When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I've seen trying to figure out some lean spikes on throttle movements on my map, but I've really been struggling. Adaptronics support used to be great (Mark Gool was the man), but lately they really seem to have dropped off... guess this is to be expected with the Haltech acquisition. If anyone here can offer a hand it would be greatly appreciated!
I'm running a Modular w/ internal wideband (side note, the readings from the internal wideband seem to be much more accurate than my Innovate MTX-L), ID 1000/1700s, street port. I'm using predictive map settings and I'm quite familiar with the settings conceptually, but would to good to have a 2nd set of eyes on them to make sure I'm not doing something stupid. I calibrated the predicted MAP values on the dyno and they seem to be very accurate as verified by logged values. One question in regards to the predicted MAP values, is it correct that the highest value should be 0 psi? What if you're on boost between shifts? Is it assumed that once u engage the next gear and re-open the throttle the delay in the ECU's pressure sensor reading positive pressure will be less than the time it takes for positive pressure to be applied to the combustion chamber again?
You can see the 2 lean spikes I'm trying to sort out here. Both look to occur because of transient throttle not kicking in.
Here's my transient throttle trigger settings:
The first lean spike shows throttle input going from 3% to 14% over 200ms. My trigger table above has sensitivity of 15%/sec at TPS of about 5%. So my actual input would be about 55%/sec. So why did AE not trigger there? Also it's weird that the lean spike lasted for that long a duration.
The 2nd lean spike, throttle went from 25% to 55% over 360ms... so that's about 90%/sec. The trigger table is set at sensitivity of 30%/s at TPS 25%. Am I supposed to be using the ending TPS % value in the trigger calculation? That makes less sense to me and would result in a trigger table way more sensitive than what Adaptronic has in their base map.
Predicted map is disabled above 100kpa for the reasons you stated. I am not sure how the TPS delta is calculated, but there must be some logic for determining "instantaneous" TPS delta and in this case it was less than what you have in your TPS sensitivity table.
Hey, thanks for the input. Agreed, I was thinking about dropping the sensitivity levels even lower to see what happens, but was hoping to be a bit more calculated about it. Based on my maths (above) I have no idea how they're calculating the trigger points based on the entered sensitivity values. Took a look at your xls, not sure if I'm reading it right, but looks like my values are already quite a bit lower than the ones your listed (the 25 and 50% throttle positions). Weird thing is, the AE triggers perfectly at some points (you can see in the graph I posted, it shows an earlier trigger), but not at other points where I'd expect it to. Makes me wonder if there's a bug in the programming.
You're right, your values are already more sensitive than mine, my mistake. Can you post up that datalog please? Don't know if I can help but it would be good to take a closer look at what's going on there.
Sure. Here's a log showing a number of pulls. The first one doesn't show the proper AE triggers, but there are a number of other throttle applications in the log that show proper AE triggers so I'm not sure what gives...
Sure. Here's a log showing a number of pulls. The first one doesn't show the proper AE triggers, but there are a number of other throttle applications in the log that show proper AE triggers so I'm not sure what gives...
There are a few places in the log where the wideband pegs lean and you're not really pushing the throttle very hard, is that what you're referring to? I can't really see why it's gone lean there, did you feel that in the driver seat?
Last edited by mrselfdestruct1994; 09-17-19 at 08:12 AM.
Ya that looks like one of the spots where if AE kicked in it should have prevented the lean condition, but I agree, it's not a very aggressive throttle input. If you look to the throttle input to the right you can see AE kicking in with what looks like a similar or even lesser throttle input.
Are the Tau and X value applicable even when AE isn't triggered (even tho you're grouped in that section of the sw)? I would think they're applicable to fuel modeling in general even when not using the predicted map values. Maybe my values are off, but I doubt they're that off.
X-Tau is always active, you can see what it's doing by viewing the calc enrich fuel % column. I don't think that lean spike in the log above is a transient fueling problem, I think it's a misfire, but I'm not sure what's caused it. In my opinion you certainly don't need predicted map activated there as I wouldn't expect the map sensor to be slower than required when you're pushing the throttle that slowly.
Hmm interesting thought re the misfire. Maybe the spark is getting drowned out by too much fuel? I saw posted on these forums that X of about 40% is recommended with ID injectors because of the spray pattern. I upped my settings to that, but have been edging them down as I feel like the afr readings look better with a lower settings. It's not easy getting the right balance with Tau and X.
Hmm interesting thought re the misfire. Maybe the spark is getting drowned out by too much fuel? I saw posted on these forums that X of about 40% is recommended with ID injectors because of the spray pattern. I upped my settings to that, but have been edging them down as I feel like the afr readings look better with a lower settings. It's not easy getting the right balance with Tau and X.
Maybe, but my first thought was some kind of mechanical problem, e.g. old plugs. X of 40% should be ok, just ensure the values you're using don't cause oscillation. Motec have some instructions on tuning X-Tau and they state that some variation from the target is to be expected so I wouldn't lose too much sleep over your results, aside from those misfires.
hmm, I'll check for fouled plugs. Sounds like a possibility, but I hope not. I'm running fancy NGK 6448 R7420-9 plugs because I was having high rpm misfire issues before.
Although, looking at the graph below, where you see the big lean spike just after throttle input, is it odd that the primary injector duty doesn't increase at the time of the throttle input and AE trigger? Seems like it goes lean because the MAP ramps up on AE as it should, but no addl fuel is being added for about 300 ms.
hmm, I'll check for fouled plugs. Sounds like a possibility, but I hope not. I'm running fancy NGK 6448 R7420-9 plugs because I was having high rpm misfire issues before.
Although, looking at the graph below, where you see the big lean spike just after throttle input, is it odd that the primary injector duty doesn't increase at the time of the throttle input and AE trigger? Seems like it goes lean because the MAP ramps up on AE as it should, but no addl fuel is being added for about 300 ms.
It's hard to say from a screenshot, put up the log and I'll have a look I would guess that you're seeing an asynchronous pulse there that is way more than the engine needs and it's causing a rich misfire. I have async turned off as I found it did not work well consistently, much like normal manual acceleration enrichment. Sometimes it's not enough and sometimes it's too much. I will go back and try tuning it again later but my car is off the road at the moment. I have found that my engine runs well without it.
Good input. That makes sense as I have had similar experience with the async settings. Sometimes it seems to help out, sometimes it seems totally off. Hadn't gone to the extreme of completely disabling it and adjusting up from there though. I'll give that a try.
I am seeing similar behavior on a modular ecu. the engine is a streetport S6 with xcessive LIM, stock 550 primaries in stock location, ID1000 primary in LIM and ID2000 as secondary in the LIM. (ID1000 atm are not used)
Now I thought 2 things can cause this. 1: the double throttle. I have the feeling if you go through a certain tps range even a small movement causes a big change in VE, as suddenly the secondary ports also come into play. depending on the rpm I would guess they either help or lose 'real' VE. and this makes that in the fuel map on a small map difference you have to write weird steep edges.
the 2nd factor is that if the secondary ports are flowing air, but the secondanry injector are not firing this causes a bad distribution of fuel in the chamber and increases the chance on misfire or poor combustion. The car I am tuning at the moment has no turbo on it yet and I can almost cover the full load with the 550 primaries. when running just the primaries and lets say AFR 12.5-13 or running similar afr with 65% fuel in the 3rd stage (ID2000) it feels like the car picks up like 30-40 horsepower.
An issue I also have is that it seems very hard to make the transition between the 550 and ID2000. I feel that anythign under 1.5 msec on the ID2000 makes for a way to rich mixture. (if the afr before with just the 550's as fine). There fore you would expect that if I write a high number in the staging 3 map things will go better, which is essentially true once you are in the area where a decent amount of fuel is flowing in the secondaries. But during the transition I always get massive rich and no matter where I put the transition, it doenst like this area. I think maybe the minimum injection qty (15 microL) for the ID2000 plays a role but increasing or decreasing doesnt help. I run 300 kPa fuel pressure delta and use a sensor to measure the rail P.
The default setting leaving all the maps on 0 which will first maximize the primaries and only then ramp in secondaries if even worse as then you end with very low quantites inejcted on the secondaries and this doesnt help the accuracy. Also the engine burns way better that when both ports are flowing air also both ports get fuel.
Sounds like you have thought through the issues quite well, the only thing that occurs to me is have you tried running the car on just the ID1000 primaries?
Also, you mentioned you tried setting a minimum fuel volume for the secondaries, but did you do the same for the primaries?
Datalog of these issues and an ecu file would be useful.
Running the car on only the id1000 in the xcessive lim i spent maybe 50-100 hours troubleshooting. The car would 4/5 stall when coming to a stop. Fuelcut or not. Spent 50 mails back and fwd with adaptronic support . Total nightmare. I asked them 2-3 times are you sure the problem is not the id1000 in the xcessive lim inj far from rotor. They assured this wasnt the issue.
So yes in the end I gave up. But yes i found out now running the car runs perfect with the stock primaries in the stock location and when just fuel comes thorugh one of the ports when the throttle is open for both it serves for a big hit in performance (as I think the mixture isnt very homogen).
thats why I figured out i am now stuck with the id2000 secondaries which i better use as soon as its possible at medium to high loads as it aids performance and combustion stability if both ports flow fuel and once the id2000 are at the 60-65% all is fine but its rhe inital ramping in which makes for a rich out.
I was able to make a log and tbh I cannot at all understand the logic for the inj staging. I tried in the end alot of settings for minimum pulsewidth and fuel volume for both the primary and 3rd stage (the ID2000 are wired as inj 5-6) in the log you can easy see that the transition from one stage to both stages results in a rich out. when you look more closely it seems the inj 1 duration doesnt go down initally. at 2.5 sec in the log you see the primary go from 6 to 5.5 msec and the secondary from 0 to 2 ( I think I had about 2 msec as min duration there) then it takes like 3 (!) seconds before it starts to reduce the primaries. This is at nearly iso load/rpm and in open loop lambda control.
Hey Ruben, I'll be looking at this later tonight and seeing if there's something on the calibration side that I can figure out for you. Thanks for the email and story. Definitely puts a lot more into perspective on this build.
We know this ECU has trigger issues with non-resistor race plugs as well as what I call "Fake Math" in which some of the table values simply do not do what they are intended for. PID values for Closed Loop Idle, Boost Control, and Closed Loop Fueling seem like they're random. Amazing how every other ECU can set and lightly alter values and show immediate results. With this, I'm seemingly attacking PID with a sledge hammer to see almost zero variance. Would be nice to reference a manual...oh wait.
Yes, also the conditions for idle PID are quite strange. In a 200 dollar ecu you can set rpm and tps threshold below which idle control turns on.
On this ecu although it seems it doesnt want to go in close loop when the vehcile speed is not zero. this means lets say you idled it with high iat, causing a higher then normal IAC duty, then drive, IAT cools down, you remove foot from pedal it keep in open loop with the last duty it had from before AND with spark from normal sparkmap not from the idle spark curve (P part of spark idle control) and rpm hangs at 1500-2000. (car has no neutral switch). It has a clutch switch so as soon you press clutch it goes in close loop and all is well. The only workaround I found until now is writing a negative spak at 1500 rpm. the car idles at -10 approx so normally at 1500 part load you would write 20-25 deg this nets double torque causing the high rpm.
Maybe there is a setting that allows a simple TPS threshold to go into closed loop idle but I didnt find it yet..
Why would I not sage in vacuum? look at an RX-8, it has 3 stages being NA an producing only 220hp.. this car has primary 550 stock FD in iron, ID1000 as inj 3-4 in xcessive LIM primary runners and ID2000 as inj 5-6 in the secondary runners
This car will be a turbo hence the decent amount of fuel available but for now its NA and I first want to get it running as good as possible before putting the turbo back on.
I can almost fuel the engine alone with the 550 primaries but in doing so it feels like 30-40hp is missing even though the measured AFR is good. when I inject 60-65% of fuel in the 3rd stage (ID2000 in the secondary runners in the xcessive LIM) it has same measured AFR but feel like at least 30hp gain. I think this is because the fuel isnt really mixing too well inisde the chamber. if you have like a mixture of lambda 0.35 coming though the primary ports and solely air through the secondary ports.
Also it should be perfectly possible running the engine on both stages as the fuel required makes for a duration > minimum on the active stages. this you can see on the screenshot at lets say 3000 rpm WOT 2 msec on both active stages make for the fuel required. the issue is that the ecu somehow takes 3 seconds before it reaches ~ 2msec on the primaries after puttung the ID2000 to 2msec and this causes overflooding. whether or not I would change the speed/load where the stages happens, it still makes for an incorrect behavior.
Bit of a long shot, but have you tried setting up the ID2000s on the injector 3-4 plugs and only running 2 stages and logging the results? It *shouldn't* make a difference, but I feel like there's been enough bugs in the adaptronic software that I wouldn't feel 100% confident that it's handling your 0% secondary staging correctly and that's possibly causing weird outputs for the 1st and 3rd stages. Based on your logs it seems like something is getting miscalculated. I have no addl insight into the inner workings of adaptronics' algorithms, but is another possibility that because of your 2msec min on the ID2000s, the ecu is calling for, say, .5msec on them, but it's getting bumped up to 2msec causing the rich condition? You'd assume that the primaries would be instantly reduced to account for this addl increase, but who knows. It's also interesting that your primaries started to reduce duty cycle at the same time that IMAP started to increase, so is the drop from some timeout or from the change in IMAP? Would have to compare against a couple other log sections to know for sure.
Good info re the effects of staging. I had my secondaries set to go 30% at 6psi, but I'll play around with the settings and see if I can feel any difference.
I'd probably agree with rx72c above in that staging isn't needed in vac. I'm not sure what size injectors the rx8 uses but I'm guessing they're a lot smaller than your ID2000s so probably much easier to control in small increments. A lot of RX8 design decisions were also done for the sake of emissions at the expense of performance and longevity.
Yes, that was also initally my plan where it not that 1-4 and 5-6 have different pin size on the ecu connector. And different inj plug on the engine side. So I first wanted it to fix it like this before messung up the rywire harness. I guess if theres no fix from the software side theres no other choice.
I agree that staging at a higher load will make the problem less severe but you will always have to cross a line in the map where you go from no to some fuel on 2nd or 3rd stage.
I have held back on giving some input on this subject as I did not feel I could give the best advice yet. This is by far the hardest most difficult part of tuning on an Adaptronic in my opinion. I have a strategy that I have found to be working very well for me.
Step 1 - Turn off map prediction (go to the tuning fuel tab, Transient corrections, basic setup, Uncheck the use map prediction)
Step 2 - go to the Evap Time you will see that the basemap has 0.200 across the board - Go to the 8000 rpm section and set it to 0.020 and then interpolate the rows to the 1000 rpm section)
Step 3 - go to the Precentage Async and bring down the async down to 30 after 160F and after 3000 rpms set it to 0
Step 4 - X Fuel pooling has a few determining factors but primarly is based on injector type I find starting around 25 is a good starting point.
Step 5 - At this point you will need to start tuning the car the Vacuum and boost areas will need to be tuned using light throttle
Step 6 - Once the fuel map is relatively tuned you can start adjustment of the X pooling and the Percentage Async - You can do this by stabbing the throttle and increasing X and Increasing the percentage Async - You will know when you have it close when your AFRs are slightly more rich than your Target AFRS.
Step 7 - You will notice that your AFRs on transient throttle will tend to be leaner than your target AFRs and it goes through the RPM range - I have found using the RPM Rate correction resolves this issue. What RPM rate is a locked rate from 0-5000 rpm rate that based on the value of the RPM rate it adds a % of your total trim value in the RPM Correction table. So lets say your RPM rate is 2500 it will had 50% of the trim value you have in the RPM rate correction table.
Step 8 - During the logging you did during tuning the vacuum and boost areas you use that data to fill out the map prediction table - Enable the map prediction and fill in the table based on the data you can get from the logs.
Step 9 - with map prediction if you have it set to long you will feel a hesitation when you are stabbing the throttle and if you have the table incorrect this will cause the same as you will see the map JUMP to the wrong location and then jump back to were it actually is. I do find having the map prediction values roughly 10% more than what the actual car has seems to work the best.
I use Megalog viewer to average out the data to obtain my RPM rate and the map prediction values.
As for setting the time for the map prediction Adaptronics scope can be used to determine this value by looking at how long it takes the map value to move based on the time in which your throttle moves.
Hope this helps you guys and if you need assistance I do offer remote tuning services and I would be happy to assist on this. www.tunedbyshawn.com
Cheers,
Shawn Christenson
Tuned By Shawn
__________________ Tuned By Shawn - Adaptronic Modular Dealer and Remote or On-site Tuning Specialist We Specialize in tuning the Mazda Rx7 Platforms with the Adaptronic Modular ECU's Website - www.tunedbyshawn.com Email - service@tunedbyshawn.com Phone - 218-330-7369
09-15-19, 08:19 PM
