Flip-Flop switching between two injectors

[AnalogKid]

Just for grins, here is a single-chip solution that eliminates the 555.

U1A is the master oscillator. I threw in some component values for R3 and C3, but they change based on the desired frequency.

R1-C1 and R2-C2 are the same delay timers as before, setting how long the FIRE-x outputs are on. In this circuit the delay time is not affected by the gate characteristics of the MOSFETs.

D1 and D2 protect the U1C and U1E inverter inputs from a small over-voltage spike, caused by C1 andC2 when pins 2 and 12 go high.

I didn't mention this before - C4 is the power supply decoupling capacitor for U1.

ak

 

[AnalogKid]

Re-reading this thread, I might have misunderstood something.

Is FIRE on for 2 ms, is FIRE + HOLD combined on for 2 ms, or am I still missing something.

ak

 

[malc9141]

Pommie Mike. You can relax now, I am getting some help.

And to AK, many thanks. It will take me some time to digest this. But I will. Of course there are things I will misunderstand and will seek more help.

M.

 

[AnalogKid]

Please answer post #22. "Ah gots to know ..."

ak

 

[augustinetez]

The problem with both the redesigns so far is none of them allow for the varying time (t) in the waveform drawing a while back.

That is, INJ (B) opens at a different amount of time after INJ (A) depending on the speed of the motor, so I'm thinking there will need to be something (4017 ??) to actually count which pulse is occurring and direct it to the appropriate set of MOSFETs.

Also, preset timing of peak to hold current switching for injectors is very power inefficient, you really need to do peak current detection to then activate the hold current.

And I think you will find total FIRE + HOLD time is ~2mS.

 

[AnalogKid]

The usual thing is to chose an injector with a large enough rate in pounds per hour to run the engine, then you would have the programing of the ECM change the time open for each desired value of performance in the engine.

OK. Combining this with augustine in #25, it sounds like what he wants is ...

1. External signal turns injector #1 on full for a short, fixed time.
2. Drop Inj#1 down to hold current for the remainder of 2 ms
3. Turn off Inj1.
4. Wait.
5. External signal turns injector #2 on full for a short, fixed time.
6. Drop Inj#2 down to hold current for the remainder of 2 ms
7. Turn off Inj2.
8. Wait.
repeat

OR, in steps 2 and6, the injector hold time is variable, and this is what controls engine speed.

ak

 

[malc9141]

@ AK
Will come back to this. Thanks so far! I must deal with separate matters for some hours. I must repeat, I am in the dark with electronics and abbreviations may fool me.

Answering post #21: If I understand the scheme, the FIRE is being held for 0.2 ms, but I can't see what the HOLD duration is. But the FIRE should be very short (eg <0.5 ms, not 2.0 ms (if I have that right). So how do we in fact vary HOLD?

=> #22: the Hold overlaps Fire. They have the same signal but the higher current is very brief (<0.0005 sec). The duration of the Hold determines the amount of fuel that goes in, ie = the accelerator. (I point out, years back, there was no info on these injectors, except a broad view. We mistakenly thought the fast response was electromagnetic - it is in fact very high fuel pressure actuated - and we made a system which worked well. Then we were told we could do it more simply with a NE555, and that worked, also. It gave a nice double voltage spike on an Oscilloscope. I had planned to use the NE555 layout and to switch the two Fire times between the two cylinder pairs, obviously with half a rev interval).

=>post #21: the Fire should be a fixed time, the HOLD should be able to be varied. But via the same delay (Pot control) for both HOLD Fets.

=> post #26: yes, nearly. The last sentence is correct. The variation in HOLD duration (more/less fuel in) controls eng speed. The HOLD may slightly exceed 2 ms. We then start to need an extra pre-Fire injection but we will leave that!

 

[malc9141]

Answering post #22: the Hold overlaps Fire. They have the same signal but the higher current is very brief (<0.0005 sec). The duration of the Hold determines the amount of fuel that goes in, ie = the accelerator.

 

[ChrisP58]

How much current do the injectors draw when feed the full battery voltage?

What is their typical inductance?

 

[augustinetez]

Depends whether it is a high resistance or low resistance injector, but it can max out at over 14A.

 

[AnalogKid]

the higher current is very brief (<0.0005 sec).

That changes things greatly. First, it means that the post #1 schematic was even more incorrect (sorry). Second, the initial high-current pulse (the FIRE stage) probably can be done with nothing more than a capacitor (depending on the timing tolerance that can be ... tolerated).

Next - the input. What is the signal coming from the engine that commands the injector to fire? Is it a series of short pulses, and the injectors alternate on each positive edge? Or is it a longer-period square wave, and the injectors toggle on both edges (one fires on the positive edge, one fires on the negative edge)? Or, something else - ?

ak

 

[augustinetez]

Here's a short article that covers several things including current draw http://www.autodiagnosticsandpublishing.com/feature/fuel-injector-testing.htm

 

[AnalogKid]

Here's a short article that covers several things including current draw http://www.autodiagnosticsandpublishing.com/feature/fuel-injector-testing.htm

Excellent. Tells me almost everything I needed to know. Thanks for the h.e.l.p.

ak

 

[malc9141]

@#31: Ok working so not on line. Will read & reply ASAP.
Just like to say: I drew the Flip-flop circuit, [the one without the dots ] but the overall pulsing circuit was designed by one of your colleagues and seemed to work on his Desktop.

Malcolm

 

[AnalogKid]

The circuit will work, but only in the sense that it will produce alternating outputs. So it will "pass" in a simulation or verification environment, in that there are no logical faults, the outputs are determined by the inputs, etxc. But it most definitely will not produce the actions described in posts 9, 11, 27, or 28.

ak

 

[AnalogKid]

=>post #21: the Fire should be a fixed time, the HOLD should be able to be varied. But via the same delay (Pot control) for both HOLD Fets.

As before, this is kinda a biggie to keep hidden ever, let alone for so long. It means there is one HOLD timer that is share between the two alternating FIRE outputs. Not difficult or complex, but one full step more complex than post #1. How complex depends on this:

Next - the input. What is the signal coming from the engine that commands the injector to fire? Is it a series of short pulses, and the injectors alternate on each positive edge? Or is it a longer-period square wave, and the injectors toggle on both edges (one fires on the positive edge, one fires on the negative edge)? Or, something else - ?

 

[malc9141]

@36. Well, most helpful member. I have not been trying to hide anything. The original diagram is there to see.

As I understand it, and what worked, was we had a foto-sensor which reflected a continuous i.r. light off a spot on the flywheel as it went around. This allowed a transistor (as I knew it) to leak to Ground. This acted on the IC to send out a current. (The Sensor circuit was meanwhile recharged via a short time delay). The duration of the current out was "the Accelerator". It split into two, one a short pulse to get the injector solenoid moving, then a longer pulse (the accelerator) which had been determined by the pot (if that is the correct word). I do not understand why the pot was positioned where it was, nor its connections with Resistors - that advice came from this Forum.
At first, we had a "belt & braces" design with much redundency because we were feeling our way. It worked fine (for our single cyl engine). But then this NE555 design was put forward and it did get the engine going but the speed control was absent. I am afraid that is the design I have posted here.
I presume it can be redesigned to work.

I repeat, I think I did say early on that Fire was short, sharp and fixed and Hold was what determined speed. But, whatever. Regarding Next - the input, the signal from the engine is a short pulse as the flywheel passes the IR sensor. Its duration is shorter, obviously, as the engine speeds up. Its minimum duration depends on the size of the reflective spot, to be exact. But it is typically around 10^-4 sec. sec.

 

[AnalogKid]

Regarding Next - the input, the signal from the engine is a short pulse as the flywheel passes the IR sensor. Its duration is shorter, obviously, as the engine speeds up. Its minimum duration depends on the size of the reflective spot, to be exact. But it is typically around 10^-4 sec. sec.

So the input to the circuit is a single continuous stream pf pulses, and you want the injectors to alternate with the leading edge of each pulse? In that case there does indeed need to be a toggle flipflop in the circuit. And based on #27, a single, variable HOLD timer (called a monostable multivibrator) that affects both injectors. That will add 1 chip to the circuit.

FIRE - do the two FIRE phases (one for each injector) have to be perfectly matched, or can they vary with respect to each other by +/-10%? If a little variability is tolerable, that *might* eliminate two of the MOSFETs. One of the car guys should jump in here - Can I drive an injector through a resistor and capacitor in parallel to get a high current spike that tapers off to a lower, continuous coil current? Or is that a 'sounds good on paper but nobody does it that way' thing?

And - where are you located?

ak

 

[shortbus=]

A picture or better description of the motor would help too. I still don't understand why 2 injectors? And what type of motor has enough room for two of them in one cylinder intake track.

What are you using for the "brains" of this? Trying to do it without some sort of brain/micro or whatever is going to be much, much harder than it needs to be.

 

[augustinetez]

He said earlier it is a 4 cyl 2 stroke.

Also he's firing 2 injectors at a time - quite a normal thing that was done before sequential injection came out.