Low impedance injector driver - HELP!

there ought to be a better way but I haven't thought of one yet.

Click to expand...

Nor have I! The problem is we're dealing with high currents storing considerable energy in the inductance of the injector, and that energy has to be dumped quickly somewhere. We can't beat the laws of physics.

Hi guys....

After some days testing the circuit as you advised me... I took the next conclusions:

It doesn't work...looking at the current through the injector it never reaches more than 1.25 Amps. So the reset never reaches and then the comparator doesn't work. I selected the FQP20N06 MOSFET to drive the injector. I noticed that the Vgs is always near 5V.Could be that the cause why the injector never get 7A? Any other idea? To fix that I had though in a charge pump to drive the mosfet...

Apart from that I found this device: BTS 410 E2 of Infineon. Looking at the data sheet, Could I use this to drive the injector instead of the circuit you proposed me?

Thank everyone in advance !!

Well it should work so lets troubleshoot it a little more. I hope you have a scope so we can see pulses.

What do you measure:
Vcc on the 555
ECU input
Collector of Q1
Trigger of 555
Reset of 555
Output of 555
Collector of Q2
Output of LM119
Gate of FET

OK. I will do these measures. As I cant pass the scope waves to the computer I will describe them for one ECU pulse.

Cheers

Just to start... I have a problem with the reset... the voltage in the reset never reach 0V. I have checked the transistor several times and its OK. It should open to pull down the voltage when the current is 0.7V....Is there any other way to pull the voltage down?

It will only go low when the current gets to 6 or 7 amps, so that may be okay. I suspect the problem is on the input side so we should take all the measurements then see the first place something is wrong.

I checked everything, and after correct some mistakes with the reset connection the systems works...more or less.... the N555 is working OK but the transistor responsible for reset the timer open before 7Amps. I measured the voltage at the base and its 800mV when the current through the coils is 2.5A. That's why the reset is made before the 7A peak current. I need to investigate that....Any idea? Thank you very much for your unselfish help.... I hope you are learning as much as me......

Hmmm. Well ohms law says that .8 volts across the .1 ohm R5 is 8 amps. So if R5 is .1 ohm the coil current should be 8 amps. So if it isn't the other current must be going somewhere. Could the zener diodes be in backwaeds? (seems like they would be very hot) or maybe the resistor is 1 ohm instead of 0.1 ohm?

PS. It's fun!

or maybe the resistor is 1 ohm instead of 0.1 ohm?

Click to expand...

If it were then the transistor base at 800mV implies a coil current of only 0.8A, but the OP is getting 2.5A. How is the 0.1 R5 terminated? Solder joint? Crimp connector? Wire wrap? Perhaps the termination resistance is adding another 0.2 or so to the 0.1?

Hi #ronv and#alec_t!

I have been quite busy the past days.... Sorry. I made some progresses, and I will explain you my results. The good news is that I have found the desired signal...measured with a current probe at the injector cable....but I made some changes......and I have some conclusions.

1. I checked everything regarding to the sense resistor and was OK, but with the circuit you proposed me, the reset signal was at 3A. I am using a protoboard with discrete elements, so maybe the resistor is changing somehow....no idea.. to solve this I used another resistor between the base and the sense resistor voltage measurement point in order to increase the voltage in this point but not in the base..... in that way with a 150k resistor I achieve 7A in the injector and the reset signal is good enough. Initially, the supply voltage to the timer was 5V and I had to pull it up to achieve enough voltage in the output to open the mosfet completely...I am using a (FQP20N06 if you want to check it)

2. I was some days trying to fix the comparator because it didn't work properly. But I was misunderstood... First I saw the output signal, and it never went up nor down.....I was confused...but then testing different values of hysteresis thresholds I understood that the problem is that the comparator I am using, a LM128 dual op amp, is not fast enough to toggle the output...so instead of this, for small threshold difference the output is constant...... I mean instead of 10V and 0V its 5V constant.....

3. The signal is 4A almost constant. But my question for you guys is: Because of this way of working is mosfet never open/close during one cycle, this means that the power is higher and also that the protection zeners never works....... Now I will try to implement the circuit in a PCB, and I am not sure if I will keep the circuit like this or I need to change something.

SORRY FOR THIS LONG MAIL.....

THANK YOU SOOO SO MUCH!

Seven amps may be a bit to much for a proto board.
The circuit as designed needs a regulated 12 volts for the low voltage circuits. This is required to fully turn on the FET and to set the current level on the comparator. We could do this with a voltage regulator like a 7812 or even a zener diode regulator running off of the high voltage.
The whole operation of the circuit depends on the voltage drop across the 0.1 ohm resistor. Its amplitude turns off the high current pulse and determines the pwm amplitude so "tweeking" it is probably not a solution. An LM119 or 319 would be good for the comparator.

Here is a theory of operation for the circuit.

A positive pulse from the ECU goes to Q1 which inverts it and creates an ~ 100 usec. Pulse via C1, R4 to trigger the timer. The timer fires and the positive output turns on M1 which in turn turns on the injector. Due to the inductance of the injector the current is slow to rise, but when it reaches ~ 6amps (0.6 volts across R5) Q3 is turned on resetting the timer and thus turning off M1. The current stored in the inductor decays through D2 and R5 to ground. When the current goes below 2 amps (.2 volts across R5) the output of the comparator at U2 goes high turning on M1 until the current goes above 2 amps. This cycle continues until the positive ECU pulse again goes low and forces the output of U2 low through Q2.

I got some LM119 to replace the lm158, but it doesn't work properly....I am connecting the ground1 to the -Vcc, is this correct? When I connect the comparator to 15V (I use a divider from the 24V), also I have connected the timer to this 15V, and then when I turn the circuit ON, neither the comparator nor the timer work! I think I am going to go back to the LM158, Any idea? Thankss!

If you are having a comparator problem try changing the value of R14. (This resistor is only necessary if the comparator is of the open collector output type).

I've decided to use the lm128... there's no a big change between both....

In order to select the voltage regulator I had thought in a switch regulator... I will drive four cylinder...in couples I mean, first two and then other two..... So the current requirement will be high....near 8A.... Do you think the sw regulator is a good idea? I decided this regulator because of the power dissipation.....

What do you think? Cheeeeeeeeeeers

The regulator is only needed for the circuit, so the current is low. The injector is still tied to the 28 volts. The comparator needs to be the open collector type - the op amp won't work.

Thr second power supply needs to be 12 volts or the hold current will be wrong.

Guys,

How do I find the LM119 in LTSPICE? I have been looking for it in my library and it doesn't appear...... Is there any library downloadable?

1insuare said:

Guys,

How do I find the LM119 in LTSPICE? I have been looking for it in my library and it doesn't appear...... Is there any library downloadable?

Click to expand...

hi,
If you join the Yahool LTS User group they have libs,, the LM119 is the same as the LM319,, the LM119 has a wider temperature range.
**broken link removed**

I sing in with my google account, but to join the group I need to make a new profile. When I attempt to create a profile the website doesn't work...grrrrrrrrrrrrrrrrrrrrr!!!!!!!!!!!!!!!!!!!!!!!!! I will try later...I need the libraries! Thank you #ericgibbs!

Can anybody pass me some libraries by mail?