ECM and stuff (sorta) explained.....

DDT

.020 Over
Joined
Jun 28, 2020
Messages
23
Location
Kingston, NH
Ride
2003 Road Glide 107"
(tried replying to other thread, but 10K character limit.....)

a little reading and i answered my own question about 3 vs 1 MAP sensor. 1 sensor connected to all 3. interesting.
i also read a little bit about how people are using the MAP sensor input to a PC-V (or whatever) to try and make the fuel mapping changes load-referenced vs tps referenced.... trying to think about whether this is a good thing or not. obviously some folks have put a bunch of effort into it and there's interest in it, so i'm going to put down some thoughts and people can feel free to pick it apart or expound on it.....

the main reason to have fuel tables referenced to load (MAP) is because it's pretty accurate way to tell the ECU what the engine needs, vs what the TPS is. say you are lugging up a hill with a lower rpm and a lower TPS for some reason, say 5% TPS..... you have the RPM, you have the TPS, you have the VE by looking up the value, and so the fuel calculation goes. but, the engine is struggling a bunch as the hill is steep and you are being stingy with the throttle. you may be at 60 kpa (i know that the R3 ECU seems to use hPA, but i think in kPA since 100 is about atmospheric, or WOT). if you have a load-based table, you're over in upper 1/2 of the map and probably asking for more fuel in that VE table. in a TPS-based VE table, the ECM doesn't distinguish between 5% throttle at this RPM whether you are going up hill or flat cruise. (obviously, you won't be staying at the 5% throttle uphill and maintaining RPM the same as cruising flat, but whatever). now, in reality what tends to happen is that at lower RPMs going uphill or with a passenger or having a headwind, your load can end up at 100kPA at throttle openings that are *far* below WOT. what i'm trying to get at in perhaps a bit of a roundabout way is that at lower RPMs, you're engine can achieve very high load, and no matter how much you open the throttle beyond that if the RPMs are staying constant (hill gets steeper, wind gets harder, etc) you aren't getting any more air (and thus don't need any more fuel). so, at lower RPMs, MAP is a good thing to use for fueling.
now i'm not a PC-V expert, i stick to flash-based tuning for the most part. piggy-back devices like PC-V work by extending (or cutting off) the pulse-width of the injector based on the internal map it has and the RPM and TPS (or MAP...) inputs. the ECM sends the injector pulse, the PC-V does a lookup based on the other inputs, and either cuts it short or extends it. what it *doesn't* do is make the pulse happen *earlier*.... why is this even a thing you might ask.... depending upon the sophistication of the ECM, it may (should) change it's behavior on injecting fuel timing depending on the operating area that it currently is working in. it's not simply sending injector pulses every other revolution and making them longer as the VE table says to (regardless of whether it's TPS or MAP)..... if it's being smart, it's timing the injection to be at a good time in the cycle as well. recall, the ECM is trying to get the right amount of fuel into the cylinder based on all the table lookups and sensor inputs..... valve timing with poppet valve engines can play havoc on this ideal that the ECM is trying to achieve..... the ECM wants all the fuel that it squirts to be actually in the cylinder when it's time to spark. a few physical things try to prevent that.... 2 of the most obvious are exhaust overlap, where some of the intake charge gets sucked out the exhaust while the exhaust valve is still open, or the piston coming back up the cylinder and pushing some of the intake charge back out the still open intake valve. so, at lower RPMs, when we've got plenty of time to get the fuel in. ex: ~2000rpm, VE = 75% with a 14.5desired AFR and 2.5% TPS in cruise, kPA in the 30s, injector pulse width is say 2.5ms. at 2000rpm, the crank is traveling 30 degrees. with my cams having 246 degrees of intake duration, we've got plenty of time to open and shut the injectors. at low RPM, the ECU wants to inject while the intake is open and the exhaust is closed. at lower RPMs the ECM is concerned with the intake valve opening crankshaft degrees. if it knows when the valve is opening (and they do), it will do a good job of getting most of the fuel into the cylinder. conversely, at higher RPMs you have way less time to inject the fuel. at 8000 rpms, you've got 15ms. (if you're smart, you're keeping the injector's duty cycle under 80%, so you really only have 12ms to get all the fuel in). what (should) end up happening is at higher RPMs, the ECM needs to figure out how early to start the injector pulse to finish when the intake valve closes. intake valve closing is known, so the ECM figures how far back to start the pulse. at high RPMs with appropriately sized injectors, you're spraying onto the back of the closed intake valve in order to get it all in. this begs the question, if you are using a piggyback tuner to extend the intake pulsewidth, at the higher RPMs when the duty cycle is close to 80%, is the injector being held open long enough so that it's now spraying onto the back of the closed intake valve? does it just average out because you've got the leftover fuel sitting in the intake now for the next cycles? when the walls of the intake get wet and puddle from the extra fuel, unless the ECM algorithm is very sophisticated (and the owner hasn't changed the characteristics of the intake system much from the as-designed system), we will have a hard time trying to control the fueling, and probably fairly impossible with a piggyback tuner. you'd probably be best off by making the injectors a bunch larger than necessary and using the piggyback to only shorten the injector pulse.

i sort of forgot where i was going with this, i have to get back to working. anyways, i'm sort of curious how well the MAP-based adjustments to fueling with PC-V are being accomplished with this ECM and isn't it just easier to use the L table instead? the L tables have plenty of granularity it looks like. combined with accurate usage of the F-L switch, i would think it'd be pretty good. maybe i'm just missing the goal of using a MAP referenced fuel adjustment add on.....
 
anyways, i'm sort of curious how well the MAP-based adjustments to fueling with PC-V are being accomplished with this ECM and isn't it just easier to use the L table instead? the L tables have plenty of granularity it looks like. combined with accurate usage of the F-L switch, i would think it'd be pretty good. maybe i'm just missing the goal of using a MAP referenced fuel adjustment add on.....

For my jobs, which most recently was creating an entirely new ECU map for the 2.5L stroker, using MAP based fuel trims created in PCV and imported into the ECU map was a method to develop the L tables correctly.
And then move on to TPS based trims in PCV again and import to ECU map in order to develop the F tables.

My goal was always to have an ECU map for my build, and not need piggybacks permanently - they're only tuning tools. This also serves the kit-for-sale well too, so one doesn't need to spend several hundred $$$ on Dynojet stuff also, just a $20 cable and free software.

Develop the ECU map, and remove the tuning tools - they're another thing to go wrong. Having said that I'm a sucker for details and information so like to run AFR (& knock, with boosted build) monitors at all times...
 
For my jobs, which most recently was creating an entirely new ECU map for the 2.5L stroker, using MAP based fuel trims created in PCV and imported into the ECU map was a method to develop the L tables correctly.
And then move on to TPS based trims in PCV again and import to ECU map in order to develop the F tables.

My goal was always to have an ECU map for my build, and not need piggybacks permanently - they're only tuning tools. This also serves the kit-for-sale well too, so one doesn't need to spend several hundred $$$ on Dynojet stuff also, just a $20 cable and free software.

Develop the ECU map, and remove the tuning tools - they're another thing to go wrong. Having said that I'm a sucker for details and information so like to run AFR (& knock, with boosted build) monitors at all times...

Yup, i get what you're doing. i suppose my question (without knowing a ton about how the R-3 ECU algorithms) would be, how do you know when to sample the MAP? MAP is changing constantly as the valves are opening and closing and the ECU is sampling it at a certain crank tooth for a particular cylinder at a particular RPM...... or is the MAP reading output from the ECU and you are reading that somehow?
 
timing....
doing anything with timing other than removing it because you've got knock sensing capability (even then it's kinda dicey) *without* a dyno (and someone who actually knows how to use it) is Bad Idea(tm).
most people with just basic mods like intake/exhaust, etc. should just forget about changing stock timing at all. the factories spend a lot of time and effort to create tables that work and are fuel efficient and safe. the difference in optimal timing for intake/exhaust changes probably isn't worth the trouble to get into for just that.
now, if you really have a need for timing changes because you've changed your compression or your cam timing or have a power adder, then yes absolutely..... go to a dyno operator you trust and get them to do it. they will be the only ones with the tools to do it right. the best you can do on the street given the available tools (and safety!) is to not turn your engine into a melted slag of aluminum, or put a window in your case where there shouldn't be one.
timing is the last thing that gets done on a dyno. it's free horsepower, as the saying goes, but even dyno operators that do it for a living sometimes just skip it for the reasons given above. (one of the reasons i went out and got my own).
typical AFR-based tuning techniques (like you will get from the average Dynojet) happens (should happen) in 3 stages..... firstly, get the VE tables right. this is the easiest part. you're creating an airflow model of the engine, from the intake to the exhaust. you know your target afr (whether it's the AFR table you started with or some safe, no change across the board AFR table), and you have your good working order WBO2 sensors and an eddy current brake dyno that you can control the engine RPM and load with. you hopefully have some good software that you can watch in real time where the engine is operating and how much data is being collected at the given conditions, and you just watch TV. you operate the throttle with your right hand, you operate the dyno brake with your left and you try and hit as many of the cells in your VE tables as you can without melting the engine (or yourself, it gets pretty hot in there). usually this is done in stages with cool-down breaks (and maybe a wobbly pop towards the end?). anyways, now you've got a good airflow model through the engine hopefully. at this point you've probably gotten about 80% of the VE table filled out (WOT can wait).
now, just because you've got a good airflow model doesn't mean you know what the best AFR the engine wants to operate at for any given conditions..... you can have some shape thought out, like lean cruise, richer acceleration, lean overrun (but not too lean, to keep that decel pop down!), and a fatter right hand side of the AFR map. you really need to use a dyno with a torque reading on it here because you are looking (ideally) what AFR gives you the best torque. live tuning capability here *really* helps. you adjust AFR in a given cell (or set of cells to try and reduce the amount of time you spend doing this) until you've gotten your best torque.

whew.

*now* you can start to try and adjust timing. hopefully you've been paying attention to the knock sensors (you *do* have knock sensors and they are properly calibrated for the bore size and positioned well, don't you?) so you haven't put a hole in a piston yet..... here's where you go through a similar process to find the best AFR as above, but now you are adjusting the timing to get best torque as well. once the torque starts to nose over, it's time to go back to peak and then subtract a little more for safety.... make sure you are on the right side of the peak though when you start! an engine that is detonating in a cylinder or 2 can produce more torque than you would think sometimes, but it won't do it for very long :) so, now you've found your peak torque, you've backed off a few degrees as well. studies show that the amount of torque increase at the absolute peak vs backing off a few degrees is minimal, and you're engine will thank you on a hot day or getting a crap tank of gas. maybe you've even gone whole hog and you've got some EGTs wired in to watch the exhaust temps while you're doing this as well. EGT will increase to a point of being too lean, and then as you get even leaner it will go down, so again make sure you are on the good side of the peak. wobble the settings you test back and forth a bit to see if you're going up when you think you should be on the torque and EGT or whether you are going down (spark too advanced is now pushing back on the cylinder as it comes up).

lots of stuff to do to get a full tune, which is why it can take awhile to get it just right. think about that when you wonder what you're getting for 3 hrs of dyno time....
 
I would not be afraid to experiment with ignition timing, been doing it for decades on all sorts of engines, no explosions. I think the factory sets an ignition timing curve to keep the engine safe in the absolute worst conditions that can be imagined, extreme ambient temperatures, worst possible gas etc and then adds cushion on top of that as a safety factor.
With a compression ratio of just 8.7:1 I think there is ample room to improve the factory timing. A dyno could tell for sure.
 
I would not be afraid to experiment with ignition timing, been doing it for decades on all sorts of engines, no explosions. I think the factory sets an ignition timing curve to keep the engine safe in the absolute worst conditions that can be imagined, extreme ambient temperatures, worst possible gas etc and then adds cushion on top of that as a safety factor.
With a compression ratio of just 8.7:1 I think there is ample room to improve the factory timing. A dyno could tell for sure.
ah! well 8.7:1 is pretty low, and yes probably fairly safe to play with a bit. (i'm used to higher CRs where it can get a little bit more touchy, the engine in my bike is 11.05:1). so yeah, 8.7 you've probably got a bit of room. hopefully you've got a a dragstrip and/or are keeping track of your mileage to figure out if you are going in the right direction.
 
Yup, i get what you're doing. i suppose my question (without knowing a ton about how the R-3 ECU algorithms) would be, how do you know when to sample the MAP? MAP is changing constantly as the valves are opening and closing and the ECU is sampling it at a certain crank tooth for a particular cylinder at a particular RPM...... or is the MAP reading output from the ECU and you are reading that somehow?

I don't know. The 5v signal wire off the MAP is piggy backed to the PCV, so I guess the PCV "knows" when to take the snapshot based on when the injector pulses and rpm?
 
I don't know. The 5v signal wire off the MAP is piggy backed to the PCV, so I guess the PCV "knows" when to take the snapshot based on when the injector pulses and rpm?
does it? or is the MAP just the X in the RPMs Y in the table at the time the injector pulses? meaning, at the time the injector pulses at low RPM, is the intake valve still closed, or is it open or do we even know when in the stroke the ECU reads MAP for a given cylinder? does PCV for the R3 use the crank position signal as strictly an RPM input or is it also using it to know where we are in the 3 cylinders' strokes?

(what are the R3s sensors for this anyways? does it have a CKP and and CPS, or just the CKP?)
 
does it? or is the MAP just the X in the RPMs Y in the table at the time the injector pulses? meaning, at the time the injector pulses at low RPM, is the intake valve still closed, or is it open or do we even know when in the stroke the ECU reads MAP for a given cylinder? does PCV for the R3 use the crank position signal as strictly an RPM input or is it also using it to know where we are in the 3 cylinders' strokes?

(what are the R3s sensors for this anyways? does it have a CKP and and CPS, or just the CKP?)

My understanding is that the ECU knows which cylinder is inducing based on crank position and snapshots the MAP at the appropriate time to give the MAP reading per cylinder (Rocket III has 1 map sensor tubes to the 3 throttles).

Yes the PCV is spliced into the crank position sensor plug so this must be where its rpm reading comes from. I'd imagine might also uses this for crank position info as the add on ignition module can be used to alter timing (ignition module also has crank input but is apparently not used for Rocket III application)

So with all this info I'd imagine the PCV knows when to snapshot the MAP signal for pressure based fueling
 
Back
Top