[LarryGrant]

I offer my advance apologies if this thread is inappropriate for the forum. I will cheerfully decist if that is requested. Further I offer my apologies for what will be a long post.

I am 1) a pilot; 2) a bit of a mechanic; and 3) an amateur inventor (under the broad use of that term). We could write a book about the legacy (certified) aircraft ignition system (magnetos) and their limitations-shortcomings and, finally, the opportunities to better serve the needs of the aircraft engine. No room for a book here, but much of the same information will come out of this thread if pursued.

Much like the spirit of Speeduino, the Megajolt boxes, designed for automotive use, are making their way into "experimental" aviation. Megajolt applications are not standardized and, therefore, decades away from being useable in certified aircraft, if ever. But there are many home built airplanes ("experimental") and in those, one can do pretty much whatever one wants as it pertains to engines and components. Going forward we'll refer to the experimental aircraft as EAB (experimental - amateur built).

Nineteen out of twenty EABs are carburetored. That's not likely to change much. We can discuss if anyone is interested in why, but the fuel side is not the thrust of this thread.

Back to ignition: the first decision block to be dealt with is do we make an improved version of Megajolt and embrace the Ford EDIS4 module and dumb coils or do we go more mainstream Speeduino and eliminate the EDIS? Consider this: the EDIS modules have a limp mode whereby the module will drive the coils at a 10* advance if there is no signal coming from the ECU. That is an attractive feature when the alternative is to lose your engine and be forced to land under unfriendly conditions. With 10* advance, the engine will not be running great but good enough to get to a landing strip.

Some EDIS4 applications just offset the VR pickup on the trigger wheel another 15* and live with the fixed 25* advance and no ECU (no Megajolt). While it is fixed timing like a magneto, this approach has significant advantages over the magneto system: much better spark at low RPM and much better spark for high altitude operations.

It might be feasable to create something super simple like an ECU with a pot controlled SAW signal to the EDIS giving the pilot the ability to advance or retard the timing (calling it an ECU might be a stretch). Maybe no tables and no lookup? The VR pickup would be mounted at TDC and the range of advance could be 0* to, perhaps, 35*.

The notion of creating a system that requires so much operator management may sound like going backwards to the good people on the forum. But I think it would be popular. This post is getting way too long so I'll explain later if there is any interest.

If you've read this far, I thank you so very much. I'm looking forward to the discussion.
Larry

[PSIG]

I and others here are very familiar with aircraft ignitions, and the use of non-standard ignitions on FAA Experimental engines. This forum may be appropriate while considering Speeduino involvement, but be aware up-front that some features or a different "limp-home" scheme will likely require modified or new code to implement. That may mean some extended development and time.

The largest drawbacks to EDIS are age (they have electronic and other components over 20 years old), and inability to command retard.

Why retard is a problem, is that loss of control should default to full-load advance, or engine damage may occur quickly. Therefore, if the ability to run uncontrolled is required (an important question in itself), then the base setting should be at full-load advance, and control retards to less advance (and often more advanced) when appropriate.

The primary advantages of controlled timing are greater part-throttle efficiency, including cruise power and economy, and reduced engine heat, wear or damage. We also have additional advantages such as spark energy, coil placement, and several others. Loss of control should be no worse than how it ran previously—if that's required. Controlled timing is good stuff.

Before going any deeper into theory at this point, I would suggest laying-down your project requirements, with justifications. No sense barking up the wrong tree for a while and taking the better path later. For example, I know electronic ignition component failures are relatively rare, and I personally know many who have run various electronic ignitions for years and thousands of hours without fail. A search of FAA records can tell you exactly how often they fail (and sometimes why), compared to magneto statistics, so you are judging on facts. Are the caution and various limitations justified? If so, to what level or on what features?

Welcome to the Forums, and if this eventually goes too far off Speeduino track, it can be moved to the Random section or another suitable site for the purposes of development.

[Chris Wolfson]

I think you should change your view on this whole ignition thing. There is no fail save ignition system, not even a magneto, as it uses only one spark plug..
The chance to blow a new car ignition system during use because of an unpredictable fault in electronics may be 10.000 to one in 10.000 hours. Just to give a number (I think the do not fail that often).
The only way to safely finish your flight without emergency landing is a second, identical ignition system, working in parallel. Two identical systems failing on one flight is near to impossible.
Should be no problem to install a second spark plug in such an engine. If one system fails, exhaust gas temperature may rise a little, so you should not adjust the ignition "on the edge", but to run perfectly on one spark. Opposed to using twin plugs for performance gains.

Even as you do not want to discus it, why use carbs instead of EFI is a question I know no answer too, as some types of aviation urgently want to use them. To make things fail safe from user induced mistakes, increase reliability and range / payload. Aim is to have only on/off and fast/ slow adjustments to make by the user / autopilot. Should help your cause too.
The one way drone is the poor terrorist's long range rocket today and in high demand. There are quite some people out there who improve them right now, with low cost EMS. What makes a drone better can not be bad for an EAB.
Carb's for airplanes are way more complicated than electronic injection systems.

Using two complete ECU systems on one engine should only need very little additional code, which doubles injection rates when AFR drops rapidly because of one side failing. Only change in engine construction would be a second plug.
If you calculate a Speeduino system at about 400-500$, a second one should be a small price for 100% safety.

There is even room for automated actions, if anything fails, as there are usually some free in and out put ports that can be used. CAN bus communication is another option you do not find in a carb.

[LarryGrant]

Chris Wolfson: Thanks for you reply.

1. I'm not against EFI - both of my automobiles enjoy this modern technology and, with it, achieve great driveability. My airplane, on the other hand, has a header fuel tank that sets well above the carb (which attaches to the bottom of the oil pan/sump) and receives fuel via gravity flow. There is no fuel pump. There is no one piece intake manifod - there are intake runners that attach to the oil pan/sump and go to the cylinder heads (kinda clever - cold air cools the oil and the warm oil heats up the fuel/air mixture for better atomization - a win/win situation). The carb has a mixture control that allows leaning the air/fuel mixture during flight. Further there is very limited space firewall forward. Depending on how bad one wants it, EFI is feasible but the effort to implement is a much more complicated project physically than just EI. The electric fuel pumps would be a point of failure and one would want to do a redundant fuel delivery system to feed the injectors. Let's focus on the EI and after that is conquered we can turn our attention to sequential EFI.
2. Lycoming and Continental airplane engines already have 2 spark plugs per cylinder. And they sport two magnetos that separately fire those two plugs plugs per cylinder. Much time has been spent over the decades in trying to time those separate magnetoes to fire at the same time. YMMV - not that easy to do. Yes, you are correct, the redundancy is present and it's a wonderful thing. My thinking, while not specifically enunciated, is to have two separate EI systems to retain that redundancy. The question is the style and sophistication of the EI. Perhaps the 10* limp mode of the EDIS is not needed or woefully inadequate? Further, ECU management of the advance throughout the RPM range (600 - 2800) would lessen the pilot workload. At the same time, some manual manipulation of advance and the WOT or near WOT performance of the engine at altitude (10K feet) provides an opportunity to zero in on certain metrics that could be retained for manual input or built into the table(s). There are more or different variables than experienced in automobiles. For instance, 100LL (104 octane low lead) aviation fuel burns slower than mogas. Further it makes use of a wide band O2 sensor difficult, if not impossible.

Most sophisticated airplane flight is primarily managed by EGTs. And, in those situations, the management is typically related to a very stable flight environment. A typical flight will be on a cross country mission and, for better or worse, will pick a higher altitude path with a constant engine RPM. At 10K feet and a constant 2500 RPM cruise, the engine is making 60% or less of its rated horse power at sea level on a 60* F day and worries about engine damage is minimal. Most pilots are going to lean out as much as possible to save fuel and extend range. With a lean mixture and little oxygen, more advance is needed. How much more? And what if the OAT is 0* F vs 30* F? Of course that could be handled by IAT considerations - but, being primarily used for EFI, can the software be programmed to have IAT impact ignition advance? And barometric pressure - mostly useful for EFI but maybe something to consider for EI. One thing that would be useful: if there was a pot type manual adjustment, a descent of 1000 feet altitude would trigger reversion back to a safe table advance - maybe it should be 500 feet.

Again, I thank you for your reply - I enjoy the mental and creative challenges. I hope we can keep this going.

[LarryGrant]

PSIG: Thank you for your reply.

You, of course, are right - I need to make a list of features needed and put that out for comment and critique. Problem is I don't know that much about Speeduino. I know that there is a Speeduino PCB that an Arduino processor attaches to. With that there is firmwire (is that the operating system?) and then public domain tables that can be manipulated by individual users. I'm getting in over my head already. Suffice it to say - I have no PCB or hardware skills-experience and my programming experience has been in IBM basic. I could probably learn C++ or whatever was needed.

Here is a start:
1. LS1 COP setup with Speeduino - self igniting (exciting) and no need for EDIS. I understand LS2/truck coil is a beast of a coil but don't think I need that kind of power. Either would be fine.
2. Hall effect sensor mated up to a 36-1 trigger wheel. Open to suggestions.
2. MAP (no TPS)
3. Barometric pressure sensor - read somewhere that the initial BP is taken by the MAP after ignition key is put in on position and before engine start. Not taken again until next key cycle. Would have to totally re-work and establish a standalone BP sensor.
4. IAT
5. POT for manual manipulation. Software would kick out of manual mode with a change in altitude of some predetermined amount (1000 ft or 500 ft).
6. A timing advance readout-instrument display on each ignition system

That's what comes to mind. Probably missing several items. Maybe we can flush them out over time.

I'll describe typical aircraft operations: it's generally going to be one of two possibilities. 1) cross country level cruise (typically at a higher altitude) with a constant engine speed; or 2) in the words of Chuck Berry - "no particular place to go" meaning that one is taking off, climbing, doing pattern exercises, descending, going 50 miles away for a hamburger, etc. - no consistent flight pattern and a situation where perfection is not required, just good performance.

Addressing no. 2, good engine performance is highly desirable, especially in climbing out. But there are so many variables, one can only hope to get the ignition advance relatively close to optimum as RPM and load changes. And getting close will satisfy almost every pilot - just about anything that is an improvement over the static advance of a magneto will be welcomed. There would be not much interest in manually tweaking the advance to the optimal degree. And the conditions are changing so rapidly, whatever was perfect a moment ago will be less than perfect in the next instant.

Addressing no 1, getting the engine perfectly tweaked out will desirable and will be a touchy-feely thing. In the absence of variable advance, pilots with EGT gauges lean until the EGTs peak and then decline by 50* - "lean of peak." So if one leans to a peak EGT of 1350*, they will continue to lean until EGTs drop to 1300*. Of course, you have 4 (or 6) cylinders and the peak is your best guest considering all cylinders. Great theory - not always easy to do. Now, add the ability to advance or retard the timing from the cockpit. If a manual adjustment was able to improve combustion in the cly head, then EGTs are going to further decline. So I could see the leaning exercise to be repeated after each ignition timing adjusted. Maybe there would typically be 2 or 3 cycles, depending on how particular a pilot is.

Thanks again for your reply - depending on how difficult it would be to do a manual tweak of the advance, this doesn't sound that wild.

[PSIG]

… Problem is I don't know that much about Speeduino. …

Ah, you are going too deep, and it can be much simpler than that. IMO, your first job is goals and requirements that the controller must accomplish. A list of jobs — not how to do them. You did that below. From that list, others with much more intricate knowledge of both Speeduino and the application of engine control can determine if one of those goals is attainable with Speeduino, or not, or another way to do it. You supply the things it needs to do, and we can fill-in the how and why.

Here is a start:
1. LS1 COP setup with Speeduino - self igniting (exciting) and no need for EDIS. I understand LS2/truck coil is a beast of a coil but don't think I need that kind of power. Either would be fine.
2. Hall effect sensor mated up to a 36-1 trigger wheel. Open to suggestions.
2. MAP (no TPS)
3. Barometric pressure sensor - read somewhere that the initial BP is taken by the MAP after ignition key is put in on position and before engine start. Not taken again until next key cycle. Would have to totally re-work and establish a standalone BP sensor.
4. IAT
5. POT for manual manipulation. Software would kick out of manual mode with a change in altitude of some predetermined amount (1000 ft or 500 ft).
6. A timing advance readout-instrument display on each ignition system

We can chat about your list, and many of the assumptions it makes. Back to the "job" list. In the meantime, I see real-time baro sensing (have that), MAP load sensing (have that), Hall CPS (no problem), IAT (yep), spark confirmation (can be added), and I'm not sure I understand #5's purpose or how you got there, if it's all automated (it is). I would respond with LS1 coils - why limit your choices? Plus coils can be "throttled" like engines if you want less spark energy for some reason. Anyway, you're also missing a few that we can fill-in as we go, based on your job list.

I'll describe typical aircraft operations: it's generally going to be one of two possibilities. 1) cross country level cruise (typically at a higher altitude) with a constant engine speed; or 2) in the words of Chuck Berry - "no particular place to go" meaning that one is taking off, climbing, doing pattern exercises, descending, going 50 miles away for a hamburger, etc. - no consistent flight pattern and a situation where perfection is not required, just good performance.

Very good. That is a job list. We can add definitions of expected and priority engine operations to bring focus, but that's a good basic outline of operation. I would suggest we use that description as the primary focus launch pad, and go from there.

Addressing no. 2, good engine performance is highly desirable, especially in climbing out. But there are so many variables, one can only hope to get the ignition advance relatively close to optimum as RPM and load changes. And getting close will satisfy almost every pilot - just about anything that is an improvement over the static advance of a magneto will be welcomed. There would be not much interest in manually tweaking the advance to the optimal degree. And the conditions are changing so rapidly, whatever was perfect a moment ago will be less than perfect in the next instant.

You are describing every engine ever made — except a stationary generator. For reference, aviation engines operate similarly to marine engines in loads and profile, and can be a good source of some concepts. We can also approach perfect pretty well, and IMO should be the goal, so that "good" is always the worst result.

Addressing no 1, getting the engine perfectly tweaked out will desirable and will be a touchy-feely thing. In the absence of variable advance, pilots with EGT gauges lean until the EGTs peak and then decline by 50* - "lean of peak." So if one leans to a peak EGT of 1350*, they will continue to lean until EGTs drop to 1300*. Of course, you have 4 (or 6) cylinders and the peak is your best guest considering all cylinders. Great theory - not always easy to do. Now, add the ability to advance or retard the timing from the cockpit. If a manual adjustment was able to improve combustion in the cly head, then EGTs are going to further decline. So I could see the leaning exercise to be repeated after each ignition timing adjusted. Maybe there would typically be 2 or 3 cycles, depending on how particular a pilot is.

Another good description, and I'm quite familiar with my aviation background. Again, the job list is the focus, as we can present different approaches based on technology and engine operation to solve these issues for you. Set it and forget it. Really. I've been there, so used to how it was done, that it's hard to see how it could be done. It's OK. As we move forward, we can explain some of this black magic so it all makes some sense. You already have the basics well in-hand.

Keep in-mind that there are now lots of EFI engines flying with full fuel and ignition control, with a lot of experience to demonstrate these options since the 1990s. Well, that's when I became involved in aircraft EFI anyway. You may benefit from visiting aviation EFI sites such as SDS EFI, where they have done this flying stuff for a very long time. Get some concepts and theory from some that have walked the walk. Then, consider if you want to transfer those concepts to a newer but more universal system such as Speeduino. Pros and cons.

Thanks again for your reply - depending on how difficult it would be to do a manual tweak of the advance, this doesn't sound that wild.

Heh, you're gonna love what you can do, once you see how it can be done. You are likely aware that topics such as this are often taboo, mostly based on liability concerns. Fortunately, the FAA provides some fairly extensive guidelines, but other than that, it's your life to do as you wish. I will not tell you what to do or not, and you are under no requirement to listen to anyone or do anything. This is the internet, after all. I can tell you how I might do it, if I were doing it. What you do or don't do is strictly up to you. As the designer of the Mini Max wooden ultralight told everyone (paraphrase from memory), "If you build this plane you WILL be seriously injured or killed", and I don't believe he was ever sued. So, consider I'm making the same statement about your engine control, whatever it is.

[LarryGrant]

PSIG: For better or worse, I'm going to try to put the EI side in place before consideration of EFI. The EI will be minimal downtime on the airplane and, for the early going, I'll retain one magneto for safety reasons. After the one EI channel proves itself, I'll add the other channel and do away with the magneto.

Are there Speeduino boards available for purchase that have different footprints? I see V2, V3 and V4. There's probably others? Setting aside the improvements in circuitry and capability, are they getting bigger or smaller? I did read through klotzy's ignition only thread. 2" x 2" may not be enough real estate but that size is attractive. With regard to the list of features, I think the LED readout (degrees of advance) and the POT (if it survives) could both be remote, which would allow, all other things being equal, a smaller PCB.

Is there one particular board that seems better suited (for ignition only) than others?

[theonewithin]

Nothing particularly suits your needs.

You would need to design something yourself if you wanted to.

Just pick any board. Say a 0.3. It's designed for prototyping.

Test. Play. Then once you decide if it's worth putting in any further effort look at other boards or make your own.

With the 0.3 just.leave put what you don't need. Like all the INJ stuff and only have 2 IGN channels populated on the board.

Good luck finding parts though. Chip shortage is making it very difficult.

[PSIG]

PSIG: For better or worse, I'm going to try to put the EI side in place before consideration of EFI. The EI will be minimal downtime on the airplane and, for the early going, I'll retain one magneto for safety reasons. After the one EI channel proves itself, I'll add the other channel and do away with the magneto.

While the ultimate performance, reliability and economy boost will certainly be when controlling both (one directly affects the other), I agree this testing and evaluation should proceed in small steps. But, it's your project. Do your thing. The example link to SDS was because they do ignition also, and have lots of info available about it. Digging may be required. There are others.

Are there Speeduino boards available for purchase that have different footprints? …
With regard to the list of features, I think the LED readout (degrees of advance) and the POT (if it survives) could both be remote, which would allow, all other things being equal, a smaller PCB.

I agree with @theonewithin, that most any (and there are many boards) will do the basic functions you are after for output hardware. I'd also suggest you get some hands-on experience with controlled ignition on... something. Anything. Buy a $500 beater than runs, and convert the ignition. Play and experiment. When you're done, pull the conversion off (save the hardware for the next one!) and sell it for $500.

While we can talk all day, hands-on is absolutely the best and quickest teacher. So many things we take for granted and even forget to tell others, because you "just know" if you've ever done it before. I firmly believe this would greatly increase not only understanding of what, how, and why, but would give you a solid knowledge base to work from as you then move much more quickly forward with your flying project.

Do you have a timing readout now? The LED readout is not a big deal, and various dash programs mean you could stick your phone on the dash for gauges and readouts, in addition to data-logging. The current timing value would only be informational, as once set-up and adjusted for best operation, it will do a better job applying best timing for current conditions than you can with a knob. That's the idea. Give somebody a knob, and let them change what should already be the optimum setting, and they will eventually mess it up. I like simple, and reduced pilot load. That's me. The pot is an interesting concept, and while I'm not fully understanding your position (given the previous and that you can change timing anytime with the tuning program), we also already have capability to input a "knob" signal for specific or global change in timing.

My first goal here is to outline the tasks (jobs) of the system, and the processes and procedures to validate results (also helping to define those jobs and methods). Tell us (me ) the reasons you want a knob, with how you would employ it. Add the "why" if possible, in order to give context.

When you top-off from avgas to mogas for the flight home, do you change magneto settings? It takes 30 seconds to switch the complete ignition tune from one fuel to another (or blended) if required, so you always have that option while you're working on other options.

Speeduino can add or subtract timing based on load (MAP) and RPM, but assumes the fuel is also controlled. As you have the ability to independently control mixture (Lambda or AFR to us), how would you coordinate for testing or operation, e.g., when fuel is at lean cruise and then the mixture is independently enriched, best lean cruise timing would then be excessively advanced.

With two separate systems firing sparks, tell us how you would experiment when the one with greatest advance would always control the outcome. I presume the magneto would be set for the minimum advance of the EI (low-power assisted landing), and the tuner would accept and calculate the power and efficiency loss of one retarded spark? Keep in-mind, that your mag-check at proper RPM or manifold pressure will of course have an abnormal drop, and where you could also use the drop as a fuel mixture indicator, that method isn't possible with two differently-timed systems. Hmm...

[Chris Wolfson]

Just to correct one broadly excepted mistake: Exhaust gas sensors work with leaded fuel.
I do not know the lead content of any avgas, but Bosch stated a life time with leaded premium (Europe) in the late 70's with about 40.000 miles. This was with narrow band, maybe check with wide band, but the principle is identical.
The nice thing: during contamination with lead the sensor output voltage stays constant, only the current gets lower over time until it is unable to drive a high ohm input.
So, if you use LL100 you should not have problems.

This myth comes from the platinium/ palladium coated catalytic converter, which would loose all effectiveness on one tank of leaded fuel. This was correct, indeed, even as one gallon of leaded gas, to limp home or the next filling station, would be tolerated by the converter. The greatest problem is or was x-contamination during transport, until the gas is actually in your tank.
The zirconium lambda sensor works on a completely different principle.
So you can use a lambda sensor at a plane, as long as you monitor it's work hours and check it's current output from time to time.