Hello all I'm new to the Arduino world and this is my first visit to this forum and have yet to buy and build my first Arduino project. The more I read and research this product and website, the more I realize I have yet to learn and how limitless this wonderful product is. I have been waiting to discover something like Arduino for a long time.
The project I have is a fuel injector tester for automotive engine performance diagnostics. I need to build a 12-15 volt driver for 8 injectors that can be energized at a variable speed (RPM) and for a variable pulse width (on-time).
To simulate variable engine speed and variable engine load I need to be able to adjust the frequency from 1 pulse per second to 60 pulses per second and be able to control the pulse width from .5 Milliseconds to 20 milliseconds (500 microseconds to 20 milliseconds)
I'm not an electronics expert but I do know that those speeds are way to fast to use relays as I have seen on some circuits on the internet. The mechanical movement of a relay is too slow.
I believe a MOSFET or similar switching transistor will be necessary. Another consideration is the collapse of the coil windings in an injector can have a kickback voltage of almost a hundred volts so some kind of protection component like a suppression diode I think would be necessary. Also two potentiometers on the board for rpm (pulse speed) and pulse width. Also according to my calculations at 60 pulses per second there can only be a maximum of 16.6 milliseconds of on time so there would be an area of implausibility with the two pots, programming of the code on the board I'm assuming would have to be written accordingly.
I don't know if I am asking in the right forum or even if the Arduino is a good option but from what I have seen so far I think I am at the right place. Any help would be appreciated or any recommendations on sources of information.
Some of these fuel injector test benches are very costly. As much as $200,000!!!!.
It would be an invaluable tool for me and not to mention make me the envy of my peers LOL. We have to send injectors to specialized facilities and becomes very time consuming and costly.
Thank you in advance and look forward to helping and returning the favour in the near future.
Hi, you could google "fuel injector driver", it will give you a range of projects an semiconductors that are suitable for injector units.
They are special drivers, as some have current control and inductive back emf components to help maximize injector speed and circuit protection.
This is VERY doable. Mosfet low side switching is the way to go.
Reading two potentiometers whilst outputing pulses based on their values is pretty simple stuff. So I can see how this would be fine for emulating the conditions in a running engine. At 60 pulses per second that would be 3600 pulses per minute (representing 7200 RPM of a 4 stroke engine) So your concept seems fine.
One thing I do wonder about though. Although we can quite easily trigger the injectors to open and close, I'm yet to see how you would detect that they have, indeed, opened and passed fuel, as intended, within expected parameters for the signals being sent to them.
I daresay, you could setup the test to begin and run a certain amount of time, then measure the fuel passed.
There's a couple of things to consider from an electric point of view. Assuming you're talking about standard petrol injectors (and not direct or diesel injectors) then they come in 2 main categories, Low-Z and High-Z.
Driving a High-Z is very straightforward and there are many standard power MOSFETs that would be fine. Low-Z injectors though are intended to be driven in what's known as a Peak and Hold manner, where there is an initial high current pulse to open the injector followed by lower current to keep it open. You can use some power series resistors and drive them like a High-Z injector, but if you want to measure them accurately (Which it sounds like you do) you'll need a specialist driver such as an LM1949.
In both cases, you do need to deal with the kickback from the injectors like you mentioned. The easy way though is to simply get a MOSFET with a nice high clamping voltage and it will take care of it all for you. Not only will this mean you can get away without diodes, but also allows the injector to kickback properly and close quickly (Using a simple passive protection circuit can mean the injector closes slower than it otherwise would). I can recommend something like a VND5N07 or to a lesser extent the VND7NV.
Either way, as far as the software is concerned, you'll need to take into account both open and close times etc, but on the whole it shouldn't be too difficult
Could you hack an engine ecu for the driver circuit? I would have thought it was possible to hack into the output pins from the ecu's microcontroller / input pins to the driver circuit. You would then have the added benefit of nice professional connections and loom etc.
Or just simulate the crank and cam signals with the arduino and feed them to the ecu.
