Day 2 Heat Tests - Some Confirmation

greg · 12th September 2003, 20:40

This is the 5th thread in a series:

Today I met up with Don (LOUV) at his office in Cupertino. Don hooked up his OBDII scanner and laptop to my car and we went out, our goal being to compare the reading from the OEM temp sensor to the one I am getting at Probe 2.

It didn't take us long. We can tell you that the OEM sensor and my sensor track the air temperature linearly with each other. The value reported by OBD was 15 degrees higher than the value my system is reporting - but please don't get excited about this - it isn't important. What is important is that they tracked each other at various temps with this constant difference, except (and this is what is important) during transitions. Don's setup can read and display the value many times per second, so we could get a feel for how quickly the air cools down upon throttle application.

The MAF's sensor has almost no lag due to its own thermal mass. When I opened the throttle, the temp it reported dropped very quickly (on the order of 15-20 degrees). My sensor began heading down, but didn't catch up for several seconds.

Conclusion 1: Opening the throttle results in much cooler air in a hurry.

Conclusion 2: The MAF's sensor reacts quickly. How important this is in terms of the engine mgmt software is unknown.

Conclusion 3: My data is good and valid for comparison of temps from place to place, but only in somewhat steady state. It points out the right trends in transitions, but by nature will tend to understate their speed.

On the way back I hooked up with a Maxima doin' 100+. (Gotta love 280) I followed at a polite distance for a while, and passed an SL55 the next lane over who apparently didn't want to play. Later, though, I had pulled over and the SL55 comes flying up, whereupon I joined again - we both left the Maxima of course, and we accelerated to an unprintable speed. The good news is that this created a longer period of time during which I was using a LOT of power. I checked when we backed off - My P2 sensor (temp near MAF) was within 4 degrees of the P3 sensor (lower inlet temp). The P3 sensor was actually reading a few degrees BELOW ambient - so I don't know what the truth is there. All I can say is that under those conditions the post-radiator air is not significantly warmer than ambient. When I yank all the sensors I will check them at some mid-range temps (like 90, 100, 110) against a standard and can correct the data. But once again, under a sustained power application, a CAI isn't going to help you much.

Need4Spd · 12th September 2003, 21:45

So, if I understand correctly, the benefits to a cold air induction system would primarily be seen at low throttle openings. And likewise, the benefits of the temp sensor relocation. While the MAF temp comes down quickly when the throttle is opened, it would appear to be reporting hotter most of the time steady state at low throttle openings. Thus, either mod would help your day to day power delivery at low to moderate throttle openings, but provide almost no advantage in all out max throttle events, such as track runs and that SL55 you come across now and then.

greg · 12th September 2003, 23:02

Quote:

Originally posted by Need4Spd
So, if I understand correctly, the benefits to a cold air induction system would primarily be seen at low throttle openings.

Yes - the longer the throttle is open, the less heat rise there is in the intake system, AND the closer the OEM-location inlet air approaches to ambient.

Quote:

And likewise, the benefits of the temp sensor relocation. While the MAF temp comes down quickly when the throttle is opened, it would appear to be reporting hotter most of the time steady state at low throttle openings.

No, it reports the true temperature of the air it sees. The air really has been warmed.

Quote:

Thus, either mod would help your day to day power delivery at low to moderate throttle openings, but provide almost no advantage in all out max throttle events, such as track runs and that SL55 you come across now and then.

There are three factors pertienent to this discussion that affect the engine's ability to make power:

1. The temperature of the air at the point of intake
2. The temperature rise on its way to the cylinders
3. The tuning of timing (and possibly mixture) that the ECU does in response to input from a sensor it believes is measuring air between the primary and secondary airboxes.

A CAI helps with #1. After around town driving or park/drive/park/drive cycles, it could help a lot, especially at lower power settings, and for short bursts of more throttle without long enough duration to help cool everything down.

The sensor cannot help with number 1.

A CAI can help or hurt with number 2 - if it provides better insulation from the engine compartment air, the inlet air will be heated less on its way to the cylinders. If it is worse, the air will be heated more. If it has more mass, it will take longer to cool so that the minimum possible heat rise occurs. If it has less mass, it will take less time to cool.

The sensor cannot help with number 2.

A CAI helps with number 3 - colder air is better - within the caveats of #1 & 2 above.

A different sensor doesn't make the air any cooler. It almost certainly does, however, tell the ECU something different than what the stock location sensor tells it. It may report a different temperature for the same true air temp, it report values too high or low at one end and too high or low at the other, but be equal in the middle, or vice versa. It might respond more slowly to temperature changes than the stock unit. In any event, this "tricks" the ECU into using different engine settings than the BMW engineers intended. It may result in a power increase in some cases and a power decrease in others. It may be safe across the range of YOUR operating temperatures, it may not. Again, my personal guess is that it is PROBABLY safe - but I'm not here to pass judgement - only to give you more facts from which to make a more intelligent decision.

Fast6 · 12th September 2003, 23:23

Thanks, Greg.
You've likely amassed more hard data in two days than a lot of aftermarket companies could shake a CAI at.
My three observations:
First, how quickly did Don's OBD readout react to the opening of the throttle?
Second, I need to investigate the sensor setup on the ol' E36.
Third, it sounds like the best solution to avoid heat soak is to abolish the use of part throttle. Works for me!

Eliot

ernie · 12th September 2003, 23:51

greg,

thanks again for going through the trouble of actually getting us some solid data! hats off to you!

ernie

greg · 13th September 2003, 01:15

Quote:

Originally posted by Fast6
First, how quickly did Don's OBD readout react to the opening of the throttle?

Don was watching the readout, so he can tell you better than me. Unfortunately he wasn't able to get the logging feature working, and we were on limited time, so we don't have a log to go back to. From our dialogue and Don's notes, I believe it was sampling at between 2 and 5 samples a second, and he saw nearly instantaneous reaction.

Quote:

Third, it sounds like the best solution to avoid heat soak is to abolish the use of part throttle.

I think hood ventilation would help a great deal. Doing this in combination with opening both sides of the main lower intake (i.e., the brake duct mod, without the brake duct channels behind) would probably make a big difference. Without the hood vents, though, if the exhaust route for the air is only down low this might not help up top much.

Chuckc · 13th September 2003, 03:19

Greg,

Here is a website you might be interested in.

http://instserv.com/prod05.htm

greg · 13th September 2003, 07:04

Quote:

Originally posted by Chuckc
Greg,

Here is a website you might be interested in.

http://instserv.com/prod05.htm

That's a nice one! Good price too. I spent a few hours last night searching the web. But the fac is - I don't have the $1K to spend (+ another $500 for thermocouples, etc.) - now, if you guys wanted to take up a collection, I'd be happy to collect the data with this machine for you..... ;0

tonyo · 13th September 2003, 07:21

Quote:

Originally posted by greg
I don't have the $1K to spend (+ another $500 for thermocouples, etc.)

Here's something a little bit cheaper.
http://www.labjack.com/labjack_u12.html

Chuckc · 13th September 2003, 10:00

Quote:

Originally posted by greg
That's a nice one! Good price too. I spent a few hours last night searching the web. But the fac is - I don't have the $1K to spend (+ another $500 for thermocouples, etc.) - now, if you guys wanted to take up a collection, I'd be happy to collect the data with this machine for you..... ;0

I used to use a very portable datalogger at a company I worked. They were about the size of a cigarette box and had no readouts; just T/C inputs and an RS232 port. (They were used for shipping studies) Thermocouple wire is very cheap. We would buy it in spools and solder the tips. I wish I had some time to devote to this as I could probably borrow all the stuff form a buddy that is still at that company..

Anyway, nice work Gregory...