injection timing discussion

[wallace]

Is the injection timing for the factory calibrations fixed or variable? Does the injection timing have modifiers like the ignition timing does?

[sdac guy]

According to the "Instructors Guide" (which describes most of the controlled functions), the injector firing timing is controlled by the sync and reference pickups in the distributor. The discussion in the manual is for 85-87 (LM/PM) operation, which is for batch firing.

"One pair of injectors is fired when the when the sync is high and the other pair is fired when the sync is low, when compared to the reference signal."

And ...

(to turn on the injectors) "The power module provides the ground at the instruction of the sync and reference inputs. The power module removes the ground at the instruction of the logic module. In other words, the injectors are turned on by hardware and turned off by software." (this allows for the software to control the injector pulse width)

Barry

[wallace]

The later ECU's were sequential injection though right? I was wondering for the sequential setups how the injection timing was controlled. Do you have the instructor guide for the later stuff? Thanks for the reply!

[ShelGame]

No, the timing is not varied from cycle to cycle. It always starts on the falling edge of the distributor shutter window. In the SBEC, there are individual cylinder trims, though. So, the PW for each cylinder is unique. The V6 does this on a bank basis as well.

[sdac guy]

There is a way to have the injector start at at different time compared to tdc. That is to advance (or retard) the distributor a few degrees and then compensate for that by moving the overall ignition timing in the opposite direction (in the cal). Of course this is an across the board adjustment to the injector timing and not a variable one.

I am told this can make a huge difference in idle quality and overall drivability.

Barry

[wallace]

So it is possible to determine the static injector timing (assuming no trim on individual cylinders) based on the position of the shutter window? If so, can you help me to determine what that would be?

[sdac guy]

So it is possible to determine the static injector timing (assuming no trim on individual cylinders) based on the position of the shutter window? If so, can you help me to determine what that would be?

Very roughly, it is the same as what the initial ignition timing is set to at the distributor (most of our cars get 12 deg). Since the two hall effect pickups are a few degrees apart and both are used for injector timing then it may not be exactly the same.

But again, very roughly, the injector timing is the same as the initial ignition timing the distributor is set at, of course, injector and ignition timing and action are on different engine strokes (which is also deduced by the two distributor pickups and the ecu logic).


Barry

[zin]

There is a way to have the injector start at at different time compared to tdc. That is to advance (or retard) the distributor a few degrees and then compensate for that by moving the overall ignition timing in the opposite direction (in the cal). Of course this is an across the board adjustment to the injector timing and not a variable one.

I am told this can make a huge difference in idle quality and overall drivability.

Barry

So... Does this mean when you install your dist/aux shaft out of phase (off a tooth or two), you are changing the injector timing, even though you properly set the ignition timing? Or are those two basically the same? In which case you would be advancing/retarding the injector timing when you change the ignition timing?...

Mike

Edit: It occurs to me that since you would be correcting for the added/retarded timing with the cal, the dist signal is the trigger for both the ignition and the injector timing... So, if the dist/ignition timing is "correct", the injector timing would be normal and the only way to really change the phasing of these would be to do as you suggested and mechanically adjust the injector timing, while "electronically correcting" the ignition timing, correct?

[wallace]

So the ignition event occurs some time BTDC at least 12* but I would think the injection event does not coincide with the spark event. If there is a trim table for each injector there must be a base that it modifies, at least I think so from my little understanding. I find this very interesting, how would you determine the base injection event wrt TDC of that particular cylinder? Easy for the spark event since we know that static timing is mechanically set at 12*. If we assume no trim then the injection timing would be the same for each cylinder so knowing one would be just as good as another. Each injection event per cylinder can occur from 0 to 720 degrees.

[zin]

So the ignition event occurs some time BTDC at least 12* but I would think the injection event does not coincide with the spark event. If there is a trim table for each injector there must be a base that it modifies, at least I think so from my little understanding. I find this very interesting, how would you determine the base injection event wrt TDC of that particular cylinder? Easy for the spark event since we know that static timing is mechanically set at 12*. If we assume no trim then the injection timing would be the same for each cylinder so knowing one would be just as good as another. Each injection event per cylinder can occur from 0 to 720 degrees.

It would be interesting to know the start time of the injection event, under some circumstances it may be advantagous to adjust the "start time" of the event, either to help idle quality and/or allow more "spray time" before the valve opens at high RPM... I'm also curious if this really has much effect...

Mike

[sdac guy]

....

Edit: It occurs to me that since you would be correcting for the added/retarded timing with the cal, the dist signal is the trigger for both the ignition and the injector timing... So, if the dist/ignition timing is "correct", the injector timing would be normal and the only way to really change the phasing of these would be to do as you suggested and mechanically adjust the injector timing, while "electronically correcting" the ignition timing, correct?Exactly.


So the ignition event occurs some time BTDC at least 12* but I would think the injection event does not coincide with the spark event. If there is a trim table for each injector there must be a base that it modifies, at least I think so from my little understanding. I find this very interesting, how would you determine the base injection event wrt TDC of that particular cylinder? Easy for the spark event since we know that static timing is mechanically set at 12*. If we assume no trim then the injection timing would be the same for each cylinder so knowing one would be just as good as another. No.

The two events happen on different engine strokes but are both controlled by the same two hall effect pickups in the distributor. So the 12 deg advance for the ignition timing also applies somewhat to the injector timing. Or in other words say your distributor is set at 12 deg BTDC, and you advance it to 14 deg BTDC. Both injector timing and ignition timing are then advanced 2 deg. BUT, they do not occur at the same time (they occur on different engine strokes).

Also, there is the width of the shutter blade and separation angle between the two pickups that affect the injector timing. The Instructors Guide (and I only have the early one) states that ignition timing is done off the reference pickup, but the injector timing is a mix of both. It states that basically, the Reference pickup gives the crank position and engine RPM, and the Sync pickup gives camshaft position. Now, since these two pickups are both stationary, and both picking up signals from a shutter window (one per cylinder), then getting the crank position is an easy thing, while the cam position must be deduced by the timing and interaction of the signals.

I am trying to state this as simply as I can as in the Guide there are nearly 10 pages of text and diagrams devoted to this particular discussion (sync & reference pickups), and another couple pages on injector control. It goes down to the transistor level ...

Barry

[ShelGame]

I am trying to state this as simply as I can as in the Guide there are nearly 10 pages of text and diagrams devoted to this particular discussion (sync & reference pickups), and another couple pages on injector control. It goes down to the transistor level ...

I think I need a photocopy...

FWIW, from what I can tell, the sync pickup is only used to determine which cylinder is at TDC, and therefore which ignition retard value to use, which injector PW to use, and which injector to fire.

The timing for the coil firing is hard coded. There is some math done to correct for the 12* base timing. But, there is no apparent reference for position when it comes to the injector timing. They just fire when the calculations are done.

[sdac guy]

I think I need a photocopy...

FWIW, from what I can tell, the sync pickup is only used to determine which cylinder is at TDC, and therefore which ignition retard value to use, which injector PW to use, and which injector to fire.

The timing for the coil firing is hard coded. There is some math done to correct for the 12* base timing. But, there is no apparent reference for position when it comes to the injector timing. They just fire when the calculations are done.Rembember, my reference is only valid for the batch firing. I have no idea how this all is handled on sequential injection. But the distributor and the cam spin at the same speed, which is one revolution for every two of the crank. But I will go through this just to make clear what I am saying.

For example lets say the distributor is set for 12* base timing and is at idle with the coolant temp plug undone so no electronic advance is added. On the compression stroke (piston going up) 12* before the piston reaches the top the spark fires. The next stroke is the power stroke (piston going down).

The next stroke is the exhaust and the piston is going up again. 12* before TDC the injector fires and the piston starts on its way down on the intake stroke.

These two events happen 360* apart at the crank, and 180* apart at the cam and distributor. What this means in the distributor is that the original shutter that triggered the ignition travelled 180* past the reference pickup, and is now at the sync pickup to trigger the injector. But the shutter opposite it that is now at the reference pickup is the one working with the shutter at the sync pickup, to trigger it.

We need to be careful when picturing all this in our heads that we don't equate a particular shutter doing all the work for a particular cylinder. In fact two different shutters (opposite each other) work for each cylinder. Since we have 4 cylinders, we have 4 shutters arranged 90* apart.

Now, again for example, lets plug our coolant temp plug back in so electronic advance is enabled. As the piston approaches TDC on the compression stroke, the spark plug fires at (12* + ElecAdv). One crank revolution later on the exhaust stroke, the injector fires at 12* BTDC and then starts the intake stroke.

The difference is that the ECU can and does adjust when the ignition event happens. But the start of the injector event is hard coded by the sync signal.

For pulsewidth control, the injector circuit is disabled (turned off) by command from the ECU.

Other electronic elements in the power module control which injector batch is fired, with each batch being fired every other time.

The sentences I have "quoted" in my previous posts are word for word quotes from the Instructors Guide.

Barry

[sdac guy]

A question I have Rob, is how does the ecu add advance to ignition timing that has not had a trigger event yet? If I have the distributor set for 12* advance and the ecu wants to add another 20* advance to that, how does the ecu cause the spark to happen 20* before the reference pickup signals the event?

This is not explained at all in the Instructors Guide, which is pretty much only how the physical electronics work.

What am I missing here?

Its sort of like I am walking across the room to turn on a light, but since it is programmed to come on 2 seconds before it is turned on, it comes on just as I am reaching for the switch. :lol:

Barry

[wallace]

Good info! So for the earlier cars the injection timing is occurring 180* cam degrees after the spark event and for crank degrees 360*?

[ShelGame]

A question I have Rob, is how does the ecu add advance to ignition timing that has not had a trigger event yet? If I have the distributor set for 12* advance and the ecu wants to add another 20* advance to that, how does the ecu cause the spark to happen 20* before the reference pickup signals the event?

This is not explained at all in the Instructors Guide, which is pretty much only how the physical electronics work.

What am I missing here?

Its sort of like I am walking across the room to turn on a light, but since it is programmed to come on 2 seconds before it is turned on, it comes on just as I am reaching for the switch. :lol:

Barry

Well, the actual coil firing doesn't happen on the distributor pickup. It happens when the dwell timer expires. The dwell is calculated and stored on the distributor pickup, though. But, there are other dwell events that happen, too. Mostly on startup, etc. I think.

The dwell calculation is based on the most recent distributor shutter period. I guess it's likely that the dwell calculation will be at least 1 cylinder old. But, that really shouldn't be a problem.

[sdac guy]

Well, the actual coil firing doesn't happen on the distributor pickup. It happens when the dwell timer expires. The dwell is calculated and stored on the distributor pickup, though. But, there are other dwell events that happen, too. Mostly on startup, etc. I think.

The dwell calculation is based on the most recent distributor shutter period. I guess it's likely that the dwell calculation will be at least 1 cylinder old. But, that really shouldn't be a problem.
Well, Duh!! Makes perfect sense! :clap:

Barry

[risen]

I
The timing for the coil firing is hard coded. There is some math done to correct for the 12* base timing. But, there is no apparent reference for position when it comes to the injector timing. They just fire when the calculations are done.

So it would be possible to delay at the end of the calculations to fire the injectors? I understand that you couldn't advance it without major work, but retarding it should be easy, no? Or is the calculation done in a ISR or somewhere else on the critical path?

I should also add that when I was writing the code for the smecstim I had to constantly remind myself that the 2 pickups aren't 180 off, they're like 160 or something (degrees measured @ the pickups, not crank). That should be kept in mind while considering this stuff, too.

[ShelGame]

So it would be possible to delay at the end of the calculations to fire the injectors? I understand that you couldn't advance it without major work, but retarding it should be easy, no? Or is the calculation done in a ISR or somewhere else on the critical path?

I'm not really sure. It would take major work to do it. On the SBEC, the injector PWM is handled by the TI interface chip. Since we don't really even know what this chip is, it would be tough to change the timing. For all we really know, it could actually be doing something with the injector timing. Has anyone every put an SBEC injector output on a scope?

On the SMEC, I guess you could. But then, it's batch fire, not sequential.

[risen]

I'm not really sure. It would take major work to do it. On the SBEC, the injector PWM is handled by the TI interface chip. Since we don't really even know what this chip is, it would be tough to change the timing. For all we really know, it could actually be doing something with the injector timing. Has anyone every put an SBEC injector output on a scope?

On the SMEC, I guess you could. But then, it's batch fire, not sequential.

I see. It's not just like a high/low signal to enable the output from the TI chip to have it do the pwm? I do have a sbec-ii from a 318 with a bad charging circuit that I can try and scope....

[ShelGame]

I see. It's not just like a high/low signal to enable the output from the TI chip to have it do the pwm? I do have a sbec-ii from a 318 with a bad charging circuit that I can try and scope....

No, all the timers are used up in the 6811 in the SBEC. The interface chip has a block diagram and pinout in the trans controller patent documents. It has a timer unit in it. So, I assume that the injector timing is done in there.

[risen]

No, all the timers are used up in the 6811 in the SBEC. The interface chip has a block diagram and pinout in the trans controller patent documents. It has a timer unit in it. So, I assume that the injector timing is done in there.
I know I'm really dragging this off topic here, but last question: that interface chip sits on the memory bus with the eeprom then? Is that how it's interfaced or is it a serial type setup?

[ShelGame]

I know I'm really dragging this off topic here, but last question: that interface chip sits on the memory bus with the eeprom then? Is that how it's interfaced or is it a serial type setup?

It's on the address/data bus. It's actually a port expander, extra memory, extra PWM control... I think it must actually be another MCU. I just haven't been able to figure out what it is. The chip was made by both TI and NS. So, it's likely one of those TMS processesors that NS made a copy of. I haven't been able to match the pinout to anything that has a currently available datasheet. Obsolete or not.

The pinout and block diagram for the chip can be found in most of the Chrysler patents for the A604 trans controller (which uses the same micro core and power supply as the SBEC).