A complete mystery with a fuel injector pulse generator

[Iawia]

I recently put together a pulse generator for my fuel injectors. I made it so that when I press a button, a pulse is generated at 12v. When I place the fuel Injector wire directly to 12v, the injectors open, when I hook up my pulse generator to it ( which pulses 12v 1 time for 1 second ), it does nothing. What's happening??? can someone help?

see attached drawing. the R3 load represents the fuel injectors.

Thanks

 

[jpanhalt]

In reference to your pulse generator, when you say, "which pulses 12V 1 time for 1 sec," do you mean it is supposed to do that or it actually does that? In other words, is relay K1 working? If not, I suspect that the PIC is simply not enough voltage to turn on the BS170 (Q1). That mosfet is not a typical logic-level device. At 5V GS, drain-source current at 5 V is about 300mA (i.e., Rds = 17 ohm).

John

 

[KeepItSimpleStupid]

Another thing to pay attention too is the operate time of your relay. You don't typically drive an injector with a relay.

 

[Iawia]

In reference to your pulse generator, when you say, "which pulses 12V 1 time for 1 sec," do you mean it is supposed to do that or it actually does that? In other words, is relay K1 working?

The pulse is working correctly, providing 12v on demand at the button push. I am using the transistor to provide signal to the solenoid in the reed relay, as not to put any more load on the pic than necessary. Pushing a solenoid with a pic I wasn't that keen on. I am using a standard NPN to do it, not the Q1 BS170.

I checked the resistance on the fuel injector and it is ~12 Ohm. Do you think that the resistance being so low has anything to do with it? What are fuel injectors typically run by if not activated by relays???

Thanks!

 

[jpanhalt]

You have got to tell us what the "standard NPN" you are using is.

The more information you provide, the better, when trying to solve a mystery. Please start with a correct schematic.

John

 

[KeepItSimpleStupid]

https://www.lmgtfy.com/?q=injector+driver

The other main issue is that your relay "operate time", e.g. 10 ms might be on the order of your injector time 10's of ms and that' s a problem.

Check what the sourcing current is available for the PIC port. Bet you don't have enough when you multiply it by the Hfe of the transistor.

FETs require very little current. Logic Level FETS have lower Vth

 

[shortbus=]

If you are applying 12V to the injector expecting it to turn on, that is the problem. All injectors that I'm aware of are supplied by constant 12V and turn on by grounding the other side at the ECM. Check out this link - https://www.youtube.com/watch?v=JTCgUY51g0s

 

[debe]

Injectors are usualy driven by transistors or FETS pulling the solenoid to ground. ( must have been typing at the same time)

 

[Iawia]

Hi Keep,

can you explain this to me a bit more? I have researched the info you asked for and it appears that the pic pins can receive a max input of 75 mA, and output a max of 25 mA. Do you think I should remove the 10k resistor ( for fast discharge of the gate ), or increase the value of the 100 ohm resistor?

-iawia

 

[KeepItSimpleStupid]

We basically don't know what your using. In the diagram which uses a FET, the 10K prevents the FET from turning on accidentally with leakage currents and gate capacitance. The 100R limits the current to the gate. The FET is likely a logic FET. I didn't look it up.

Your injector might be known as a high impedance injector. From what I read, I think, the low Z, require the current to be cut once activated.

I did this to a few solenoids I used at work for shutters although for different reasons. One was a rotary solenoid that wasn't activated for long periods, so that was a traditional kick + drop to a lower voltage. Without the drop, the solenoid would burn up. Without the kick, it would not open. The other one was because there was a 1 mm expensive piece of glass nearby and I wanted slow opening, so I used a PWM chip with soft-start.

 

[Iawia]

Hi Keep,

The npn transistor i am using is a 9014 (attached). When I remove the 100 ohm resistor off the board and replace it with a wire directly to the relay it works. However, it will pulse LEDs that I hook up to it, but it will not pulse the injector. I think you are right Keep, there is something more to these injectors than just a mere pulse.

Im so confused with this thing. argh.

 

[KeepItSimpleStupid]

The Hi-Z injectors are constant current and driven at 1 AMP. **broken link removed**

take 25 (mA) * 100 (hfe, but not min) ; that's 2500 mA (fixed error)
The xistor is only good to 100 mA

It's like trying to push a car with a bicycle. OK, not that bad, but it won't work.

See what happens when you provide better info?

 

[Iawia]

Ok I am beginning to see the issue here ... I think.

what is hfe?

So, what you are saying is that the 9014 npn cannot handle the drain of the pulse current + solenoid switching current?
Those being computed at i = 12 v/150 ohm (solenoid) + 5 v/ 100 ohm (pulse) = 210 mA

Sorry I am still learning. That would make total sense why this is not working. A good exercise in power requirements.

 

[KeepItSimpleStupid]

Hfe is the term for current gain. Ic = ib * Hfe
Hfe varies all over the place. Some transistors even though the same type will have different suffixes to denote the current gain range.

When a transistor is used for a switch, the term Vce(sat) basically tells you that the transistor will drop that amount of voltage when fully turned on.

So, if you need 1 Amp of current (doesn't matter the voltage at this point) and you have a transistor with an Hfe of 100 say, you will need a 0.01 Amps of base current to turn it on. Technically 1A/(12V - Vce(sat). I ignored that for now.
100* 0.01 = 1 Amp. You have to plan for the lowest Hfe in this case.

I made a math error above which I fixed: 25 mA * 100 Hfe = 2500 mA or 2.5 Amps BUT Ic of the transistor is only rated for 100 mA, so it won't work.

The base resistor should be computed by say R< (5V - Vbe)/Ib where 5V is the Pic output voltage.

Your transistor does have a range of Hfe depeding on the suffix of A B C D
hFE 60 ~ 150 100 ~ 300 200 ~ 600 400 ~ 1000

See the bottom of the datasheet.

 

[Ian Rogers]

I have researched the info you asked for and it appears that the pic pins can receive a max input of 75 mA,

Are you sure.... All pics can sink OR source 25mA max.... I think you'll find that 75mA is for all 6 pins...

Also the bs170 fet might not be able to power the relay coil... As KISS mentioned.. Although it can handle 500mA, you need a higher voltage on your gate to drive the relay.

Ditch the 100Ω resistor, it may be enough..

 

[jamie_s]

driving an injector from a relay is not optimal, eventually the relay contacts will fail due to the inductive load.

your much better off driving the injector the way it is intended, using a FET (or even a TIP120 or sim.) in place of Q1 as a low side swtich to the injector.

 

[jpanhalt]

Also the bs170 fet might not be able to power the relay coil... As KISS mentioned.. Although it can handle 500mA, you need a higher voltage on your gate to drive the relay.

Did you see post #4? He is not using a mosfet. In post #11 he revealed that he is using an NPN that is limited to 100 mA. I am not sure an accurate schematic has yet to be posted.

John

 

[jamie_s]

try this:

 

[KeepItSimpleStupid]

R1 is unknown. The IRFZ44 stands a chance. There are two bipolar transistors and one FET.

Alphabet soup. Yum!

 

[jamie_s]

R1 is unknown. The IRFZ44 stands a chance. There are two bipolar transistors and one FET.

Alphabet soup. Yum!

I am not sure an accurate schematic has yet to be posted.

John

Well i posted a schematic that will work.
R1 is dependant on what transistor or mosfet [Q1] the OP has available to them, any of the listed numbers for Q1 will work ok, seeing as its only a driver for testing the injector.