Mass Air Sensor

[75-1138604505]

I am considering building a DIY flow bench somwhere in the 500 CFM range. Has anybody here tried using an automotive Mass Air Sensor to do flow testing? They are a dime a dozen at the wrecking yards. All that would be required would be the signal voltage/air flow map for the unit being used. This could be mapped using a known accurate flow bench.
Kurt

[bruce]

It has not been a very widely discussed idea. There is some posts on my old forum archives which can be found via a google search

maf site:tractorsport.com

this will turn up a few posts on the subject.

I have been playing with a Ford sensor to use on my dyno project to measure incoming air to my carb, but right now its just been a "back burner project" bout as far as I have gotten is testing one on my flowbench to prove it actually works. The Ford sensor outputs a voltage not freq.

On a side note, not everyone that views a subject can post a reply. Only registered members may post, so out of those 25 views maybe only a handful are able to post a reply. Not all forums are open to public viewing either.



Edited By bruce on 1138845547

[Thomas Vaught]

Well I guess I will be the first to answer your question. Yep I built a couple of mass air benches.

The benches used Ford sensors too. 70mm stuff.

What I had to do was get an accurate "transfer function for the meter and did this by having the meter flowed on Hatachi's mass air meter stand. (You could get really close by just doing the same deal on a professional or an accurate home built bench).

I needed a 5 volt reference voltage to drive the thing and a good volt meter to read the voltage per mass flow value.

Originally I used a Ford computer to power the thing driven by a battery but later used a custom 5 volt power supply. House current will cause errors even with some fancy electronics to get to the 5 volt deal. With the computer industry using 5 volts a lot the problem is basically resolved.

I took cfm points every 5 cfm then generated a best fit curve that would allow a 1 cfm readout from a excel spread sheet.

The mass air meter fit into a long pipe between the vacuum motor box and a plenum chamber.

Today you could get a round truck filter housing with a built in mass air meter that might be better.

Hope this helps.

Tom V.

[Tom]

Hi Kurt,

I am using mass air sensors.

First element one is a Bosch (found in 3rd gen. F-bodys and C4 Corvette)
I have replaced the element into a 90mm pipe.
Accuracy is <3% from measuring point, repeatability <0,5% from measuring point.

Second element is a ABB sensyflow (reference meter for dynometers)
Accuracy is <1% from measuring point, repeatability <0,25% from measuring point.
Do not ask for the price. (I have this one from work for calibration use)

I would suggest you to by a air mass sensor and calibrate it on a bench near your place like Thomas Vaught said earlier.

Tom

[75-1138604505]

Very cool guy's
Thanks for the reply's. I was hoping for a positive response like that. I think it would make a great DIY project. During my browsing I found a map on the net for a 94 mustang. I will attempt to post the pic. The map is in KG/HR vs Voltage. I used the following formula to convert to CFM.
KG/HR x 2.2 = LB/HR
LB/HR / 60 = Lb/Min
LB/min x 14 = CFM
The graph tops out @ 930 KG/HR and if I did my calculations correctly it should be capable of 477.4 CFM. Does this sound correct?
Kurt

[Mouse]

This has been talked about several times in the past. One problem with a typical automotive sensor is that it samples only one spot of the bounded path. This requires a very long lead-in and lead-out to prevent the possibility of "fooling" the flow bench. You can test your design by rotating the sensor while flow testing a section of pipe located at the edge of your test stand. If your readings change, go back to the drawing board. If they stay good, move the pipe to the other side and repeat test. Keep moving the pipe to different locations and repeat test.
This is what I call balance. Good balance means the readings will not change no matter where you place the section of pipe on your test stand or test port.

Your probably asking " so why are they used in cars if they are not so good for a flow bench?". The answer is because the air input in a car is relativly static. It is settled by a nice air filter and it only changes in volume. A flow bench is more dynamic in it's air input. The location of the air input can change (balance test) because the valve of a cylinder head will not always be in the same place on the test stand, and the lift of the valve will cause certain levels of swirl and turbulence as it changes height.
In a car, the sensor is right after the air filter, nice clean. straight air flow. In a bench, is is usually located last, not the best place, requiring a long lead-in and lead-out for best results.

These sensors are great for measuring mass, but they are not the best choice for a flow bench.

John

[86rocco]

[2seater]

I am having less luck with using a MAF sensor than some of the others. Actually I am testing a Hitachi MAF sensor as used on Buick 3800 engines to establish a baseline for modification. This sensor is an indirect flow heated wire type where there is a main air flow passage and a smaller aux. passage where the sensor resides. This sensor also generates a square wave frequency rather than a voltage. I know what the frequencies are supposed to be at various air flows based on weight (grams per second), but the air flow volume (pitot bench) is up to 27% higher than the MAF signal indicates for the mass of air. I am using a figure of 36.5 grams per cubic foot of air for the calculation. The MAF is powered by the factory ECM and the MAF is mounted to the test port in the bench. I have checked the bench for leaks, the manometers for accuracy and the Excel spreadsheet formulas for velocity and cannot find any errors large enough to explain. Any thoughts.

[Tom]

[75-1138604505]

[quote]Today you could get a round truck filter housing with a built in mass air meter that might be better.
Thomas,
Do you have model year in mind that I could have a look at?
Do you know if any of the fuel injected Ford 460's came with MAF or were they all MAP?


[quote]First element one is a Bosch (found in 3rd gen. F-bodys and C4 Corvette)
Tom,
Does this unit give a DCV signal or PWM signal?
Kurt

[75-1138604505]

No only that, the maf sensor isn't the only means a car has of determining air flow, on the input side, it knows the rpm, throttle position and perhaps manifold pressure and on the output side, it has an oxygen sensor to provide continoius feedback.

86rocco,
You will not see both a MAF sensor and a Manifold pressure sensor on the same EFI system. The system that uses the pressure sensor in the manifold is reffered to as a speed density system. Speed density systems do look at a lot of input varibles to determine air flow.
Kurt.

[Tom]

@ kurt

it is a DCV.

My car has both MAF & MAP => you will found that setup on turbo engines
MAF for fuel map
MAP for boost / ignition map

Tom

[Mouse]

An old power secret for the late model Mustang 5.0 is to rotate the meter to a certain position.

You cannot make a linear correlation between mass and cubic feet. Converting mass to cubic feet is not an easy task.

Many production automotive mass meters are not linear and need to be mapped. Flow Performance can perform this service at a nominal fee for you.

This looks like an interesting project. Keep us posted.

John

[86rocco]

[PFM]

Kurt,

I was going to sit back and watch but... I gave this a try 10 years ago with a Mustang hot wire deal and had minimal success. The trouble was at the low flow rates, the units just did not work well at low flow rates. The flow sensors just do not get used on a car at 30 CFM to control the EFI. I worked it for a while from say 100 CFM to 200 CFM it worked OK, above that very good. I was not happy with a flowbench the would not do well accross the range I expected to use it. I was using a battery source and 12 to 16 bit A to D converters minimal filtering and I was still not happy. Keep us all informed if you push on.

PFM