MAF based Methanol Injection with PWM Valve

Using Arduino Project Guidance
1

Good day community!

I'm currently a car tuning enthusiast, a computer programmer by employment and DIY runs in my blood!

I'm about to embark on a Car performance project which will utilize arduino's nano controlling the flow of methanol with a fast PWM pulsating valve.
For those who are not aware how methanol injection works:

Water/methanol injection boosts a car's performance by introducing into the cylinder a water/methanol mixture that acts as a cooling and anti-detonanant additive (From the water) as well as a boost in octane and oxygen (from the methanol)

But just like the gasoline electronic fuel injection system in our cars, the methanol i plan to spray will need to be varied with load (a measure of rpm and airflow)

So the system should work like this:

  1. Plumb a tubing from a 200psi recirculating bypassing pump from back of the car to the front, where a PWM valve and a misting nozzle will deliver the fuel.

  2. Have the Arduino computer use the car's AFM (Air flow meter) to determine how much air is being used and therefore be able to modulate how much water/methanol we will inject into the cylinder.

  3. In the case of a low fuel (low methanol) situation, the arduino should stop pumping liquid and send an output current to a pin that will tell the car's ECU not to run boost above a certain point
    --- (don't worry about this part, we have it figured out, the tuning too. I can go in details if you want)

I have one group of question for the community so far. When reading the 0-5V output from the car's sensor (wired in parallel), would it change the voltage going to the car's ECU after the location i tap into it?
Also would a pull down/pull up resistor be needed? If so how would this affect the reading going to the car's ECU? Wouldn't it change it to one extreme or the other?

Thanks for reading, I await your ideas

2

you have two options, read the sensor directly or get the data from the ECU.

your request was to read the sensor directly, fine.

any electrical connection will have some effect, the idea is to have as little effect as possible.
by using an op-amp, you can add a load with a 100k ohm resistor. this means the the load would be about 0.005mA or some such.

without knowing how the sensor is working, there may be no effect, the sensor may just output power to compensate, or this may be a sacrificial load because of the sensor.

obviously, the way to tell is to run the motor on the bench and connect and disconnect your new sensor to see if there is any measurable change.

I would expect that you will see something like the 7th or 8th decimal place change a couple numbers.

by adding an industrial op-amp. this is an op-amp circuit that has multiple stages. one is the primary stage, another allows you to off-set the value, another allows you to expand the range,
the idea is that you for all intent, isolate the reading on the first stage and pump up the power.
the next will allow you to set the span by controlling the amount of gain.
the next offers the offset of the zero so that the 0 reading is at the zero you want, which is usually 0.

I know, it sounds funny, but if you have a 4-20 reading, the equates to 1-5 volts. or 80% of the full scale of the Arduino input. by manipulating the zero point, you would drop that 1 down to zero and then have 0-4 volts. and with that gain adjustment, you could spread the span to have 0-5 volts. adjusting both the zero and span if often very desirable.

in another life, the techs called it slant and elevation. the span is the slant. you adjust that to get the ends to have a separation, in your case might be 200 CFM or some such, and then adjust the elevation of the slant on the graph so that the zero point is at zero.

just to beat the dead horse, most op-amps are designed to go from -20 volts to +20 or(supply rails ) some such spread could be much different, but positive rail to negative rail. that way zero is a point on the slope that is a point with power.
some are offered as rail to rail, meaning they can output a reliable signal AT the ends of the power supply values. really just NEAR the ends of the rails.

we live in a 0-5v world and at 0, there is no power. most single end, or (0 to voltage) op-amps get near to 0 and near to 5, but at these extremes on a 5v supply are either without power or completely saturated. you have to pay lots more for an op-amp that can get close. the old phrase of getting close costs money, how close can you afford to go ?
for 15 cents, you can get an op-amp with great specs, can handle from +15 to -15v (dual supply requires -rail ) and be rock steady at 0 and 5v. or, for $12, you can get a single ended, rail to rail that will do almost the same thing.

FAIRCHILD SEMICONDUCTOR KA4558 from Vertical for 15 cents

to the are you out of your F#Q#$# mind ??!! $832 each....
http://www.digikey.com/product-detail/en/PA52A/598-1436-ND/1762035

for the op-amp, giving it =12v and 12v eans that your span could be -12 to +12, you can see that span stage is needed badly. and since that span is across the range, zero is not a frill, but required in order to get a useable signal.
since you live in a world that has something near 14-ish volts, give or take a couple volts, .... getting 5 volts is both easy because you have much more available, and hard, because it is so noisy.

just cause it's 5 am....

an industrial op-amp will often have two front ends. these both look at the signal. one uses the ground and reads how high the signal is, the other uses the high signal as the reference and reads how low the ground is. this would create a perfect mirror curve thereby eliminating any ill effects in either method of reading.

please note that this post is very casual and in broad generalities. the sloppy engineering is to aid in comprehension that seems to be inversely proportional to adherence to technical terms.

3

Wow, that's a whole lot to consume :o

4

donperry:
Wow, that's a whole lot to consume :o

op-amps can be a whole line of study.

I'm hoping the reading was easy.

5

My father was telling me the same thing, to use op amps. Guess I'll have to go read up on that and figure ...

Hey, by the way, did I tell you the sensor is already outputting a voltage range 0- 5v?

I realize the op amp will amplify low current/voltage but is it necessary if the voltage is already high?

6

Don't re-invent the wheel..
Or at least part of the wheel.

http://forum.arduino.cc/index.php?topic=95037.0

That post may help with getting the MAF information you need.

Jason.

7

That will not work as my ODB2 connection will be occupied with a tuning software. Nice solution to read sensor readings though.

8

SO i figured after reading and video watching (thanks to dave ^) that I need an Op amp to function as a buffer. Thing is it must have a low voltage threshold to pick up on the maf.

So let me go over, Dave or anyone can answer:

  1. The sensor already outputs 0-5v, this means i won't need to amplify anything, right?
  2. An op amp buffer could work in this case?
9

if the sensor is putting our 0-5v, then connect it to an analog input on your arduino.use a 10k resistor to connect that signal to the arduino.

if the signal is weak, and the readings on the MAU are effected, then you need to up the resistance. this will delay the speeds of response for the arduino ADC.

the op-amp would take an odd signal and convert it. we use 4-20 for long distances on industrial sensors ans as was shown, you need to change both the zero and span when converting to voltage.

I would try the 10k to see if that does the trick.

from what I have read, you can go higher, but that slows down the speed of the reading, so there might be some sweet spot for your project.