When the fan motor get started?

[Guest Markino]

Hi guys,

do you know the "exact" coil temperature when the fan motor get started?

And what about the warm up temperature? When (coil temperature) our motor could be considered okay?

[Englishman 165]

[djsb 493]

This is probably controlled by the EMU. The Engine Coolant temperature Sensor turns on the over temperature warning light at about 100 degrees centigrade so I'm guessing that the fan should come on maybe at about 90 degrees. I cant find anything in the manual about this so it may not even be a fixed temperature and could depend on various factors programed into firmware.

[Guest Markino]

Capture.PNG

Thanks!

That means that the fan motor starts at which temperature?

[Guest Markino]

This is probably controlled by the EMU. The Engine Coolant temperature Sensor turns on the over temperature warning light at about 100 degrees centigrade so I'm guessing that the fan should come on maybe at about 90 degrees. I cant find anything in the manual about this so it may not even be a fixed temperature and could depend on various factors programed into firmware.

Can I ask you where did you get the info about the light on at 100 degrees? Thanks

[DaveM59 1,608]

Applying a bit of logic and theory, water boils at 100 degrees C at normal atmospheric pressure of 1 bar or 14psi (roughly) and the cooling system is rated at 16-20psi (18psi nominal) so a total pressure of around 32psi. At 32psi the boiling point is raised to 123 degrees C and so if the stat is fully open at 95 degrees a mid point for the fan to cut in would be about 110 degrees, but as the fan is less efficient than a good blast of air from movement at speed it probably comes on a little lower at maybe 100-105 degrees, and to avoid getting a boil on and give you some advance warning the warning lamp would be on at about 110. If it were any lower with a stat temp of 95 to start with the warning lamp would be coming on at every set of traffic lights on a hot day, but the fan will cycle on and off in such circumstances and it is designed to. It would also give you enough notice of a leak which would result in a drop in pressure too.

 

Accurate figures are quite probably not made known as you cannot measure the temps in a real world situation and they would all be taken into account in the design of the system and programming of the EMU.

 

Older systems with no engine management would use a bi-metallic switch on the radiator tank to trip the fan and they would operate at much lower temperatures as they are not measuring the hottest part of the cooling system which is around the cylinder head. Some cars still use a similar system.

[Guest Markino]

Applying a bit of logic and theory, water boils at 100 degrees C at normal atmospheric pressure of 1 bar or 14psi (roughly) and the cooling system is rated at 16-20psi (18psi nominal) so a total pressure of around 32psi. At 32psi the boiling point is raised to 123 degrees C and so if the stat is fully open at 95 degrees a mid point for the fan to cut in would be about 110 degrees, but as the fan is less efficient than a good blast of air from movement at speed it probably comes on a little lower at maybe 100-105 degrees, and to avoid getting a boil on and give you some advance warning the warning lamp would be on at about 110. If it were any lower with a stat temp of 95 to start with the warning lamp would be coming on at every set of traffic lights on a hot day, but the fan will cycle on and off in such circumstances and it is designed to. It would also give you enough notice of a leak which would result in a drop in pressure too.

 

Accurate figures are quite probably not made known as you cannot measure the temps in a real world situation and they would all be taken into account in the design of the system and programming of the EMU.

 

Older systems with no engine management would use a bi-metallic switch on the radiator tank to trip the fan and they would operate at much lower temperatures as they are not measuring the hottest part of the cooling system which is around the cylinder head. Some cars still use a similar system.

 

 

Thanks for the great explanation!

I'm asking that question because I'm trying to do a little board with Arduino Nano that will display the coil temperature taken directly from the bike sensor....

I'm making some tests and it seems works! Leaving the bike in idle the fan motor has been switched on when on my display the coil temperature was around 105 degrees C... could be a reasonable value for you?

djsb wrote that the led of high temperature turns on at 100 degrees C ant it seems strange to me.

I've also read about a guy that added a temperature sensor and he wrote that the fan motor starts for him at around 80 degrees C (normal summer temperatur around 70 degrees C)... another strange value form me so I'm looking to some documentation so that I can check if my project works or not.

[CFB 4,602]

The normal running temperature on my car is 90*c

Wouldn't it be the same on the NC?

[DaveM59 1,608]

In the spec sheet Andrew posted above it states a fully open stat at 95 degrees so if the fan came on at less than that the two would be fighting each other. Some fans are stepped speeds and come on at a low speed initially and step up as the temps rise, but I don't thing many bikes tend to have this complexity.

The warning LED will light when normal operating temp has been exceeded and attention is needed, you wouldn't want a warning light instructing you to stop the engine lighting up unnecessarily would you?

Most vehicles will slowly lower the temp when the fan cuts in, but it rarely quite goes back to normal until you get some forward motion and proper air flow again.

Getting an accurate measurement of the internal coolant temp is difficult and without a gauge as standard to give you some indication of what the normal running condition is, even if you tapped a feed from the temp sensor and calibrated your electronics from there, unless you did it on the move you would be incorrect, as if you wait for the fan to come on at idle, that is over normal running temp.

I have to admit the lack of a temp gauge on the integra does worry me a bit as other bikes I have had suffer damage if they don't fully warm up, as well as if they get overheated so you need a visual indication that the system, thermostat and fan are all within tolerance, and simply having a danger lamp if it gets overheated seems a bit 'cheap'. I'll have to keep telling myself 'it's a Honda - they don't go wrong'.

Edited April 6, 2016 by DaveM59

[Guest Markino]

In the spec sheet Andrew posted above it states a fully open stat at 95 degrees so if the fan came on at less than that the two would be fighting each other. Some fans are stepped speeds and come on at a low speed initially and step up as the temps rise, but I don't thing many bikes tend to have this complexity.

The warning LED will light when normal operating temp has been exceeded and attention is needed, you wouldn't want a warning light instructing you to stop the engine lighting up unnecessarily would you?

 

So, in your opinion our fan motor will starts for sure at a temperature greater then 95 degrees, correct?

The warning led at 100 degrees seems strange to me, isn't it?

[djsb 493]

In the workshop manual there are details of how to test the DC resistance of the Engine Coolant temperature Sensor at 2 set temperatures, 40 degrees and 100 degrees Celsius. The DC resistance at these temperatures are 1.0-1.3K Ohm and 0.1-0.2K Ohm respectively. I assumed that these two temperature where chosen to represent Normal and Abnormal temperatures of the coolant. However the EMU can use a resistance anywhere between these 2 range of values to decide when to light the high coolant temperature warning light and likewise the cooling fan. This information is not described in the manual. 

[DaveM59 1,608]

If the temp sensor is a basic variable resistance device you can probably power any car/bike accessory water temp gauge from it and adjust the needle to sit centrally at a normal running temperature, then at least you can watch it rise from cold so you know you are actually warming up and don't have a dud thermostat, and also see it rise and hear the fan cut in then watch it drop again to know the failsafe is OK as well. Probably no need for anything bespoke or particularly annotated accurately.

Edited April 6, 2016 by DaveM59

[embee 7,288]

The NC uses a coolant temperature sensor as an input to the ECU for the usual engine management functions (fuelling etc). The ECU also controls the fan switching using the same sensor input, there is no separate fan switch as usually used on older bikes.

 

Temperature sensors are usually thermistors, with this kind of characteristic (example only)

Because the ECU controls the fan, it could (guessing here) also use other parameters such as vehicle speed and engine speed for the logic. I suspect it probably doesn't, but it could. There is little point switching on the fan if the bike is doing 100kph for example.

The thermistor is in the thermostat housing before the stat itself, i.e. the hottest part of the system. People fitting accessory type temperature gauges usually don't get the sensor at this point, so the values they see are not the same as the max coolant temp. Also the sensor must be immersed in flowing coolant, attaching it to a metal surface (e.g. thermostat housing) will give a false reading.

 

As an example for switching temperatures, the 650 Deauville manual says the high temperature warning switch (discrete switch) which is in the stat housing closes at 112-118C, and opens again (falling) at "below 108C". The fan switch is in the cold side radiator header tank so not the same conditions as the thermostat housing, but FWIW it closes at 98-102C rising, and opens at 93-97C falling. For thermostat housing sensing I'd guess about 5degC on top of those temps, so maybe fan on at 105C rising and off at 98C falling (ballpark guess, I don't have exact figures for the NC but that's about right).

Typical 50/50 glycol/water coolant at 1Bar gauge/cap pressure has a boiling point of between 125-130C.

[Guest Markino]

Thanks for all the information guys!

At the moment I connect one of the two wire of the ECT to my board based on Arduino Nano and I can read the value in Volts comming from the sensor.... I found a forumula on the web but I'm not sure the output temperature value is correct: the fan motor leaving the bike in idle starts at arount 105 degrees C.

 

Do you know how to translate/calculate the Volts comming from the ECT into a temperature? If the sensor is a variable resistor what is constant? The current putting on the resistor?

[DaveM59 1,608]

You have lost me now when you start calculating resistors and circuit voltages and what's a Nano? 

I just know a bit of theory of how the bike 'thinks'. My X9 also uses a sensor to ECU (they call it that not EMU) for the FI and the ECU controls the fan but the sensor has 2 separate parts inside it, one to feed the ECU and one for the gauge on the dash. They are different type of sensors embedded in one casing and fitted to the water jacket on the cylinder.

 

If you tapped into the NC's sensor, would that effect the output sent to the EMU and cause it to adjust the fuelling incorrectly? The fact that Piaggio use a dual sensor suggests they felt it needed to be kept separate although the two parts operate at totally different voltage/resistance levels, so it could be that the ECU and gauge needed different levels of input. The NC. sensor is a lot more conventional it seems. Beats me why they didn't fit a temp gauge though. It's an important instrument.

Edited April 7, 2016 by DaveM59

[embee 7,288]

I'm not sure how the ECU treats the thermistor resistance, I'd guess possibly the principle is in series with another fixed resistor (in simplest terms) as a voltage divider circuit, but that's a guess. As for calibration, that's entirely up to the ECU engineers, you'll have to find a way to calibrate it yourself. If the voltages you're reading are something maybe in the 0-5V range this is quite likely I'd suggest.

 

Typically you'd use the thermistor in a pot of boiling water at 100C as one "fixed point". If you also do it at ambient, say 20C against a suitable thermometer, it'll give you a good temperature spread and one calibration point is very close to the working temp (around 100C) so accuracy should be good.