Ad Widget

Collapse

Announcement

Collapse
No announcement yet.
X

The definitive Katana EFI swap thread

Collapse
  • Filter
  • Time
  • Show
Clear All
new posts

  • TheSteve
    replied
    Alright, here goes:

    Megatune (current as of todays date):
    http://www.msextra.com/manuals/Downl..._Installer.exe
    Includes MSnS-E firmware

    MegaLogViewer:
    http://www.efianalytics.com/MegaLogV...tup_v2.951.msi
    Youll need Java to run it, if you dont have it installed already. http://www.java.com/en/download/manual.jsp

    Added the files below.

    Heres the sample datalog from one of my rides to work: Sample Katana Datalog.zip I apologize for the somewhat annoying process to download this one. I'll find a better way later.
    Attached Files
    Last edited by TheSteve; 09-19-2008, 02:03 AM. Reason: Adding files

    Leave a comment:


  • robw
    replied
    That is some awesome home engineering, Mad Props

    Leave a comment:


  • TheSteve
    replied
    I'll post up the config files and everything you need later today. The current tune is very good for cruising but theres a flat spot at full throttle somewhere around 5-6k rpm that will need to be fixed. For a 750, the tune should be pretty close unless you're running better cams. Start with my VE tables, but change your ReqFuel for the larger engine. Try 125% of my current value. That should get you running pretty well.

    Later I'll post up:
    Current version of Megatune with MSnS-E support
    Latest MSnS-E firmware to burn onto the MS
    MegaLogViewer with a few of my own logs for reference
    My katana general config file
    And a pair of tune files for 600s and 750s. One will be pretty basic, the other will be with overrun settings, etc. Same one I'm running now.

    Leave a comment:


  • jwilliams
    replied
    For TheSteve

    Hello,

    Well due to unforeseen $1500 air-conditioner/furnace repair I have had to put my MS EFI Kat plans on hold until during the winter some time.

    I was wondering for those who are going to try this if you could provide your current config files or such that could be loaded into MS so we get the results of your work? I assume there is some global config file and some map files that can be shared and loaded into any MS to get the same config?

    I'll probably need to tweak it since I beleive you have a 600 and mine is a 750.

    THanks, Jeff

    Leave a comment:


  • Theeun_Holy
    replied
    well done!

    Awesome, I just might have to do this sometime, I like the smoothness of EFI.

    Originally posted by TheSteve View Post
    I got 46MPG last tank, even with lots of 85MPH freeway riding and a good bit of redline pulls and power tuning. Maybe once I'm done I'll reach my original goal of 55-60MPG after all...

    Leave a comment:


  • TheSteve
    replied
    Thanks! Next mod is probably going to be an ignition advancer and possibly a regear. I really want to get a solid MPG figure first so I can see for sure if the ignition advancer will help or hurt mileage.

    Leave a comment:


  • ATOMonkey
    replied
    Good job!!

    Leave a comment:


  • TheSteve
    replied
    WOW. I just changed it to hybrid a-n mode, estimated a new tuning table (new table has a more triangular shape than the square SD table) and took it for a test run. It runs amazingly! I was expecting to make it around the block and no farther with the guesstimate tune but I actually was able to get about 10 miles without issue before I ended up back at the house. Got some high speed freeway pulls as well. It fixed the problem completely. Now twisting the throttle past the previous choke point just makes it pull harder and harder! Theres one flat spot that I noticed, so I need to reshape the table some.

    If you look at the screenshots of the tuning part of the guide youll see the VE table. Now instead of the Y axis being KPA it's now TP. The reason it becomes essentially triangular is that at low RPM, 20% throttle will still fill the entire 600cc engine with air. So the left side of the table only has a few cells before airflow maxes out and all cells higher up the Y axis are identical. I guessed it pretty damn well it seems. Unfortunately it messed up my AFR target table by changing its Y axis to TP as well. So the resulting datalog shows it chasing phantom AFR targets. I just corrected the table and I'll be trying it out shortly.

    I noticed with Alpha-N mode that everything on the bike just seems to work better. Smoother idle, very linear power delivery at high RPM, etc. Off idle throttle response is insane. Before if the bike got into one of its moods you'd have to slip the clutch very carefully or it stalled. Now it revs smoothly and powerfully right off the line.

    I got 46MPG last tank, even with lots of 85MPH freeway riding and a good bit of redline pulls and power tuning. Maybe once I'm done I'll reach my original goal of 55-60MPG after all...

    Leave a comment:


  • TheSteve
    replied
    From the naked eye it would look the same. Speed density (the one with the MAP) and hybrid alpha-n both need a map sensor to run. To see which one it really is, unplug your TPS. If you can still ride the bike its probably SD. If not, or if theres no high end, its HAN.

    The more I read about the hybrid setup the more excited I'm getting over this! It looks to be the answer to everything I'm having trouble with (high load and occasional idle stallout). Very excited, I'll start playing with it tomorrow even if it means no nap between class and work. I'll just sleep at work.

    Leave a comment:


  • Newbie2it
    replied
    My k4 GSX-R runs Map as far as I am aware of.

    Leave a comment:


  • TheSteve
    replied
    Eh, after a nights worth of research I'm not gonna bother with the restrictors. The MAP signal isn't really bouncing around, it just isn't there in the first place. This is common with ITBs. I'll be running the Extra code's Hybrid Alpha-N mode. It basically uses the Alpha-N tuning table and divides the final value by KPA/100. So instead of being throttle dependent without regard to load, this uses a bit of everything.*

    So you might be asking yourself why I didn't notice this before. If you recall, the last time I actually went for a full power run was with the HEI in 8 cylinder mode. It still coughed a bit, but made decent power. Much more than now at least. I have a feeling that it was still messed up the same way, but in trying to rough tune the map (as opposed to now where it's just fine tuning) the wideband had almost FULL control over the final pulsewidth. So it choked out, recalculated, and instantly richened it up by a large margin. Now its adjusting in much finer increments, for a more stable ride.

    I might try some alpha-n tuning tomorrow. I'm working 6 days a week now and going to school Mon thru Fri so free time is at a premium. And most of the time I'm too exhausted to do anything but sleep.

    Is anyone interested in seeing the datalogs from these runs? I can post them if anyone wants to actually SEE what I'm rambling about.

    I am curious how any of the modern fuel injected bikes handle this. I know some have two sets of throttle plates (GSXR) while others have computerized drive-by-wire throttles (R6 I think). I think these are only to overcome accel enrichment issues though. I know that the GSXR has a MAP sensor, so it is for sure using one of the same modes MS is. Not alpha-n though, as it wouldn't need the MAP sensor in the first place. Probably a hybrid of the two. Do any of the EFI bikes run MAF type sensors to measure air flow instead of pressure?

    *The MS2 Extra code (only available with the MS2 cpu) has a mode that is FULL Speed Density based at low load and either full or hybrid Alpha-N based at high load. This would be ideal. You may want to take that into consideration when ordering your MS kit.

    Leave a comment:


  • ericej2344
    replied
    From what i've been reading the restrictors would be a good place to start. You might be getting the right vacuum, but not the right reading because of the bouncing effect.

    Leave a comment:


  • ATOMonkey
    replied
    Vacuum line restrictors are there to keep the carb tune from showing 0-1-0-1-0 type of vacuum signal.

    Couldn't hurt.

    Leave a comment:


  • TheSteve
    replied
    So I've been riding the bike for a couple weeks now tuning mainly for economy. Yesterday I started working on the high load area of the table, also known as the fun part. I've had the Kat up to about 95 at this point, but I haven't gunned it lately. With the trusty laptop in tow I hopped on the freeway and started tuning for hard acceleration. I slammed the throttle open and the bike promptly started choking and smoking up a storm. I let off and slowly reapplied the throttle and got the same effect. I did this a few more times just to ensure I got numerous datalogs of it and turned back.

    Looking at the logs, I noticed something that I've been crossing my fingers and hoping to not have to deal with: At higher speeds (60+) while cruising, my MAP reading is 92kpa (atmosphere is 100 at sea level, usually 94 at my elevation). This shouldn't be, as I'd suspect I'm at 30% throttle while cruising at 60mph. In other words, there should be plenty of vacuum registering. When I peg the throttle, as expected, I rise to 94kpa. This is a strange concept to me at least, but I'll try to explain it. The engine is a 600, so it should be trying to pull 600ccs of air in every cycle. Running at 50% throttle it'd get 300ccs of air, but in trying to pull more a vacuum would be created. This is the vacuum MS uses to calculate load. At 100% throttle there should be no vacuum and the engine should get its desired 600ccs of air each cycle. Right now between half and full throttle (300 vs 600ccs, in this example) there is only a 2kpa difference. It should be at least 30 to 40kpa. So even though the throttles are holding back air from entering the engine, no real vacuum is being created. Seems that even though air FLOW is being regulated, PRESSURE is not. Very strange to me, especially at these rpms (5000-8000). Above that it seems to behave normally, I think.

    So what might be causing this? I have a couple ideas. If you look back at the pics, the vacuum nipples are very close to the throttle plates. Maybe they're too close to where the throttle is cracked open, and are in a natural high pressure zone. Perhaps if they were on the sidewall where the linkage comes through it would be in a less turblent area and get true pressure. Maybe at cruising speed/load there is so little vacuum signal that its stealing air from an adjacent cylinder and screwing up the pressure signal. Id say the throttles flow too well, but even still if there was no vacuum the engine would think you're at full throttle which is untrue. If I open up the throttle theres plenty more power on tap until I hit a certain point when flow increases but pressure does not.

    Watching the log, the pressure does behave as it should at lower speeds and while cruising at 45-50. It's not till slightly higher loads where it appears to be at WOT all the time. I've heard this can happen with ITBs because of their flow rate and small vacuum chamber. I could try running the MAP sensor to a single cylinder and seeing if it improves things. Idle will suffer though.

    MS also has another tuning algorithum: Alpha-N. Alpha-N is different in that instead of using the MAP sensor, it relies solely on RPM vs throttle position. It's designed for engines with giant lopey cams where theres near zero vacuum at idle because of cam overlap. Also works wonders on ITBs so I hear. Drivability can suffer as its harder to tune for varying loads (towing for cars, riding 2 up for bikes) and effective WOT (what throttle position shows atmospheric pressure for a given RPM) is hard to find. There is a hybrid mode, but I'm not familiar with it yet.

    Lastly, I could try running restrictors in the vacuum lines going to the MAP sensor. It would be the same way the carbtune does it. Without the restrictors I'd figure the slides in the carbtune would bounce all the way up then all the way down without dampening. With the restrictors it averages out the high vs low pressures and stays pretty calm. I could try this, but at high RPM dampening shouldn't be needed. Might be worth a shot, otherwise it's looking like Hybrid Alpha-N might be the way to go if I ever find any documentation on it.

    Leave a comment:


  • Contrldkaos
    replied
    Originally posted by TheSteve View Post
    The sensor is run straight to the MS, yes. To monitor the actual oil temperature I'd actually go with the oil pan mounted sensor as thats immersed in the oil. I didn't need the accuracy to do startup condition monitoring so I didn't bother. You can't T one sensor off to two gauges though. The split signal will alter the voltage causing neither unit to work right.
    so an oilpan insert will work ok then. sounds good thanks a ton.

    Leave a comment:

Working...
X