741 bad as a comarator, too sensitive to noise? Is there a certain comparator that you would recommend?

 

I'm sorry, I tried understanding what you mean by common emitter, but after looking at http://www.allaboutcircuits.com/vol_6/chpt_5/12.html, I'm still not getting what changes need to be made. Do you mean I should place the relay coil in series with the collector and the +12 volts, get rid of the 47 ohm resistor and then connect the emmitter to 0 V?

Yes, that's correct.

The 5 volt reference will be off of the PCM on the car (there is a dedicated 5 volt reference according to my wire diagram of the vehicle). Is there any reason you would produce a voltage reference rather than use the vehicle's 5 v reference? (I'm trying to keep things as simple as I can).

The external ref is fine, a dedicated ref would probably be a little more stable. Add a small capacitor (0.1uF) between pin 2 and 0V and also pin 3 and 0V to improve stability.

I would use a comparator, but this is just what I have on hand. What makes the 741 bad as a comarator, too sensitive to noise? Is there a certain comparator that you would recommend?

If the 741 is working Ok then leave it as it is. I'm just not a big fan of the device.

The only other thing might be to modify the circuit to incorporate some hysteresis to prevent the relay clicking in and out too frequently when the input is near the reference voltage.

 

I agree with Blocco that a 741 op amp is a poor choice.
You should not use an op amp as a comparator; ie: open loop, or no feedback. This causes the output circuitry of the op amp to be in constant saturation, which will lead to early failure due to heat. Comparators have an output that's designed to be operated in that mode.

Also, a typical 741 that you might buy in an electronics store is not rated to operate in the severe environment that a vehicle presents; you'll need to check the datasheet for your particular version to find out. A standard version is rated for 0°C to 70°C, you need one rated for at least -40°C to 125°C.

I suggest that instead you use something like an LM2903 dual comparator, which is basically the same as an LM393 dual comparator, but is rated for automotive temperatures (-40°C to 125°C).

Running the engine at a lower temp and turning on your fans at a lower temperature will result in lower fuel economy, and higher emissions. There is really not a pressing reason to turn on your fans early, or changing the operating temperature of the engine. If you are running your A/C, the fans will be on anyway.

If your engine has been broken in while operating at a higher temperature, it will not run as efficiently at the colder temperatures; the components won't expand to their "normal" dimensions. Besides, the ECU will map for a richer mixture due to the engine being cold.

 

I don't want to shill a product, but a Thermodisc controller is just the thing. Only one moving part and no electronics - http://www.thermodisc.com/productDetail.asp?ProductID=89.

 

Check with a dealer service department, or a friend who may have the appropriate computer service tool.

There are some parameters in a cars computer "brain" that can be "tweaked" slightly, one being the threshold for the cruise control, or an adjustment of the cruising fuel mixture................

It's possible they might be able to redux your onboard computer to turn the fan on at a lower temp, without jumping thru all the hoops.

 

Check with a dealer service department, or a friend who may have the appropriate computer service tool.

There are some parameters in a cars computer "brain" that can be "tweaked" slightly, one being the threshold for the cruise control, or an adjustment of the cruising fuel mixture................

It's possible they might be able to redux your onboard computer to turn the fan on at a lower temp, without jumping thru all the hoops.

Yes, I've looked at reprogramming, but this vehicle uses an OBDI computer with a rom chip that can't be reflashed, thus reprogramming it is difficult to say the least. To make matters worse, it's sort of a hybrid computer GM used in it's transition between OBDI and OBDII, so aftermarket products are just not there. Many turn to manual switches to control the fans, and those who want to adjust fuel mixture turn to aftermarket units such as megasquirt (I'm not wanting to go quite that far).

I chose this as a project to learn a little more about circuits and come up with a useful toy.

 

I suggest that instead you use something like an LM2903 dual comparator, which is basically the same as an LM393 dual comparator, but is rated for automotive temperatures (-40°C to 125°C).

Running the engine at a lower temp and turning on your fans at a lower temperature will result in lower fuel economy, and higher emissions. There is really not a pressing reason to turn on your fans early, or changing the operating temperature of the engine. If you are running your A/C, the fans will be on anyway.

Thanks Blocco and Sgt! You make an excellent point about the op amp that I just didn't think about, it's definitely not robust enough for this project just based on the environmental conditions. I'll pick up a LM2903 comparator. Is that something Radio shack would have, or will I have to order from Digikey?

Re: lower engine temp. I know that I'm getting a little less fuel efficiency from the 180 thermostat rather than the 195 due to a more viscous oil, but I'm running 5W-30 anyways which is pretty light stuff. The 180 thermo is a common replacement to help prevent knock retard in these 60-deg 3.4L v6's. The 180 also still allows the engine to run in closed loop mode, so the engine is still monitoring the AF ratio closely. Problem is the fans don't turn on until 220, so when stopped at a light or in traffic the temp goes up to 220, but then when the car is moving, the temp drops to 180. I'm trying to prevent this fluctuation.

I'll post my final circuit when I pick up the comparator and get it working on the breadboard.

 

I'll pick up a LM2903 comparator. Is that something Radio shack would have, or will I have to order from Digikey?

Digikey or Mouser. Radio Shack only carries "commercial" temp grade semiconductors. The only comparator they stock is the LM339 quad, with a temp range from 0°C to 70°C.

Re: lower engine temp. I know that I'm getting a little less fuel efficiency from the 180 thermostat rather than the 195 due to a more viscous oil, but I'm running 5W-30 anyways which is pretty light stuff.

It's not just the oil viscosity; it's how the moving parts "mesh" together at various temps.

The 180 thermo is a common replacement to help prevent knock retard in these 60-deg 3.4L v6's. The 180 also still allows the engine to run in closed loop mode, so the engine is still monitoring the AF ratio closely.

I see. Have you de-carbonized the cylinders? Carbon buildup on the pistons/head/valves is a big cause of knock/preignition.

Problem is the fans don't turn on until 220, so when stopped at a light or in traffic the temp goes up to 220, but then when the car is moving, the temp drops to 180. I'm trying to prevent this fluctuation.

That's not so good!

I'll post my final circuit when I pick up the comparator and get it working on the breadboard.

OK. You should plan on having some feedback from the output of the comparator to the noninverting input to provide hysteresis. Otherwise, your fan will likely cycle on and off fairly rapidly.

 

I don't want to shill a product, but a Thermodisc controller is just the thing. Only one moving part and no electronics - http://www.thermodisc.com/productDetail.asp?ProductID=89.

That would be a nice simple fix. This project is more about getting there than the solution itself. I'll keep it in mind though!

 

Okay, I'm a little overwhelmed at the options for a comparator. Seems like there are some choices that I'm just not sure how to make out the differences. Will this one work okay, it's LM2903? http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=497-1560-5-ND

data sheet: http://www.st.com/stonline/products/literature/ds/2470/lm2903.pdf

Here's the latest diagram:

 

Blocco:

I tried wiring up the circuit with a common emitter, but the LED just stays on. Here's the circuit diagram for the common emitter that I tried. What am I missing?

If it's working as it is without having a common emitter, am I loosing that much by leaving the way it is in the original version?

Thanks!

 

Are you still using the 741? I don't think the output of the 741 will get close enough to 0V (i.e less than 0.6V to allow the transistor to switch off.

Check the voltage on the output of the 741, you can always add a diode or two in series with the base of the transistor if it's too high.

And another thing can you attach your schematics a something other than pdf? Sorry, but I hate pdf.

Just noticed, to add hysteresis the op-amp needs to be used in inverting mode with the feedback from the output affecting the reference voltage by pulling it in the opposite direction to the input. You've got positive feedback which will affect the value of the thermistor as seen by the engine ECU.

 

Are you still using the 741? I don't think the output of the 741 will get close enough to 0V (i.e less than 0.6V to allow the transistor to switch off.

Yes, I'm still using the 741 for the breadboard experimentation...

Check the voltage on the output of the 741, you can always add a diode or two in series with the base of the transistor if it's too high.

I think I'll just wait until I get the op amp to try a common emitter, though if I have some time before that I'll give that a try.

And another thing can you attach your schematics a something other than pdf? Sorry, but I hate pdf.

I prefer .jpg myself, the .pdf was strongly suggested by one of the moderators of this site in one of the "newbies read this post first" threads.

Just noticed, to add hysteresis the op-amp needs to be used in inverting mode with the feedback from the output affecting the reference voltage by pulling it in the opposite direction to the input. You've got positive feedback which will affect the value of the thermistor as seen by the engine ECU.

Yes, I noticed that too, I am considering just running it without any resistor for feedback hysteresis, relying instead on the thermal hysteresis that will occur with the time it takes to cool and heat the coolant. If I connect feed back to the reference voltage, I'll be affecting the reference voltage as seen by the ECU too. I guess I'm back to what you were saying earlier about generating my own 5-V reference. What chip would I use to generate my own 5-V reference? I know I can use a voltage divider circuit with the 12 v supply from the car, but the battery charging cycle will cause some fluctuation.

 

Hysteresis will be necessary due to the extreme amount of electrical "noise" under the hood (bonnet for you UK'ers). You really won't be able to get by without it.

To minimize the impact on your ECU (which would be a good idea) you could use an automotive-grade opamp as a noninverting buffer. You might look at LM2904's, which are dual opamps for automotive temp ranges. Just make sure you are ordering one in a DIP package so you don't have to fiddle with etching a circuit board to even test your circuit.

To answer your prior question, the LM2903N is the DIP version, that should work just fine. It's inputs will work nearly down to ground. They're pretty inexpensive; you might as well buy 10 of them so that you'll get the price break, and have several spares and supplies for other projects.

 

Yes, without hysteresis the relay will be probably chatter. For an easy 5V reference use a 7805 voltage regulator.

A good format for posting schematics is .png

 

Once you get this working as an ON/OFF type of thing, you might consider the next step; using PWM to slowly increase the speed of the fan as the temperature rises. That'll decrease noise, power consumption, and wear/tear on your fan. Your engine temp will be more stable, too.

But for the moment, let's keep it simple and get it working.