Post by ski on Sept 23, 2016 21:45:45 GMT -5
I will start by telling you this is a fairly long, somewhat technical post that will take some time to get through but I believe it is worth it. I dislike half assing things so I tend to look at all possibilities, aspects, drawbacks and advantages before starting a project. I take it all into consideration before jumping into action so I apologize in advance for the length of the post.
There have been a few ideas thrown around about this, but no one has yet to do it, or, from what I have seen, properly address it. I have a bit of experience in adapting fuel injection to engines that didn't have it. I thought it would be nice to start a post, point some things out, spitball some ideas and thoughts and really start working on this in a reliable, cost effective manner. So, let’s have a go.
Let's start off with the idea that has been put fourth for using a 4 stroke EFI set up. 4 stroke EFI systems won't work. The software and one piece of hardware, the reluctor, won't allow it to. The software could be rewritten though chances of that are slim to none and that leaves us with the reluctor. The rotary engine is unique in that all three actions of the internal combustion engine occur at the same time on each crank revolution within milliseconds of one another. A 4 stroke completes one combustion stroke every two crank revolutions. A rotary engine completes two combustion strokes every two crank revolutions because the rotor moves at 1/3 the crank speed. Mathematically, this gives us a 998cc displacement engine. Power wise, it's more like 750cc as I'm sure we all know this doesn't ride like any 500cc bike though it certainly doesn't pull like a 1000cc bike either. However, I digress...
All EFI systems (factory or aftermarket) use a crank trigger set up to determine TDC except for some very early versions of TBI that used a reluctor (Hall effect sensor) built into the distributor and the distributor rotates at 1/2 crank speed. The reluctor, or trigger wheel for these systems is either 36-1, 60-2, 24-2, 6-1 or 4-1. The first number refers to the total number of evenly spaced teeth the wheel WOULD have and the second number is the number of teeth that were removed from the total, maintaining the first numbers equal spacing. This missing tooth is viewed by the ECU as an indexing point. I did some preliminary math, though I didn't delve too far into it, and found that there wasn't a variation of any of the reluctor counts that would work with 4 stroke software and maintain a proper TDC count for the rotary allowing for properly timed intake and combustion events. Let's say you use a 36-1 reluctor and remove other teeth to fake the software into thinking TDC is occurring more frequently. The software still needs to see a specific number of teeth (30, 12, 6 or 4) in between the TDC windows for proper ignition and injection timing. Given the ratio difference between the 4 stroke and rotary thermodynamic cycles, there is no possible combination of missing teeth vs actual teeth that will result in a correct TDC trigger for every revolution of the rotary engine using 4 stroke software. So that idea, unfortunately, is out.
Second, is the kit offered by Ecotrons for rotary engines. Here I'm going to have to operate on speculation as it's been two weeks since I emailed them with questions regarding this system and have had no response. They stated their kit was already in use on several rotary engines. I asked them what engines it was being used on and what type of rotary it was designed for. Every production rotary engine save for ours and one or two others runs two spark plugs and multiple rotors. This presents some obstacles when using set ups designed for those engines with ours. Twin plug rotaries use a different ignition map. One leading spark and one trailing spark from separate coils to achieve proper spark duration long enough to ensure full ignition of the AF mixture. Single plug set ups like ours use CDI. Standard spark gives you one, short spark duration, whereas CDI gives you one long spark duration or multiple sparks per single ignition event.
Depending on the amount of the control the user is given by Ecotrons software, this can be overcome by disabling the trailing spark, retarding the leading spark and using a CDI box (which they offer). Assuming we can do that gets us passed one issue. The second would be if it's designed for multi rotor applications. Most commonly used was the twin rotor engine. The rotors were staggered 60 degrees out of phase. This puts the apex seal of one rotor in line with the center of the rotor face on the companion rotor. Now you have to contend with an extra injection and two extra ignition events per revolution. The solution: You simply don't hook up the circuits for the other coils and injector (providing the software doesn't monitor them). The optional solution is to install load resistors in those particular circuits to fool the software into thinking they're functioning correctly. If we get passed these issues, this is a viable system. If those issues are non-existent, then we've hit pay dirt. I'll cover more on Ecotrons later.
The last is 2 stroke injection which is a very close mirroring of rotary injection. There are issues here as well. Kits you would get for 2 strokes to use on the rotary are going to require a huge amount of live tuning to get the injection and ignition timing correct, and a lot of that is going to have to be done with a mock set up off the engine just to get it close enough to run the engine at all. Once that is done, you're going to need hours of dyno tuning to dial it in. That last part applies to any EFI kit we use. Rotary engines fire the injectors 270-330 degrees BTDC depending on application and engine speed. This number is somewhat close to 2 stroke timing and most software allows that number to be adjusted quite a bit so it's not really an issue. Two stroke injection systems and all others inject fuel only since 2 strokes are premix for the most part. Most injection systems run 40-50psi for fuel pressure. This is well above the pressure of our oil pump for oil injection so now we have a huge issue with injection as our engines require the oil or catastrophic engine failure occurs. This leaves us with 4 options for oil injection. 1) Premix every tank. I don't know about you, but I really don't want to carry a quart of oil and measuring cup so I can fill up every 120 miles. 2) Build a siphon style manifold that will pull the oil into the fuel stream. I don't see this as viable for a couple of reasons. Fuel pressure is constant in the system. The fuel under pressure is always going to seek out the path of least resistance to exit the system. I believe using a siphon style manifold is just going to cause fuel to flow into the oil tank, at least according to basic physics. I do know that high pressure systems causing a vacuum with a low pressure draw do work with air, but I'm not sure if that will transpose to liquid due to the actual flow rate which is much lower with EFI. 3) Increasing the pressure of the oil injection pump. Based on a preliminary inspection of the related mechanical components, I don't think it can easily be accomplished without major modifications to those parts. We could eliminate the mechanical pump and go electric but we would have to install a pressure switch that would cut power to the ignition system if the pressure falls below a preset level to avoid engine damage in the event of pump failure. 4) Create a reservoir upstream from the pump where oil is injected. This reservoir would mimic the current float bowl of the carb and it would need to be roughly the same volume as the bowl in the carb. I like the reservoir idea over the pressure increase via electric pump idea as I dislike adding more moving parts to fail.
Now we have to address the return fuel issue. There can't be any return fuel unless you're premixing because now you're returning mixed fuel back into the tank. We can't feed return fuel into the reservoir mentioned above because you'd be changing the mix ratio by returning mixed fuel into the chamber with fuel that is being mixed. I do have an idea for this. Use an inline FPR (fuel pressure regulator) after the pump and have the return fuel sent back to the inlet side of the pump and have a check valve installed upstream from that point to stop it from flowing back into the premix reservoir. The pressure introduced to the inlet side of the pump via the return line should eventually reach zero since the engine is using fuel and we're just cycling the leftovers to the pump inlet. When the pressure at the inlet does reach zero, the check valve will open allowing mixed fuel from the reservoir to the inlet side of the pump. I would need to build a bench test version of this fuel system to verify functionality. We do have to run some type of return though. Running a fuel pump continuously at full pressure dramatically shortens the life of the pump and makes it run hot.
Moving on, we'd need a throttle body and there are things we need to consider. Having a primary and secondary port, correct air flow is critical. The reason we have to different ports is that at very low RPMs, air velocity is much lower than a comparable piston engine so you need a smaller bore to maintain correct velocity to pull in the fuel and atomize it or idle quality and low end suffer or become non-existent. As the RPMs increase, so does the velocity and volume so the smaller bore is no longer sufficient to feed enough air and fuel so the secondary port valve opens. You cannot remove one or the other without suffering a loss at the bottom or top end. Using a throttle body with a standard butterfly/throttle plate will most likely cause issues for us on the bottom end. When the throttle plate begins to open, air is passing the plate around the top and bottom of the circumference and because of its design and flow characteristics, it creates turbulence which reduces the velocity. On a piston engine with the low end velocity being higher, this really has little to no effect but on our engine; I believe this will cause stumbling and bogging issues. What we need is a variable venturi throttle body. IE: Something with a slide, either square or round. Those of you who have tried to fit other carbs to the bike and have fought tuning issues may benefit from that thought and those who have used the Mini Cooper carb with no issues already know how well the slide carbs work on these bikes, especially those with a vacuum actuated slide. That's because the slide creates a variable venturi opening that is always roughly the same shape even though the size of the opening changes. The opening has little to no turbulence and that translates to the intake velocity remaining constant through the opening and into the bore. Luckily for us, these vacuum (load) actuated throttle bodies already exist for motorcycles and are fairly easy to come by plus they already have the injector we need. We also need to consider the throttle body diameter. I read a post here where someone added the diameters of our bores together and stated we'd need something like a 50mm bore. That's not how it works. lol Calculate the surface area of our primary and secondary bores, add them together then divide by pi to get the equivalent single bore diameter which happens to be 36mm. This is probably a bit bigger than we actually need as it will flow a higher volume of air than our carb does. Why would it flow more air if it has the same opening surface area? I know someone is going to ask. lol It's because we have an air horn in each bore of our carb for drawing in and atomizing fuel. This physical protrusion reduces the area of the bore and more or less creates an obstruction reducing the volume of air that can pass through. No obstructions in a throttle body.
Now let's talk actual control systems. I finally got a response from Ecotrons and after three weeks of emails with their tech advisor whose second language is English, I've come to the conclusion that using them will not be an easy road. Language aside, I'm not sure their advisor truly understands how fuel injection systems work. One main point I could not get her to grasp was that we can't use a crank trigger with our application which is required with all (including theirs) but a very few systems to trigger injection and ignition timing. We have no external access to the crank and internal access would require teardown and modification to fit a reluctor and the trouble of locating or creating a pickup sensor that can survive an ongoing oil bath at high temps. We can use the flywheel for the alternator! Right? Nope. The magnetic field generated by it would wipe out any signal generated by the reluctor and sensor and the system would not fire. We do have one place to mount a trigger. Since the distributor turns at crank speed, we can mount it in the distributor housing. Yay! We're saved! Nope. Crank reluctors for injection systems need to be fairly large, starting at 4" or so to maintain a clean signal to the ECU. That won't fit in our housing. The reason they have to be this size is so the spacing between the teeth of the reluctor is large enough to create a clean high/low signal to the ECU for it to determine crank speed and position. As the RPMs increase, the actual time of the high/low pulses becomes smaller. If the teeth are too close together, you get to a point in the RPM range where the signal becomes a blur and is unreadable so the system shuts down, The signals generated by reluctors or hall effect sensors are generated magnetically or by blocking a window between two poles of a sensor. Because both are generated electrically (analog signal), there is a finite response time for the sensor to sense and trigger the high/low pulse. We could make the wheel smaller and reduce the number of teeth which would increase the time in the high/low pulse. Something important to mention here: The more teeth, the more accurate the system is with timing the injection and ignition events. Fewer teeth, not so much. Timing is pretty crucial with our engines. Misfueling or detonation (even minor) on a rotary engine leads to catastrophic engine failure much faster than with a piston engine. Not a gamble I'm willing to take even though it could work. There is a trigger system that would work and fit in our housing. Using an optical sensor with a slotted trigger wheel dramatically cuts the response time in the high/low switching as these systems are digital and hall effect type systems are analog. Keep in mind we're talking milliseconds here and at high RPMs, a few milliseconds here and there makes a big difference. Optical systems have been employed by GM and Nissan but never really caught on as integrating them into the engine in a place that is oil and dirt free is a challenge. Good thing we have a more or less sealed distributor housing.
So we have a solution for a trigger system which is good and bad at the same time. Ecotrons system and every other 'kit' system I've looked at will not fire with an optical trigger without circuit modification. So now what? Enter Megasquirt ECUs, the Microsquirt system in particular. Megasquirt has developed EFI controllers in an open source forum. This means that they allow anyone access to their software and functionality of the hardware to constantly evolve the system, debug and add features. They will fire on damn near any signal. Ecotrons and everyone else use proprietary software because they don't want to play nice with the other kids. That approach is slowly dying in a digital society that has realized if you open source things, problems are fixed faster and nifty little features get added without costing the company more money. I bring this point up because I had the chance to play with Ecotrons tuning software and didn't find it as user friendly as I'd like. Add to that, their tuning software manual isn't as clear cut as it could be for people who have never tuned before, and because of that, any questions or issues with tuning have to be directed to them. Did I mention English as a second language? Megasquirt has a huge online community in several forums with loads of applied knowledge including their own website and techs. Their manual is better written, the software more user friendly and support is virtually unlimited. Yes, it's just a controller (ECU) and you have to wire it yourself and piece together the appropriate sensors but they've created a system that virtually anyone can build and use. GM sensors tend to be the preferred as they are cheap and easy to come by, but the system will use others. What happens when/if Ecotrons goes out of business? You're left high and dry with a system you can't service. I think we have enough of that with our bikes already. Megasquirt will control both the injection and ignition systems. The best part of that, it will control whatever ignition system you choose to use. That means you can use the stock CDI setup, Jess's electronic set up or one that I recently found. It's a universal MSD (multi spark discharge, not the conglomerate name) CDI system. Unfortunately, it's a build it yourself system so I will have to build one and bench test it for a few weeks at full RPM and output to verify it's durability. It was designed for V8 applications at a max of 10K RPMs, so it should do just fine with our bikes.
Last is cost. Ecotrons wanted $1000 (minimum starting point) for a system we could use on our bikes not including the CDI box which is another $100. After playing with their tuning software, I suspect they will have to tweek it to allow us to properly tune the system which I can only assume will be an additional cost. Granted you are getting everything with the Ecotrons system minus everything we'd have to fabricate to adapt it, but that's still pretty hefty. I'll break down the cost of using Microsquirt and getting all the parts you'll need minus the fabricated parts.
Micrsquirt ECU: $300.00
Microsquirt 30" pigtail harness $80.00
Inline fuel pump $50.00
Reservoir for fuel/oil mixing $20.00
Fuel pressure regulator (FPR) $100.00
Check valve (for fuel line inlet) $10.00
Optical sensor (trigger) $20.00
Optical trigger wheel $20.00
Throttle body (used, come as a set of 4 from eBay) $50.00
Throttle position sensor (TPS) $20.00
Manifold absolute sensor (MAP) $20.00
Baro sensor (for altitude compensation) $20.00
Engine coolant sensor (ECT) $20.00
Total: $730 (US dollars. Sorry, I was too lazy to figure exchange rates)
Still pricey, but that number is more or less fixed though you could throw in another $20.00 for things like fuel lines, clamps and other misc bits. Fabrication wise for either system, there will have to be an adapter for the throttle body to our manifold and another for air filtration. Misc brackets and relays. Tuning time is a whole other story depending on cost and experience of the tuner. Whether its $750 or $1000, that's still a pretty good chunk of change for most though I believe the benefits will offset the cost.
So I've thrown everything out there. I welcome questions, ideas, counterpoints, rebuttals or comments on things where you think I've just completely derailed. I've already sourced the necessary parts and am looking to start picking them up to give this a go. I'm happy to share my results and any insight or help is welcome. Post or PM me with anything you like. Ski
There have been a few ideas thrown around about this, but no one has yet to do it, or, from what I have seen, properly address it. I have a bit of experience in adapting fuel injection to engines that didn't have it. I thought it would be nice to start a post, point some things out, spitball some ideas and thoughts and really start working on this in a reliable, cost effective manner. So, let’s have a go.
Let's start off with the idea that has been put fourth for using a 4 stroke EFI set up. 4 stroke EFI systems won't work. The software and one piece of hardware, the reluctor, won't allow it to. The software could be rewritten though chances of that are slim to none and that leaves us with the reluctor. The rotary engine is unique in that all three actions of the internal combustion engine occur at the same time on each crank revolution within milliseconds of one another. A 4 stroke completes one combustion stroke every two crank revolutions. A rotary engine completes two combustion strokes every two crank revolutions because the rotor moves at 1/3 the crank speed. Mathematically, this gives us a 998cc displacement engine. Power wise, it's more like 750cc as I'm sure we all know this doesn't ride like any 500cc bike though it certainly doesn't pull like a 1000cc bike either. However, I digress...
All EFI systems (factory or aftermarket) use a crank trigger set up to determine TDC except for some very early versions of TBI that used a reluctor (Hall effect sensor) built into the distributor and the distributor rotates at 1/2 crank speed. The reluctor, or trigger wheel for these systems is either 36-1, 60-2, 24-2, 6-1 or 4-1. The first number refers to the total number of evenly spaced teeth the wheel WOULD have and the second number is the number of teeth that were removed from the total, maintaining the first numbers equal spacing. This missing tooth is viewed by the ECU as an indexing point. I did some preliminary math, though I didn't delve too far into it, and found that there wasn't a variation of any of the reluctor counts that would work with 4 stroke software and maintain a proper TDC count for the rotary allowing for properly timed intake and combustion events. Let's say you use a 36-1 reluctor and remove other teeth to fake the software into thinking TDC is occurring more frequently. The software still needs to see a specific number of teeth (30, 12, 6 or 4) in between the TDC windows for proper ignition and injection timing. Given the ratio difference between the 4 stroke and rotary thermodynamic cycles, there is no possible combination of missing teeth vs actual teeth that will result in a correct TDC trigger for every revolution of the rotary engine using 4 stroke software. So that idea, unfortunately, is out.
Second, is the kit offered by Ecotrons for rotary engines. Here I'm going to have to operate on speculation as it's been two weeks since I emailed them with questions regarding this system and have had no response. They stated their kit was already in use on several rotary engines. I asked them what engines it was being used on and what type of rotary it was designed for. Every production rotary engine save for ours and one or two others runs two spark plugs and multiple rotors. This presents some obstacles when using set ups designed for those engines with ours. Twin plug rotaries use a different ignition map. One leading spark and one trailing spark from separate coils to achieve proper spark duration long enough to ensure full ignition of the AF mixture. Single plug set ups like ours use CDI. Standard spark gives you one, short spark duration, whereas CDI gives you one long spark duration or multiple sparks per single ignition event.
Depending on the amount of the control the user is given by Ecotrons software, this can be overcome by disabling the trailing spark, retarding the leading spark and using a CDI box (which they offer). Assuming we can do that gets us passed one issue. The second would be if it's designed for multi rotor applications. Most commonly used was the twin rotor engine. The rotors were staggered 60 degrees out of phase. This puts the apex seal of one rotor in line with the center of the rotor face on the companion rotor. Now you have to contend with an extra injection and two extra ignition events per revolution. The solution: You simply don't hook up the circuits for the other coils and injector (providing the software doesn't monitor them). The optional solution is to install load resistors in those particular circuits to fool the software into thinking they're functioning correctly. If we get passed these issues, this is a viable system. If those issues are non-existent, then we've hit pay dirt. I'll cover more on Ecotrons later.
The last is 2 stroke injection which is a very close mirroring of rotary injection. There are issues here as well. Kits you would get for 2 strokes to use on the rotary are going to require a huge amount of live tuning to get the injection and ignition timing correct, and a lot of that is going to have to be done with a mock set up off the engine just to get it close enough to run the engine at all. Once that is done, you're going to need hours of dyno tuning to dial it in. That last part applies to any EFI kit we use. Rotary engines fire the injectors 270-330 degrees BTDC depending on application and engine speed. This number is somewhat close to 2 stroke timing and most software allows that number to be adjusted quite a bit so it's not really an issue. Two stroke injection systems and all others inject fuel only since 2 strokes are premix for the most part. Most injection systems run 40-50psi for fuel pressure. This is well above the pressure of our oil pump for oil injection so now we have a huge issue with injection as our engines require the oil or catastrophic engine failure occurs. This leaves us with 4 options for oil injection. 1) Premix every tank. I don't know about you, but I really don't want to carry a quart of oil and measuring cup so I can fill up every 120 miles. 2) Build a siphon style manifold that will pull the oil into the fuel stream. I don't see this as viable for a couple of reasons. Fuel pressure is constant in the system. The fuel under pressure is always going to seek out the path of least resistance to exit the system. I believe using a siphon style manifold is just going to cause fuel to flow into the oil tank, at least according to basic physics. I do know that high pressure systems causing a vacuum with a low pressure draw do work with air, but I'm not sure if that will transpose to liquid due to the actual flow rate which is much lower with EFI. 3) Increasing the pressure of the oil injection pump. Based on a preliminary inspection of the related mechanical components, I don't think it can easily be accomplished without major modifications to those parts. We could eliminate the mechanical pump and go electric but we would have to install a pressure switch that would cut power to the ignition system if the pressure falls below a preset level to avoid engine damage in the event of pump failure. 4) Create a reservoir upstream from the pump where oil is injected. This reservoir would mimic the current float bowl of the carb and it would need to be roughly the same volume as the bowl in the carb. I like the reservoir idea over the pressure increase via electric pump idea as I dislike adding more moving parts to fail.
Now we have to address the return fuel issue. There can't be any return fuel unless you're premixing because now you're returning mixed fuel back into the tank. We can't feed return fuel into the reservoir mentioned above because you'd be changing the mix ratio by returning mixed fuel into the chamber with fuel that is being mixed. I do have an idea for this. Use an inline FPR (fuel pressure regulator) after the pump and have the return fuel sent back to the inlet side of the pump and have a check valve installed upstream from that point to stop it from flowing back into the premix reservoir. The pressure introduced to the inlet side of the pump via the return line should eventually reach zero since the engine is using fuel and we're just cycling the leftovers to the pump inlet. When the pressure at the inlet does reach zero, the check valve will open allowing mixed fuel from the reservoir to the inlet side of the pump. I would need to build a bench test version of this fuel system to verify functionality. We do have to run some type of return though. Running a fuel pump continuously at full pressure dramatically shortens the life of the pump and makes it run hot.
Moving on, we'd need a throttle body and there are things we need to consider. Having a primary and secondary port, correct air flow is critical. The reason we have to different ports is that at very low RPMs, air velocity is much lower than a comparable piston engine so you need a smaller bore to maintain correct velocity to pull in the fuel and atomize it or idle quality and low end suffer or become non-existent. As the RPMs increase, so does the velocity and volume so the smaller bore is no longer sufficient to feed enough air and fuel so the secondary port valve opens. You cannot remove one or the other without suffering a loss at the bottom or top end. Using a throttle body with a standard butterfly/throttle plate will most likely cause issues for us on the bottom end. When the throttle plate begins to open, air is passing the plate around the top and bottom of the circumference and because of its design and flow characteristics, it creates turbulence which reduces the velocity. On a piston engine with the low end velocity being higher, this really has little to no effect but on our engine; I believe this will cause stumbling and bogging issues. What we need is a variable venturi throttle body. IE: Something with a slide, either square or round. Those of you who have tried to fit other carbs to the bike and have fought tuning issues may benefit from that thought and those who have used the Mini Cooper carb with no issues already know how well the slide carbs work on these bikes, especially those with a vacuum actuated slide. That's because the slide creates a variable venturi opening that is always roughly the same shape even though the size of the opening changes. The opening has little to no turbulence and that translates to the intake velocity remaining constant through the opening and into the bore. Luckily for us, these vacuum (load) actuated throttle bodies already exist for motorcycles and are fairly easy to come by plus they already have the injector we need. We also need to consider the throttle body diameter. I read a post here where someone added the diameters of our bores together and stated we'd need something like a 50mm bore. That's not how it works. lol Calculate the surface area of our primary and secondary bores, add them together then divide by pi to get the equivalent single bore diameter which happens to be 36mm. This is probably a bit bigger than we actually need as it will flow a higher volume of air than our carb does. Why would it flow more air if it has the same opening surface area? I know someone is going to ask. lol It's because we have an air horn in each bore of our carb for drawing in and atomizing fuel. This physical protrusion reduces the area of the bore and more or less creates an obstruction reducing the volume of air that can pass through. No obstructions in a throttle body.
Now let's talk actual control systems. I finally got a response from Ecotrons and after three weeks of emails with their tech advisor whose second language is English, I've come to the conclusion that using them will not be an easy road. Language aside, I'm not sure their advisor truly understands how fuel injection systems work. One main point I could not get her to grasp was that we can't use a crank trigger with our application which is required with all (including theirs) but a very few systems to trigger injection and ignition timing. We have no external access to the crank and internal access would require teardown and modification to fit a reluctor and the trouble of locating or creating a pickup sensor that can survive an ongoing oil bath at high temps. We can use the flywheel for the alternator! Right? Nope. The magnetic field generated by it would wipe out any signal generated by the reluctor and sensor and the system would not fire. We do have one place to mount a trigger. Since the distributor turns at crank speed, we can mount it in the distributor housing. Yay! We're saved! Nope. Crank reluctors for injection systems need to be fairly large, starting at 4" or so to maintain a clean signal to the ECU. That won't fit in our housing. The reason they have to be this size is so the spacing between the teeth of the reluctor is large enough to create a clean high/low signal to the ECU for it to determine crank speed and position. As the RPMs increase, the actual time of the high/low pulses becomes smaller. If the teeth are too close together, you get to a point in the RPM range where the signal becomes a blur and is unreadable so the system shuts down, The signals generated by reluctors or hall effect sensors are generated magnetically or by blocking a window between two poles of a sensor. Because both are generated electrically (analog signal), there is a finite response time for the sensor to sense and trigger the high/low pulse. We could make the wheel smaller and reduce the number of teeth which would increase the time in the high/low pulse. Something important to mention here: The more teeth, the more accurate the system is with timing the injection and ignition events. Fewer teeth, not so much. Timing is pretty crucial with our engines. Misfueling or detonation (even minor) on a rotary engine leads to catastrophic engine failure much faster than with a piston engine. Not a gamble I'm willing to take even though it could work. There is a trigger system that would work and fit in our housing. Using an optical sensor with a slotted trigger wheel dramatically cuts the response time in the high/low switching as these systems are digital and hall effect type systems are analog. Keep in mind we're talking milliseconds here and at high RPMs, a few milliseconds here and there makes a big difference. Optical systems have been employed by GM and Nissan but never really caught on as integrating them into the engine in a place that is oil and dirt free is a challenge. Good thing we have a more or less sealed distributor housing.
So we have a solution for a trigger system which is good and bad at the same time. Ecotrons system and every other 'kit' system I've looked at will not fire with an optical trigger without circuit modification. So now what? Enter Megasquirt ECUs, the Microsquirt system in particular. Megasquirt has developed EFI controllers in an open source forum. This means that they allow anyone access to their software and functionality of the hardware to constantly evolve the system, debug and add features. They will fire on damn near any signal. Ecotrons and everyone else use proprietary software because they don't want to play nice with the other kids. That approach is slowly dying in a digital society that has realized if you open source things, problems are fixed faster and nifty little features get added without costing the company more money. I bring this point up because I had the chance to play with Ecotrons tuning software and didn't find it as user friendly as I'd like. Add to that, their tuning software manual isn't as clear cut as it could be for people who have never tuned before, and because of that, any questions or issues with tuning have to be directed to them. Did I mention English as a second language? Megasquirt has a huge online community in several forums with loads of applied knowledge including their own website and techs. Their manual is better written, the software more user friendly and support is virtually unlimited. Yes, it's just a controller (ECU) and you have to wire it yourself and piece together the appropriate sensors but they've created a system that virtually anyone can build and use. GM sensors tend to be the preferred as they are cheap and easy to come by, but the system will use others. What happens when/if Ecotrons goes out of business? You're left high and dry with a system you can't service. I think we have enough of that with our bikes already. Megasquirt will control both the injection and ignition systems. The best part of that, it will control whatever ignition system you choose to use. That means you can use the stock CDI setup, Jess's electronic set up or one that I recently found. It's a universal MSD (multi spark discharge, not the conglomerate name) CDI system. Unfortunately, it's a build it yourself system so I will have to build one and bench test it for a few weeks at full RPM and output to verify it's durability. It was designed for V8 applications at a max of 10K RPMs, so it should do just fine with our bikes.
Last is cost. Ecotrons wanted $1000 (minimum starting point) for a system we could use on our bikes not including the CDI box which is another $100. After playing with their tuning software, I suspect they will have to tweek it to allow us to properly tune the system which I can only assume will be an additional cost. Granted you are getting everything with the Ecotrons system minus everything we'd have to fabricate to adapt it, but that's still pretty hefty. I'll break down the cost of using Microsquirt and getting all the parts you'll need minus the fabricated parts.
Micrsquirt ECU: $300.00
Microsquirt 30" pigtail harness $80.00
Inline fuel pump $50.00
Reservoir for fuel/oil mixing $20.00
Fuel pressure regulator (FPR) $100.00
Check valve (for fuel line inlet) $10.00
Optical sensor (trigger) $20.00
Optical trigger wheel $20.00
Throttle body (used, come as a set of 4 from eBay) $50.00
Throttle position sensor (TPS) $20.00
Manifold absolute sensor (MAP) $20.00
Baro sensor (for altitude compensation) $20.00
Engine coolant sensor (ECT) $20.00
Total: $730 (US dollars. Sorry, I was too lazy to figure exchange rates)
Still pricey, but that number is more or less fixed though you could throw in another $20.00 for things like fuel lines, clamps and other misc bits. Fabrication wise for either system, there will have to be an adapter for the throttle body to our manifold and another for air filtration. Misc brackets and relays. Tuning time is a whole other story depending on cost and experience of the tuner. Whether its $750 or $1000, that's still a pretty good chunk of change for most though I believe the benefits will offset the cost.
So I've thrown everything out there. I welcome questions, ideas, counterpoints, rebuttals or comments on things where you think I've just completely derailed. I've already sourced the necessary parts and am looking to start picking them up to give this a go. I'm happy to share my results and any insight or help is welcome. Post or PM me with anything you like. Ski
I have no problem with running 2 staged injectors. This is exactly how Mazda do it. 
