Fuel Injector Driver.

Hi,

I am looking to use an Arduino to pulse some Fuel Injectors off a motorcycle and measure the fuel they produce in a given time period. These are 12v high impedance injectors, from some research I have done it seems that they could be controlled with an N-Channel power mosfet as a low side switch.

The solenoid in the injectors is between 11 and 13 ohms, so it will take just over an amp to open one. I am looking for some help selecting the correct mosfet(s) for this circuit. I could control each injector individually or control them all together, this would determine how many mosfets I need.

I was looking at this MOSFET here: Datasheet

Will this MOSFET work with 0 - 5v on the gate?

Below is the circuit I plan to use, the switch represents the signal from an Arduino digital output pin, the Lamp is the injector.

That is a good MOSFET choice for your use. It will fully turn on at 5V gate voltage.

You are missing the fkyback diode on the injector. A high reverse voltage is generated when the injector is turned off. Without the flyback, the MOSFET can be damaged.

A resistor between the Arduino output and the gate will limit the current to charge the gate capacitance, protecting the Arduino output. 180 - 220 Ohms. Not required but a good precaution.

Don't forget you need a controlled high pressure fuel supply, and atomised fuel is extremely (as in extremely) combustable - so this can be a pretty hazardous experiment and a little 12v spark would make it even more exciting .

You can look on line for a given injectors flow rate characteristic ,which is easier. Some injectors ( usually diesel) have a coding for their individual performance ( often loaded into the ECU when an injector is changed)

If you don't use fuel remember a liquid of different density and viscosity will flow at a different rate

lookie

Hello!

I'm curently doing very similar project.
I was planning to use four independent mosfets and steering them with one PWM signal.
The problem is that for one connected mosfet it works perfectly, but if I connect more mosfets it didn't.
Anynone know what is the problem?
I'm attaching the circuit

onlybike:
Hello!

I'm curently doing very similar project.
I was planning to use four independent mosfets and steering them with one PWM signal.
The problem is that for one connected mosfet it works perfectly, but if I connect more mosfets it didn't.
Anynone know what is the problem?
I'm attaching the circuit

Perhaps someone could help if you described what not working perfectly actually means.

Paul

Are those modules the ones with the IRF520 MOSFET? If so, be aware that they are not logic level MOSFETs and will not fully turn on with a 5V gate signal*. The FQP30N06L MOSFET that the OP linked in their first post is a much much better choice.

*See the Rds(on) spec and Figure 3 in the IRF520 data sheet.

How to post an image.

Paul_KD7HB:
Perhaps someone could help if you described what not working perfectly actually means.

Paul

Sorry, you are right, should do that.

If only one mosfet is connected to singal, injector open and close with the frequency of PWM signal. When both mosfets are connected (by gate of course) they both not working, are still closed.

groundFungus:
Are those modules the ones with the IRF520 MOSFET? If so, be aware that they are not logic level MOSFETs and will not fully turn on with a 5V gate signal*. The FQP30N06L MOSFET that the OP linked in their first post is a much much better choice.

*See the Rds(on) spec and Figure 3 in the IRF520 data sheet.

That is good point, thanks!
But did very similar circuit on bradeboard with IRLZ44N, and it behaves exaclty the same.

Can you post a data sheet for the Generator PWM? What are the specs of the 12V power supply (Volts, Amps)? What are the specs of the injectors?

Those modules have no flyback diodes, the diodes need to be added.

Have you tried to monitor the 12V supply while the circuit operates to make sure that the 12V supply is steady 12V?

groundFungus:
Can you post a data sheet for the Generator PWM? What are the specs of the 12V power supply (Volts, Amps)? What are the specs of the injectors?

Those modules have no flyback diodes, the diodes need to be added.

Injector are high-ohm ones (14,5 ohm). Calculated current is about 0.8A and measured one is the same.
In cicrciut "urz.1" and "urz.2" means injector + flyback diode - I should wrote that.

Couldn't find manufacturer PDF, but here is specification of PWM generator:

I have test it with multimeter and works fine - 50% duty cycle gives 2,5V, 100% gives 5V.
Even withount PWM generator one mosfet works correct just by switching by connecting jumper-wire to 5V and.

12V comes from 33A industry power supply.

onlybike:
Injector are high-ohm ones (14,5 ohm). Calculated current is about 0.8A and measured one is the same.
In cicrciut "urz.1" and "urz.2" means injector + flyback diode - I should wrote that.

Couldn't find manufacturer PDF, but here is specification of PWM generator:

Adjustable Square Rectangle Wave Signal Generator Module 2 Channel PWM Pulse Frequency 1Hz-150KHz

I have test it with multimeter and works fine - 50% duty cycle gives 2,5V, 100% gives 5V.
Even withount PWM generator one mosfet works correct just by switching by connecting jumper-wire to 5V and.

12V comes from 33A industry power supply.

What is the current limit on the supply set to?

What happens if you use a wire to jump two injector MOSFETs to 5 volts, at the same time?

Paul

hammy:
Don't forget you need a controlled high pressure fuel supply, and atomised fuel is extremely (as in extremely) combustable - so this can be a pretty hazardous experiment and a little 12v spark would make it even more exciting .

You can look on line for a given injectors flow rate characteristic ,which is easier. Some injectors ( usually diesel) have a coding for their individual performance ( often loaded into the ECU when an injector is changed)

If you don't use fuel remember a liquid of different density and viscosity will flow at a different rate

lookie

Hi,

I have a regulated fuel supply, I can either use the fuel system that is on the racing sidecar that this engine is connected to, I am mainly worried about finding 4 similar injectors (in fuel flow rate) at present. If I find the injectors are very different I will be making something that can supply cleaning solution to the injectors so it can be pressured through while being pulsed by the Arduino.

I have been testing the injectors today, manually connecting and disconnecting a 12v circuit to the injector, but obviously they can be pulsed much faster by an Arduino and I can also measure the difference in flow rates if the pulses are precisely controlled,

These are petrol injectors, of a 600cc motorbike engine. I have been looking for the specs of them but have been struggling to find out what flow rate they are rated at, and what duty cycle they are run at.

I take your point about the danger, I will keep that in mind.

I have ordered some of the MOSFETs I linked in my original post, they should arrive on Monday so will start experimenting. I have some of the 1N4007 diodes and so will be adding a flyback diode.

onlybike:
Hello!

I'm curently doing very similar project.
I was planning to use four independent mosfets and steering them with one PWM signal.
The problem is that for one connected mosfet it works perfectly, but if I connect more mosfets it didn't.
Anynone know what is the problem?
I'm attaching the circuit

If you are using the same PWM signal, can you not connect all 4 injectors to 1 MOSFET? Provided it can handle the current of 4 injectors.

A couple of things that might help:

  1. You can use hardware grade Mineral spirits for testing. It is very close to Stoddard solvent which is the industry standard for fuel system testing (not for every test but for 90% of the testing). Any error you might see between the Mineral spirits and gas is likely to be below your ability to measure.

  2. Unless this is going to be a short test, injectors like to see a full voltage pulse to open then a reduced current to stay open. see the below driver.

fuel injector driver

JohnRob:
A couple of things that might help:

  1. You can use hardware grade Mineral spirits for testing. It is very close to Stoddard solvent which is the industry standard for fuel system testing (not for every test but for 90% of the testing). Any error you might see between the Mineral spirits and gas is likely to be below your ability to measure.

  2. Unless this is going to be a short test, injectors like to see a full voltage pulse to open then a reduced current to stay open. see the below driver.

fuel injector driver

Hi,

I believe that is true for low inpedance injectors but I am trying to control high impedance injectors

groundFungus:
That is a good MOSFET choice for your use. It will fully turn on at 5V gate voltage.

You are missing the fkyback diode on the injector. A high reverse voltage is generated when the injector is turned off. Without the flyback, the MOSFET can be damaged.

Alas with a freewheel diode like this the injector probably won't run fast enough. You need rapid turn off as
well as rapid turn on, so the reverse voltage on switch off must be allowed to rise, perhaps to 12V or more.

For this use a 12V zener and diode combination across the coil.

But best of all the datasheet for the injectors ought to specify the drive parameters for operation.

MarkT:
Alas with a freewheel diode like this the injector probably won't run fast enough. You need rapid turn off as
well as rapid turn on, so the reverse voltage on switch off must be allowed to rise, perhaps to 12V or more.

For this use a 12V zener and diode combination across the coil.

But best of all the datasheet for the injectors ought to specify the drive parameters for operation.

Like this?

The service manual for the motorcycle does not mention the part number of the injectors and so I am struggling to find a datasheet for them.

Nearly - gate driver chip is probably needed for accurate switch timing, but the zener snubber is right sort of thing.

BTW there are specialized injector driver chips for example LM1949 - datasheets for such chips are probably worth reading.