|
Post by arttu on Sept 19, 2014 5:25:18 GMT
Hello all EFI users. Has anyone tested how injection timing on sequential systems affect to power or engine behavior otherwise? I did recently some limited testing on a dyno and though the results weren't very surprising or dramatic they were still quite interesting. There are two injectors sets on my engine. The primary injectors are on the throttle bodies quite close to the head and the secondaries are on the back of the plenum aimed to the intake trumpets. The primaries are injecting always and the secondaries come up when the boost rises above 0.4 bar which happens around 5000 rpm at full throttle. The primaries are about 250cc and the secondaries about 600cc so the majority of the fuel comes from the secondaries when they are active. Before dyno experiments I had adjusted timing for primaries with seat of the pants method by looking for stable idling and good throttle response. On the dyno I did just one experiment with primary timing by changing original value by 100°. For my surprise there wasn't any change on power or torque but throttle response got clearly worse. Since there wasn't any change on power to any direction I didn't see much point to continue experiments with this. The secondary timing on the other hand had some effect to power. For the starters this was set to default value 360°. In addition to that I tried 240° and 180°. 240° gave the highest power and 360° lowest, difference was about 3%. Since time was limited I quit experiments at that point. Most likely there wasn't much additional gain available since the peak was at middle value. Though the differences were small they are most likely real. Repeatability between pulls without changes was quite good, usually they were within 1hp. Here are the dyno graphs with different secondary settings: The Megasquirt 3 ECU that I'm using allows setting timing as table with RPM and throttle/MAP but I was using just single values over the whole table. Based on the dyno graphs there doesn't seem to be much need to alter timing with RPM. Of course there is probably some gain available by optimizing the timing further for different RPMs and load levels but I guess the gain would be very minimal compared to required effort. Comments?
|
|
|
Post by slingy1052 on Sept 19, 2014 21:25:35 GMT
Can't really help Arttu, but interesting. Have to say though, 300hp is very impressive result.
|
|
|
Post by arttu on Sept 20, 2014 18:50:34 GMT
Thanks. I'm not completely sure how accurate those hp numbers are since that was a homemade dyno. But it should be pretty close to what commercial dynos usually read.
I haven't seen much facts how injection timing affects to top end power. On the MS forum there was one apparently quite professional tuner who shared his experiences about this. His findings were quite close what I saw. He was talking about some engine masters competition engine that he tuned and if I recall correctly he was able to squeeze out few percents of extra power by tuning the timing.
|
|
|
Post by katana on Sept 20, 2014 21:05:49 GMT
Logic says to make power you've got to get the fuel into the cylinders. Airflow assistance to the injected fuel is low at lower rpm's so getting it right here timed to match valve opening won't be the same at higher rpms as the moving volume of air will assist the fuel getting to the valve quicker so i'd guess shorter timing would work better. The car manufacturers who use VVT systems pay a lot of attention to this due to moving valve timing up to 60-70 degrees and complete combustion is essential for gas mileage and emission reductions. Any measurable difference is a result but you may see bigger gains at lower rpm's?
|
|
|
Post by arttu on Sept 21, 2014 6:12:36 GMT
Yes. Before I also thought that gains would be biggest at lower rpms. But reality doesn't seem to match to that At least in my case. Of course my testing was very limited but I think I would have seen some change at low rpm if there was something to gain. In the dyno graphs the primary injector timing was altered by 100° between the 295hp curve and the rest. But effect on throttle response was quite clear at low rpms. It sounded like the engine would needed much more acceleration enrichment with less optimal timing.
|
|
|
Post by katana on Sept 21, 2014 11:10:35 GMT
Yes, but was the injection timing a fixed number ie. 100 degrees ATDC or mapped to vary based on rpm? That could make the difference? Its like cam timing, swing it around and you'll move the torque but not significantly improve it as valve lift is fixed so airflow is pegged. I seem to remember my old Motec M8 had mappable injection timing based on rpm and load - never really played with it but does obviously have benefits!
|
|
|
Post by arttu on Sept 22, 2014 6:04:13 GMT
The timing is mappable based on rpm and load but I was using just a single value for the whole table. But if changing that fixed timing value doesn't make any difference at any rpm then variable timing probably wont change the situation. Right?
|
|
|
Post by slingy1052 on Sept 22, 2014 8:20:54 GMT
So, if the injector timing has little effect, is the only real advantage of going sequential, (as far as this topic goes,) having a longer dwell time for coil charging? If so, seems like extra work for not much gain
|
|
|
Post by katana on Sept 22, 2014 10:25:48 GMT
The timing is mappable based on rpm and load but I was using just a single value for the whole table. But if changing that fixed timing value doesn't make any difference at any rpm then variable timing probably wont change the situation. Right? I honestly don't know but if using a fixed number throughout you could be missing the areas where different timing at, say, lower rpm's or load is required. Not knowing how you dyno your motor, it's always been a gripe of mine that dyno operators seem to strap down, warm up, up through the gears and then WOT to obtain the graph and map on that basis whereas there are about 9000 other operating points that aren't looked at. Without step loading the engine and optimising at each step point it is hard to rule out the benefits at this point. Regarding the question of sequential fueling / ignition - unless you are looking for maximum low down torque / minimal fuel consumption it isn't worth it, but it does have advantages so if it can be used - why not......it costs nothing other than a sensor, COPs and some wiring?
|
|
|
Post by katana on Sept 22, 2014 10:26:22 GMT
The timing is mappable based on rpm and load but I was using just a single value for the whole table. But if changing that fixed timing value doesn't make any difference at any rpm then variable timing probably wont change the situation. Right? I honestly don't know but if using a fixed number throughout you could be missing the areas where different timing at, say, lower rpm's or load is required. Not knowing how you dyno your motor, it's always been a gripe of mine that dyno operators seem to strap down, warm up, up through the gears and then WOT to obtain the graph and map on that basis whereas there are about 9000 other operating points that aren't looked at. Without step loading the engine and optimising at each step point it is hard to rule out the benefits at this point. Regarding the question of sequential fueling / ignition - unless you are looking for maximum low down torque / minimal fuel consumption it isn't worth it, but it does have advantages so if it can be used - why not......it costs nothing other than a sensor, COPs and some wiring?
|
|
|
Post by arttu on Sept 22, 2014 14:32:28 GMT
So, if the injector timing has little effect, is the only real advantage of going sequential, (as far as this topic goes,) having a longer dwell time for coil charging? If so, seems like extra work for not much gain Well, for maximum power/torque benefits of sequential injections are small indeed. On the other hand it's a bit relative if for example 3% increase is small or big gain. Of course it's a small increase in absolute scale but if you have already used all other possibilities to increase power then it might be a significant improvement. And it's kind of free power that doesn't require increasing boost or rpm with all negative side effects. But I think the main benefits are elsewhere, like improved throttle response, reduced fuel consumption and emissions. The latter two are probably the main motivation why car/bike manufacturers are using sequential. One nice side feature for average gearhead is ability tune injection balance between the cylinders. In my case I noticed especially more stable idling when I moved to sequential. Also I would say that low down torque improved as well, I mean below 2000rpm and so. Unfortunately it isn't easy to measure that low rpms on the dyno. Throttle response improved slightly too but difference to batch injection isn't huge if both have well tuned acceleration enrichments. On ignition side the main benefit is more time for dwell like you said. And also smaller current draw that might be useful in some cases. Overall the benefits are probably too small to justify upgrading the whole engine management just because of them, for most of the people. But if you are for example upgrading the ECU any ways and going sequential needs just a cam sensor and some wiring then it's worth of it. But it's good to remember that these new adjustable parameters also need someone to tune them to get any benefit. That's often the biggest effort
|
|
|
Post by arttu on Sept 22, 2014 14:49:38 GMT
I honestly don't know but if using a fixed number throughout you could be missing the areas where different timing at, say, lower rpm's or load is required. Not knowing how you dyno your motor, it's always been a gripe of mine that dyno operators seem to strap down, warm up, up through the gears and then WOT to obtain the graph and map on that basis whereas there are about 9000 other operating points that aren't looked at. Without step loading the engine and optimising at each step point it is hard to rule out the benefits at this point. My point was that if timing affects to torque at some certain rpm point, lets say 3500rpm for exapmple, then shifting the timing for the whole rpm range should change torque at 3500rpm too, to up or down. And that didn't happen in my case. Of course my testing was very limited and there might be some timing value that I didn't try that gives noticeable improvement. And like I mentioned above my feeling is that torque below 2000rpm improved when I changed to sequential. In this case it was completely up to me what to run in the dyno since it was me twisting the throttle . This time I did some fuel table mapping for partial throttle areas but I simply hadn't time to try for example injection timing for anything else than WOT. General problem with tuning is that required tuning time and effort tends to explode when you add more parameters to tune. At least for me this kind testing, result analysis and tuning based on the analysis tends to take time. It's surprisingly small number of things that you can get tested during one day dyno session.
|
|
|
Post by katana on Sept 22, 2014 17:12:08 GMT
Absolutely agree - Dyno time cost someone money down the line!
|
|
Teresa O.
Velociraptor
Posts: 37
Bikes: Trickframed 1127, Turbo 14 and many more
Reg: May 31, 2013 18:51:42 GMT
|
Post by Teresa O. on Sept 23, 2014 5:57:31 GMT
actually the whole story about MPFI injection strategy works a bit different than most people might thinkt.
From a combustion stability point of view, you want to inject on a closed intake valve when idling or at low speed part-throttle. This permits the fuel to have a lot of time to evaporate on the warm valve and homogenize with the air befor the intake stroke. This decreases the events of statistic combustion iregularities...
At high rpm WOT you have to overcome two issues. Minimum time for evaporation and homogenization (thats why we like secondary incectors far away over stacks) , paired with the wish for maximum load. Besides the obvious factors on load like cam-timing aso., your WOT performance benefits when you manage to make your fuel evaporate in the combustion chamber (Injecting in intake-stroke). The enthalpy increases density and with that, the potential to increase load. Thats also one of the few benefits of modern DI engines...
Cheers, Philippe
|
|
|
Post by nitro on Sept 23, 2014 11:51:10 GMT
I tried to change timing at idle. When getting better (richer) emission value the throttle response becomes shitt. I think the longer evaporisation in the intake port when hitting the closed intake valve makes richer better combustion. The injection in the intake stroke makes better throttle response.
At higher rpm / load I did not check, coz have no possibility to compare.
|
|