fixed dwell question

[Keithg]

I do not have a jimstim, so I have not been able to determine what is happening on a bench, but there is some distinct difference with how MS2/E outputs fixed dwell and multiple spark outputs and how MS1/E does it.

In MS1 and MS2, with 75% dwell for example, are the same with a single spark output. The line is hi for 75% of the cycle and then lo for 25%. When multiple spark outputs are used in MS1, the line is hi for the whole time rand then lo for the 25%. In MS2/E, I cannot confirm this, but it appears that it is hi for 75% and low until the next.

James put in the 'time after event' for me to run my DI cassette on my Saab. It runs fine, but I am trying to see if I can drive it more similar to how the OEM does it. With MS1/E I can run it with 75% fixed dwell and it runs reasonably well. So I tried with MS2/E and put 75% in the box. No fire. What I think is happening is that multiple lines are held down simultaneously (overlapping on the lo) so the cap never charges and it cannot fire.

As best I can tell, the OEM holds the line hi except when firing. The duration of the 'fire' is:
rpm.........501...........1 200.........3 999........6 000....... 7 500......8 499
volt
15.00.......10.0.......... 4.0........... 2.0........... 2.0........... 2.0........... 1.5
14.00........10.0......... 4.0........... 2.0........... 2.0........... 1.9........... 1.4
13.00........10.0......... 4.0........... 2.0........... 2.0........... 1.7........... 1.3
12.00........10.0......... 4.0........... 2.0........... 2.0........... 1.5........... 1.3
11.00........10.0......... 4.0........... 2.0........... 2.0........... 1.3........... 1.3

As voltage decreases, it shortens the discharge time once the cap can no longer be fully charged. At low RPMs it discharges a long time for 'multispark' help in idle misfire prevention (like EDIS).

If the operation were similar to MS1 where it held the line hi except when firing, this scheme may be a bit easier to realize. Why is there a difference between the way MS1/E and MS2/E do this?

[jsmcortina]

I do not have a jimstim, so I have not been able to determine what is happening on a bench, but there is some distinct difference with how MS2/E outputs fixed dwell and multiple spark outputs and how MS1/E does it.

In MS1 and MS2, with 75% dwell for example, are the same with a single spark output. The line is hi for 75% of the cycle and then lo for 25%. When multiple spark outputs are used in MS1, the line is hi for the whole time rand then lo for the 25%. In MS2/E, I cannot confirm this, but it appears that it is hi for 75% and low until the next.

Keith,
I'm confused. By my maths 75% + 25% = 100%.
So if "the line is hi for the whole time rand then lo for the 25%." then it is high for 75% and low for 25% ?

And if "it appears that it is hi for 75% and low until the next."
Then it is high for 75% and low for 25% ?

So, what's the difference?

James

[Keithg]

Ah, I was not clear.

The difference comes into it when there are more spark outputs. MS1/E and MS2/E appear work identically with a single spark output.

With a single spark output, all is well. I know this was written specifically for a single spark output, but I am trying to run it with 4 spark outputs. I cannot verify this with MS2/E, but it seems that this is what is happening:

The basic difference is what the code does with the line on the 'other' spark events. With MS1/E, it went back hi after 100% had been reached, with MS2/E, the line stays lo until it comes back to its 'dwell' period.

Lets look at a 4 spark output COP install. Spark A has a 75% dwell. If spark A stays low during the other spark events (ms2/E operation), my CDI cannot charge for the other cylinders because Spark A stays low for 3 spark events and does not come back hi again until it cycles back to spark A. With MS1, it would be hi for 75%, then go low for the 25% then go back HI and stay there until the next time spark A was to fire.

I cannot verify this on the bench, but surmise that this is what it does because I get no spark what so ever when I try to run a fixed dwell of 75% on the car and with MS1/E, it will fire up and run with 75%.

KeithG

[jsmcortina]

MS1 is broken then. 100% is supposed to be the time between sparks on a channel, not between sparks overall.

James

[Keithg]

Would that operation make sense, though? I mean, what possible component selection with multiple spark events would want fixed dwell to be calculated based on only that one spark output regardless of the others? (not being difficult or confrontational, just curious.)

Ok, then let me ask another question as clarification on how the code works...

If the 'time after spark' is used and it is 2ms, yet there is not enough time to hold the line low for 2ms plus the 2.5ms required to recharge the cap before spark B fires, for example, it will not care. Spark A will be held low for 2ms regardless of when spark B fires. Is this correct?

For my CDI unit (and maybe others, I do not know) the capacitor at the heart (single cap for 4 spark outputs) needs to be fully charged. The way it remains charged is to have the spark lines held hi. Once the trigger line goes lo, (any line) it discharges the cap and it needs to recharge for the other event. Is there a way to get the operation I outlined in the first post with some minor code modification?

KeithG

[jsmcortina]

Would that operation make sense, though? I mean, what possible component selection with multiple spark events would want fixed dwell to be calculated based on only that one spark output regardless of the others? (not being difficult or confrontational, just curious.)

A Bosch ignitor.

James

[Keithg]

Would that operation make sense, though? I mean, what possible component selection with multiple spark events would want fixed dwell to be calculated based on only that one spark output regardless of the others? (not being difficult or confrontational, just curious.)

A Bosch ignitor.

James

So, you could use 2 bosch ignitors (-139) like this to drive 2 coils on a W/S install? Hmmm.

KeithG

[jsmcortina]

Would that operation make sense, though? I mean, what possible component selection with multiple spark events would want fixed dwell to be calculated based on only that one spark output regardless of the others? (not being difficult or confrontational, just curious.)

A Bosch ignitor.

James

So, you could use 2 bosch ignitors (-139) like this to drive 2 coils on a W/S install? Hmmm.

KeithG

That's the theory yes. The ignitor is expecting a fixed duty input like it would receive from a windowed hall distributor, so the duty as seen at the ignition module needs to be preserved regardless of the number of spark outputs.

The last time I worked with modules was HEI4 before I'd added dwell control to MS1 (remember no dwell??), so my experience is somewhat limited.

James

[Keithg]

Makes sense...

So, back to my previous question... To get some 'dwell' curve similar to the table earlier in the thread would be quite a bit of work I am guessing?

KeithG

[jsmcortina]

Why is the "dwell" longer at higher voltages?

To achieve times as you suggest would require a specific lookup table to be added.

James

[Keithg]

Why is the "dwell" longer at higher voltages?

To achieve times as you suggest would require a specific lookup table to be added.

James

It is longer because this is the 'discharge' profile of the CDI system. As long as all lines are held hi, the cap is charged. With higher voltage, its recharge time is shorter and a larger portion of the cycle can then be used for discharge.

The 10ms is for idle multispark near idle. The bulk of the range is 2ms. When revs rise and voltage drops, the discharge time needs to be reduced. The compromise is a slightly shorter spark duration at higher revs.

I currently run 1.6ms across the board. I will try 2.0 and see what happens. I am guessing it will fall flat at higher revs. This is great information.

As far as spark is concerned, of greater interest to me is the WS crank and COP run followed by the ability to use my 'combustion sense' output lines instead of a second cam trigger.

Thanks!

KeithG