Any general discussion around the firmware, what is does, how it does it etc.
#45450
Bounty is gone. It was refunded as no one picked it up.

Use a MAP sensor and feed it into a VR sensor as suggested. Should work well enough.

It is what I will be doing as new wiring loom I am putting in will allow for the "stock" sensor to be wired in.

Hopefully it works so I can improve my spark energy.
#45452
Several have used MAP pulses to signal cyl#1. However, keep in-mind it may be "less-useful" on some setups such as ITBs, which commonly lose most of the MAP signal as throttle increases. This loses the ability to re-sync. That's a separate issue to little signal while cranking, possibly requiring non-cycle-sync start with transition to cycle-sync. I say it that way as that doesn't mean it has to be a sequential setup to use cycle signal.

Alternatively, investigation of other signal sources with a MAP or pressure sensors are still possible, and I have seen modified throttle body signal placement (more of an air velocity detection), crankcase pulse on MAP sensor, and exhaust pulse sensing, for examples. There is a lot to explore, and thanks for your willingness to look at this general option! It would certainly be simpler than poking a hole for a sensor on the valve train or in the case to read a reduction gear. 8-)

David
#45557
Great to see this post have been ressurected, i have also posted a simmilar thread recently, trying to list what i have learned so far to assist the cause.

-We can use map sensor as a indexer for phase detection
-Map sensor data can be converted to a square wave form at its lowest point via a schmitt trigger, which can be fed to the cam sensor input. which than ecu will see it as a regular cam sensor output.
-microsquirt needs this cam data for only couple of turns of the crank. than it knows the "phase" since sequence of the phases will not ever change while engine is running. so map sensor's inability to produce good data at high rpms is not a problem. I assume speeduino works the same cause i can not think a good reason for any ecu to need constant cam phase detection signal. (Need more info here)
-Considering map sensor will also be used for metering instrument(for motorbikes and all ITB vehicles) on ITB mode (which still negate problems regarding high rpm map problems by switching itself to alpha-n at some point). We need to split map sensor output to feed 2 inputs on the ecu, one is cam input which will be fed trough schmitt trigger another one is fed to map input for ITB mode. This signal siplitting part of things is another part that i have no clue. how we can splitt the signal without degrading it or how much of a degradation will be acceptable etc. Really looking forward on oppinions about those.
#45591
I was wondering if it would be possible to set the second trigger in the missing tooth decoder - to "map", and then skip resync.
This could possibly be a feature request. And sorry guys havent been able to play around with either tunerstudio or the code, since my first speeduino is arriving next week :?
And maby skip a revolution before reading map so there enough vacuum in the intake.
Just a thought :ugeek:
#67039
Quote "Map sensor data can be converted to a square wave form at its lowest point via a schmitt trigger, which can be fed to the cam sensor input. which than ecu will see it as a regular cam sensor output. "

How?

I looked into the CD40106B chip but I dont see how you design the circuit to produce a square wave when the map is at its lowest point.

If you instead just take a second dedicated map from cylinder 1 on a twin cylinder directly into a signal conditioner for the cam, how do you tell it to trigger ? is it as simple as falling edge or rising edge, then no schmitt is even needed ?

I read some of the speed code and see it syncs on cranking, so I would assume if rpms are greater than crank there is then no further reason for the cam signal since it is running ? If so the MAP hack would seem to be a good option to establish sync ?

Just how to do it ?
#67041
usatracy wrote: Tue Feb 06, 2024 4:30 am Quote "Map sensor data can be converted to a square wave form at its lowest point via a schmitt trigger, which can be fed to the cam sensor input. which than ecu will see it as a regular cam sensor output. "
That isn't strictly true. A typical logic level schmitt trigger would need to drop below 0.4V and then go above 3.4V to generate a square wave which, with a typical 2.5bar sensor, would mean you would need to hit boost during cranking.
You could use a comparator circuit with hysteresis but it would need some fine tuning to get it to switch reliably and you may need to shift the reference dynamically based on the average MAP reading; its much simpler to do this in firmware.
#67054
I totally understand, I am thinking that I should consider moving my engine dev to microsquirt, as it seems to support this MAP sync on startup then ignore sync after ten revs, (from what I have read anyhow).

Might sound stupid to some, but with sync I am not burning up my coils (wasted spark) with a duty cycle double what they need to run the engine, one set of coils is the cost difference between the two ecu choices. Being an aircraft solution, weight is limited so choosing heavy coils to increase heat dissipation and life is not an option when the existing coils might work fine if operated properly.

From threads I have reviewed a lot of this goes back 5 or more years with no resolution or interest in sight, including a seemingly low interest in 2 cylinder applications.

Rotax dropped all support and manufacture of 2 stroke aviation engines. The Brigss 23 horsepower twin has been used for decades in the EU and now in the USA. It has achieved 70 hp sustained in racing applications and shows hope to cover the 38 to 40 hp needs for ultralight aircraft in the usa where 2 strokes are soon to be under epa attack.

To that end I have been redesigning the BS 23 and using the work of several people internationally and domestically from their proven BS 23 solutions. I decided EFI was a core piece to achieve the higher power and longer TBO needed for a good option for Part 103 ops. I initially decided that speeduino might be a good fit, but I am questioning that now.

Since I am vested in the NO2C, for now, I will continue on that path, and consider if it really is going to meet the requirements, or, if the solution should be with a more mature and supported COTS solution, albeit at higher cost. I am paying 1700 dollars for high flow billet heads for a 1700 dollar engine, so paying more for a COTS ECU isn't a dis\qualifier for my build, if it provides the functionality the other will not.

THANKS ALL :)
#67061
usatracy wrote: Tue Feb 06, 2024 7:12 pm… Might sound stupid to some, but with sync I am not burning up my coils (wasted spark) with a duty cycle double what they need to run the engine
You should not be using dwell that is anything but correct for the current voltage, else something is wrong. Correct dwell does not injure the coils at any duty cycle. Literally millions upon millions of every type of engine using waste-spark shows this rather clearly.

usatracy wrote: Tue Feb 06, 2024 7:12 pmFrom threads I have reviewed a lot of this goes back 5 or more years with no resolution or interest in sight, including a seemingly low interest in 2 cylinder applications
From what I have seen, the technical interest is the same for any number of cylinders. The interest may be low for a complex and possibly unreliable solution to a standard requirement. Most use the well-proven and reliable established methods for cycle signals, by many means. If you wish to forge a new path, i think that's awesome! But, be aware you are an explorer, and the adventure priority is yours.

usatracy wrote: Tue Feb 06, 2024 7:12 pmthe usa where 2 strokes are soon to be under epa attack.
Gasoline engines (any type) rated under 25 hp are in bills in California and Washington (perhaps others) to be outlawed next year. It's not all from the top.

usatracy wrote: Tue Feb 06, 2024 7:12 pmI initially decided that speeduino might be a good fit, but I am questioning that now.
Question everything. :lol: IMO seriously, use what you like, and not what you don't. It's your project with your goals and judgment. I see no clear benefit to sequential in an application like this (and I am well versed in FAR 103 operations), and Speeduino is fully capable in multiple modes - but again, that's your call and no one else's. Do your thing. 8-)

I uploaded logs and last tune and dyno plots look[…]

you could set 'skip revolutions' to 3, then it wou[…]

Hi, I am trying to assign Signed values to the x-a[…]

blitzbox

I've finally figured out why MAP and Lambda weren'[…]

Still can't find what you're looking for?