Page 1 of 7 123 ... LastLast
Results 1 to 10 of 65
  1. #1

    Hacking the Ultra LX ECU

    I bought my first ski - a 2013 ultra lx.
    For my living I reverse engineer and modify engine control units on cars.

    I read about the timing mod, and that seems like a really crude way to add timing.
    Since I have all the tools, the first step was to read out the ECU:


    After I got the dump, I looked around and found some information on a Kawasaki ZX10R motorcycle.
    The engine control unit structure was similar, the factors and offsets also quite similar.

    The control unit contains a bunch of ignition maps. Because the ECU runs alpha/n the axes are RPM and TPS.
    There are 3 sets of 2 maps that are different, and after that the last map is just repeated like 8 times over and over.

    Below are a few pictures of the stock ignition maps:



    After that there are a few "trim" maps. I have no clue when they are applied - maybe when engine is hot or intake air is hot.
    There are a few things to notice about them:
    1. They are calibrated in a very lazy way - ignition gets lower with load, not higher. But because the jetski has such a narrow operating point - everything between 5000 - 9000 below 50% throttle is just not calibrated but interpolated with small values.
    2. There is something particular done to kill engine performance. Notice the ignition at 7400 rpm. Then look at 7600 and 7800 rpm. The ignition timing in most of the maps is dropped by a whopping 5-6 degrees. Maybe they did that so that when you hold the engine WOT for a long time and it gets hotter, it does not knock on bad gas - I don't know. But this kills performance.

    For my first test I added 2-3 degrees of timing through the range, and made the "cutoff" rpm after which timing is dropped 8000. This means adding over 7 degrees at 7800 rpm and 9 degrees at 8000 rpm. It will probably not be possible to find the correct timing without an engine dyno - I only own a 4wd car dyno - I can't use that to measure the jetski. I don't think this engine will be knock limited on 98 RON (93 MON).

    There are also a few fuel maps, the interesting ones have TPS as axis, the other ones have MAP.
    With the fuel maps the problem for me is, that there is no easy way to put a lambda sensor into the exhaust without having it sloshed with water and immediately destroyed. I need to find some way to do this so that I can calibrate the mixture. Right now again as a first test I just added 2% of fuel above 7600 rpm. This means pushing the lambda from 0.9 to 0.88 for example - but I have no idea what it is running. It might like less fuel or more fuel to make more power.

    I uploaded my first version of the map and tested it out.
    My RPM on WOT on the sea with the same conditions (light chop) went from 7650 RPM to 7850 RPM consistently. My indicated to speed (sorry, I am in Europe) went from 84-85km/h to 87-88km/h. I am out on the sea, and there is some chop/waves, on still water the absolute numbers are probably better.

    The problem with the "ignition" mod that is done on here, which tricks the engine speed sensor, is that you are advancing the timing through the entire range. And 6 degrees at 3500 rpm is double as much as on 7000 rpm. I am not sure it needs that much timing at low rpm. Certainly I didn't do it and i picked up the top speed just fine.

    I need to make some more tests, maybe try advancing the timing on the top end by another couple degrees and see if it gets faster or not. I might attach a knock monitoring device also to the engine and make sure I don't damage anything.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	1559682740734.jpg 
Views:	327 
Size:	75.5 KB 
ID:	449308   Click image for larger version. 

Name:	map1.png 
Views:	310 
Size:	23.6 KB 
ID:	449309   Click image for larger version. 

Name:	map2.png 
Views:	316 
Size:	27.6 KB 
ID:	449310   Click image for larger version. 

Name:	map3.png 
Views:	323 
Size:	20.9 KB 
ID:	449311  
    Last edited by prj; 06-10-2019 at 01:02 PM.


  2. #2
    Any chance you could throw a logic analyzer on the serial line going to the gauge (i am 95% sure it uses a k-line transceiver) and log packets. I am currently trying to borrow an ultra lx ecm to do this with. I want to build a standalone board using an stm chip to drive the gauge assembly. I would love to know what the datastream looks like going up to the gauge.

  3. #3
    steve45's Avatar
    Join Date
    Feb 2008
    Location
    Texas
    Posts
    6,503
    +1
    602
    Interesting! I wish I knew more about this stuff, but I appreciate you posting it. Where did you get the cables & software?

  4. #4
    Quote Originally Posted by steve45 View Post
    Interesting! I wish I knew more about this stuff, but I appreciate you posting it. Where did you get the cables & software?
    Well, I do have tools because I do this stuff for a living...
    They are too expensive for the end user - the one I used to read and write the ecu costs 4500 EUR + yearly subscription of 800 EUR. Not affordable. There exist Chinese knockoffs that might work.

  5. #5
    My name is Sean and I am addicted to STXs smokeysevin's Avatar
    Join Date
    May 2008
    Location
    Houston
    Posts
    4,585
    +1
    339
    Quote Originally Posted by prj View Post
    Well, I do have tools because I do this stuff for a living...
    They are too expensive for the end user - the one I used to read and write the ecu costs 4500 EUR + yearly subscription of 800 EUR. Not affordable. There exist Chinese knockoffs that might work.

    I have a pisaini serial suite but I was never able to figure out the software side to view the maps, only to transfer. Pretty neat, I was also never able to hook up to the ultra ecu's, just the stx

    Sean

  6. #6
    Quote Originally Posted by smokeysevin View Post
    I have a pisaini serial suite but I was never able to figure out the software side to view the maps, only to transfer. Pretty neat, I was also never able to hook up to the ultra ecu's, just the stx

    Sean
    The software I used is WinOLS. It will detect the map locations, dimensions, and axes on a Melco ECU, but there will be no factors or what the maps do... That comes down to experience and you can disassemble the dump with IDA Pro.

  7. +1 by:


  8. #7
    My name is Sean and I am addicted to STXs smokeysevin's Avatar
    Join Date
    May 2008
    Location
    Houston
    Posts
    4,585
    +1
    339
    Cool, I actually used both of those a few years ago to do that. I neever got past there though.

    Sean

  9. #8
    I modified the timing curves a bit more, especially at 7800 and beyond.
    I also tried to reduce the fueling a little bit.

    Will test today and see if there is a noticeable difference.

    I might get access to a SX-R that has the required modifications to add a lambda sensor, so I can see what kind of fueling it's running.
    Tbh I already had a few "tunes" for the SX-R as we flashed some maps that customers supplied on their jetski-s.

    I looked into them, and they were pretty bad. A few it is clear the person who made it had no idea what they were doing (10% to the whole ignition map, why?).
    The only one that looked okay was where 2% fuel was pulled at WOT and 2 degrees of timing added. But that probably isn't even possible to feel because for me to increase the rpm by 200 I had to add 6-7 degrees after the "dropoff" line at 7.4k..

  10. #9
    Tested richer and leaner fuel.
    No noticeable difference. Going to leave it on the richer side at WOT, as I'm planning to mod the intake, and if that gives anything then it will push the mixture leaner again.

    So far after 3 iterations I gained a solid 200-250 rpm. It might be possible to add even more timing but without an engine Dyno I don't feel like taking chances... Good enough as a stage 1 tune.

  11. #10
    My name is Sean and I am addicted to STXs smokeysevin's Avatar
    Join Date
    May 2008
    Location
    Houston
    Posts
    4,585
    +1
    339
    Did you monitor afr or just do an incrimental bump of the fuel table?

    Sean

Page 1 of 7 123 ... LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. 2011-12 ultra 300 ECU work correctly on a 2013 ultra 300?
    By jetdave56 in forum Kawasaki PWC Performance (4-stroke)
    Replies: 13
    Last Post: 06-19-2014, 03:37 AM
  2. Ultra 300x ecu?
    By jeff007 in forum Kawasaki PWC Performance (4-stroke)
    Replies: 9
    Last Post: 07-05-2013, 07:16 PM
  3. Ultra 250x ecu
    By danieletrono in forum Kawasaki PWC Performance (4-stroke)
    Replies: 2
    Last Post: 08-28-2009, 01:48 AM
  4. Replies: 8
    Last Post: 12-04-2007, 07:55 AM
  5. hacking tivo
    By Lurker in forum Sea Doo Open Discussion
    Replies: 2
    Last Post: 03-13-2006, 12:17 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •