Mitsubishi 6G72 page
Installing VEMS into a Mitsubishi Galant 2.5 V6 turbo, engine code 6G72 (Cyclone V6 engine firing order: ...)
Timing info on MembersPage/PeterJensenMitsubishi
Apparently supported by any nonancient fw (use >=1.2.31 for other reasons) with proper config
One-click 6G72 trigger config
- start from some 6 cyl engine config
- VemsTune (2017-03-22 or newer) primary trigger dialog, bottom, select 6G72
- for now please verify as below
- Please capture triggerlogs (falling/falling is a must, others are theoretically not needed; Never forget to burn+reboot after trigger enable/disable or polarity change; capture triggerlog and CHANGE FILENAMES TO BE DESCRIPTIVE) and publish
Old method, configure manually (or verify):
- start from VT primtrig dialog bottom, EVO configlet
- after applying EVO6-7-8 configlet change number of ignition outputs to 6 cyl, and trigger tooth reference table 0,1,2,3,4,5 (in entries 0,5,4,3,2,1)
- Adjust so that primtrig and sectrig polarity are both falling: sectrig falling to PRIMTRING FALLING time is how engine sync works
- EVO valid sectrig => primtrig angle 0.. 84.4 deg (or 78.8 or even 73)
- 84.4 deg actually means 56 crankdeg for 6 cyl = 84.4/180*120 (VT assumes 4 cyl, validate might also show some tooth_wheel warning, neglect that for 6G72)
- only 1 place where sectrig to primtrig measured <56 crankdeg: 25 crankdeg at TDC cyl4 => sync !
- [l006_202-5000_10_50_1000_1000.wav] generated by command./signalgen l006 202-5000 10 50 1000 1000 ) has same feature, good for testing (note that some soundcards might invert, so record triggerlog and invert polarity in audacity if needed).
Trigger pattern is a bit unusual. Trigger wheel is like this:
It's very similar to the Lancer EVO trigger, but 6 cyls instead of 4. MembersPage/Fero/MitsubishiEVO
Apparently Peter Jensen had experience with this engine back in 2010, but now I cannot open his old triggerlogs, or interpret those configs with currently accesible VemsTune versions:
http://www.vems.hu/files/MembersPage/PeterJensen/mitsu/
Some hints to prepare?
Scope and/or Bench-test with the following in mind:
- start from VT primtrig dialog bottom, EVO configlet, key (verify after setting) is
- primtrig normal coil with sectrig subaruspecial = EVO 6-7-8
- bit6 of: config_secondary_trigger & _BV(strig_subaruspecial)
sync condition
GOOD SOLUTION:
- sectrig falling to primtrig falling < 53 deg is true for ONLY ONE event (25 deg from 100 to 75)
- => so no firmware mod needed apparently, just configure "EVO" and than change to 6cyl and primtrig falling'''
NOT GOOD
- sectrig (falling) to primtrig (rising) time < 80/180 * period'''
- threshold is 80 deg for 4 cyl (where period is 180 crankdeg)
- threshold is 53 deg for 6cyl (where period is 120 crankdeg)
- Is that good or not for this 6 cyl trigger ?
- on this V6 : 45 degrees in 3 occasions (=> exact same condition not good for V6)
Historical note: Marcell first didn't notice that primtrig falling is the simplest and perfect soltion, he implemented a CYCLONEV6 fw mod (in effect when "EVO" with >4 cyl configured; 1.2.36 version, but not published): at primtrig (falling), NO sectrig (falling) pulse seen since last primtrig
- but looking at the pulsetrain again, it seems now that happens twice (NOT good)
Proceed path: get scopeshot(s) or angular positions numerically (or verify Peter's)
- can be captured with VEMS configured to simple coiltype trigger + strobe if you know what you're doing
- but better scope, or get a free USB soundcard from Marcell and fry that (NOT your notebook's mic input :-). Mix the primtrig and sectrig with DIFFERENT resistors (because most USB soundcards have only 1 channel input; only output is stereo).
- timing is everything here: precise info needed to make a good "signalgen" wav for audacity benchtesting (results are not valid if sectrig relative pos is different than real)