Z3 2.3 fan delete report with some data

[Schreier]

One of the biggest tips I have heard concerning the Z3 besides minding the rear subframe, is to delete the primary fan that is driven off the water pump. The primary concern is that the blades of the fan will, on a long enough timeline, develop fractures and explode, potentially taking any number of essential engine cooling components with it.

Since eventually I would like to put the Z3 through its paces on a dyno, and generally would like to keep catastrophic incidents to a minimum, the fan delete seemed like a simple and cost-effective modification with little downside. I grant that a new, factory fan and fan clutch would likely keep the risk to a minimum, but even that minimum of risk and investment seems to me worth less than simply deleting the fan and getting on with enjoying the car.

Having said that, I'm very much in favor of going about things as scientifically as possible. There are many posts of fan deletes going very well, and a few posts reporting some overheating issues. So to study this the best I could, I bought an OBD Link LX and the Torque Pro app for my old Android phone, in order to view cooling info in real time wile driving and to log data for later analysis. Given the weaknesses in many BMW cooling systems, it could be a prudent investment even if you're not considering the fan delete.

My car is a 2000 Z3 2.3 roadster, which of course has the M52TUB25. A word of caution: this is not a full-blown study on the issue with many samples. I can only speak for my own experiences, so make of that what you will, but I can not take responsibility for whatever you decide to do with your own vehicle. But for what it's worth, here's what I found.

Prologue: Fan and fan clutch removal

This was slightly more difficult than advertised, but I assume only because it was my first time doing a fan delete, and probably the first time the fan had been removed from the car. The nut holding the fan clutch to the water pump was quite rusty, but the screw/nut on the water pump was pristine. For the task, I used this 32mm fan clutch removal kit, and a hammer. And this YouTube video of the process was extremely helpful (as was the Z3 forum you are now reading).

The fan clutch removal kit doesn't actually fit over any of two bolts on the water pump, no matter which way you turn. Rather, you hook one bolt with the far end of the pry bar, and rest the prybar over the adjacent bolt. In this fashion, you can prevent the water pump from rotating while you work on getting the fan clutch off. You'll know you got a good fit when you can get the water pump to rotate just ever so slightly by applying pressure to the pry bar.

If you're like me, you're likely to find this arrangement rather difficult. The bar and the bolt on which it is resting are quite narrow compared to the force required to move the fan clutch nut. It might slip frequently until you find the best position for everything. But the thing that really helps the pry bar hold fast is when you attach the wrench to the fan nut. You can sandwich the pry bar between the wrench and the water pump to hold everything in place while you work.

Several whacks with the hammer on the handle of the wrench should get the fan clutch nut loose. You'll know success when the wrench drops a bit. Remember that this is a counter-rotating thread, so it's right to loosen. So you will swing your hammer from left to right to hit the wrench appropriately.

With the fan clutch loosened, use the wrench and then hands to loosen it the rest of the way, and it will drop off the water pump. Screw it back on just a little bit to hold it in place, while you remove the fan shroud. The fan shroud is only held on by two plastic expanding rivets, which are at the top and either side of the tank. To remove the plastic rivets, use small needle-nose pliers to work the center cap out. The cap and shaft of the plastic rivet will come out together. Sometimes the back of the plastic rivet will come out as well, while other times it will stay in place. Either way, gently remove both those items. Pull up on the fan shroud to take the shroud and the fan and fan clutch out all at once.

I decided against re-attaching the fan shroud. Its design is to optimize the air flow that the fan is pushing. But without the fan, I'm not sure that it serves any purpose.

Here's what the fan looked like, in case you were wondering. I'm fairly certain it's original to the car and has never been replaced.

20170103_210810.jpg
20170103_210816.jpg
20170103_210827.jpg
20170103_210837.jpg

[Schreier]

Pre-fan delete data


I used Torque to log data from the Z3's OBD port before I removed the fan to obtain a baseline. In a previous test where I idled the engine until warm, I saw the coolant temp plateau at around 197F, while the ambient temperature was around 45F. Unfortunately I didn't log that data, and I don't know long it was sitting before I hit the plateau. I did, however, log a trip where the car was idling stationary for the first 15 minutes or so, at which point the coolant reached 188F and I took off driving in 38F ambient air. Here's the graphed data from that trip.

Stock 2000 BMW Z3 2.3 trip at 38F.jpg


As you can see, the temp hovers near, but rarely exceeds, 200F no matter the conditions.


Post-fan delete idling


After removing the fan, my first test was to idle the car to see when and where temps would stabilize. Ambient air temperature was much warmer on this day, so this test was done in 62F ambient air.

fan delete warm up.jpg
I didn't notice a substantial difference in warm-up times or temps, or the maximum coolant temp, even with the 62F air. In 15 minutes of idling, the coolant temp was 194F. So at 15 minutes of idling, the fan-delete is 6F (3.3C) warmer despite ambient air temps that were 24F (13.3C) warmer.


Do also notice the very end of the data, however. At 15 minutes, I got bored and figured I had collected enough data, but just as I was preparing to shut down the car, the temp increased past 200F.

stock v fan delete.jpg
So, in a side-by side comparison, how do these idling temp curves look? They bot are plotted here, and have been fitted with power function trend-lines. Functions are y1 = 20.282*x^0.3158 and y2 = 23.729*x^0.298 for the stock and fan-delete trend-lines, respectively. This means the fan-delete warmed more aggressively. The fan delete trend-line had a higher R^2 than did the fan-delete, meaning the fan-delete better matched the characteristics of a power function. However both have generally high R-squared values.


Post-fan delete cruising


It's useful to know if the car is going to overheat by simply idling in a garage, but it's not the best real-world example of what the car might do in traffic. Therefore, I took another trip and logged data along the way. Here's what the trip looked like without the fan.

bmw z3 fan delete trip.jpg
Yep, pretty much the same as with the fan, despite the warmer ambient air temps. The car really likes to be at 200F. But also note the substantially higher intake temps.


The engine seemed to be "freer," as well, without having to rotate the weight of the fan and fan clutch. Sound is a very subjective thing, but I notice more satisfying engine sounds without the whoosh and whir of the fan muddying things. I also have to wonder if deleting the fan and fan clutch puts less strain on the water pump, potentially prolonging its life.


Afterward


Something that I did not log that, in hindsight, I really should have, was log the temps just as I was pulling back into the garage. During this whole experiment, I was not aware if the auxiliary fan was functioning properly. I knew when I first received the car a month ago that the aux fan would work if the air conditioning was activated. However, I didn't know whether it would be activated through temperature, and I wanted to find out once and for all.


I pulled into the garage, and let the car sit and run. It took maybe five minutes for the temp to get up to 210F, and the aux fan would not turn on. I take this to mean that the temp switch for the aux fan is dead. I checked the wiring on the temp switch, and it is wired correctly (with the brown wire in the middle). Turning the A/C on immediately turned the aux fan on and got the temperature to around 194F in short order.


It's nice to know I do have the aux fan available when I need it, but a little frustrating that it can't turn on without also turning on the A/C compressor.


Takeaways


The fan delete seems like a harmless mod that at best will help prevent you from catastrophe, and at worse will bump your coolant temp up a couple of degrees in certain conditions. I would not recommend this modification unless you have some means of digitally monitoring coolant temperatures besides using the factory gauges, or if you don't have a reliable aux fan. Having an accurate temp reading on these cars is just a wise thing to do regardless.


Long-term, I plan to keep the fan deleted and attach one of those nifty nuts to make sure the bolt off the water pump isn't naked. I also plan to get the aux fan temp switch replaced. But for now, I am a very satisfied customer.

[Vintage42]

... let the car sit and run. It took maybe five minutes for the temp to get up to 210F, and the aux fan would not turn on. I take this to mean that the temp switch for the aux fan is dead. I checked the wiring on the temp switch, and it is wired correctly (with the brown wire in the middle). Turning the A/C on immediately turned the aux fan on and got the temperature to around 194F in short order...

Why not wait longer than five minutes and let the temperature go 5 or 10 degrees higher? That would not be overheating, and might be the point where your temp switch turns the fan on.
My 1.9 only has one fan, electric, and it does come on at 210F.

[Schreier]

Why not wait longer than five minutes and let the temperature go 5 or 10 degrees higher? That would not be overheating, and might be the point where your temp switch turns the fan on.
My 1.9 only has one fan, electric, and it does come on at 210F.

My understanding was the aux fan is meant to turn to its low setting at 91C/196F, and its high setting at 99C/210F, so I thought I was risking the car by going past 210F. But I'll give 215F a try and see what happens. Thank you for the info here.

[rf900rkw]

There is a reason your temp is so steady. The sensor on an M52TU does not measure the engine coolant temp coming out of the engine, it measures the temp going in. The computer influenced thermostat mixes the incoming coolant and the sensor checks the resulting temp. Closed loop control. This is fine as long as you understand that it is different than all of the earlier Z3s and you can not compare numbers with the early M52 or the 4cyl M44 owners that invariably jump into these threads.

On the fan. The fan cools the radiator. It does not respond to engine temp. At all. The sensor is in the radiator. It should trigger at 91C, which is below the nominal thermostat target temp of 92C. Barely. At no load idle, engine temp will hold steady as long as the radiator effluent is below, say, 90C. Thus, in this case we can infer that the engine inlet temp climbing to 99C means the fan should have come on. Inoperative fans can be a failed switch, but most often is caused by low coolant flow at idle, often an air pocket. A good first test is the heater. Does it continue to blow hot at idle, or does it cool down? If it wimps, it surely is an air pocket in the head.

[Vintage42]

... I used Torque to log data from the Z3's OBD port...

... My 1.9 only has one fan, electric, and it does come on at 210F.

My understanding was the aux fan is meant to turn to its low setting at 91C/196F, and its high setting at 99C/210F, so I thought I was risking the car by going past 210F. But I'll give 215F a try...

... you can not compare numbers with the... 4cyl M44 owners that invariably jump into these threads...

Can we not compare the temperature display that we both get from our OBD ports? That display may not be engine inlet temperature from the radiator, or engine outlet temperature to the radiator, but surely it is coolant temperature at some same place in the engine for both of us. Call it OBD temperature.
My electric fan comes on at 210F OBD temperature, and I would be curious to know at what OBD temperature his fan comes on.

[rf900rkw]

His thermostat is 92C, yours is 95C

His fan switch is 91/99C, yours is 80/88C.

His waterpump push the coolant into the back of the head, the block sees no forced flow. Yours pushes the coolant into the block and bleeds it to the head.

His temp sensor measures the coolant before entering the engine, yours measures after it has been through the head.


No you can not compare.

- - - Updated - - -

Further your fan does not come on at a set engine temp. It comes on when the radiator reaches 80C. The engine temp has NO bearing on this, other than it has to be above 80C.

[cngizbleevng]

Don't know if this is helpful for everyone, but there are ways of displaying coolant temp on your gauge cluster, at least for some of our cars.

http://www.bimmerforums.com/forum/sh...ghlight=hidden

[rf900rkw]

Here is a video I did a while back that shows the relationship of engine temp, radiator temp, and fan speed.






The car is a supercharged E39 M62TU. This makes a nice demo because it has both an engine outlet and a radiator outlet sensor, as well as a computer control of the fan. Before you freak, the stock thermostat for the M62 is 105C.

[Schreier]

There is a reason your temp is so steady. The sensor on an M52TU does not measure the engine coolant temp coming out of the engine, it measures the temp going in.

Ah, thank you, Randy. Also the video is super handy. I can see what you mean. I didn't know where in the system the temp was being taken. This is good to know. Would it be wise to invest in a radiator outlet sensor and track it?

... we can infer that the engine inlet temp climbing to 99C means the fan should have come on. Inoperative fans can be a failed switch, but most often is caused by low coolant flow at idle, often an air pocket. A good first test is the heater. Does it continue to blow hot at idle, or does it cool down? If it wimps, it surely is an air pocket in the head.

I'll do the heater test at idle and get back with some results. If this does suggest an air pocket, should I follow your advice on your post here to bleed the system?

With a cold engine, remove the cap, fill the system, and squeeze the hoses. Keep filling and squeezing. You have to fill the entire block... before you start the first time.

One last question; where should the tip of the coolant reservoir "bobber" be?

[rf900rkw]

I should probably mention that video was done on a 100 degree afternoon right after a solid romp around the block. The engine is happily gulping coolish coolant from the radiator, replacing it with hot stuff. As the radiator runs out of cooled coolant, the fan is ramped in quickly pulling the temp back down. Since the fan in those cars is quite powerful, and the computer programming is expecting the fan to be auxiliary, it tends to be a bit behind and and overshoots.

The E46 version of our engines has the same PWM feature. "One of these days" (tm) I'm going to implement it on the wife's 3.0.


As far as coolant level, I float the top of the stick just above the neck, per BMW, on customer cars, and closer to the bottom of the neck on my own.

[cngizbleevng]

I float the top of the stick just above the neck, per BMW, on customer cars, and closer to the bottom of the neck on my own.

Is there an advantage to the lower coolant level you prefer?

[Vintage42]

Thread with answers in Post #6 and 8:
http://www.bimmerforums.com/forum/sh...-Coolant-level

[Schreier]

Thanks for hunting down that post, Vintage. For ease of use, here's posts #6 and #8 quoted below:

The cold line is a maximum. The siphon pipe from the top of the radiator goes to the bottom of the expansion tank. The coolant level must stay above that pipe opening or you'll back air into the head. As long as there is enough coolant in the tank to keep the pipe covered, you're good. I run below the COLD line on my personal cars, and right at it on customer cars. Never above.

If you overfill and let it blow out the excess, that excess will be anything over 2bar, 30psi. The next heat cycle, it will pull back to the 2.0b. When properly filled, the system runs at 1.2bar or so, depending on the model/thermostat. If you underfill slightly, you can get a S52 down to 1.0b.

[rf900rkw]

While it applies to the entire range, that series of posts were about the M/S52 tank. It, like the M44, is below the coolant level in the head, so the urge to overfill is greater.

[Schreier]

After squeezing the hoses to the best of my ability, and still not having the aux fan kick on of its own accord, I decided to take further steps to burp out any air that may be in the coolant system. So I tried the "spill-less funnel" technique.

burping.jpg

There's a cloth belt wrapped around the base because this "universal" kit didn't seal completely. Some coolant did leak, but I think I did the procedure effectively. I turned the dial to heat inside the car, but did not turn the fan on. Started the car.

It seemed like some bubbles came out when I squeezed the upper and lower radiator hoses while the car was running, but it may have just been the added pressure causing the fluid to burble. Otherwise, the fluid just kind of circulated around.

I did notice a few small burps as the engine warmed up (maybe related to the thermostat moving?). Maybe some burps came out as I throttled between 2K and 4K RPM. The level definitely rose as the engine warmed, and the level didn't drop when the thermostat opened as I thought it would.

The lower radiator hose was cold until the coolant temp got up to around 200F, at which point it began to warm noticeably. But otherwise no stream of bubbles or anything. I shut everything down when the temp rose to 210F.

Questions:

The aux fan didn't turn on throughout any of this. If I did actually burp the system successfully -- and this seems as thorough as it gets -- shouldn't the aux fan turn on, if its temp sensor and relay were functioning normally?

What should I try next?

20170118_195838.jpg coolant bobber.jpg

And finally, is this the appropriate level for the coolant bobber? The bobber only has two bulges, located very close together, and it doesn't look anything like the diagram impressed on the top of the reservoir.

[Schreier]

More updates

Covered up the water pump's now-naked fan stud with this nifty fan stud nut. It weighs 24 grams, and I think it's made of some kind of heavy duty nylon. For some reason the product page doesn't show it's for Z3 use, but if you look at the list, it fits the M52 (and quite a few other motors).

C2-gjjOVEAAjAb7.jpg

before
C2-gkY-UUAAwW5E.jpg

after
C2-glikUEAEp6jQ.jpg

Also purchased the Bentley repair manual for this order. So far, seems great. I can see where it falls short on certain information, but it is a great guide and still worth every penny (especially for a newb like me who doesn't have much wrenching time under his belt).

The guide gave me a few ideas of what to look for next, in order to get the aux fan working as it should. Going to give one last shot at coolant bleeding with the technique described in the manual, then move on to replacing the temp sensor and/or relay.

[Schreier]

Success!

Or something like it?

I haven't done anything to the coolant system since the last post when I burped it, but I took her out for a drive today. When I was done with my trip, I parked the car and let her warm a bit.

Coolant temp was about 197F (91.6C) when I pulled into the garage, and at about 213F (100.5C) the aux fan kicked on all on its own. That seems to be the key temperature, at least for my car.

Now I don't know if that's a result of my burping procedure, since I never let it rise to 213F before. But whatever the case, the temp switch and aux fan seem to be working.

[K&B.21]

Yep, 97 z3 2.8 fan delete (Open hood remove clutch fan and close hood.)

1) Open hood look at it. I decided to not remove the shroud as it is a one pc gig and I didn't want to drain radiator and or remove radiator.

2) Make a simple tool out of a 1/8th inch by 8 inch metal stock. Drill a hole at one end that will go over the pulley bolt and the length of the metal extends to the next bolt without hitting the fans nut, this will lock the water pump so the pulley doesn't rotate while unscrewing the fans nut.

3) Place home made tool in position on "your" right side two pulley bolts, place fan nut wrench on fan nut, hold home made tool placing force to the left, place force on the fan nut wrench to the right and the fan will unscrew.

4) Wiggle and giggle and wink at the fan a few times as your working the fan up and out of the fan shroud making sure not to damage the radiator. The fan will come out of the top without removing the shroud. "As long as you wink at it!"

5) Check pulley bolts and tighten if necessary.

6) Close hood.

7) Drive car, I went out to a close back road and got her up to 120 for about 5 miles. (I live out in the sticks where I can do this.)

8) Park car and let it idle for around 10-13 minutes and the temp gauge will start to rise "Mine went to one bar past the center" No where near the red. The electric fan started and the temp gauge went down back to center, and the fan shut off.

9) Enjoy the car now.

Yes, open hood remove clutch fan close hood. Dead on accurate.

I am not afraid of the temp rising as it is normal as long as it does not go into the red. That's why there is a scale and a red section. It never went near the red so it never overheated. All is good.

Thank you guys for all the info!!

[Tigershark48]

Going into my second summer with viscous clutch fan deleted. No overheating and aux fan typically goes on when I get home and backing it into the garage. Nice to have the engine rumble under the hood minus the old vacuum cleaner noise of the deleted fan drowning it out.

[BimmerBreaker]

I am not afraid of the temp rising as it is normal as long as it does not go into the red. That's why there is a scale and a red section. It never went near the red so it never overheated. All is good.

Thank you guys for all the info!!

The temperature gauges on these cars are buffered. Meaning top dead center encompasses the entire operating range if of the vehicle. If the gauge moves at ALL above TDC, the car is overheating. No ifs ands or buts about it

You should at the bare minimum invest in a lower temp fan switch. A little above TDC in controlled circumstances means massive overheating is only "stuck in traffic on a hot day" away...

One bar past the center is actually pretty frikkin hor on these motors...

[ProductUser]

The temperature gauges on these cars are buffered. Meaning top dead center encompasses the entire operating range if of the vehicle. If the gauge moves at ALL above TDC, the car is overheating. No ifs ands or buts about it

You should at the bare minimum invest in a lower temp fan switch. A little above TDC in controlled circumstances means massive overheating is only "stuck in traffic on a hot day" away...

One bar past the center is actually pretty frikkin hor on these motors...

My temp guage reports a tad right-of-center (looking at guage dead straight). Validation: Using the Torque app, the reported water temp remained well within normal operating temps.

[BimmerBreaker]

Then your car is overheating slightly, the needle wasnt reset right after cluster work or you are more perceptive than most and are noticing a very slight movement within it's buffer range

Either way he said his went "one bar past center" which means that hashmark at the 3/4 location which is WAY too hot

[ProductUser]

First, I can only say the car has had no cluster work for the 10+ years I've owned the car. Second, others would argue that I'm not that perceptive

I don't trust gauges and perfer to see data streams over a range of driving conditions. <= function of my profession