Got the MS running with the 7 tonight. Several times around the block trying to get the bugs out. Not easy setting up the ve map, just working in the lower left corner for now.

The ve table tuning with the 3D map was helpful to get idle to 3k working reasonable. Guess since I'm running by myself I'll be stuck with datalogging and VexMe. The keys could be better utilized. I think MegaTune should use the function keys more, like for start/stop datalog, and in the realtime tuning screen--they're much easier to find while your driving.

So I'm guessing these numbers are going to go over 100% ve at the high loads--right? Idle is in the 46-54% range.

Also, does anyone know how the staged injection is going to affect the ve map?

TIA,

Scott

 

staged shouldn't affect the VE map at all except maybe right where it kicks in... you might have to make that spot richer to tune out the slight lean spot before the fuel from the secondaries reaches the engine.... or you can just move that spot up to where the air velocity keeps it from being a factor. I'll eventually get around to writing a feat

If you have a wideband controller (or even if you don't), it might be worthwhile to use autotune... it'll make the low-load tuning a lot easier (if you have a narrowband) or you can tune the whole map that way (using a wideband and the AFR table).

On tofuball's NA, we go over 100% VE at about 5500-6000 rpms at 13.2:1, and it doesn't come back down until around 8500. At an idle of around 800 rpms, tofuball is right around where you are for VE numbers.

 

Yea I had the staging off delta at 100 and the 2ndarys were cycling, causing it to run lean. I moved it to 300 rpm--that should be better.

I've been looking for autotune, how do I run it? Is it part of MSTweak3000?

 

Nevermind, I found your thread.

 

Scott,

Congratulations! You're on your way now. If it's drivable, go driving and datalog the entire trip. Then run the datalog through VexME, MSTweak, or whatever. There are a number of datalog processors floating around now. The VE table itself should fall into place very quickly. Working out the warmup, after-start, accel, etc can take a bit longer.

I would recommend TPS-based staging, rather than rpm. It is possible to get a stumble with the rpm method, as in gentle acceleration, you may reach the staging point prior to having adequate airflow in the secondary runners due to the progressive TB action. In my install it was noticable, but not severe by any means. Using tps-based staging avoids this problem.

Not sure where you are looking for auto-tune info, but the details are here:

http://www.msefi.com/viewtopic.php?t=11314

Just stay in the low-power areas of the map if a NBO2 is used. The auto-tune active range can be configured to do just that.

Happy tuning!
Roger.

 

Thanks for the advice Roger.

Funny about the secondary transition stumble, that's part of the reason I want to get away from the stock ECU--the dreaded 3800 rpm hesitation. I hope the TPS based staging works. I'm wondering how the tps-based staging works when you have low rpm and stomp on the throttle?

As far as tuning, I'm using the LC-1 WBO2, so far it has worked well. I've always been concerned about the sensor failing due to heat, but that's why it's mounted at the far end of the DP. FWIW, I've recorded EGTs up to 1500 F with the stock ECU, hopefully I can bring that down a couple hundred degrees using MS.

Have you guys noticed the rotary likes to idle very rich? Mine seems to be happy around 11.5.

Sure I'll have more Qs soon. Thanks Again,

Scott

 

tofuball's car is happy idling around 12:1.

I use rpm based staging, but I raised the staging point to around 4400, and that seems to have gotten rid of most of the stumble problem... (I can see a lean spot in the logs, but I don't feel anything).

I started to say in my previous post in this thread that I'm going to try to write a feature that lets you delay the amount of time before the primaries cut their pulse-width to their staged value so that you can tune out the stumble by tuning the amount of time it takes for the fuel from the secondaries to reach the engine... You might end up with a lean spot in some situations (I could make it MAP dependent too though, and make a small table if there's room in the flash). Either way I think this is a good solution to that problem, but I won't have time to work on it until after I finish the rx8 wheel decoder.

Ken

 

if you could ( i know... im trying not to be that guy but...) i'd be willing to give you some $$ for your map! i Just need it to start up and give it to my tuner! PA PA PA Please!
aaron
Rx7pro@aol.com is my email and im aware its not finished but id still like to have a map that will make my cae half way run.

 

you really don't need to get a map from someone else, megatune will generate a map that can get you started. If you're going to have another guy tuning it, and you'll be tuning with a wideband, then generate a map and just go with that. Ignition map I can understand, but fuel is pretty easy to tune.

 

Yea I don't mind. But this map is waaaay far from done--getting me around town ok but stay in the vacuum and below 5000 rpm--even then your application will need tweaking for the vacuum side. All this does is save you half a day getting it roughed in.

I'll shoot it over to your email.

Scott

 

I definitely think it would help to see other car's maps. Not necessarily to drive with but to look at trends--at least once you get a portion of the map done you'll know what directions things should be heading. I know I would have liked to have something to go on starting out. The map that the online autogen tool created really sucked--it had like 19's down low by idle--now it's doing really good at about 46-50. But you're right it's not needed to have someone's map.

But I also have concerns about posting maps and then someone blows their engine. In the end it's just better to share the information with the usual disclaimer.

 

meh... I must've forgotten to say on here that when you generate your first map, it's good to specify an engine size of 228 cubic inches... that'll get you started pretty easily.

 

ok ill try both ways! sounds good (and ive blown my share of motors... im not afraid of another ahaha sad realy...
aaron
5+

 

Scott,
Congrats on getting it running. Im still a ways off from that point,mostly due to me learing the PC/programming end of things. Muy has also been a great help. I have a few question and hope you dont mind me asking here. First Scott are you using the stock BAC idle control valve with a V3 board???? If you are how you hooked it up would be great (if the Ztx688 transistor worked and how so). Second I agree seeing another persons maps would be helpful. And last feel free to laugh at my expense I downloaded megatune2.25 beta and MSNS extra but am a bit lost. Do I need to export a file and burn that to the ECU to get the same screens Muy has? As it is I built the board and stimulator V3 board HC908 chip and verified it works and solved the USB adaptor problem. Now Im frustrated getting the required info into the MS itself. Call me stoopid but I'm more of a gear head and this is new to me. Thanks in advance,Shawn

 

I do plan on using the MS idle control but right now I'm still using the stock ECU for idle control, emissions, and spark. If you hooked up the ztx688 correctly I don't have any doubt that it will work right. Thus far I've had no problems with the MS hardware side. Once I get fuel working good, I'll try the BAC idle control, and then maybe spark.

I think exporting to the ECU and the screens MTB has are two different things. I think he is using a very new version of code--maybe even an alpha version. I'm using MT 2.25b514 with the 024S12 firmware. I'm not sure if my version even has a working Rotary ignition setup. I also have a few screens that are different than MTB. In general you make changes in MT and the changes need to be saved in the MT file AND downloaded to the firmware--which means you should be connected to MS for the download.

Gotta go to work now, HTH,

Scott

 

ok, first you'll both want to use 025k or higher. Trailing doesn't work correctly before that. Second, you'll want megatune 2.25b614 or higher... autotune doesn't work with any version before that.

For the BAC valve, I need to update the FAQ with a proper v3 circuit.

The screens I showed in the FAQ were for 025v2 (the version I added my idle advance feature to).

Last, in order to get the proper things to show up in megatune, you need to take the msns-extra.ini file from the firmware you downloaded, and put it in the installation directory for megatune in the mtCfg/car1 directory. (I think there are other places you can put it but this is where I always put it).

 

Thanks again Muy I'm working on figuring it all out. Seems to me the easy part is building the board and wiring the car. The software is where it gets really confusing. Think its time to inlist the help of a more experienced friend before I loose it entirely. Actually lookin at this MS box and it would make a really nice wheel chock.

 

Well I'm getting the fuel map tuned in and overall I like the results--the response is much smoother and pulls hard enough to feel the improvement. I've got decent fuel up to about 12 psi and 6000 rpm. However, there are a few bumps I've run into and thought I'd post it up for discussion.




First the primary/secondary transition is still apparent in the data but thus far I have not really felt it--other than what appears to be a little leaning out prior to the secondaries kicking in. To some degree the symptoms can be eased by adjusting the fuel map. Notice that vexme has substantially increased the ve values in the 4500 rpm bin--a result of leaning after transition. A software solution that fades the secondaries in would be optimum and shouldn't be hard to implement. If one of you guys would like to work on it I'd be glad to define a transition program that would do the job.

Second, I think I'm running into the end of what the tuning tools will support. Vexme is giving a large variation in ve up at higher boost levels and I have no way of giving it multiple log files. It would be nice if it could use a whole bunch of files like mstweak does. I need to look into mstweak more but it appears to only target one afr across the whole range--I need it to target the afr map instead. For tuning under boost, the autotune feature is just to experimental at this time.

I think I'm gonna just slow down on MS tuning and make sure I know what I'm doing--just read in the 2nd gen forum about some guy blowing his engine for like the fifth time (knocks on wood).

Also does anyone know of how to turn off the EGO switching above a certain MAP value. It appears the MS-II has the option but not MS-I--again I'll run into the problem of MS adjusting the fuel in the boost region. Maybe I'm just over reacting but I was hoping to use ve like an open loop map and then let the MS optimize on 14.7 afr during cruise only (in the vacuum region)--any ideas?

Scott

 

there is a way to turn off closed loop above a certain tps or certain map value, I'll have to take a look around...

I just got back from a tuning run with tofuball; we were using the autotune feature in megatune, and all I have to say is that it works great! I don't think I'll touch vexme or mstweak again. It was pretty cool to tell tofuball to hold on one vertex, and then sit there and watch megatune adjust the map until it's perfect.

One of these days I'll work in some code for transitioning... I was going to implement a user-configurable delay before cutting the pulse width on the primaries... and possibly make it into a 1x10 map or something (if I can find enough flash to do it with). The Y axis would be MAP and at each point you specify a delay...

That's the idea anyway.

Ken

 

Scott,

Glad to hear you are making progress! Are you running fuel+spark control, or fuel-only at this point?

I'd suggest trying the tps-based staging, as it'll improve that potential lean issue during the transition. With rpm-based staging, you'll pass through that staging point many times during a typical logging run. If only a couple times are at reduced throttle, the resulting momentary leaning will throw off VexME or whatever tuning tool you are using.

Here's a question: What do the other ecu mfg's do with respect to staging. I chatted with a guy recently who said his Haltech was set up to run map-based staging, and ran primaries only until 100 kPa or so. Not sure on injector sizes.

I coded a progressive staging mode as well, which looked really neat on the scope. It slowly reduced primary pulse width while increasing secondary pulse width over 256 ignition events. This was way before the MSnSExtra days, so it'd need to be merged with more recent code, but I'll send it your way if you like.

As for the EGO switching, I'd suggest you limit the control authority to +-5% at this point. That'll keep the ecu from doing something nasty if it gets a faulty EGO signal. There is EGO feedback cancellation built-in to the TPS, but that may be over-ridden by the AFR target table if in use. Check the MSnSExtra docs, or if necessary dig through the .asm code to see for sure.

Is your target AFR table in this thread somewhere?

Roger.

 

Hi Roger,

Thanks for looking this stuff over--I know I'll get their sometime. Man this tuning is alot of work and it's getting old real fast. I know I just need to slow down and take a little break--yea right.

Right now just fuel.

I think I will give TPS staging a try--at first I was skeptical about it, but after thinking about it more, MS will still inject the same fuel, just through 4 injectors instead of 2. Any suggestions for S4 type TPS settings?

I don't know what other mfg's use--but I use this type of fader logic to smooth out flight control guidance commands and it works pretty good. Looks like your idea is the same thing and should work fine.

Nope--but here it is.

Obviously I don't have the cruise afr's right. If I can get the AFR targets set in MSNSE then the ve table won't matter too much--right? Then I hope to disable ego above a certain kpa. That way I'll have a nice open loop table for acceleration and then let the AFR targeting do it's thing at cruise. That's the plan but I don't know if the software will let me do it.

Pretty sure I'm gonna nix the 220 kpa and 8500 rpm bins, both are way more than I want to go. Guess I was getting greedy. Don't get me wrong, it's running pretty good alright--pulls the best I've ever seen for this car. Last tank only got me about 200 miles so I need to kick in the ego pretty soon.

TIA,

Scott

 

I'd swear it was for MS-II but then again I was probably lost in the swamp of information.

Maybe I'll give it try and see how it works, but the only documentation I find on it is in one locked thread over on msefi--is there any better information on how to set it up?

Look at renns idea--sounds very simple but should work just fine.

Scott

 

Rex, Thanks for posting your VE map, Im a couple steps behind you. I finally reconfigured the files the proper way and have the MS ready to be tested in car...finally. Rex are you using the S4 narrow range TPS? I have adapted a GM TPS to my throttle body and will be using a full range TPS. Not a easy thing to do but can take pics and do a write up if anyones interested and fairly good at fabricating. Im also wondering Rex if you had to leave the mass air flow installed for your stock ECU to work properly for the spark??? Please let me know on the MAF as I said Im just a step or two behind you. I did not use the MS as a wheel chock was tempted but didnt. Thanks in advance.

 

I don't have a spare TB with stock S4 TPS, so I can't check the mechanical action of the TB against the TPS movement. Given the narrow range of that TPS, it may no be usable for proper injector staging if it reaches full-span prior to the secondary throttle plate opening. For test purposes you could try putting in a TPS value a couple ticks below full-span.

The VE table is critical. Get it right, and then use the AFR target for trimming the last couple percent only. That way in the case of sensor failure, you've still got a nice running car.

Go to General -> Lamda AFR Settings, and set the 'Change EGO limit above' to 100 kPa, 'Change EGO limit +-' to 1. That'll limit EGO correction to 1% when boosting, so it's effectively disabled at that point.

In cruise you should be able to run 15.5:1 or so. I ran a bit leaner, but there was a bit of surging noticed at cruise.

Have fun,
Roger.

 

renns, I got rid of the lean surges on cruise at 15+:1 afr by switching to using mSec instead of ignition events for the ego steps... I'm using 200 mSec right now, and I notice no surges.

Also, for TPS-based, there's a chance that at higher revs, that might cause the duty cycle when you're using primaries only to be too high.... have you looked into this? say at lower kPa values above 5000 rpms? This might not be a concern in normal driving, or in drag racing, but if you're racing on a road course, and you're at 6000 rpms but at 50 kPa sustaining your speed through a turn, this might be too much for the primaries. That's why I've been reluctant to try tps based, and have wanted to do my mappable delay instead. This also might only be a concern for NA cars... as judging from the map above, turbo guys don't expect to be off boost that often.

Last, there is a setting that specifically turns off Closed loop above a certain TPS or MAP point (from the FAQ):