Electric Fan Control

I doubt that there is any resistor in series with the fans in the original design. Running the fan off an automotive electrical system you would just choose a fan that had the desired requirements at the supply voltage. But it's an easy check. Just measure the voltage across the fan motor when running. If it's 12 V (more like 14V) or there about, no resistor.

But like I said before, why ,mess with something that has worked well for 20 years. Ok maybe for a race set up but for the street?

 

Yes, this is exactly what BMW E23-s had. All you need is an additional (lower) thermo-switch (or replace the existing one with a dual temp thermo-switch), additional relay and a resistor with a value determined according to how fast you want the fan to run at the low speed (0.5 or 0.6 Ohm in the case of the BMW fan). The important thing is using a robust resistor (the BMW one is ceramic, about 1 in. square, 3 in. long) and it has to be mounted close to the fan blades to get cooling.
Image Unavailable, Please Login

Image Unavailable, Please Login

Image Unavailable, Please Login

 

Before modifying things, are you certain your car is operating optimally? These cars are now 20+ years old and sensors and switches may no longer be operating as intended. I would consider replacing the thermostat, thermoswitch and temp sensor in the valley before doing much else. Even in stop and go traffic on 90+ degree F days I never saw my temp gauge rise beyond 1 o'clock. At round the 11 year mark, I did replace all those parts as "preventive maintenance" on my 97 F355 spider.

 

I think he kniws that he just wants fan or fans to run at half speed to be quiter and also try to keep engine closer to optimum temp.
Fans will likely run closer to continuous rather than all or nothing

 

That being said I like mine the way it is but to each their own

 

If one wants to run both fans at half speed with a reserve of full speed of both, if required, then I think the best way is to go Jaguar (series/parallel) plus a dual temp switch (like the one I mentioned earlier). No resistors, no waste heat i.e. energy efficient. Used jag fan relays are plentiful on Ebay, ~ GBP 10 from the UK, more expensive from the US. The wiring-in would be quite simple.

Image Unavailable, Please Login Image Unavailable, Please Login

 

You don't need any resistors for series/parallel operation of the fans. In series, the fans themselves are resistors to each other. The Jaguar relay (module of relays) design is different from your design so there is no patent infringement:

Image Unavailable, Please Login

Somebody else also designed a similar circuit:

Image Unavailable, Please Login

 

Keep in mind roth it is not very oftain that both fans turn on, and especially at same time. You will obviously have to change how they activate.

 

What I am concerned about is using resistors to slow down the fans. These resistors emit a lot of heat and would significantly contribute to global warming. When the fans are run in series, each runs at half speed and each represents a resistor to the other, i.e. the energy that would be wasted in a resistor (converted to heat) runs a fan! Very environmentally friendly invention by Jaguar (or Ford?). However, Chevrolet LS1 has the same type of twin fans speed control so I don't know who was first.

 

I've been having a think about this problem today and I think that the whole concept of controlling the fan speed more precisely than on the original vehicle is a bit pointless...

Lets think about it all thermodynamically:

It's the engine thermostat that has a higher influence over the engine core temperature than the radiator fans. The engine thermostat controls the engine temperature proportionally by restricting coolant flow.

The radiator fans only INCREASE THE EFFICIENCY of the radiators if there is insufficient natural airflow to maintain an acceptable radiator coolant exit temperature. This happens normally when the car is running at speed and the fans do not operate as the engine thermostat maintains engine temperature control.

I can apply this to the 348 because this is the workshop manual I have in front of me.

On the 348 the radiator coolant exit temperature is maintained at 76 to 85'C by the thermoswitch. (everything I write below is related purely to the 348 but I believe the F355 is very similar as is pretty much any car from the same era).

So the manufacturer has decided and tested that in a credible worst case scenario the engine needs 85'C engine coolant inlet temperature in order to maintain a engine coolant exit temperature below 115'C (the temperature by which the engine temperature high lamp is illuminated by the engine thermoswitch) this would be during low speed start stop driving / reckless burnouts etc in high ambient temperatures.

When the engine thermostat is fully open (96'C) then the fan thermoswitch will come into play if the radiator exit temperature is not maintained below 85'C by natural airflow. Once the radiator exit temperature is brought back into control to 76'C by the fans (which they can do easily in all credible scenarios) then the engine thermostat will take over to maintain the engine core temperature within the acceptable range. It's this relationship between the components that work to ensure the engine temperature is controlled within the acceptable range determined by the manufacturer.

As far as I'm aware the laws of thermodynamics have not changed since the cars were built therefore the acceptable range in the mid 90's will still be acceptable now.

As stated previously by Miro the engine temperature fluctuation is perfectly normal and fine control isn't necessary and is the reason why Joe public have to be 'fooled' by modern electronic engine temperature gauges to combat unnecessary wasted time in false warranty claims due to people over-thinking what an accurate temperature gauge is telling them.

Again on the 348 the oil radiator has its own thermoswitch and is controlled between 80 and 90'C in an effort to control the oil temperature independently of engine core temperature under heavy engine load (in my experience this only comes into play on track). Essentially the oil is cooling the piston crowns with spray jets and dissipating that heat thorough the engine block and oil radiator but under heavy loads needs to be brought in control by the oil radiator fan. It's worth checking this system is working properly for track use and keeping an eye on oil temperature on the gauge. The one time I took my car on track this fan system wasn't working and has probably never operated from new. It limited my time on track to around 15 minute stints before I decided oil temperature was too high.

If the goal of this project is to achieve a more accurate radiator exit temperature than the original specification of 76 to 85'C (+-2'C) then personally I would use a general purpose PT100 thermistor to measure the engine coolant inlet temperature (that coolant exit temperature from both radiators) and use that as an input to an industrial P.I.D. (Proportional, Integral and Derivative calculation) process controller with self adaptive tuning of the PID parameters (think Eurotherm PID process controller) you can then use the output of the PID process controller to control the fan relay and this will then use a very clever and precise PID control loop to maintain a very precise engine coolant inlet temperature taking into account all the thermal time constants of the radiator fans and the engine thermostat response time etc. All this can be set to 80'C and piggy backed on top of the existing system to provide a dual redundant backup in case something in the new PID system fails.

The PID process controller can also be used in conjuction with a PWM fan speed controller as Miro describes if fully proportional fan speed is desired and this would provide a VERY NICE control system that will use only the necessary fan speed to maintain the correct temperature rather than a more crude but perfectly acceptable on-off fan control. (Miro I always read your posts with interest I think they are the most interesting and useful on here .

The reason car manufacturers do not go for this fine temperature control is because it is BEYOND REASON and simply not worth the capital investment cost to achieve a very nice but much higher level of control than what is needed.

SO,

Roth with respect, I guess the point of all this is that if you are concerned the engine temperature is fluctuating an excessive amount then I would look to other possible causes. An airlock in coolant system pipework would cause the symptoms you describe and I would suggest ensuring the system is bled sufficiently, or other possible blockages in the system like silted up radiators, low coolant level, engine thermostat sticking etc. This effectively means that coolant flow is restricted and the engine thermostat cannot do its job and the system is asking too much of the radiator fans. This is a potentially dangerous situation if allowed to continue and will eventually result in an engine over-temperature condition which if the engine over-temperature switch fails or the lamp in the dash is blown then these are the last ditch engine protection measures and the engine could end up being destroyed. I don't believe the Motronic systems on the 348/355 have any kind of limp mode or thermal protection feature to protect the engine unit from over temperature condition or engine temperature sensor implausibility. If you want 100% thermal engine protection then that is another project altogether!

The reason I write this lengthy post (it started short) is that I fear that in trying to achieve one thing, which is in my opinion beyond reason, it could essentially mask a real problem that could risk causing a complete engine meltdown. If the engine temperature gauge is not noticed and the engine over temperature warning indication either fails or is not noticed, then the engine will probably keep running until the block / heads are warped and/or sensors or plug caps start melting off and there is smoke or fire.

If the condition of the coolant system can be confirmed and the engine is operating happily within its operating range then with respect I would leave the radiator fan control system well alone and I honestly wouldn't give it any more thought, but if you do decide to build something for the sheer fun of it then by all means build away let us know the results



Regards,

Dave

 

It all ends up as heat my friend. Yes, fans is series would be more efficient that fans in parallel, each with series resistors. But even with a series resistor a fan with resistor generates less heat than a fan running full speed.

 

I have been thinking about modifying my 550 cooling fan actuation myself. On the 550 there are two sensors which sends signals to a fan control ECU. Which in its turn activate normal relays in the fuse box to start the fans. Both fans always starts and the second fan starts about a second after the first one. If there is a problem with the sensor the second sensor will also activate the relays but at a higher temp. I had to replace both my sensors begause they made the fans start way too late and the car was over normal temps. Now after I did that the car still overheats even if the fan system worls as it should but it only overheats if the outside temp is about 25 degrees Celsius or more and the car is in stop and go traffic or travels at less than 40 kph. which in fact is quite often. When I have driven in really hot temps in France and Italy at around 35-40 degrees Celsius the cooling is not sufficient at all and many other 550 owners experience the same. The cooling fan shroud only covers about 70% of the radiator area but on the 575 the shroud is redesigned and covers the radiator completely.

My odea of modification is to take the power from the relays (and small gauge original wiring that becomes hot) to run a set of extra relays (big for continious operation) and new big gauge wiring directly from the battery via one fuse for each fan. That way, the fans will operate normally, but with two manual switches, I can control the second relays manually to activate one or both relays to add cooling for example when it is obvious that the traffic will clog up or after a very hard sprint or if I suddenly have to stop the engine in special circumstances. The way I want it to work is to have a yellow LED diode on the output pin on each relay to be 100% certain that the fans get power and run. Instead of both fans coming on at times, I can let one fan run all the time and make the coolant be less likely to become really hot. This will strengthen the performance of the aircondition too because the condenser and its fluid will always have a lower temperature. The 550s need AC air even if the outside temp is not hot because the exhaust and general heat from the engine and under the car is significat.... So the colder the AC air is the better. I will also replace the pulley on the waterpump to a smaller one used one the 550 Barchetta which was build later and change to a 575 fan shroud with more efficient 575 fans to further improve the cooling. To add an acoustic warning for over normal temp would be a good thing because if you forget to check the needle then it might be too late. The way my 550 is now, the fans cant get the heat out quick enough but when driving 50 kph or more it always stays cool. The problem is however if I get the car hot, and it is hot outside, while running full AC to keep the cabin cool and standing still, the needle goes alnost to the red and if I then stop it at that temp, the temp continue to rise as the exhaust heats up the heads so the temp goes out of control. But with two manual switches then the temp can be more chill in these conditions as the coolant can faster cool down so when you start up the coolant is much colder. One fan continiously running is better than two (more noisy) going on and off all the time and the needle goes from 75-120 degrees C all the time. You get really stressed about the neddle fluctuation!!


Sent from my iPhone using FerrariChat

 

If you have temp varying between 75 and 120 C you have bigger problems than fans.

 

Yes, a fan (motor) with a resistor running at half speed will generate less heat then the same fan motor running at full speed. But you have to add the heat generated by the resistor to the heat by the fan motor at half speed which, in total, is more than that same fan motor running at full speed. This is because a brushless DC electric motor converts most of the electrical energy to mechanical energy. There will be a small amount of heat generated by the fan motor as it cannot convert 100% of the electrical to mechanical but about 90% so about 10% electrical energy is lost in heat and friction (say all 10% is heat). Not accurate and very simplified, just for crude illustration, a 20A fan motor at full speed will lose 10% of 20A to heat, i.e. 2A is converted to heat (2 "heat Amps"). The same fan motor with a resistor will "try" to pull 20A but it only gets 10A as the resistor "eats" 10A by converting it to heat. Of the 10A that reaches the fan motor, 10% is lost to heat so 1A is converted to heat in the motor. Resistor 10 "heat Amps" plus motor 1 "heat Amp" = 11 "heat Amps" as compared to 2 "heat Amps" of the motor at full speed. And you get full speed whereas the other case gives you only half speed. I am sure there are experts amongst us in this field who can provide more accurate calculations than my crude presentation.

 

The 355's operation is somewhat different as the operation of its two radiator fans is staggered - only one fan comes on at a lower temp point (switched on by a thermo-switch on the radiator) the other fan comes on at a higher temp point, as signalled by the ECU.

In the case of the 348 (which I have) it is simpler - both radiator fans come on simultaneously, controlled by the thermo-switch on the radiator. When I bought my 348, the first two things I noticed and did not like were:

1. The original thermo-switch has a too wide on-off range (85C-76C as you mentioned). I was wondering why it takes ages for my fans to go off, i.e. run until the temp gauge needle is unnecessarily way down. When I got a new thermo-switch under the relevant part number from Eurospares, I noticed that it is 88C on, 82C off. Much better now, the fans cycle as one would expect. Looks like Ferrari have changed their mind here and the original "wide range" thermo-switches have been discontinued.

2. Both fans coming on at the same time, not a good idea. Huge load spike onto the alternator on start-up, probably some 100-120 Amps (a 20A fan will draw up to 3 times that, for a short period of time, on start-up). This is probably what used to kill early alternators. I installed "soft start" controllers at each fan so they now slowly ramp up to full speed (over a few seconds) when switched on by the thermo-switch avoiding the "big bang" of Amps. Another way to avoid the high momentary load would be to install a timer which would switch the second fan a few seconds after the first.

 

One switch two fans.

 

But that wasn't the point I was making. Ultimately, the sum of all the energy put into the system to spin the fan and pump the air ends up as heat. But yes, you are correct. I had a brain fart. Heat dissipated in the motor is I^2 x Rdc where Rdc is the DC resistance the motor which is much less than the resistance at some RPM. Assuming that 1/2 speed requires 1/2 current, then the series resistance would have to equal the motor resistance at 1/2 speed, which would still be much more than the DC resistance.

 

I think this needs to be my new sig