Is this project doable?

Projects General Guidance
1

Hello. I am looking to possibly make a controller that takes the input of 6 Piezo Injectors and converts them to another voltage to power 6 additional low impedance injectors. The Piezo injectors operate at 100+ volts. I am going to need to determine the PWM on the injectors and then cross reference them to how low impedance injectors operate. So if a Piezo is operating at 100% and variables X, X, and X are set, then I need to put the low impedance injector in the same state. I know this is going to be testing to see how they operate. It will basically monitor all 6 piezo injectors. Once one of them start flowing fuel, it should start low impedance injector to flow as well. So 12 injectors, 2 for each cylinder.

Here is a little diagram I drew up.

[Album] imgur.com

I am looking to get input on if this is possible and what tips you have. Or if I should not attempt this as it is to hard.

2

What calculation would tell you the piezo is operating at 100%. !00% of what?

Paul

3

Well if I get in the car and give it lots of load then I can monitor the injector to see how it reacts. I could also monitor it at idle. I can monitor anything on my car from boost, fuel trims, fuel pump psi, wgdc, load, etc. So I could check all of that against how the injector is acting.

Also from what I have read, the fuel injectors are capable of supporting over 1000hp. It is just getting fuel into the injectors that we run into. So im sure the injectors won't actually be working at 100%.

4

BUT! You ask the Arduino to compute 100% without giving anything to base that calculation on!

Paul

5

Two key issues you have to work out.

  1. your inputs.
    There's the physical problem of the "over 100V". What exactly is this signal? What is the voltage? What current is used? How would you connect a sensor to these things?
    The data problem: what actual signal do you get, and what constitutes 100%, or 60% for that matter.

  2. your outputs.
    What exact output do you expect, in terms of voltage and current required? How are these signals going to be connected to the actuators? What does constitute "100%" here?

6

Hi,
OPs image;

fsh8joV.png
fsh8joV.png1303×616 6.51 KB

Tom.. :slight_smile:

7

Hi,
If you are going to ram extra fuel down your engine, how are you going to ram the extra air required to make the mixture in the compressed cylinder explode?

Piezo injectors sound like electronic diesel injection, definitely not the sort of thing to play with.

Or are you playing with a direct injection petrol/gasoline engine and putting a low pressure injector in the inlet manifold like earlier technology?
Again I would ask, how are you going to ram the extra air required to make the mixture in the compressed cylinder explode?

How do you envisage driving your low impedance injectors?
What exactly is your project?

Can you please tell us your electronics, programming, Arduino, hardware experience?

Tom.... :slight_smile:

8

Reading the Origional post, he is using piezo injectors.

they need a PWM signal. that is a series of on and off at a high rate with durations of on and off times.

what needs to be known is the frequency of the pulse train

one has to look at the start of the pulse, the duration of ON and duration of OFF, then the next pulse.

dy definition, 100% of a PWM is ON.
by definition, 50% is ON 50% of the time, and off 50% of the time.
by definition, a piezo has to vibrate so 100% is the maximum frequency that the device needs to operate at maximum.
so 100% is a little confusing, but knowable from the data sheet

Since we do not care what voltage, amps, etc are being used that is not really important. We have the technology to monitor a signal and isolate it regardless of those things.

what IS important is the frequency of the PWM

as I see the question :
can the Arduino monitor 6 individual PWM signals, calculate, then output 6 more at timing that would not make the engine suffer. ie : response would be in real-time.

it might be better to use discrete components and not try to run software. that way the speed of response will be orders of magnitude better.

9

dave-in-nj:
Reading the Origional post, he is using piezo injectors.

they need a PWM signal. that is a series of on and off at a high rate with durations of on and off times.

what needs to be known is the frequency of the pulse train

one has to look at the start of the pulse, the duration of ON and duration of OFF, then the next pulse.

dy definition, 100% of a PWM is ON.
by definition, 50% is ON 50% of the time, and off 50% of the time.
by definition, a piezo has to vibrate so 100% is the maximum frequency that the device needs to operate at maximum.
so 100% is a little confusing, but knowable from the data sheet

Since we do not care what voltage, amps, etc are being used that is not really important. We have the technology to monitor a signal and isolate it regardless of those things.

what IS important is the frequency of the PWM

as I see the question :
can the Arduino monitor 6 individual PWM signals, calculate, then output 6 more at timing that would not make the engine suffer. ie : response would be in real-time.

it might be better to use discrete components and not try to run software. that way the speed of response will be orders of magnitude better.

The voltage is also a factor in how much fuel is sprayed. I don't know the max voltage these injectors take. And I don't know how much it takes to open them half way. But I planned to sort of monitor it to determine how much volts the injector gets when I am pegged out at 7k RPM. It will take some trial and error with some logging to determine how much volts will hold the injector all the way open.

So I will need to get the volts. This will tell me how much of the injector nozzle is open. I also need to get the pulses. How long each injector sprays, and how many times. [3ms][off][3ms][off], etc. So if I can take the input of the voltage I think it should tell me the pulses. [110v][off][110v][off][110v][off]. So 3 pulse at 110v. Then After I have the data showing how much fuel comes out of the injector in that pulse, I can create a table to convert it to the low impedance injector.

I think you know what needs to be done as you pretty much nailed it. But in terms of discrete components, can you elaborate a little more? This would be my first project in this sort of stuff. I have never constructed my own chip like this and I am not familiar with all the hardware stuff for the electronics.

10

IF you are not recording the signals with a recording oscilloscope, you have no idea of any of the parameters you are suggesting.

Paul

11

Paul_KD7HB:
IF you are not recording the signals with a recording oscilloscope, you have no idea of any of the parameters you are suggesting.

Paul

A recent product that just came out for our car is a kit that allows us to run E85 and 93 octane blends. AKA flex fuel. One of the ways it works is by taking in a voltage from the tmap sensor. The TMAP sensor is a sensor that basically reads how much boost (psi) is getting put into the engine. And then it scales it on 0-5v. So if I was running lets say 10psi of boost, the signal may be at 1.85v. As I continue to go up in boost up to 20psi, the volts will continue to go up.

This flex fuel kit will read that signal and change things in the engine to account for more boost. Is this a similar concept or different?

I am trying to understand this, and thank you for your help.

12

Also, what about building a oscilloscope into the chip?

13

pysical:
The voltage is also a factor in how much fuel is sprayed. I don't know the max voltage these injectors take. And I don't know how much it takes to open them half way. But I planned to sort of monitor it to determine how much volts the injector gets when I am pegged out at 7k RPM. It will take some trial and error with some logging to determine how much volts will hold the injector all the way open.

So I will need to get the volts. This will tell me how much of the injector nozzle is open. I also need to get the pulses. How long each injector sprays, and how many times. [3ms][off][3ms][off], etc. So if I can take the input of the voltage I think it should tell me the pulses. [110v][off][110v][off][110v][off]. So 3 pulse at 110v. Then After I have the data showing how much fuel comes out of the injector in that pulse, I can create a table to convert it to the low impedance injector.

I think you know what needs to be done as you pretty much nailed it. But in terms of discrete components, can you elaborate a little more? This would be my first project in this sort of stuff. I have never constructed my own chip like this and I am not familiar with all the hardware stuff for the electronics.

And I don't know how much it takes to open them half way.
This will tell me how much of the injector nozzle is open

You don't seem to understand that the injectors are fully open or fully closed.
They are not held half open, they need the full pressure and flow of fuel for a particular time to meter out the required amount of fuel.

What electronic test equipment do you have to analyse the engine?
Because you will be trying to monitor a high voltage signal in a capacitive circuit, I'm not sure how the injector driver circuit or the injector itself will react to any load.

A recent product that just came out for our car is a kit that allows us to run E85 and 93 octane blends. AKA flex fuel. One of the ways it works is by taking in a voltage from the tmap sensor. The TMAP sensor is a sensor that basically reads how much boost (psi) is getting put into the engine. And then it scales it on 0-5v. So if I was running lets say 10psi of boost, the signal may be at 1.85v. As I continue to go up in boost up to 20psi, the volts will continue to go up.

The reason you could not run flex fuel without the kit is probably because of the fact that the engine will knock and that ECU does not have the range to adjust.
As you point out the TMAP is read and its signal monitored, but in conjunction with the engines knock sensor, it sounds like the kit is just to let you run flex fuel, any extra power I don't know.

Reading signals from TMAP sensor is nothing like reading injector pulse timing

But what has this to do with your extra injectors and added fuel feed?

Tom... :slight_smile:

14

pysical:
Also, what about building a oscilloscope into the chip?

http://www.instructables.com/id/Girino-Fast-Arduino-Oscilloscope/

That is an incomplete project and will not be of any help to you as it has no display or input voltage scaling or timebase calibration etc etc.
Tom... :slight_smile:

15

TomGeorge:
And I don't know how much it takes to open them half way.
This will tell me how much of the injector nozzle is open

You don't seem to understand that the injectors are fully open or fully closed.
They are not held half open, they need the full pressure and flow of fuel for a particular time to meter out the required amount of fuel.

What electronic test equipment do you have to analyse the engine?
Because you will be trying to monitor a high voltage signal in a capacitive circuit, I'm not sure how the injector driver circuit or the injector itself will react to any load.
The reason you could not run flex fuel without the kit is probably because of the fact that the engine will knock and that ECU does not have the range to adjust.
As you point out the TMAP is read and its signal monitored, but in conjunction with the engines knock sensor, it sounds like the kit is just to let you run flex fuel, any extra power I don't know.

Reading signals from TMAP sensor is nothing like reading injector pulse timing

But what has this to do with your extra injectors and added fuel feed?

Tom... :slight_smile:

For one thing, by applying a little less electricity, the piezo crystals expand less so the injectors can open partway. A smaller opening means a longer injection time, which is beneficial when trying to accurately inject a tiny amount of fuel, such as when a car is nearly coasting.

https://www.caranddriver.com/features/tech-department-piezo-fuel-injectors-explained

The voltage matters on how much they open.

The reason I brought up the flex fuel kit was because I thought it was similar. It is reading a signal 0-5v and then acting in real time based on that.

My idea behind my project was to read the injector signal and act on it in a way to fire another injector in the similar ratio. The additional injectors wouldn't begin to operate until the piezo are operating harder. So no reason to spray fuel from the additional injectors at idle.

If 0 = Idle and 100 = full throttle at 7k rpm. I could get the pwm at idle, the pwm at full throttle, and then find a middle ground to use for activating the additional injectors.

As far as the Oscilliscope. If I buy one, monitor the injectors at different rpms/loads/etc. Then I can build a table.

At X RPM, X Load, X Boost, X etc the injector is firing at X pulses, X Voltage.

Then I can begin to spray the additional injectors when the piezo injectors are starting to tank and can't keep up.

16

I think I am going to put this project on the back burner until I learn a little more about all of this.

17

Hi,
From your link;

For one thing, by applying a little less electricity, the piezo crystals expand less so the injectors can open partway. A smaller opening means a longer injection time, which is beneficial when trying to accurately inject a tiny amount of fuel, such as when a car is nearly coasting.

You are looking at near full throttle, at that point you will not have part injector opening.
As stated its used when TINY amounts of fuel are needed. So your extra injectors will be shut off anyway.

At 7000RPM you need fast injector response to get fuel quickly into the cylinder, part opening will only slow that proceedure down.

I watch with interest as to how you progress.

Tom.. :slight_smile:

18

pysical:
I think I am going to put this project on the back burner until I learn a little more about all of this.

Good Idea..