Yea...cuz from what I read (the link Jim posted earlier)

"Because a Speed Density system still has no sensors that directly measure engine airflow, all the fuel mapping points must be preprogrammed, so any significant change to the engine that alters its VE requires reprogramming the computer.

By contrast, Mass Air Flow (MAF) systems use a sensor mounted in front of the throttle body that directly measures the amount of air inducted into the engine. The amount of current needed to heat the wire is proportional to the mass of air flowing across the wire.

MAF systems are much more flexible in their ability to compensate for engine changes since they actually measure airflow instead of computing it based on preprogrammed assumptions."

So, I'd say you have it backward, since MAF is much more sensitive and thus likely to react (at all, vs. Speed Density) to actual changes and fluctuations.

 

Actually, he's correct (no need to post a smartass remark, Jim). Think "inertia". As I stated in my other posts, a MAF sensor is most accurate when the airflow is fairly stable-- rapid fluctuations in airflow drop its accuracy, so they are usually mounted in an application where there is a huge plenum, airbox, or chamber between the sensor and the actual air entering the engine from the intake ports. They cannot be mounted directly "in front of the throttle body" because the airflow through that area is not consistent through the cross-section of the intake tract. Thus, the problem is that it takes time for the airflow passing through the sensor to catch up to the speed of the actual airflow entering the combustion chamber (like when the throttle plates are suddenly opened/closed). Air has mass just like anything else, so changes in an engine's air demands will take time before they affect the amount of air coming into the plenum/airbox/chamber. However, changes in manifold air pressure are nearly instantaneous from airbox to intake port. Nevertheless, this still doesn't get over the fact that MAF sensing can be more accurate over a wider range of engine demands.

I should reiterate here that a MAF-based system still uses fueling "maps" just like a MAP-based system; it's just not approximating those needs without actually seeing the airflow into the engine.

My original thinking was that Mazda possibly went with the MAP-based system on the FD because of this lag time, since the margin for any inaccuracies that can have catastrophic effects on piston engines is much larger. But there's no getting past the fact that MAF systems are much more expensive, and that is probably the reason the FD wasn't equipped as such.

 

Kento, I understood your points, including the slow reaction to certain variations in airflow, yet the MAF still being more accurate over a wider range of engine demands. Yet lemme ask you this...the specific reason why I was interested in MAF vs. SD was I thought MAF sensors would sense slight fluctuations in airflow and thus be able to immediately react by providing the adequate fuel necessary (yes, from a predetermined table). The advantage of this (or what I was hoping for) was you'd considerably reduce the possibility of detonation from sudden boost spike. Obviously, this is based on my understanding that since a SD system approximates, and that it's slower to react than an MAF based sensor, you're very prone to running lean during a sudden boost spike. What's your take on that? (I hope that made sense...)

 

The MAP-based/speed density system is not "slower to react"; as stated before, changes to air density are nearly instantaneous throughout the intake components. The problem is that it's approximating the airflow since it can't "see" it. Conversely, a MAF sensor needs to have the airflow consistent through the cross-section of the intake tract where it's mounted (otherwise some obvious fueling inaccuracies will result). A boost spike is not a "slight fluctuation" in airflow; a MAP or MAF will both sense it, although a MAP will probably react quicker due to the aforementioned inertia of the airflow. The big advantage with the MAF-based system is its flexibility and accuracy.

 

I'm not gonna give you a crash course in physics or fluid dynamics over the computer Jim. Go read something or take a college course if you wanna play with the big boys alright.

 

That's funny, coming from someone born in 1984... shouldn't that be "play with the little boys"?

 

You don't have to be old to know what your talking about, and you don't have to be young to be stupid.

 

that was good hehe. well said my son

 

You haven't yet given any evidence that you know what you're talking about, so the jury is still leaning towards stupid.

Sorry, but you'll have to do more than throw around a phrase like "fluid dynamics" to impress me.

 

I didn't read much of this thread, I just noticed Jim has 7611 posts....damn

Tim

 

Kento seems to agree with me Jim. Oh and Tim some people spout crap all day long.

 

Are either of you engineers?

You're forgetting the throttle position sensor. The computer knows PRECISELY when the throttle closes on a MAF-based system.

 

Look, both of you leave me out of your arguments, please. I'm only posting here to give reliable knowledge and information. I'm not here to get caught up in *****-jousting.

Yes, but it still doesn't know exactly how that will affect the fueling at that point, because of the fluctuations and differences in airflow. There's no way for it to determine the actual airflow into the engine at various throttle positions and rpm in real time. That's the reason for the TPS; the ECU uses that to approximate airflow and fueling, just as a MAP-based system would.

 

YOU DID IT AGAIN!!!

 

Were you looking for some more attention?

 

A speed-density system has NO CLUE how much air is entering the engine at ANY time. It's ALWAYS an approximation based on sensor readings and programmed maps. The proof is that the FD can't handle mods which significantly increase airflow into the engine without having the ECU reprogrammed to compensate while a MAF-based system like the Supra's can. Based on that, which system is more sensitive to fluctuations and differences in airflow?

A MAF sensor can monitor changes in airflow, even during a boost spike, while the speed-density system has no clue what's going on since it has no concept of air volume. It can only read the pressure of the spike and hope to compensate based on its preprogrammed maps.

 

That's EXACTLY my understanding of MAF vs. SD...and everything I've read seems to support that. So Kento, could you explain exactly why you disagree, and how the above is possible?

 

Sigh. Did I ever say that a MAP-based is "more sensitive to fluctuations and differences in airflow"? Nope. I only said that a MAP sensor can react to an event such as a boost spike quicker due to the aforementioned inertia. I never said that a MAF sensor can't sense fluctuations and differences in airflow; only that a MAF-sensor needs to have consistent airflow, because rapid fluctuations in flow at the location of the MAF sensor give it problems, due to the "hot film" sensor's inability to ensure that the airflow that it is reading is exactly the same as the airflow heading into the engine. I only stated the advantages and disadvantages to each system; not that a MAP/speed density system is superior in all aspects. Just because the Supra and other turbocharged cars happen to use a MAF system does not automatically guarantee that the same will work on the rotary. And yet, on that same token, I'm not saying that it won't work, either. There's just a lot of backyard engineering going on here, and I'm only pointing out the differences between each.

Think about the reason for the existence of a TPS on a MAF-based system. Again, I'm not saying the MAP/Speed Density system is better, but if the MAF system was so sensitive and perfect, why would it need to know the position of the throttle plates?

 

Please read this again, especially the last sentence...

 

No, that was mostly for the benefit of our resident fluid dynamicist.

So you run a little rich. Seems to work for Lingenfelter.

 

I agree with Kento. If you read my original post I just said that the SD system can react quicker. The MAF system is more accurate, I said that in the beginning. I never said either one was superior I just said that some people are now opting to convert from MAF to SD. I never said the MAF could not read a boost spike I just said that the SD system could read it and respond faster. Each system has its own pros and cons and I would appreciate it if you would read my original post carefully instead of trying to argue with something I never said.

 

So how does a speed density system react quicker when it relies on the same sensors (MAP, temp, TPS) that a MAF-based system has?

And I asked you to prove that if you didn't simply make it up. You have yet to do that.

Maybe you should re-read your own post.

 

IF the SD system could react faster than the MAF, the question is, is it even worth it? Because we're all agreed that the SD system *approximates*, while the MAF will respond *accurately*?

It's my belief (and plz correct me if I'm wrong) that it's the approximation of the SD system (not it's speed) that causes you to run lean during a boost spike, simply because it doesn't realize you ARE spiking. And I also believe that a spike would cause a large enough change in airflow for an MAF to immediately recognize the situation, and thus correct *accurately* for it, thus potentially saving your engine.

 

The point is that all the THEORY about how quickly a MAF-based system reacts seems to be SOLELY based on the ability of the MAF sensor to monitor air volume past the sensor.

What's being missed is that a MAF-based system does not consist solely of a MAF sensor. It has the same complement of sensors that the speed-density system has. Unless someone here has specific information about the operation of a MAF-based engine management system to support their claim, all the CONJECTURE about how quickly a MAF-based system reacts is pointless, because a speed-density system is nothing more than a SUBSET of a MAF-based system.

It does realize you are spiking, if solely based on the reading of the MAP sensor. The problem is that you have to have accurate fuel maps for anything up to and including the pressure level of the spike and that's not usually the case unless you have aftermarket engine management and have tuned at those pressure levels.

If you haven't, any fuel maps your ECU might have are probably approximations at best, or worse, your ECU doesn't have maps past a given pressure level and "flatlines" past that point with the same amount of fuel being injected for anything above X psi. That's how people run lean during boost spikes with the stock FD ECU. Of course, we must also take into account the ability of the fuel system to deliver the required amount of fuel, which is another problem with the FD.

The spike would be accompanied by an increase in the volume of air passing the MAF sensor, so you're right, to a certain point. "More accurately" would probably be the best phrase, but you still have to have maps for the conditions you're experiencing.

 

It doesn't use the same sensors. They both use the TPS and air temp sensor but the MAF system uses, you guessed it, a MAF sensor which I believe stands for Mass Air Flow. The SD system on the other hand uses a MAP sensor which i believe stands for Manifold Absolute Pressure. Sorry if I remembered the long version of the abbreviations wrong but I think they're right. The MAF sensor works by measuring the air in a small part of the intake plumbing. This reading is actually only a small percentage of the air that flows through the pipe. The SD system works by comparing the pressure inside and outside of the manifold. It can obtain an approximation (I"m not sure how) of how much air is flowing through the manifold. The two systems are similar but use two different methods of measuring airflow and calculating the amount of fuel to inject. This is where the get there differences from. But just for a second lets say your right and there could be no difference in there reaction time. Why then would you expect a difference in accuracy and if there is no difference between the systems what is the point of converting or even having this conversation. Oh by the way just take my word I know what I'm talking about. If I don't know or am not sure I'll let you know.