Prelude: I have a Arduino Duemilanove. I am new to the language of C. back in the mid 90's i had a little experience with visual basic, and have forgotten most of it since. I dont have any math skills past geometry (i know, i know...) I am a Ham radio operator with a general class licence and build lots of circuits, and have been building electronic circuits for 20 years (never stop tinkering).
Project: i want to make a flowbench based on a ford mass air flow sensor from a 5.0 ford. it has an output of 0-5v. However, the curve is NOT linear. there is a few websites with the plot data so i may be able to get some formula from. I also want to implement a General Motors MAP sensor (manifold absolute pressure), which also has the output from 0-5V. The flow source of air will be a large shop vac.
Outcome: i want to be able to estimate actual CFM (cubic feet per minuite) of airflow threw the mass air sensor, and print it on the computer screen.
Problem: since i am completely a newb at the programming, and math bit, i am somewhat lost. I've been looking at examples and can input 0-5v, and print it using a formula, but it is about 100mv off. so i guess the first task is to be able to make a voltage meter from 0-5v and print it to the screen accurately. Then add the equasion to output mass air flow to cubic feet per minuit, plus pressure from the GM MAP sensor.
Hi, if those sensors both put out 0..5V then Arduino is easy..
BUT the actual accuracy of the Analog input readings is proportional to the actual Vcc on the arduino processor chip. If your values are 100 mv off, read both you voltage and the Arduino Vcc at the chip and see what's happening.
If absolute (not relative to Vcc) accuracy is important you can add an external voltage reference and connect it to VREF and program arduino to use it..
As a motor-head, I like your idea! Have you found examples of other people using a MAS for a flow bench before?
If you need high resolution you can get an I2c 24bit ADC....analog to digital converter that gives finer voltage measurements. You will have to come up with a reliable external voltage reference with it also.
If you can post a datasheet on the MAS someone may be able to help you more.
The MAF sensors I've looked into are nowhere near so simple. They are basically a hot-wire anemometer -- wikipedia has a useful write-up.
The wire element has an electrical resistance that changes with temperature. You have an electrical circuit that feeds power into the hot wire to achieve a set resistance value. The amount of power needed is a function of (1) the amount of air flow, and (2) the temperature of the air compared to the temperature of the hot-wire. The electronics required are not super complicated, but calibrating it all is likely complicated, and you do need a temperature sensor as well.
Simpler is a rotating cup or windmill with an optical sensor. Then you just have to count pulses.
All the MAF sensors I've run into as an auto mechanic the hot wire control mechanism has been built into the sensor, and it just outputs 0-5v (actually 0.5-4.5, much like a TPS) for air flow, and another signal (voltage ranges vary) for air temp.
Thanks to all your interest! both devices output 0-5v.
RE: cyclegadget- i agree the resolution of the stock analog input is not as exact as i would like, so i want to integrate your idea. As for reliable voltage source, nothing that a voltage regulator and pass transistor cant fix.
i have so far built my bench with the downtube that will go to the sensor (4inch) with adapter to 3inch (the sensor size). so i am set for the hardware side.
im not willing to spend much more than about 10 dollars on extra parts (married... ugh.... u who are married know what im talking about...)
I think I remember something about these sensors being calibrated to suit the relevant bore of hose that they are located in. Does your rig match the original, intended hose diameter?
wildbill- thanks for that, so next is to try to put that into the program.
Can anyone suggest a chip to measure voltage more precise and communicate to the arduino?
Also, before the mass air to the vacuum source i am making a equalizer made from lots and lots of straws in a pipe, thus to even the flow of air (and get rid of turbulence) before the sensor
daveg360- yes i am going to use the original hose that came off the mass air sensor to the engine (thumbs up)
ok so i just found a HUGE flaw in my experiment. when the vacuum source is completely blocked before the mass air sensor (like blocking it with your hand), the vacuum freaks out the Mass air sensor into thinking it is flowing air. so my theroy of this is because it is in a vacuum, vacuum = nothingness, meaning sucking heat away from everying, meaning pulling heat off the heated element on the sensor... at least thats my thinking. so i thought of putting the sensor on the exhaust of the vacuum. well the vacuum is not sealed so it actually flows air when the inlet is blocked....
ok so now im thinking about using the manifold absolute pressure sensor. i will calibrate the program with a known orifice size to know flow.... at least thats my thinking. any thoughts?
You might be getting electrical noise from the vacuum motor. I am pretty sure it has brushes like your 110volt hand drill would. It could be produce a lot of interference in your circuits. If you can add distance from your circuits to the vacuum unit you may be able to fix the problem.
In refrigeration, vacuum can make liquids "flash off" and therefore produce cold but, in your case I can't see how your sensor would be loosing heat from a vacuum condition.
But wouldnt you be better off with a MAF off a much smaller engine?
I hate talking imperial but 5L is a 302 ci. Back in the day a common carb would be a 650cfm.
I very much doubt your vacuum cleaner would flow anywhere near that resulting in huge reduction in resolution.
But I am guessing you already had the 5L MAF right?
Packhorse has a good point. You might be better off with a maf off of a compact car. I am not sure where you could get one off a smaller engine than maybe an Ford Aspire or maybe Escort.
