Forums » General Electronics Discussion

I know next to nothing about electronics but I'm trying to build a variable speed driver for the fuel injectors on a small turbo jet engine I've built. What i need to achieve is a 12-15v or earth pulse of 2-5ms variable between 3hz (ish) and 3khz (ish). I have built one using a 555 timer output through a mosfet but a square wave is no good and I don't know how get just a short pulse output? I also need to be able to switch drive of 1-4 injectors on the same pulse! Someone please help me its the last piece of the puzzle!

"need to be able to switch drive of 1-4 injectors on the same pulse"

Please clarify what you mean by this.

More data on the injectors (operating voltage/current/resistance/inductance) would help too.

What circuit protection (against back emf, short circuits etc.) do you require?

How accurate do your pulses need to be?

Do you need some form of feedback for speed control?

Hello Hip. How much current do you need?

Hi, the injectors are 12-15volts, 26mA, 11.1ohm and I don't know about inductance. They are normally earth triggered but think that matters too much as polarity is not important. The "1-4" thing is I will need to be able to switch each one on and off as required or have 2 or 3 or 4 on at the same time but all on the same variable frequency. I would like to do this electronically but i do have a back up plan of a variable speed motor driving a hall generator then mosfet that or something but that's a little over complicated and bulky! Grateful for any help, Thanks!

Hmm..

"the injectors are 12-15volts, 26mA, 11.1ohm"

Er, 12V across 11.1Ohms = 108mA.

12V/26mA = 461Ohms

Please reconcile: they can't both be correct at DC.

"They are normally earth triggered "

That description is too woolly. Do you mean they are supplied by a high side positive 12V to 15V supply with a switch between their low side ends and ground?

Like this:

where Rload represents your injector?

Yes supply 12V -15V and switched low side to ground! I really don't know very much ab out this stuff, those are just the readings i took from injector with my multimeter and 12V dc

This is the circuit that i built to try and do the job as a tester (just 1 injector). It performs at the correct frequency ranges but it has a square wave 50% duty cycle output! the one i built has a IRF1405 MOSFET

This should do what you want:

It has only simple manual injector selection. Did you need some form of electronic selection mechanism?

You will need to do your own sums, selecting the appropriate Ct1, Rt1 and Rt2 values to get the frequency, Thigh and Tlow values you want.

Be sure to decouple the 555 properly as close as possible to the device vcc and gnd pins (refer to device datasheet for detailed info).

The example shows vcc as 5V. In practice you can make this 12V - 15V with no modifications.

Note that the CL 555 model only works properly at vcc = 5:

About the CL 555 timer.

The high level voltage at the OUTPUT pin of the CL 555 timer model does not track the vcc supply. It is fixed at 5V.

The CL model uses a fixed internal supply of 5V and ignores the voltage on the vcc pin. This means that the internal switching thresholds are fixed at 5/3V and 10/3V.

This is incorrect and can be misleading.

The circuit below correctly models the output swing of the real 555 timer although it does not accurately model the output stage behaviour vs. load currents.

Simulate > Time Domain > Run Time-Domain Simulation

https://www.circuitlab.com/circuit/9fu6wh/555-timer-as-astable-multivibrator-oscillator-model-error/

There’s a bug report about this here:

https://www.circuitlab.com/forums/support/topic/5yc7a6mu/bug-report-high-level-voltage-at-output-pin-of-cl-555-timer/

and a lengthy discussion about some of the other “faults” on the CL 555 timer model here:

https://www.circuitlab.com/forums/support/topic/e4wyttj4/bug-report-555-timer-model-vcc/#comment_3885

:)

Thank you very much! That's awesome, I'll build that soon. Thanks again!!

Donations gratefully received.

:)

Beware: there's no short circuit protection in the MOSFET drivers.

You probably need to at least put a fuse in the vcc line.

Check it with dummy loads first (resistors, LEDs or incandescent lamps). You'll probably need a scope to see < 10ms pulses at low prfs.