Long and technical: thoughts on cold air intakes and temp sensor relocation - The Unofficial BMW M5 Messageboard (m5board.com)

greg · 8th September 2003, 06:51

This is the second in a series of related threads:

In my previous post, I said I no longer believe the M5 is breathing the coolest possible air. Now I want to talk about some implications.

First, I believe the temp sensor co-located with the driver side MAF reads an accurate representation of the actual air temperature entering the engine.

Note that the volume of air that goes through an engine is tremendous. A rough estimate uses (RPM/2)*Displacement - in our case, at 5000 RPM, that is 25,000 liters of air per minute. That is a HUGE amount of air. Now consider Don (LOUV)'s picture of the MAF:

(For the full res version of the above pic and the rest of Don's story, go here

Common sense should tell you that the conduction of heat from the plastic to the thermistor in the middle of that wire is itty bitty teeny tiny negligible compared to the cooling effect of all that air rushing by. Heat soak? I don't think so. But is the air the engine breathes hotter than ambient air temp? Emphatically, yes!

So why do people feel more performance after installing the UFMotorsport mod?

I believe it is NOT because it gives the ECU a more "accurate" temp reading. I think the OEM one in fact gives it a more accurate reading - after all, it IS in the flow of air the engine receives.

In the relocated temp sensor case, the engine management computer is simply being "tricked" into thinking it is getting colder air than it actually is. The location UFMotorsport recommends is not in the flow of air to the engine! It is in the pre-radiator, unwarmed air entering the grills up front.

Is this OK? Maybe. But that would imply that BMW was just dumb - and although I've certainly seen cases where that is true, I think it is a dangerous assumption to make without understanding why they might have done things the way they did.

Were BMW engineers simply stupid? I doubt it. Pressed for time, maybe. Making tradeoffs to meet emissions? Even more likely. Ever see those pictures of "test mule" cars with all that instrumentation? One of the key things they do is tune the "maps" which (among many other things) tell a standard engine management computer (like Bosch) what to set the mixture and timing to for this specific model car when the temp is X, the MAF is Y, etc.

Note that it doesn't matter if the temp sensor says "130" or "45320" when the incoming air is 98 degrees f. It only matters that the associated value in the map properly tunes the engine at that temperature. So they could have easily changed the map to lean the engine or advance the timing at a given temperature. They wouldn't have even had to ask the mechanical engineers to move the temp sensor - just remap it! Why didn't they? Perhaps they were concerned about engine longevity?

I wondered, so I sanity-checked my assumption with Wayne at Powerchip. Surely they could change the temp-->timing and temp-->mixture maps too - why didn't they?

Quote:

[Wayned responded:1.____The powerchip software does not currently, nor ever has changed the temperature compensation map._

2.____We cant see any reason to change the map_values in the software to 'trick' the car, without solid, factual, long term data (including 110 F plus testing) to support the idea.

It could also be that the reasons for the factory map have nothing to do with safety, longevity and performance, and that "tricking" the ECU is a perfectly reasonable way to get more power.

I'd rather feed it colder air in the first place!

It looks like the M5 was designed to allow cooler air to make its way to the engine, but that these routes were blocked (sort like the vestigial brake ducts.) Review these pics:

1. Note that the OEM intake ducts head straight toward the lower center large "M" grille opening - only to be baffled away from the incoming air. Removing the side baffles (the ones through which the UFM sensor is to be mounted) would allow plenty of nice cool air into the engine compartment, and straight to the intakes.

2) Note the plastic caps ahead of the radiator that, when removed, would let some cool air "leak" to the intakes also. Are they plugged up in other 5's?

So I see lots of opporrtunity for clever, easy cold air intake modifications. So again, let's at least be aware of the implications.

Clearly, BMW engineers went out of their way to control the airflow around and through our cars. I think I understand why. The engine is water cooled, and cooling the water efficiently is the best way to keep it cool. The less air they have to use to do that, the better - because air rushing through the car creates more drag than air being diverted around it. You may have noticed there isn't a lot of hot air exhaust area in our cars either - the underbody panels, seals and ducts make it so. This benefits us in drag reduction (think fuel economy, think top speed) and aerodynamics (think about that planted feeling you love at 150mph.)

But the price paid is much less airflow through the engine compartment, and therefore high ambient temperatures there. I can live with high engine compartment temps as long as nothing melts. And I don't believe there is much of a heat soak problem, either at the much discussed intake plenum or at the MAF/temp sensor, as said above.

So why not let the engine breath cold air, independent of the air used to cool the engine? Because by drawing its intake air from the air that has already passed through the radiator, most of that air ends up finding its way out through the exhaust pipe! So we need to take less in, vent less out - and we get lower drag and better aero.

Another reason to feed the engine this radiator-warmed air is emission controls. A cold engine makes much more emissions than a warm one - so one goal of the system is to ensure it warms up as soon as possible.

So now you know why I think they did it the way they did. It also means there may well be power to gain by ducting cold air to the engine intakes. I don't believe this should be done without also considering how to then get rid of all the post-radiator air the engine used to suck up. If you don't let it escape, pressure builds and less air flows through the radiator - so you could actually create an overheating problem.

On the other hand, I know we have some CAI users in hot climates, and I haven't heard of overheating as a problem. Maybe the engineers went overboard in deference to emissions, fuel economy, or extremely high-temp environments. And maybe they set the engine a little on the rich side for teir own comfort (The Powerchip guys believe the engine is usually a tad richer than it needs to be.)

SO - let me conclude. This is meant as food for thought. I am NOT slamming the UFMotorsport product - and in fact I like those guys. I may well install one on my own car - I'm putting that off because I have a backlog of other stuff I promised to evaluate.

I just sure would like to talk to the guys who engineered our cars. Directly. I haven't found anyone in BMWNA with whom an informed and intelligent conversation about this kind of stuff can be had. Would a real M Engineer please stand up?

Took me WAY too long to write this. I'll have to save the Pong install doc for another time.

TCM · 8th September 2003, 07:04

I find this an intersting questin as well. I agree that the BMW engineers are not stupid. They designed the intake this way for a reason. I personally feel it is for emissions as you do Greg. It is the only thing that makes sense to me as the M5 runs rich as is. Will be interesting to see what we can conclude after this thread finally comes to a close. Thanks for initiating it

!!

Need4Spd · 8th September 2003, 07:16

So, bottom line is: (a) the air that the engine receives is NOT the same temp as ambient air; (b) the most likely reason for purposely sealing off the intake from ambient air is to control/reduce emissions; (c) there MAY be a side benefit to sealing off the intakes in that insuring flow through the radiator maximizes cooling efficiency, and possibly reduced air pressure inside the engine compartment and enhanced aerodynamics; and (d) re-locating the temp sensor ala UFM kit to measure ambient air will make the engine THINK it's getting air that's colder than it really is.

I'm in agreement that this bears more thought about what to do if anything now that we're armed with this knowledge. It seems the best/safest thing to do is make sure that whatever temp air the engine receives is also the air that you're measuring, so installation of the UFM kit should be accompanied by a mod to allow the engine to receive cold air. Simply pulling out the plugs/baffles that force the air into the radiator seems the easiest way to achieve the cold air part, and if that's so, then assuming you can gain access, it seems the best place to which to relocate the temp sensor would be on/at/near the air horn/pickup which receives the incoming, pre-radiator air.

Seem right? Jump in, people.

BTW, looking at the directions for cleaning the MAF sensors, why do you have to remove the air cleaners? Why can't you just unplug the harnesses and unscrew the MAF sensors and remove them w/o touching the air cleaners?

greg · 8th September 2003, 07:30

Quote:

Originally posted by Need4Spd
it seems the best place to which to relocate the temp sensor would be on/at/near the air horn/pickup which receives the incoming, pre-radiator air.

well, I'm jumping the gun, 'cuz I can't document the rest until tomorrow, but Pong and I agreed - we want to measure the temp the engine is getting. We mounted the sensor right in front of the intake trumpet (still warm air). This is the view from above, TiAg bumper at the bottom. We fashioned the bracket out of some scrap I had, and fastened it to a radiator frame support with a cupla sheet metal screws.

chunpng · 8th September 2003, 07:42

Greg,

Thanks again for a thought-provoking thread. And even more thanks for spending your precious afternoon working on my car.

Here's yet another monkey wrench in the mix. The ECU is supposed to program the air fuel mixture for optimal perfromance. Since the only mixture that matters is the amount of oxygen/fuel right before detonation, thus the only appropriate measurement of air temp. is that of air right after it hit the cylinders. The fact that we are measuring temp. in the air-flow meter, or right at the air inlet trumpet, really is no assurance that the detonation air-fuel mixture is correct.

The only known fact is that, if the temp. sensor sends a low temp. reading to the ECU, the ECU would send more fuel to take advantage of more oxygen in the cool (dense) air. If the air temp. at the detobation point is higher (i.e. less oxygen) the car will burn rich. If by luck the air temp is low, we will have more detonation energy as there is more fuel sent. In any case, we will only have better performance.

Of course, the remaining question is if the engine actually burns rich, is it harmful to the engine?

Now you engine gurus please feel free to chime in.

Thanks.

CP

1C00lVC · 8th September 2003, 08:17

One more thing to consider when locating an air intake and it is air pressure. When a car travels at 80 mph, different spots in and around the car experience different barometric pressure due to airflow and aerodynamics.
For example, I would assume that because the shape of the front fender, there would be some vacuum in the wheel well when that car is traveling at high speed, and maybe there is higher pressure in front of the radiator. and maybe it is better to get a bit more pressure on the intake, at the cost of a few degrees of heat... Just a thought...

Chuckc · 8th September 2003, 08:30

Quote:

Originally posted by chunpng
Greg,

Thanks again for a thought-provoking thread. And even more thanks for spending your precious afternoon working on my car.

Of course, the remaining question is if the engine actually burns rich, is it harmful to the engine?

Now you engine gurus please feel free to chime in.

Thanks.

CP

Very interesting thread. Excellent work! I believe that if the car runs too rich, you can get cylinder wall washing from the excess fuel. I have no idea at what FA ratio this would occur.

MIB · 8th September 2003, 10:09

Enter BeastPower GTR M5 Vented Hood.

And Yowll thought we were crazy.

Hey Greg,

Great thought provoking post. The Vented hood gives a place for the now unused hot air.

I'm just waiting for someone to say it may make the motor too cold and it won't run efficiently.

M5Don · 8th September 2003, 11:01

This has become quite the interesting thread. Sorry for not posting sooner, but i've had a lot of catching up to do since our Bay Area road trip.

Let me start out by saying that I will never post "speculations", or anything unsubstantiated. Guesses and questions are fine, but misinformation can be dangerous, especially when presented to a group as technical and analytical as this one!

First of all, to answer the discussion regarding wind chill:
The variable resistor temp sensors used in this application are not affected by wind chill. If the temperature is 45F, the sensor will see it as 45F regardless of wind speed. The sensor does not have nerves, skin, nor does it perspire. "Wind chill" is what 45F "feels like" because of wind conditions. Want to test this? Hook up a resistance monitor (most multimeters will have this) to the connectors on the sensor. you will notice that the numbers will change as you heat the spike with your hand and cool it off by blowing on it. Let go of the sensor and let it stabilize to room temp (readings stop changing). now turn on a fan aimed at the spike (room must be closed). If the fan is pushing un-heated or cooled air to the spike, the numbers will remain the same. ** Warning** hair dryers with a "cool" setting does not work, as the "cool" setting still blow slightly warmer than ambient temperature.

Secondly, regarding the position of the relocated temp sensor:
The reason why the location of the sensor is at the bottom, to the side is because the M5 in stock form protects the sensor from flying rocks, dirt, and other debris. While wind chill does not affect the sensors readings, excessive dirt and damage of the spike by flying debris will. It also keeps the sensor away from excessive heat coming off the radiator.

As far as heat soak is concerned:
this is a problem on the E39 M5. The Z8 has the S62 with the same setup (air temp sensor inside left side MAF) and does not seem to have the same issues. The powerloss due to the heat soak is not nearly as evident in the Z8. Why? Because the radiator sits a few inches (considerable) further from the MAF. The engine compartment is also considerably larger. The software from what I have seen is the same for both cars with this engine.

The 740 and 540 (M62) also has the stock air temp sensor in the MAF. But why do these cars not benefit from the relocated sensor? The engine compartment temp does not reach the same temps as the E39 M5. Another reason? The MAFs on the M5 sit directly over the cylinder head and upper radiator hose, with less than 0.5" clearance. The MAFs on the S62 cars are offset by at least 2.5" The point? E39 M5 has a tighter and hotter engine compartment than the Z8 and most other late model BMWs.

The covers that are pictured that can be removed: not all E39 cars have this setup. The 528 and 525 both have intake pipes that duct into this hole (left cover is off, with an intake pipe attached, while the right cover remains in tact.) The 540 has an intake pipe into the right hole, while the left cover remains in tact. All E39s share the same plastic cover in front of the radiator, so I assume that the 2 covers exist so that the same cover can be used in other E39s.

Other cars have the temp sensors located inside the air boxes. This setup is fine, because the sensor is located far away from the engine and radiator.

Now one last thing: Are BMW engineers stupid? of course not. But do they think of everything? Usually not. When a defect is found after the car has been shipped and sold, what should they do? Recall, reprogram, or retrofit every car? how about the cars that have not been sold yet? should they change the software maps? what if the change relates to a molded plastic or metal part? should the machinery and process be adjusted for the rest of the cars? Sometimes, but many times it is not. If this was the case, BMW engineers must be stupid for the clutch, door seal, cupholders alone. Make the argument that the clutch and the cupholder issues are user-error problems. Why not reengineer the door seals to match those found on the E38 (never a problem), then recall the cars for the retrofit? Instead, the part was updated, replaced upon request only through the factory warranty, and still the car sounds like its suffering from excessive body flex, even with the new felt-inlayed part. If every last detail was treated with utmost importance, the cost of the car would be astronomical.

BMW's Aftersales and Engineering Department is there for this reason. Very few oversights render a repair, recall, or redesign in the middle of the production run. Most of them get Service bulletins instead, kind of like a release note in the software world. If this issue even got an SIB, the "Warranty status" field would say "Information Only". (BTW, I've been checking to see if it actually did get an SIB) Bugfixes for every little issue is impossible for cost reasons alone.

I hope this answers a few questions that have been brought up in this thread. I don't condone "hack-job" mods that threaten the integrity of the car. What the UFM IATS kit does is nothing new. It has been done before. What we did was improve the hardware and verify the results, then offered it at a fraction of the cost.

BTW, the instructions advise the removal of the bumper. Bumper removal is easier than it sounds for those who have actually tried it. It also allows easier access to the recommended mounting location and makes it easier to neatly route the wires to the recommended, least intrusive places. For those of you without a lift, you probably noticed that the main plastic center cover under the car will not come off unless you have VERY tall jackstands...

We have not tested the sensor in other locations extensively enough for us to include it in the installation instructions.

Thanks for reading!

ELEVENS · 8th September 2003, 15:33

greg:

This is what I have maintained all along... and got flamed for it when this temp sensor relocation first came up.

In a nutshell: "BMW engineers are not stupid, the stock location best represents the actual temp of the air that goes into the combustion chamber, and moving the sensor to a colder location "tricks" the ECU into providing an apparent performance advantage."

That said, I put a temp sensor relocation kit on my Vette years ago, and saw performance increases. Later, when I began programming my own chips, the location of the intake air sensor was less critical.

Finally, I will be buying the kit for my M5.

Does it work? Apparently. Why does it work? Opinions vary!

-Bill