Using PCM to control aftermarket electric radiator fan?

[pagrey]

No SPAL that fits a TJ moves a fraction of the air. This is basic HVAC stuff.

 

[pagrey]

There is a simple reason that in 20 years there isn't a recommendation for an electric fan that fits a TJ, there isn't one. Get over it.

 

[SkylinesSuck]

Pagrey, shut up and move on. Go regurgitate your internet "knowledge" somewhere else. You constantly spew your (often incorrect) opinions all over the board and present them as obvious fact when they are anything but. We were specifically discussing potential switchable outputs from the factory PCM, not looking for your thoughts on electric vs mechanical fans. Not only are you wrong, but also off topic. This is basic forum stuff.

Steel City-- disappointing results but thanks for trying and sharing. I'm curious if the PCM fixer guy has any input here (don't remember his handle at the moment). I've asked the guy I'm trying to have tune my PCM about it several times but he hasn't spoken to the subject. Maybe he had the same results as you?

 

[Steel City 06]

It is possible I am wrong at my guess that it would be on connector C3. Looking at the VCM Scanner, it is clear the PCM is trying to set a fan speed. It will actually constantly adjust a target fan percentage. I just need to find what it is trying to set. Maybe I will repeat the test on C1 and C2.

The more I think about it, the more I suspect there has to be an output somewhere. Why would the PCM be programmed to set a fan speed if it had nowhere to set it?

 

[eastbloc]

Take my suggestion as from someone who only had little exposure to electronics in computer org class.

In case it is a feature without an output, which would be puzzling since scanner being able to read is already a response to an output, look where does it get parameters for the input. Something is sending a signal to it, and as a result it is responding to it by adjusting target fan percentage.

In the end, PCM is a middle guy that takes data in and sends data out. I am sure that PCM can be bypassed by a knowledgeable one.
I have spoken to guys studying electrical engineering and robotics when I was in college, there are so many smart controllers out there for pretty much everything. For someone knowledgeable in this field this should not be an issue.

 

[Steel City 06]

Take my suggestion as from someone who only had little exposure to electronics in computer org class.

In case it is a feature without an output, which would be puzzling since scanner being able to read is already a response to an output, look where does it get parameters for the input. Something is sending a signal to it, and as a result it is responding to it by adjusting target fan percentage.

In the end, PCM is a middle guy that takes data in and sends data out. I am sure that PCM can be bypassed by a knowledgeable one.
I have spoken to guys studying electrical engineering and robotics when I was in college, there are so many smart controllers out there for pretty much everything. For someone knowledgeable in this field this should not be an issue.

I agree with this. SPAL does in fact sell PWM controllers for their fans. However, they operate solely based on coolant temperature, and scale fan shaft speed linearly with temperature.

I could easily build a system using an Arduino or similar to drive the fan based on multiple inputs, in this case vehicle speed, coolant temperature, A/C high side pressure, etc.

However, if such a system already is built into the PCM, it seems like a no-brainer to use it. Plus, I would expect the PCM to be far more reliable than an Arduino or other PLC.

At a certain point, it could even become counterproductive if I spend too much engineering a system to run an electric fan. The ultimate goal is to increase fuel economy (and thus range) and simultaneously increasing power output. At a certain point, there comes better ways to achieve that.

Though I would strongly be tempted to play with an Arduino and make something work for the fun of it.

 

[Apparition]

Has anyone asked Mark @Wranglerfix ?

 

[AMS417]

I agree with this. SPAL does in fact sell PWM controllers for their fans. However, they operate solely based on coolant temperature, and scale fan shaft speed linearly with temperature.

I could easily build a system using an Arduino or similar to drive the fan based on multiple inputs, in this case vehicle speed, coolant temperature, A/C high side pressure, etc.

However, if such a system already is built into the PCM, it seems like a no-brainer to use it. Plus, I would expect the PCM to be far more reliable than an Arduino or other PLC.

At a certain point, it could even become counterproductive if I spend too much engineering a system to run an electric fan. The ultimate goal is to increase fuel economy (and thus range) and simultaneously increasing power output. At a certain point, there comes better ways to achieve that.

Though I would strongly be tempted to play with an Arduino and make something work for the fun of it.

If you find an electric fan that can provide enough CFM and cooling without destroying or requiring an upgrade to the stock alternator I’m interested. So far nothing fits the bill. I’m considering going back to engine driven next summer. Can’t run my A/C on the trails without getting hot.

 

[Steel City 06]

If you find an electric fan that can provide enough CFM and cooling without destroying or requiring an upgrade to the stock alternator I’m interested. So far nothing fits the bill. I’m considering going back to engine driven next summer. Can’t run my A/C on the trails without getting hot.

SPAL makes a few fans that might fit the bill.

In their brushless line, the SPAL 30107101 is the most powerful offering. This fan is designed to pull up to 2400 CFM, and is also designed to pull air through a significant number of heat exchanger rows. Unlike the cheaper brands, it is very good at pulling air in spite of the pressure drop across the radiator. It pulls about 32 amps at top speed, and requires a 50 amp fuse. The 30107090 is a toned down version of this fan.

In the brushed fan line, the 30102803 is the most powerful 16” fan. It’s only rated to around 2,000 CFM under most conditions, but still can pull very near that even with multiple radiator cores. It also consumes right around 31 amps at full power. The 30102802 is also rated to around 2000 CFM, but the airflow drops off a lot more quickly with increasing head than the 30102803.

Note that while the brushed fans are cheaper, it may actually be cheaper overall to go with a brushless fan. The brushed fans are just a simple fan and brushed motor. If they have power, they run. They have no speed control unless you apply amplified PWM to the power terminals. Meaning you have to spend a large amount of money to get a controller capable of supplying 30+ amps of 12V PWM to the power wire, plus a driver to drive the controller. This generally does not make economic sense.

The brushless fans, unlike the brushed fans, have internal relays and internal speed controls. Simply hook the positive and ground to e battery, and feed it either an analog or digital (PWM) signal of a few milliamperes. It does the rest for you. So if the PCM is capable of PWM, literally all you do is wire the fan straight to the PCM. If not, you can drive it by any number of means.

Just as important as the fan is the shroud that you run. A crappy shroud will make even the best fan underperform. The shroud must have a shape that allows the fan to pull air across as much of the surface as possible, and allow air to bypass the fan at highway speeds. Flaps are a great addition to low-profile shrouds.

If you truly need the highest CFM, a clutch fan may be more appropriate. The explorer 11-blade and heavy duty clutch swap is a great place to start.

 

[Steel City 06]

In regards to the Explorer fan I mentioned, here are the details I wrote for another thread:

The explorer 11-blade fan and HD fan clutch actually seems to have dropped the A/C temperature noticeably, particularly at the highest blower motor speeds when the vehicle is idling or moving at low speeds. The change isn't very large, but I can stay cooler on warm days. The hotter the outside temperature, the more noticeable the difference.

By increasing the airflow across the condenser, you effectively increase the amount of heat rejection. Assuming the A/C was already working properly, you are not going to condense any additional refrigerant. The condensed refrigerant should already be flowing out as a liquid. But the additional airflow provides an additional amount of subcooling, lowering the condensed refrigerant's temperature before it enters the orifice tube, resulting in a colder low side/evaporator temperature. The effect will be most noticeable on the hottest days.

It is possible to overcool the condenser too far, to the point that high side system pressures drop and less fluid is forced through the orifice tube. But in normal operating conditions this will not be a problem.

To do this mild upgrade you need two parts:

Amazon.com: Hayden Automotive 2794 Premium Fan Clutch: Automotive​

Buy Hayden Automotive 2794 Premium Fan Clutch: Clutches - Amazon.com ✓ FREE DELIVERY possible on eligible purchases
www.amazon.com

F87Z-8600-EA - Genuine Ford Base No. #8600 Fan Assembly​

Ford parts #F87Z8600EA, Base parts number #8600, $42.29 online at FordPartsGiant.com. Wholesale-priced and fast-delivered Ford parts Fan Assembly.

www.fordpartsgiant.com

The fan and clutch are based off the Ford Explorer and directly bolt in. You may need to pick up four small metric bolts for the fan clutch to fan attachment. If you cross reference the parts, specifically look for the 11-blade fan for best performance.

 

[pagrey]

Pagrey, shut up and move on. Go regurgitate your internet "knowledge" somewhere else. You constantly spew your (often incorrect) opinions all over the board and present them as obvious fact when they are anything but.

I've done the math on airflow, you haven't. You also probably don't have a clue about the J1850 bus or our PCM but your posts still seem super helpful. Nobody needs you to manage this site or your insight on my posts. Go on a personal rant, might need a keyboard timeout, seems like a good life lesson.

 

[AMS417]

In regards to the Explorer fan I mentioned, here are the details I wrote for another thread:

The explorer 11-blade fan and HD fan clutch actually seems to have dropped the A/C temperature noticeably, particularly at the highest blower motor speeds when the vehicle is idling or moving at low speeds. The change isn't very large, but I can stay cooler on warm days. The hotter the outside temperature, the more noticeable the difference.

By increasing the airflow across the condenser, you effectively increase the amount of heat rejection. Assuming the A/C was already working properly, you are not going to condense any additional refrigerant. The condensed refrigerant should already be flowing out as a liquid. But the additional airflow provides an additional amount of subcooling, lowering the condensed refrigerant's temperature before it enters the orifice tube, resulting in a colder low side/evaporator temperature. The effect will be most noticeable on the hottest days.

It is possible to overcool the condenser too far, to the point that high side system pressures drop and less fluid is forced through the orifice tube. But in normal operating conditions this will not be a problem.

To do this mild upgrade you need two parts:

Amazon.com: Hayden Automotive 2794 Premium Fan Clutch: Automotive​

Buy Hayden Automotive 2794 Premium Fan Clutch: Clutches - Amazon.com ✓ FREE DELIVERY possible on eligible purchases
www.amazon.com

F87Z-8600-EA - Genuine Ford Base No. #8600 Fan Assembly​

Ford parts #F87Z8600EA, Base parts number #8600, $42.29 online at FordPartsGiant.com. Wholesale-priced and fast-delivered Ford parts Fan Assembly.
View attachment 278213 www.fordpartsgiant.com

The fan and clutch are based off the Ford Explorer and directly bolt in. You may need to pick up four small metric bolts for the fan clutch to fan attachment. If you cross reference the parts, specifically look for the 11-blade fan for best performance.

Thanks, probably what I will do. My current fan (Mishimoto) pulls enough amps it sets my check gauges light on occasion due to low RPM / low amp production in the stock alternator. Alternator is new. Engine is supercharged with a header so lots of heat. Also have a aux trans cooler in front of the condenser with PSC cooler mounted in the fender vent. Overheats with AC on at low speed while climbing. Runs 225 or more at 75mph on the highway. Also need to put a 195 t stat back in, PO said he installed a 165. Has yet to boil over but it runs above the 215 I am familiar with.

 

[TurboTJ]

In addition, I've seen quite a few internet posts claiming that a mechanically-driven fan is 100% efficient in its power conversion. There are two problems with this logic:

1. For a fan driven with a viscous clutch, this is simply not true. A significant portion of the power taken off the water pump shaft is lost as heat. More details later.
2. For a fan solidly connected to the water pump shaft, this is technically true. However, as per the fan affinity laws, the fan requires a stupid amount of power to run at high RPMs. More on this later.

Regarding Problem #1, a fan clutch is designed to slip both when locked and unlocked. According to Newton's 3rd law, the torque on the fan clutch at the water pump end must equal the torque at the fan end. (Otherwise the fan clutch would start rapidly spinning until it explodes.) However, we are concerned with power, and not torque. Power is proportional to the rotational speed of a shaft multiplied by the torque on the shaft.

A typical fan clutch states that it slips 20% when unlocked and 80% when locked. I've never seen anyone state at which engine speed this occurs at, so I'll take the conservative approach and assume this occurs at high idle (1000 RPM). For simplicity for this example, I'm also going to assume that water pump RPM equals engine RPM. If you change this RPM value based on a different pulley ratio, you will note that it does not affect the rest of the calculations. In fact, you could substitute in any variable of your choice and it will ultimately not change the final answer of the following calculations.

So how much power does an unlocked fan + clutch pull at 1000 RPM? We'll have to make a few assumptions here. Imagine if you were to take a typical household box fan and duct tape it behind the radiator. Due to the restrictions of the radiator, condenser, and grille, it would have to be roughly at 50W power consumption. 50 watts isn't much in the realm of things. Way lower than our electric fan at full power. But in order to spin this fan at the unlocked speed, the clutch slips to 20% of shaft speed. Torque on both ends of the clutch is the same. But power is proportional to torque times RPM. So where does the extra power off the water pump go? 80% of the energy leaving the water pump is lost as heat in the fan clutch! This means we are using 250 watts, 200 of which is being wasted as heat! That's 1/3 of a horsepower we are using to drive this fan and clutch.

But what if the fan locks up? Surely it becomes more efficient! Assuming the 20%/80% rules apply here, the fan spins at 4X the speed when the clutch locks up. According to the fan affinity laws, power scales with the cube of shaft RPM. Head increases with the square of shaft RPM. Airflow scales linearly with the shaft RPM. The head assumption is perfectly valid in this case, as the force required to pull air through the radiator increases with the square of the airspeed as per the drag equations. But in order to spin our fan at 4X the speed, we need to provide 64X more power! Hence our fan is suddenly demanding 3,200 watts, or roughly 4.3 horsepower! In addition, the fan clutch is still burning up 20% of the energy as heat, while passing 80%. So our fan + clutch combination is now burning 4,000 watts (5.4 HP) of engine power! And we have 800 watts being converted straight to heat! (Ever wonder why fan clutches have so many heat fins?)

So what happens when we start revving our engine higher? If it were a mechanically locked fan, power would increase by the cube of the input RPM, and we would be burning an insane amount of horsepower at redline! Fortunately the mechanical clutch helps us out here, but perhaps by not as much as we might like.

The torque transmitted by a fluid coupling is proportional to the square of the slip RPM. Hence, the faster it slips, the more torque it can transmit. Or in other words, as torque increases, slip RPM increases by the square root of the applied torque. Given the fan affinity laws, the torque on the impeller shaft is proportional to the square of the RPM driving it. So (to my actual surprise), it turns out that the slip ratio between input and output speed is actually roughly constant!

So what happens when we double engine RPM to 2000? Our fan/clutch combinations are now taking 8X as much power! Our unlocked fan is now taking 2000 watts (2.7 HP) of power, and our locked fan is taking 32,000 watts (43.2 HP)!

That said, few fan clutches will remain locked past a few thousand engine RPM. In addition, as the fan clutch oil heats up, the fan begins to lose viscosity, and the fan slip ratio actually increases. Realistically, you will never see a clutch fan take more than about 40 HP off the engine when locked at high RPM.

But our unlocked fan still takes a lot of power. Even if we assume the slip ratio doubles to 90% (10% shaft speed), at redline (5300 RPM), that fan is pulling 4,650 watts, or 6.2 horsepower! 90% of that loss immediately becomes heat! And if somehow it is still locked, you could see a loss as high as 40 horsepower! (These numbers line up well with a lot of dyno results you can find on various racing forums.)

So there are really two things that prevent the engine clutch fan from being efficient mechanically. (1), a large amount of power is converted to heat in the clutch. You've probably noticed that a fan clutch has many heat fins. All that heat is wasted energy. (2), with a lack of control over the input shaft RPM, the power demands increase rapidly and unnecessarily. The fan has to be sized to cool at idle, and as such at high RPM it is way overpowered.

Meanwhile, the brushless electric fan always has a loss coefficient of about 50% (roughly 60% alternator efficiency multiplied by the 85% efficiency ratio of the electric motor), and only uses the power it needs to when it needs to.


TL;DR Fan clutches are great and save us a lot of fuel over directly driven fans, but they are still very inefficient and a well set-up electric fan will absolutely dominate the clutch fan in terms of power (and as a result, fuel) conservation.

Perhaps your mechanical fan uses no power at all when you are at speed. The radiator probably slows the airspeed down somewhat, but the 103 ft/sec hurricane blowing through it at 70 mph may be enough to windmill the fan. It may actually be assisting the engine, not that this is free because of the corresponding aerodynamic drag.

 

[Wildman]

For us 4 cylinder folks the Ford Taurus 2 speed fan was the go to CHEAP option. It freed up some engine power by getting rid of the mechanical fan plus it gave tons of room between the engine & radiator.

Steel City,
I have the SPAL PWM and I REALLY like the fact I can turn my fan on at 50% speed at a lower temp and then at 100% when it hit a set temp.

I wonder if it depends on the year of the rig/PCM as to if it has fan outputs? Sort of the same thing with like the O/D light on the dash but if you don't have a PCM from a year that had a auto tranny with O/D can you turn it on?
I'm just shooting from the hip here.

 

[freedom_in_4low]

Love this thread.

How do we know the airflow of the stock mechanical fan?

Just curious, but what did you check for on the pcm pins?

I haven't messed with arduinos before, but would it be possible to read the fan speed % off the bus and produce a 0-10V or PWM signal from that? Then you could at least use the on board logic from the pcm and just use the arduino as a translator.

 

[TurboTJ]

Love this thread.

How do we know the airflow of the stock mechanical fan?

Just curious, but what did you check for on the pcm pins?

I haven't messed with arduinos before, but would it be possible to read the fan speed % off the bus and produce a 0-10V or PWM signal from that? Then you could at least use the on board logic from the pcm and just use the arduino as a translator.

I was thinking of something similar. Use some sort of magnetic or optical sensor to measure the fan RPM. If the fan RPM is less than the engine RPM the engine is driving the fan. If the fan RPM is greater than the engine RPM the fan is driving the engine.

 

[Steel City 06]

Love this thread.

How do we know the airflow of the stock mechanical fan?

Just curious, but what did you check for on the pcm pins?

I haven't messed with arduinos before, but would it be possible to read the fan speed % off the bus and produce a 0-10V or PWM signal from that? Then you could at least use the on board logic from the pcm and just use the arduino as a translator.

CFM estimates for a clutch fan are very rough, as it very much depends upon engine RPM. I'm primarily comparing to comparable fans used in sports cars and classic car retrofits. They can be anywhere from 1000 up to 5000 cfm depending upon fan design and driven RPM. The Explorer fan is most likely at the top end, but the TJ stock fan is likely half that.

When I checked for relay operation, I checked for a ground or other changing voltage at every empty hole on the C3 connector. First, I measured DC voltage between the battery ground and pin. The voltage would drop to zero if it was grounded. I confirmed all of those by comparing it to the positive battery terminal.

When I checked for PWM, I checked each open pin relative to both positive and ground for an AC voltage, Frequency, and Duty Cycle. I actually did pick up a rare frequency right at 60Hz (which is clearly suspicious), but I was able to replicate it with a wire hanging in free air (it was just noise).

I idled the engine and used a piece of cardboard to intermittently block the radiator to cycle the coolant temperature through each fan state bin. The changing temperature clearly resulted in a different PCM target fan state per VCM Scanner, but I could find no corresponding change in any of the pins.

Each set of tests was done with a different fan mode enabled in the PCM software. Different modes did in fact have different target fan speed percentages, so the PCM is calculating something.

Note that I only tested the C3 connector.

 

[Steel City 06]

As to the Arduino sampling from the PCM, sure that is absolutely possible. But regardless of whether or not the Arduino is the chosen method, I need to figure out the output can actually be pulled from the PCM. As of right now I have no idea how to do that. Obviously it is possible in some method, given that I'm actively reading it with VCM scanner.

With a standalone Arduino I could sample directly from the coolant temperature sensor, the A/C pressure sensor, the vehicle speed sensor, and any other sensor I want and build whatever program I want to control a fan 0 to 100%. That would obviously be the most efficient solution, but that would take significantly more work to complete.

 

[dbbd1]

In regards to the Explorer fan I mentioned, here are the details I wrote for another thread:

The explorer 11-blade fan and HD fan clutch actually seems to have dropped the A/C temperature noticeably, particularly at the highest blower motor speeds when the vehicle is idling or moving at low speeds. The change isn't very large, but I can stay cooler on warm days. The hotter the outside temperature, the more noticeable the difference.

By increasing the airflow across the condenser, you effectively increase the amount of heat rejection. Assuming the A/C was already working properly, you are not going to condense any additional refrigerant. The condensed refrigerant should already be flowing out as a liquid. But the additional airflow provides an additional amount of subcooling, lowering the condensed refrigerant's temperature before it enters the orifice tube, resulting in a colder low side/evaporator temperature. The effect will be most noticeable on the hottest days.

It is possible to overcool the condenser too far, to the point that high side system pressures drop and less fluid is forced through the orifice tube. But in normal operating conditions this will not be a problem.

To do this mild upgrade you need two parts:

Amazon.com: Hayden Automotive 2794 Premium Fan Clutch: Automotive​

Buy Hayden Automotive 2794 Premium Fan Clutch: Clutches - Amazon.com ✓ FREE DELIVERY possible on eligible purchases
www.amazon.com

F87Z-8600-EA - Genuine Ford Base No. #8600 Fan Assembly​

Ford parts #F87Z8600EA, Base parts number #8600, $42.29 online at FordPartsGiant.com. Wholesale-priced and fast-delivered Ford parts Fan Assembly.
View attachment 278213 www.fordpartsgiant.com

The fan and clutch are based off the Ford Explorer and directly bolt in. You may need to pick up four small metric bolts for the fan clutch to fan attachment. If you cross reference the parts, specifically look for the 11-blade fan for best performance.

Would you consider this as a cooling system upgrade or an A/C system upgrade or neither?

 

[Steel City 06]

Would you consider this as a cooling system upgrade or an A/C system upgrade or neither

Technically both. I didn't need it from a cooling system perspective, but it absolutely helps that as well.