The two sensors in the AFM allow the ECU to calculate the mass airflow of the air entering the engine. As Henrik said, the system is a closed one, so the mass flow cannot change at any point in the system. It's physically impossible. Yes temperature changes, but the air can't go anywhere else, so it's mass flow remains constant, however as it's density changes with temperature, its velocity must change because of the constant mass flow. And the ECU is perfectly aware of the temp and pressure at the maniflod courtesy of the air temp sensor in the TB and the manifold pressure sensor. While the pressure sensor doesn't affect fuelling (only ignition timing), the temp sensor allows the ECU to alter fuelling to respond to the temp changes caused by the turbo and IC.

 

An air temp sensor isn't going to measure coolant temp

 

There is. It's needed because the turbo and IC dramatically change the air temp depending on how the car's being driven, even though the temp sensor in the AFM remains reletively constant (because it's measuring ambient air temp).

 

The Earth's atmosphere can also be considered a closed system, but I doubt that the air where I am right now has the same properties as the air where you are. IMO it is more accurate for you to calculate the air mass where you live than it is for me to measure the air mass where I live and try to apply it to your location.

What you are forgetting is that a vehicle intake system has dynamic properties. As the air travels through the system, it will compress and expand in different areas, heat and cool in different areas, get caught and released in different areas, absorb and release moisture in different areas, etc. Just because a given amount of air mass goes through the AFM, it doesn't necessarily mean that the same amount feeds into the rotors at that exact same time.

To be honest, I'm not sure what all the stock ECU measures in regard to engine tuning, or how it weighs different data to make changes to the engine tuning. However, if the AFM were so darn accurate at estimating the mass airflow at the manifold, then there would be no need for an additional temperature sensor on the intake elbow. Think about it.

While looking up something for a friend, I found a passage from Corky Bell's Maximum Boost book which may help explain at least one issue that I seem to have trouble conveying to you folks.
“It is distinctly possible to upset the basic throttle response if an engine is equipped with an airflow meter positioned too far from the throttle body. Opening the throttle causes a low-pressure pulse to be created that travels upstream toward the airflow meter. The time it takes this pulse to reach the flow-meter and cause it to react is indeed the delay in throttle response. Typically, such a pulse must travel from the throttle body to the intercooler, through the intercooler, back to the turbo, then to the flowmeter, in order for the flowmeter to register a response. It is not until the flowmeter receives this pulse that the air/fuel ratio can change to account for new load conditions in the engine.”
Like most of the Corky Bell issues, this is simplified, but hopefully you will begin to understand part of what I am talking about.

 

Evil, I understand what your saying, but really all your doing is pointing out the drawbacks of using an AFM - these are higher order effects. I think you'de agree that after the system has stabilized from a rapid throttle change, the AFM will give a reasonably accurate measurement. Even building boost should be a relatively slow event compared to stomping the pedal.

The only point I was trying to make is that what an AFM, a T sensor and an atmospheric pressure sensor you can calculate the mass entering the system, you had indicated that this was not possible. With a speed density system you need a model for the VE of the engine, the T sensor, atmospheric pressure and MAP sensors.

To cover the pros/cons of each system is a completely different discussion.

Henrik
87TII

 

Airflow meters are the most accurate way to measure the engine's airflow (that's why they are by far the most common method used by manufacturers), but as Evil's pointed out, aren't as quick to react as a MAP sensor. That's why systems more modern than the FC's use both an AFM and a MAP sensor to calculate fuel requirements. The AFM is the primary sensor, but the ECU uses the MAP sensor to compensate for rapid throttle movements, similar to the way the accelerator pump works in a carb.

 

So whats the verdict should you be able to put the afm after the turbo or after the intercooler????

 

wouldnt the flapper thing go all the way open under boost?

 

No... boost doesn't change the volume of air... (well, it does, but not in the same sense... it compresses more air into the same space) the flapper doesn't care what the AIR PRESSURE IS (which is how it would impact boost) it just cares how FAST the air is flowing through a given space. Note, that is a generalization... it truly isn't that simple, there IS some impact... but not that much. Suffice it to say that I (and probably you too) don't have the background in math and physics to make the answer any more clear.

Brad

 

The AFM is calibrated for use at atmospheric pressure. The flapper door will open proportional to the mass of the air and the square of the velocity of the airflow. When you pressurize the flow, the mass remains the same but the flow decreases so the AFM will measure progressively lean as boost rises. You can probably do the change, but you need to fix the fuel problem.

As an analogy, think of a garden hose. Turn the water on to a fixed rate (fixed mass flow) and imagine spraying the flapper. Now pinch off the hose opening and spray again. In which case does the flapper open more - the second one of course. In both cases you've got basically the same mass flow but your readings are way different due to the change in velocity. For a turbo system the intake velocity will be greater before the turbo than after; ie the flapper will open more before the turbo than if placed after.

Edit: I haven't tried moving the AFM and the above is just based on my understanding of the physics involved.

Henrik
87TII

 

Someone shoudl do some tests on this... if it would work well just before the TB this would be a veru good mod to aid throttle response and allow proper atmo vented BOV operation.

 

alright we need a gini pig who's up???

 

It was already done in 1988. The 10AE with relocated AFM, stock internals, and a bunch of HKS add-on goodies did 191.115 mph at the Bonneville Salt Flats. Power was estimated at 390bhp. For those wondering, the AFM spring was re-adjusted.

 

daym yo lets do this shyt then

 

I don't think it's worth the trouble given the low price of a standalone EMS now. Even if you make minimum wage, you could work weekends for 2 months and buy a full-function Microtech EMS which would work much better than a jury-rigged AFM, not to mention that it would be easier to tune.

Web viewing isn't so great, but I think you can read the stats on the Bonneville RX-7 if you download the picture to your computer and read it with one of your graphics programs. Note that it had a whole lot of HKS EMS add-ons in addition to the modified AFM. It would be much easier and cheaper to use a standalone EMS, which was not available at the time this RX-7 was built.
http://www.geocities.com/evilaviator/bonneville

 

hi guys
just reading your posts through quickly
just wanted to note that on my E15ET i successfully fitted an FJ20ET AFM on it. Worked fine, but i had to adjust the spring though.
On another note, someone said something about the AFM open all the way on WOT, if relocated behind the IC. It was discovered by those at n12turbo.com that the E15ET's AFM flap is fully open at 3000rpm. I would suspect it would be similar on the rx-7s.
Also does any of these US computers, (haltech, wolf, motec) have cold air idle sensors? I'm currently running microtech, it does the job, but quite embarassing stalling while getting out of the parking lot!

 

I guess that would work fine if the particular engine's max airflow occured at 3000rpm. However, if the max airflow is higher than this, then I don't see how the engine could run properly unless it went into Speed-Density or Alpha-N mode after 3000rpm. What am I missing here?

Just keep your foot on the gas! That's what I do with my chokeless 1Gen.

The Wolf3D 4.0 has an idle speed control solenoid PWM lead, as well as water temperature correction (similar to a carb choke). I'm sure that the other latest and greatest EMS products on the market also have similar functions (Haltech E11, TEC3, etc.). You may also want to check into the new Microtech LTX series of EMS products to see if they may function better than what you currently have.