MS2/Extra pre-3.4 alpha 9 released

[ol boy]

How will sequential mode work on a 2 stroke? Does the code know to only go from 0 to 360 degrees?

[LAV1000]

How will sequential mode work on a 2 stroke? Does the code know to only go from 0 to 360 degrees?

You could use semi sequential on a two stroke.
Real sequential works with 720 and needs a cam pulse.

[ol boy]

I figured as much, just want to make sure before I go and have 2 more magnets placed in the prop hub.(4-1). Funny thing though.... I can see the intake/compression stroke pulse on the MAP line to give an indication of where the engine is at in it's phasing.

[LAV1000]

Ultralight engine ?

To get more profit of the ms unit, try to get more pulses/rev.
Ignition will be better timed.

[ol boy]

nope, this engine http://www.desertaircraft.com/engines_d ... Page=DA-70

It currently has a single flying magnet, but id like to add 2 more magnets to get a 4-1. I'm sniffing exhaust gasses from each exhaust pipe so i can see the imbalance in fuel distribution which is corrected by moving the batch fire percent angle around. The problem is that the desired angle to get equal fueling moves. The pick up is at 90* BTDC and my percent starts at 25% at idle, 2000 rpm (TDC), and progresses down to 0 (90* BTDC) at 5500 rpm then to 90% (126* BTDC) at 6000 rpm. I can do the math and convert the trigger offset and percent to degrees BTDC.

I just need to make sure it's going to work before drilling holes!

[LAV1000]

Checked my settings and it seems possible to use a table for injection timing.
This is on 3.3.2 firmware.

Don't know if it actualy works haven't test it, just runnig fixed timing.
At the moment I am rebuilding the injection system.

Running this on a motorcycle and using an optical sensor to get crank rpm.

[ol boy]

It works. added the extra 2 magnets. My math was pretty close. all good!

[racingmini_mtl]

Don't know if it actualy works haven't test it, just runnig fixed timing.

It does work and has worked since the day it was added. This is essential for some applications as the one here but also other completely different ones (siamese input ports being the reason this was implemented initially and where tuning the timing as a function of rpm and load is an essential part of getting the correct AFR).

Jean

[ol boy]

What's your thoughts Jean on using the map signal as a can input but using only basic trigger for crank input. I have some good looking map pulses measured inbewteen the throttle valve and reed block.

[racingmini_mtl]

The basic trigger signal would be a problem because the sequential code requires more than 2 pulses on the crank signal. But the 4-1 signal would work.

Jean

[LAV1000]

What's your thoughts Jean on using the map signal as a can input but using only basic trigger for crank input. I have some good looking map pulses measured inbewteen the throttle valve and reed block.

Maybe I don't understand it.

You got a 2 stroke twin cillinder engine.
Every 360 degree cyclus you need to have to fuel injected and ignition.
This is for 1 cillinder the other is 180 degrees retarded.
Semi sequential works this way.

Why do you want to use MAP signal as if it was a cam signal, there's no need to do this.
Only on a 4 stroke (720 degree) you need a signal so the megasquirt knows in wich of the 2 x360 it is.

[ol boy]

The cam/map signal could be used as an extra timing point on a single tooth crank pick up. No need to do the missing tooth wheel. Just an idea... I've noticed that the pulse does move around with rpm quite a bit.

[LAV1000]

I've noticed that the pulse does move around with rpm quite a bit.

And that's why you use the timed injection.

Might be better to find a way to get a toothed wheel on that propshaft.
This way you can also control ignition, I assume you only controlling fuel at the moment ?

[ol boy]

No full fuel and ignition control with a single pick up. I added the 2 extra magnets on Friday and ran it. Runs very well. I was able to get both holes running about the same CO% throughout the rpm range.

[Buelliholic]

I've had some troubles to map the "Sequential Siamese" with "Dual values" settings (timing degrees) to reality on my 2-cylinder 45deg oddfire engine setup. Reading the manual http://www.msextra.com/doc/pdf/Megasqui ... Up-3.3.pdf section 4.1.2- and looking at the image below the header "Siamese-port Timing", for me not clear what should be the base reference point for the "Injection Timing #2" especially when the "Oddfire First Angle" is something else than zero.

After doing some scoping, at the moment for me it looks like the base reference point is always/only "Cylinder#1" => e.g. when calculating degrees for the Intake valve closing moment for the "Cylinder#2", need to take the "Oddfire First Angle" into the calculation: should it be like that or is it a bug?

Siamese-timing.JPG

---
4.1.2 Injector timing

For sequential mode, the timing value is with respect to TDC on the power stroke when both valves are
closed (same as ignition).

For sequential siamese mode, the timing is with respect to TDC on the intake stroke which is the start of the
injection window for the outer cylinders. And the timing 2 values are only used in sequential siamese mode
with the first timing value being for the outer cylinders (1 and 4) while the second is for the inner cylinders (2
and 3).

4.1.3 Sequential/Semi-sequential Timing

The image below shows how the timing values relate to the engine cycle. The timing value of 0 degrees
corresponds to TDC between the compression and power stokes. (The same as ignition timing.)

Positive values are for timing in advance of this point up to 360 degrees which corresponds to TDC between
the exhaust and intake strokes with 180 degrees being BDC between the intake and compression strokes.

Negative values are for timing retarded from the 0 degree point with -360 degrees being TDC between the
exhaust and intake strokes and -180 degrees being BDC between the power and exhaust strokes.

Siamese-port Timing

The siamese-port modes use a different reference point in the cycle due to the position of the injection
window. The image on the right shows how the timing values relate to the engine cycle. The timing value of 0
degrees corresponds to TDC between the exhaust and intake strokes.

Positive values are for timing in advance of this point up to 360 degrees which corresponds to TDC between
the compression and power stokes with 180 degrees being BDC between the power and exhaust strokes.
---

[racingmini_mtl]

That sequential siamese mode was designed to be used with a 4 cylinder even fire engine with 2 intake ports and a single injector per port (it does work with staging but that's still a single logic injector per port).

Since it was never the intention of using it with a 2 cylinder odd fire engine, this was not tested nor was the code designed to deal with this specifically. It might work or it might not but I'd have to go through the code to see it that may work as-is or not and what would be needed to make it work. It has been a very long time since I looked at that code so this is not something that can be done quickly.

Jean

[Buelliholic]

ok, good news if the feature will be fixed/changed/added later. For me it seems working ok - just need to be aware of the "Oddfire First Angle" effect.

A bit surprised that there are no one else asking/playing with the "Sequential injection on an 'Oddfire First Angle' motor" before, as it's quite a common engine type - especially motorcycles. Or is there some other way around (settings/tricks in TunerStudio) to achieve the "Oddfire ignition and Sequential injection on a V-type motor"?

BTW: is the same feature also on other Microsquirt MSextra code versions?

Thanks Jean!

[jsmcortina]

BTW: is the same feature also on other Microsquirt MSextra code versions?

As Jean says, this is a long established feature since 3.0.x

James

[racingmini_mtl]

BTW: is the same feature also on other Microsquirt MSextra code versions?

It has been there since the 3.0 version.

Jean

[Buelliholic]

That sequential siamese mode was designed to be used with a 4 cylinder even fire engine with 2 intake ports and a single injector per port (it does work with staging but that's still a single logic injector per port).

Since it was never the intention of using it with a 2 cylinder odd fire engine, this was not tested nor was the code designed to deal with this specifically. It might work or it might not but I'd have to go through the code to see it that may work as-is or not and what would be needed to make it work. It has been a very long time since I looked at that code so this is not something that can be done quickly.

Jean

re-reading what you're saying: is the "staging" something unreliable sort of method? what is the risk if...?

I've given some über high rpm tests on JimStim and injector pulses seems to stay alive, however, exact timing changes against injection timing degrees on tables neither injection lengths not scoped yet.