Megasquirt Rotary EAE tuning
Rotary EAE tuning
Hi all,
Does anyone have experience with EAE and the rotary? I have decided to make getting it working a part of my project goal. I was trying using the default settings and on any throttle change there was always a lean-misfire. I first changed zero'd out temperature and rpm corrections. And then I changed the shape of the Added to walls co-efficients, without changing the end points, which is what the mega manual says to do. This didn't really seem to help, so I started increasing the values across the board. It seems to be slightly better, but there is still a lean spot on throttle changes. I wasn't sure if I needed to get the mapdot predictor going because I have twin 50mm downdraft throttles. My tpsdot AE (old style) was quite strong and sensitive.
Also a question specifically for Ken, how fast is EAE at responding to engine changes, as I understand it, it must read the map pressure change before it can compensate, whereas tpsdot is injecting the extra fuel the cycle after the tpsdot is triggered.
I also read somewhere about using a larger number of injections per cycle, but I'm not sure if thats a good idea for my car with 550cc primary injectors.
Overall I like EAE because it seems to catch the return to idle better than without because it seems to inject extra fuel as soon as it sees the rising map signal, basically the engine is returning to idle slowly near idle speed rather than dropping lower and then coming back up slowly.
Does anyone have experience with EAE and the rotary? I have decided to make getting it working a part of my project goal. I was trying using the default settings and on any throttle change there was always a lean-misfire. I first changed zero'd out temperature and rpm corrections. And then I changed the shape of the Added to walls co-efficients, without changing the end points, which is what the mega manual says to do. This didn't really seem to help, so I started increasing the values across the board. It seems to be slightly better, but there is still a lean spot on throttle changes. I wasn't sure if I needed to get the mapdot predictor going because I have twin 50mm downdraft throttles. My tpsdot AE (old style) was quite strong and sensitive.
Also a question specifically for Ken, how fast is EAE at responding to engine changes, as I understand it, it must read the map pressure change before it can compensate, whereas tpsdot is injecting the extra fuel the cycle after the tpsdot is triggered.
I also read somewhere about using a larger number of injections per cycle, but I'm not sure if thats a good idea for my car with 550cc primary injectors.
Overall I like EAE because it seems to catch the return to idle better than without because it seems to inject extra fuel as soon as it sees the rising map signal, basically the engine is returning to idle slowly near idle speed rather than dropping lower and then coming back up slowly.
MegaSquirt Mod
Joined: Sep 2004
Posts: 4,721
Likes: 1
From: Maryland
I've not started tuning EAE on the stock 13B. For the primaries I'm sure very little wall-wetting occurs just looking at my accel enrichments and how little are required. I think accel enrichment with the primaries just accounts for the fact that MAP response, and therefore VE table lookup lags a bit behind actual conditions.
That said, I know some does occur, so I'm working on ways to make EAE more compatible with staged injection and to make it respond quicker.
As that code stands now, it's not really suitable for use with staged injection. I'm working on it, but getting it right will likely require tuning EAE separately for the primaries and secondaries. Right now when you stage, the secondaries will blast a huge amount of fuel to both wet the walls and get enough fuel to the engine, but because the primaries and secondaries are tuned with the same curves, when the primaries start to drop their pulse-widths, EAE will drop the pulse-widths faster than it should since tuned for the secondaries, a lot of fuel would be coming off the walls of the intake. What you want during staging is for the secondaries to blast that huge amount of fuel initially, and the primaries scaling down to lag a bit behind the secondaries coming on... but you don't want the primaries to drop artificially fast due to EAE compensation. So that leaves you with needing 2 separate sets of curves.
In addition, because of several factors, including the fact that it's based on MAP, EAE as it is now will always respond slower to quick throttle movements than normal AE does. Increasing the number of squirts can help the required fuel get to the engine a bit quicker, and the MS2 should be able to handle the lower pulse-width as long as it isn't running up against the opening time of the injector.
After 2.1.0 is done, and I've had some time to work on ms3's ignition, my next work is going to be focused on making EAE more responsive... One idea involves switching to alpha-n on quick stabs when off boost (for turbo engines). The idea there being that if the VE is correct it can be used with PV=nRT (ideal gas law) to figure out a predicted MAP to feed into EAE, and in addition, the alpha-n looked-up VE can be used in place of the MAP one during the transiant. I still need to do some testing with this to see if it's going to work, but according to my reading, it should, and it's based on some of the more complicated model-based ideas I've been reading.
I know mike_robert had some good luck tuning EAE on his 12A with ITBs back when I first wrote the algorithm, and it worked pretty well once tuned with a 20v 4age engine (5 valves per cylinder, and ITBs stock), so it should work for the 13B as well, but I still need to solve the responsiveness and staging problems.
Ken
That said, I know some does occur, so I'm working on ways to make EAE more compatible with staged injection and to make it respond quicker.
As that code stands now, it's not really suitable for use with staged injection. I'm working on it, but getting it right will likely require tuning EAE separately for the primaries and secondaries. Right now when you stage, the secondaries will blast a huge amount of fuel to both wet the walls and get enough fuel to the engine, but because the primaries and secondaries are tuned with the same curves, when the primaries start to drop their pulse-widths, EAE will drop the pulse-widths faster than it should since tuned for the secondaries, a lot of fuel would be coming off the walls of the intake. What you want during staging is for the secondaries to blast that huge amount of fuel initially, and the primaries scaling down to lag a bit behind the secondaries coming on... but you don't want the primaries to drop artificially fast due to EAE compensation. So that leaves you with needing 2 separate sets of curves.
In addition, because of several factors, including the fact that it's based on MAP, EAE as it is now will always respond slower to quick throttle movements than normal AE does. Increasing the number of squirts can help the required fuel get to the engine a bit quicker, and the MS2 should be able to handle the lower pulse-width as long as it isn't running up against the opening time of the injector.
After 2.1.0 is done, and I've had some time to work on ms3's ignition, my next work is going to be focused on making EAE more responsive... One idea involves switching to alpha-n on quick stabs when off boost (for turbo engines). The idea there being that if the VE is correct it can be used with PV=nRT (ideal gas law) to figure out a predicted MAP to feed into EAE, and in addition, the alpha-n looked-up VE can be used in place of the MAP one during the transiant. I still need to do some testing with this to see if it's going to work, but according to my reading, it should, and it's based on some of the more complicated model-based ideas I've been reading.
I know mike_robert had some good luck tuning EAE on his 12A with ITBs back when I first wrote the algorithm, and it worked pretty well once tuned with a 20v 4age engine (5 valves per cylinder, and ITBs stock), so it should work for the 13B as well, but I still need to solve the responsiveness and staging problems.
Ken
Last edited by muythaibxr; Jan 1, 2009 at 03:02 PM.
Hi Ken, thanks for your reply.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
Senior Member
Joined: May 2002
Posts: 386
Likes: 0
From: FL
I'm on it - I just have to find the msq on a laptop that is now defunct. I had EAE working nicely with minor changes to the default settings. This was with the 1/08 beta, I think . I have pretty much the same TB setup as you except the mainfold is the wraparound so has the 2 90 degree bends from TB to port - lots of wall to wet. The default settings didn't have quite enough adheres-to or sucked-from value but I recall that I changed it maybe 10-15%. I had simple MAPdot AE running very well for a long time but that did take a lot of time and effort to get to that point. EAE was running about as perfectly after 20 minutes. I need to go and find a 2.5" hard drive adapter so I can pull that info of the dead laptop's drive and will get the .msq to you shortly thereafter. First startup of 6 month long turbo addition to my 12A is coming very soon so I need those values for myself as well though EAE may not help me as the current project has staged injectors. We'll see...
-Mike
-Mike
Hi Ken, thanks for your reply.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
I couldn't agree more. My brother has the same manifold and throttle body , and this is a steady problem, especially considering that it's turbocharged and the manifold was "polished" using 60 (?) grit sanding rollers, so you can imagine how much fuel sticks to the walls. It makes tuning horrible with the ms1......absolutely horrible, lol. Needless to say, I got soo discouraged with everything, that the car just sits (since september or so).
I tried getting the ms2 to work reliably with it, but I kept having problems with corruption, etc., so I gave up after a month of messing with it. I'm just waiting for a writeup for it (the ms2) at this point.
Trending Topics
MegaSquirt Mod
Joined: Sep 2004
Posts: 4,721
Likes: 1
From: Maryland
Hi Ken, thanks for your reply.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
#1 I won't be having the problem of primary and secondary injectors being in different locations. As I said I'm using a Weber manifold and a twin 50mm downdraft throttle body such as this.
I think wall wetting will really help my tune because with normal AE I need a rather strong and sensitive throttle pump, and on a throttle change its not uncommon to see the AFR go rich and take a bit to come back as the fuel comes back off the walls.
I probably should hit mike_roberts up for his values to see how they work for me.
As I said before, with staged injection, you pretty much need different parameters for primary vs secondary, and right now, the code is not set up for that.
The best thing to do to tune EAE is to start with slow throttle changes and get those working first. Quick throttle changes are the most difficult part to get right with EAE. You may end up needing to use the lag compensation feature.
Ken
Back from the dead, but has EAE been worked out with staged injection? I'd like to try it out, as well as the table feature for staging since I still get stumbles in certain places.
For EAE, does the % (kPa) refer to fuel load %, so 15 psi would be 200%? Could you post up some of the EAE curves that you've got running on your 7 as a starting point? Thanks!
Thread
Thread Starter
Forum
Replies
Last Post
Daua118
SE RX-7 Forum
25
Dec 5, 2021 04:15 PM
