Starting to adjust EAE and decel settings please comment
My tune felt good, and then I enabled the AEA. It immediately felt better. The power comes on buttery smooth if that makes sense.
The downside is that I notice a slight hesitation between shifts. I thought it was my shifting, but now I think it is cutting a little fuel as I shift. I am removing a couple tenths from "suck from walls" to see if this is the issue. I also have not turned on lag compensation or overrun fuel cut. Do my starting point settings look ok? I don't want to go backwards turning these on, so if you have a jumping point, please post. MSextra doc is a little vague about the overun fuel cut. My TPS accel based enrichment seems to work fine, but I thought I would throw it in the mix for review. Thanks. https://www.miataturbo.net/picture.p...pictureid=1411 |
I started playing with EAE lately and it's getting better, but still going lean as shit between shifts. I know it's not my overrun settings because A) I have a 1.8 second delay on it and it doesn't take anywhere near that long to shift and B) datalogs don't show my injector PWs totally dropping out.
Notice that the EAE tables only go down to 30 kpa. I know on overrun I will pull 14-15 kpa when in gear, but I'm not exactly sure how low the MAP is going when I clutch in. Needz moar datal0gz. It might be possible that it's below 30 kpa and I might need to add cells below that to the EAE tables. I don't know, but I'm not messing with it for at least the next few days because in this cold weather my car barely even warms up. There are some threads on msextra that might help. Some people occasionally type "EAE" when they really mean "normal AE" (the old TPSdot/MAPdot type) but you can usually figure out what they mean. http://www.msextra.com/forums/viewto...?f=101&t=28038 (long) http://www.msextra.com/forums/viewto...?f=101&t=32000 http://www.msextra.com/forums/viewto...p?f=91&t=35230 |
I started playing with EAE lately and it's getting better, but still going lean as shit between shifts. I'm going to mess with this, then turn on lag compensation, and then turn on "over rune fuel cut" This way I should be able to pinpoint my trouble areas, and tune the problem. Thanks for the links. I searched this forum for a while last night, but there isn't one particular section where I found what I needed. |
You need a good long tuning session on a stretch of open road to tune your EAE, it's nothing we'll be able to look at and say it looks good.
Read the tuning manuals for it: http://msextra.com/doc/ms2extra/MS2-...anual.html#EAE |
Originally Posted by Braineack
(Post 668820)
You need a good long tuning session on a stretch of open road to tune your EAE, it's nothing we'll be able to look at and say it looks good.
Read the tuning manuals for it: http://msextra.com/doc/ms2extra/MS2-...anual.html#EAE I just don't want to be too aggressive with the EAE if I should be tuning first with the lag compensation. I've been reading that the EAE is more for fine tuning, which leads me to believe I should turn on lag compensation first and adjust for lean tip in with this. My lean to recover time is very rapid, which based on the msextra doc, suggests that my ATW and SFW are close to an ideal ratio. The overrun fuel cut has me a little confused as well. It seems like this could start causing problems during shifting. I guess the kpa needs to be set lower than what it reads between shifts, but above the min. kpa it sees at throttle let up. It asks for rpm ranges and I'm not sure what these should be. I want it to be high enough to recover the idle. What rpm do you guys have the fuel come back on? I was going to try to use the defaults 1500 and 1100. Ideally I would take a day and tune this. Unfortunately this is not a luxury I have at the moment. I'm logging to and from work and making my adjustments in between. Test, bench adjust, test, repeat. It's slow, but is working, and that is why I want as much input as possible, to help speed the process. |
I have fuel cut above 1700 and return below 1600. Delay of 2 seconds.
The problem I was having with the fuel return at 1100 was the delay between when the RPMs hit 1100 and the time it takes for the injector PW to kick back in. It was causing idle droop and I was relaying on a large dashpot value to compensate, but then a hanging idle under normal conditions. I was going over logs and there was a few ms of time that past between when the CL idle was kicking in and there was still no PW. raising the rpm to 1600 solved the issue...fuel kicks in a littler earlier but it allows everything to stabilize as you come to a slow stop. If you have a good tune, there should be little notice of when the fuel even kicks back in if it does when coasting at that low rpm or if you like going uphill in 2nd gear without throttle input. |
Here is a piece of my datalog. The good news is that the AFR on all three shifts follows the same pattern. This means that it is tunable and due to my settings.
What I see: On throttle lift the AFR remains stable, then richens, AE comes on as throttle is applied, AE cuts off and results in huge lean spike. Keep in mind that this is during warmup so the AFR is going to be richer than normal. https://www.miataturbo.net/picture.p...pictureid=1412 https://www.miataturbo.net/picture.p...pictureid=1413 Keep in mind that shifts with this tune this morning feel pretty good, but I can feel that there is improvement. the datalog seems to suggest this. How would you combat this datalog? Lag compensation is off, as well as fuel overrun cutoff. |
I see around 16:1 when i stomp on it, for a blink of an eye. I dont feel it at all. Couldn't get much better so that's good enough for me, much better than when I was on ms-I and basic enrichments. Honestly, the on throttle response is much better than both my Altima and Prelude.
|
My overrun rpm is set for above 1800. You're not saving a lot of fuel by setting it much lower. Kpa below 20 I believe (don't have it in front of me) because it's lower than what I should see with the clutch in for shifts, but higher than the 14-16 kpa I see during overrun. TPS below 0.8%.
I believe the EAE section of the manual says that lag compensation is specifically for cars running sequential injection, so if you're still running batch I'd leave it off. It just adds to the # of squirts while keeping the total amount of fuel the same, to get a little fuel in quicker for cars that normally inject only once per cycle, so the response is a little quicker during sharp blips. That's the theory anyway. I no haz sequential yet so I haven't messed with it. Tuning while looking at logs... it's helpful to have your AFR Target plotted on the same graph as your WBO2 AFR, so you can see how they differ when you accelerate or lift. The kpa ranges where they start diverging is where you need to adjust ATW or SFW for accel or lift. That's what I've been doing anyway. Unfortunately I realized I've been doing this shit with normal AE still enabled, so I will get to redo most/all of it. |
file > save graph as png or jpeg
knock off that blurry and complicated screen capture BS. |
I see around 16:1 when i stomp on it, for a blink of an eye. I dont feel it at all. Couldn't get much better so that's good enough for me, much better than when I was on ms-I and basic enrichments. Honestly, the on throttle response is much better than both my Altima and Prelude. "it's helpful to have your AFR Target plotted on the same graph as your WBO2 AFR, so you can see how they differ when you accelerate or lift" knock off that blurry and complicated screen capture BS. |
Originally Posted by djp0623
(Post 668938)
How do I divide the target AFR by 100 so that it shows up as 14.70 instead of 1470?
Remove the quotes but keep the brackets. |
you also want to makesure they scale the same, min max at 0-20
|
Originally Posted by djp0623
(Post 668938)
yes sir. If I ever figure out how to successfully trim the data, or have a log file that is less than the mas allowable files size, then I will include it.
http://y8spec.com/megasquirt/images/complog_running.png ... Incidentally, the "16:1 AFR Spike" is not supposed to be there. I'd look at whatever setting is responsible for low RPM + high MAP + high TPSdot fuel adding. I'd check first the TPS Based AE and put some significant values (50%?) for TPSdot values that equate to "stomping the pedal". |
Originally Posted by y8s
(Post 668969)
i dont think you understood my suggestion... there's a direct way to save an image file of the current graph you're viewing within MLV. it results in this:
http://y8spec.com/megasquirt/images/complog_running.png ... Incidentally, the "16:1 AFR Spike" is not supposed to be there. I'd look at whatever setting is responsible for low RPM + high MAP + high TPSdot fuel adding. I'd check first the TPS Based AE and put some significant values (50%?) for TPSdot values that equate to "stomping the pedal". I really don't understand what you are saying about the 50% TPSdot values. My TPS AE tableis at the top of the page. I will readily admit that I do not understand how this works, or how to adjust it. I am currently reading the MSextra manual, bur the reference tables seem to be to megatune, and I have never used it. I have no clue as to what to adjust on the TPS based AE since there is no correlation to kpa or TP on the graph. Edit: The best I could come up with would be to increase PW adder values by about 11% since this is the % difference between 14.3 and 16 afr |
Actually, lag compensation is meant for those *not* running seq, but is not necessary with later firmware revs b/c you can use a little bit of normal AE in the quick-blip situations to compensate for the lag that lag compensation was originally written to compensate for.
The trick to getting EAE working right everywhere is to make sure your VE table covers the areas you don't normally drive very long (all the way down to just above cranking RPM, and all the way into overrun). This is because EAE's purpose is to make sure the amount of fuel specified by the VE table is actually getting into the cylinder during transients. If the VE table is wrong or doesn't go into the operational range you're having an issue with, you won't be able to tune the bad behavior out no matter how much you mess with the EAE settings. The base problem in those cases is the VE table not the EAE settings. Ken |
Originally Posted by muythaibxr
(Post 669032)
Actually, lag compensation is meant for those *not* running seq, but is not necessary with later firmware revs b/c you can use a little bit of normal AE in the quick-blip situations to compensate for the lag that lag compensation was originally written to compensate for.
The trick to getting EAE working right everywhere is to make sure your VE table covers the areas you don't normally drive very long (all the way down to just above cranking RPM, and all the way into overrun). This is because EAE's purpose is to make sure the amount of fuel specified by the VE table is actually getting into the cylinder during transients. If the VE table is wrong or doesn't go into the operational range you're having an issue with, you won't be able to tune the bad behavior out no matter how much you mess with the EAE settings. The base problem in those cases is the VE table not the EAE settings. Ken |
Originally Posted by muythaibxr
(Post 669032)
Actually, lag compensation is meant for those *not* running seq...
Sounds like I need to add a 14 kpa row to my VE table.
Originally Posted by djp0623
(Post 668972)
I really don't understand what you are saying about the 50% TPSdot values. My TPS AE tableis at the top of the page. I will readily admit that I do not understand how this works, or how to adjust it.
When I attempt adjusting EAE again I will put the TPSdot threshold high enough so that normal AE is never triggered. Then once I have EAE pretty good I will add back just a small amount of normal AE to take care of the quick stabs. |
Got it, thanks. I think I am trying to tune out some things with EAE that can be handled with AE
|
Ok, so after a lot of reading in the provided links.
EAE is for mainting afr on throttle lift, and AE is for eliminating the rich or lean spike when the throttle is applied? Yes? I'm not sure when to start adjusting adhere to walls. I guess when the values for suck from walls get too low or high. Anyone find a ratio/relationship between the two that seems to be standard despite the setup? This would make it easier to adjust both curves. I'm done for the night. I'll see what my bench tuning produces tomorrow. |
All times are GMT -4. The time now is 09:55 PM. |
© 2024 MH Sub I, LLC dba Internet Brands