N52 cold weather operating temps

[hoveringuy]

Yeah, it's cold, but I still want to drive the e30 occasionally....

Took it to work the other day after completing my undertray, which is amazing and makes a shocking difference in highway noise level.

So, the car hardly warmed-up with a 32F ambient temp. I couldn't even get it to the 1/4 mark on my temp gauge, which sits at 75C for me. The usual suspect would be a stuck-open thermostat so I pulled that when I got home and it' was fine, still looks new.

Next, I had the tune revised to make sure I had stock cooling settings, remembering that I had Nando set my minimum pump speed to 40% so I would get flow through the heater core. That also did nothing.

The final solution was to block 60% of the radiator and now the temps are PERFECT, sits between 95-100C all the time.

So, I think the issue is that when the thermostat gets close to it's 97C opening temperature the force holding it closed is much weaker and the thermostat can be unseated by enough coolant flow. Watching INPA, I could see that when I mashed the throttle coolant pump speed increased significantly and the temperature also dropped immediately. There seem to be lots of "N52 won't warmup in the cold" threads on Google.

The other thing I will note is that warming the engine seems to take much longer since I added a coolant-to-oil heat exchanger. I've got 6 qts of cold oil that need to be warmed along with the 2 gallons of coolant.


The undertray. Easy on, easy off, and it's much, much quieter now. I'm hoping for another 2-3mph at the end of the straight.




The N52 cooling system. Thermostat works exactly like every other thermostat except it's in an external place now and it also has an electric booster (mapped thermostat). Electric cooling pump varies speed with load, not just rpm.



This is the warmest it got.



steady-state 65mph on the freeway after 35 minutes. Still not warm, but pump is running somewhat fast.



Final solution (for now). This is like a trucker putting the vinyl flaps over their radiator. That remaining 40% is more than enough!

[Gaarath378]

Reminds me of the active grills new BMWS come with that everyone seems to hate. See people take them off and deal with a fault bc of it or sometimes chop it so it can still rotate but not actually block anything, and then they complain when their engine takes forever to bring to operating temp. Wonder how hard (and practical) it would be to retrofit a system like that.

[Panici]

Took it to work the other day after completing my undertray, which is amazing and makes a shocking difference in highway noise level.

That undertray looks great! Something you fabbed up yourself?
How much of a faff is it to remove for service?

Final solution (for now). This is like a trucker putting the vinyl flaps over their radiator. That remaining 40% is more than enough!

Basically how I ran my 328i E36 for most of a winter. It had a thermostat that wouldn't close all the way (which has since been fixed).
Sometimes the simplest solutions are the best.

[MrBurgundy]

I bet that pump is running all the time.

N52s run hot, for efficiency. It will throttle the pump to make sure the car stays where it wants to be. Sometimes, not running at all.

IDK how your reservoir/return is setup, but I wouldn't be surprised it that pump came on as soon as you started the car. The throttle input should not dictate the water pump speed. Only the temp will, or should rather.

Your pump should not be running at 50 percent if the coolant temp is 60c, theoretically, it would be barely running to get the car up to temp as soon as possible

[hoveringuy]

I bet that pump is running all the time.

N52s run hot, for efficiency. It will throttle the pump to make sure the car stays where it wants to be. Sometimes, not running at all.

IDK how your reservoir/return is setup, but I wouldn't be surprised it that pump came on as soon as you started the car. The throttle input should not dictate the water pump speed. Only the temp will, or should rather.

Your pump should not be running at 50 percent if the coolant temp is 60c, theoretically, it would be barely running to get the car up to temp as soon as possible

If the car is warm and I re-start the car the pump will be at 0% until it hits 95C. At cold start the pump runs at 20% which is normal for N52's. All other times, the pump runs proportional with load. That's all normal behavior.

The interesting thing is that there are lots of hits for "N52 overcooling" and such where people have replaced their thermostat multiple times and concluded that it's just the ECU with too much pump speed mapped with load, and no consideration for ambient temps.

We need more people like Striker01 and adam.nonis to get some miles on their cars so we can compare behaviors.

[hoveringuy]

Nando and I were just chewing this over and we talked about the radiator outlet temperature sensor. The earlier E90's with the magnesium valve covers had these sensors and the LC1 N52N from about 2008 and on with the plastic valve covers did not.

I do not have one.

The "radiation" outlet temperature is still active in INPA and the ECU just calculates it as 7 or 8 degrees lower than the inlet temperature. That's probably accurate on a warm day, but on a cold day with very little radiator flow I'm sure ice water is coming out of the radiator.

I have ordered an outlet sensor, not a big deal to add it if I need to, but I will wire it up to see if the ECU recognizes the input to start with and if it affects how it cools. I'll put it in a cup of icewater or something. Could be fun!


On the freeway the other day....



Parked in my garage just now....

[Striker01]

Nando and I were just chewing this over and we talked about the radiator outlet temperature sensor. The earlier E90's with the magnesium valve covers had these sensors and the LC1 N52N from about 2008 and on with the plastic valve covers did not.

Should be easy to get it flashed to look for that again. I was getting a code for not having it and 22RPD got rid of it.

Would be nice to find a threaded one instead of the o-ring fitting type.

[Striker01]

I got a variable resistance tool and did some testing today. Confirmed that resistance value correlated to ECU temp readings in INPA are the same for both the Coolant Temp (CT) and Radiation Outlet Temp (ROT), so we can use the threaded head temp sensor in our radiator; if the thread size is the same. Both hoveringuy and I have one on the way to verify.

Keep in mind my water pump minimum is set to 27.7 so I have heat.

It appears that the CT and ROT only control the operation of the cooling fan and don't affect the water pump very much, if any at all. As I adjusted the ROT, the only change was the cooling fan came on. Pump speed did not change.





I disconnected the CT sensor and used the variable resistance to simulate CT changes. Pump speed didn't vary with CT changes, even when I set it above 121C which I didn't get a picture of. The pump speed changed from 27 before I started the test, not during. It changed as I varied RPM but was stable when I ran this test.




I don't think this did anything to help Steve troubleshoot his car but it's interesting data. It helped me out because I think 22RPD just deleted my ROT trouble code but didn't change the ECU to calculate it based off of CT.

I let the car actually get to 114 and the fan never came on until I simulated a hot ROT. Once I set the ROT to 142 the CT dropped fast, all the way to 85 but the pump speed stayed constant.

I wonder if I can connect both sets of wires through the resistance tool at the same time to see if pump speed increases when both are hot, thoughts? I don't want to damage the ECU.

[hoveringuy]

Resolved! I found the root cause of my issue and it wasn't the thermostat, coolant pump speed or any of that. My bleed hose was just too big...

I had moved the reservoir to the passenger side for better packaging and moved the nipple to the same side, which is the cold side of the radiator. I had very strong flow through that hose, small as it is, so that ice-cold coolant was flowing through the reservoir and then directly into the thermostat where it was mixing with the coolant returning from the mickey mouse fitting. The reason it was getting colder when pump speed increased was not because it was unseating the thermostat, but just creating more flow through that bleed hose. A lot more..

The fix was just a flow restrictor in that bleed hose with a 2mm orifice so now it bleeds air the way it's supposed to but doesn't screw-up the cooling. Warm-up is also faster, naturally, although the oil cooler has increased that time a bit.




I also added the radiator outlet temp bung while I was at it. (thanks Striker01 ) The functionality is built-in to the ECU; if it detects the sensor you get actual outlet temp, if it doesn't see the sensor it gives a figure that 7C below the inlet temp. It adds ZERO functionality other than having a valid figure in INPA, I look forward to being able to see radiator inlet vs outlet temps. Maybe I could use the functionality to log my differential temps??




Finally, I tried adding an aux pump to my heater circuit. This was $20 on Amazon and is a VW unit that I didn't spend too much time integrating it because I wanted to see what it would do first. Remember, in an E90 when you select heat on the climate module it asks the ECU to bump-up the coolant pump speed from 20% to 85% to move enough coolant to give heat. At 20% pump speed, even 40%, the heat is tepid. So....

I got the engine up to temp and my vent temps were vaguely warm. I turned-on the pump and within a few seconds it went to "IT BURNS US!". Yeah, it gets crazy hot now!.

I'll need to put some effort into finding the pump's "forever" home now. It's small and absolutely silent so I will workshop some ideas.

The ultimate long-term solution is that I'll need to find a Z4 climate module and decode the CAN traffic to find the one that the ECU is looking for to put it in heating mode. If I could do that, a tiny circuit would generate the CAN signal to boost the coolant pump.