Hi folks,
I'm trying to figure out a basic electronics principle for a simple variable resistance circuit. This concerns an automotive coolant temperature sensor circuit.
Buss voltage is sent through a variable resistor, which varies with coolant temperature, and the resultant voltage is sent to the computer. When the voltage reaches a certain value, the computer turns on the radiator cooling fan.

Resistance lowers as the temperature increases.
Buss is 14.0v at low coolant temperature (high resistance, ~1300 ohms), the fan is not running. Computer sense voltage is low.
At 840 ohms (223 deg. F) the sense voltage reaches 5.4v and the computer turns on the fan.
I want the fan to turn On at a lower temperature. 840 ohms @210 deg. F. If I add a resistor in series, the fan will come on later(a higher coolant temperature).
I'm thinking I should boost voltage for that circuit. Not sure how to do that. Maybe use a boost converter in series?


Altering the coolant sensor is not possible.
I've replaced the sensor 3 times, they all have the same temperature/resistance value. Verified on the car and bench tested.
I've had the car for 30 years and only recently when the original sensor failed did I have this different fan operation values.
The fan is activating at the correct resistance value when I substitute a decade box for a test.
I have verified the temperature gauge operation is correct using a decade box. Gauge uses a separate sensor on the engine.

 

If I add a resistor in series, the fan will come on later(a higher coolant temperature).

Hi Eover,
Welcome to AAC
Have you actually tried adding a series resistor, say 100R?
If yes, what change in fan ON temperature did you measure.
E

Note: reason for this test is to see the Temp change, from that change for that resistor we could work out the value of a parallel resistor

 

Have you considered adding a resistor in parallel with the sensor?
You want to shift the threshold temperature by about 6%. So you need to lower the effective sensor resistance by about 6% (the change of resistance of the sensor with temperature can be approximated as linear for such a small change). A 12k or 15k resistor in parallel with the sensor should do the job. Tweak the value as necessary.
Depending on how the computer interprets the sensor output, the fan might run on for a bit longer than before.

 

Hi folks,
I'm trying to figure out a basic electronics principle for a simple variable resistance circuit. This concerns an automotive coolant temperature sensor circuit.
Buss voltage is sent through a variable resistor, which varies with coolant temperature, and the resultant voltage is sent to the computer. When the voltage reaches a certain value, the computer turns on the radiator cooling fan.

Resistance lowers as the temperature increases.
Buss is 14.0v at low coolant temperature (high resistance, ~1300 ohms), the fan is not running. Computer sense voltage is low.
At 840 ohms (223 deg. F) the sense voltage reaches 5.4v and the computer turns on the fan.
I want the fan to turn On at a lower temperature. 840 ohms @210 deg. F. If I add a resistor in series, the fan will come on later(a higher coolant temperature).
I'm thinking I should boost voltage for that circuit. Not sure how to do that. Maybe use a boost converter in series?


Altering the coolant sensor is not possible.
I've replaced the sensor 3 times, they all have the same temperature/resistance value. Verified on the car and bench tested.
I've had the car for 30 years and only recently when the original sensor failed did I have this different fan operation values.
The fan is activating at the correct resistance value when I substitute a decade box for a test.
I have verified the temperature gauge operation is correct using a decade box. Gauge uses a separate sensor on the engine.

Have not tried that. I could try that with the decade box.

 

Hi Eover,
Let's know the changes, and we can advise a parallel resistor value.
E

 

Have you considered adding a resistor in parallel with the sensor?

Didn't think of that. I'm at work, now. Let me get back to this thread in a couple of hours. Thanks for the reply.

 

Many Cars are supposed to run a Coolant-Temperature over ~200F normally.

The Engine's Thermostat will still stay closed until somewhere around ~200F.
This is to improve Fuel-Mileage and reduce Emissions.

Why are You trying to change the Temperature ?
.
.
.

 

You are correct for most modern vehicles. I'm just trying to get it back to the way it was. I"m familiar with it's operating temps/summer/winter/A/C on, etc. I've owned it since it was almost new (1994). OEM spec. is 180F thermostat. That's what it always had. I do know that these particular engines are prone to the rear cylinder warping/cracking or blown gasket if gotten hot too many times.

I could probably live with it as is. However, I'm also curious about electronic circuits and want to learn more. I know just the basics of DC analog.

 

"Gotten Hot" is a relative term that doesn't mean much.
To warp/crack a Block, or blow a Head-Gasket,
you'll have to get into some severe Detonation,
and just ignore the heavy rattling until it won't run anymore.

If all parts and Fluids are in good condition,
going ~30-degrees past Thermostat-Opening is a routine event.

How do You "know" what "the way it was" is ?
.
.
.

 

Good morning,
I'm sorry for letting the thread languish over the weekend but, I was unable to access a computer. Did not get to make any tests on my car.

Alec_t
Thanks! That's very helpful information to give me a starting point. I can do some tests using the decade box beginning with those values.

LowQCab
With this series engine, they are susceptible to cracking and/or blown gaskets, on the rear cylinder head, due to repeated operation near the far end of the operating scale. 240F. (without experiencing detonation). I know this from a model specific online forum and from personal experience. I have two identical vehicles of this model. One blew the rear cylinder head gasket due to the condition I"m trying to rectify.

I know what normal is from 29 years of owning and driving this vehicle. It used to be my daily transportation. Also, information from the factory service manual states various temps. and ohms.

I'm seeing about 40F over design temp. and the coolant gauge indicating in a range it never did before.
(as stated earlier, I did verify the accuracy of the gauge).

Could the engine survive with this operating condition?
Possibly. But, why concede to this? Not only that, the fix appears to be very simple.

 

"" Could the engine survive with this operating condition?
Possibly. But, why concede to this? Not only that, the fix appears to be very simple. ""

It depends on why You think it's too High.

The quickest and best way that I know how to determine what's going on
is use an Infa-Red-Temperature-Gun, they're cheap at Harbor-Freight.
Drive the Car for a good 15 minutes to get it thoroughly warmed-up,
the measure the Cylinder-Head-Temperature near where the Head meets the Block,
and away from the Exhaust if at all possible.
Do not measure the Temperature of the Valve-Cover.

If it's over 180F, flush-out and replace the Coolant, AND, replace the Thermostat and Radiator-Cap.
Just do it anyway, even if You think it doesn't "need it".
Use ONLY the exact type of Coolant recommended by the manufacturer,
( You still haven't stated the manufacturer, Model, and Engine-type ),
If the wrong type of Coolant has been used previously,
your entire Cooling-System could be hopelessly clogged-up with nasty black-goo.

The next step is to check every Plastic-Panel under the Engine and surrounding the Radiator.
Every Plastic-Panel serves an important purpose in the Cooling-System's functions.

If all Plastic-Panels are in place, and the Head still exceeds ~180F,
You may have to replace the Radiator,
especially if the wrong type of Coolant was ever put into the Engine,
and it's now all gooped-up inside.

Until your entire Cooling system is verified to be in top-notch-operating-condition,
don't mess with the factory designed Cooling-Fan operation,
if it actually does help anything, you'll just be covering-up the real problem.

There is a possibility that You have a 2-Speed-Radiator-Fan, and the "Low-Speed" is not working.

If You have a continuously Variable-Speed-Fan, then the problem is probably listed above.

If your Cooling-System is full of "Black-Goo",
let me know, and I'll provide the only procedure that I've seen work,
( 99% of commercial "Radiator-Flushes" will NOT cure this condition ).
( Usually it's best to replace the Radiator, rather than trying to clean out the Black-Goo ).
.
.
.

 

The point is: the cooling fan is coming on at a higher temperature than it ever did before I changed the sensor. By any metrics that I can find, the fan On/off is not operating as factory designed. Not covering anything up. When the fan finally does turn ON, the coolant temp. visibly cools back down to about 190F because of hysteresis.

Typically the fan would turn On at about 210F. (which is the 30F spread you mentioned). Now it's 223F when the fan actuates.
Maybe I should have explained that the coolant temp. climbs above historical normal when the car is sitting at a stop light. Which is normal behavior for any vehicle with no cooling air flow. If I move down the highway, the temp. returns to the 180F mark. It will remain at 180F on a 95F day while cruising down the highway.

 

The point is: the cooling fan is coming on at a higher temperature than it ever did before I changed the sensor. By any metrics that I can find, the fan On/off is not operating as factory designed. Not covering anything up. When the fan finally does turn ON, the coolant temp. visibly cools back down to about 190F because of hysteresis.

Typically the fan would turn On at about 210F. (which is the 30F spread you mentioned). Now it's 223F when the fan actuates.
Maybe I should have explained that the coolant temp. climbs above historical normal when the car is sitting at a stop light. Which is normal behavior for any vehicle with no cooling air flow. If I move down the highway, the temp. returns to the 180F mark. It will remain at 180F on a 95F day while cruising down the highway.

Its definitely a Ntc type sensor, so has it gets hotter the resistance goes lower, so putting a resistor in parallel with the sensor will tell the computer the the temperature is hotter than normal, try a variable resistor first ( say 4K7)to see if you can get it to work at the temperature you need, then substitute it for a fixed resistor if it works.

 

Did not have time to perform any actual work on my car.
However, theoretical analysis:
@210F the measured sensor resistance is 960Ω. I need it to be 840Ω at that temperature.
By adding a 7000Ω resistor in parallel, I get an effective resistance of 840 ohm.
As if I had a single resistor of 840Ω.

840Ω is less than either value of the two resistors.
I'm not sure what is actually happening. Electricity takes the path of least resistance so some voltage is bled through the higher Ω resistor, reducing the effective resistance? Not sure.
I have done some reading here on Allaboutcircuits to help my understanding but, I"m actually more confused.
Anyone care to help explain this?

 

Electricity takes the path of least resistance

In the case of two paths in parallel it takes both paths. The path with least resistance carries more current than the other path.
Put a 6,8k resistor in parallel with the 960Ω to bring the effective resistance down to 840Ω.

 

Electricity takes the path of least resistance

That old saw, which is not literally true, has led a lot of people astray when trying to analyze circuits.
When electricity has more than one path, it takes all the paths, but more electricity will go down the paths with the lower resistances.

So say, for the sake of argument, that you have a 1kΩ and a 2Ω resistor connected in parallel.
So if you apply 1V across the parallel connection the current through the 1k resistor will be 1mA and the current through the 2k resistor will be 0.5mA, giving a total current flow of 1.5mA.
The equivalent resistance is then 1V / 1.5mA = 666.6Ω, the same value if you use the parallel resistance formula.

Make more sense now?

 

Right! the electricity will go down both paths. Didn't mean to imply it goes down one path to the exclusion of the other. That's why I said "some voltage is bled through the higher resistor". Which is a poor use of terminology. Should have said "current" , not voltage?
Mathematically I do see that effective resistance is lower with parallel resistors as opposed to the single one. Just seems odd, the effective resistance is lower than either one the actual resistors.

 

Just seems odd, the effective resistance is lower than either one the actual resistors.

Not odd at all. The total current flow is always more than the flow in just one path.

 

Just seems odd, the effective resistance is lower than either one the actual resistors.

Since the current of resistors in parallel is always higher than either resistor by itself, and resistance is a measure of current for a given voltage, the resistance of the parallel set must always be lower.
So what would you expect the equivalent resistance of parallel resistors to be that would not be "odd"?

 

The odd part: Adding more resistors reduces total resistance. Adding more restrictors is less restrictive than a single restrictor. I realize my thinking is superficial and not looking into the details of the circuit.

I'm actually opening more channels for current.
Would this be a correct analogy: You have a large lake (voltage potential), a sluice from the lake 10 feet wide flows 100 gpm (current). Another parallel sluice is opened, 5 feet wide and flows 50 gpm. The total flow downstream is now 150 gpm. Since current is increased and voltage remains the same, mathematically resistance must decrease.