_______ 
(_____  )
     /'/'
   /'/'  
 /'/'___ 
(_______)
         
         

 _   _ 
( ) ( )
| |_| |
|  _  |
| | | |
(_) (_)
       
       

 _    _ 
( )  ( )
`\`\/'/'
  >  <  
 /'/\`\ 
(_)  (_)
        
        

 __  __
 \ \/ /
  \  / 
  /  \ 
 /_/\_\
       

 _       
| \    /\
|  \  / /
|  (_/ / 
|   _ (  
|  ( \ \ 
|  /  \ \
|_/    \/
         

IMPORTANT: enter the case-INsensitive alphabetic (no numbers) code AND WRITE SOME SHORT summary of changes (below) if you are saving changes. (not required for previewing changes). Wiki-spamming is not tolerated, will be removed, so it does NOT even show up in history. Spammers go away now.


Visit Preferences to set your user name
Summary of change:
*Wiring complete,sensors working,i can see the gauges in Megatune,but there is no RPM signal yet (can wrong calibration (60-2 instead of 5 window) cause this?) * primary_trigger=FF (or FE if falling edge is used). The primary_trigger=01 is for multitooth (eg 60-2 missing-tooth) setup * secondary_trigger=02 must be used. But hey, this is appears on almost every pages of yours ! * TPS reads 10% after calibration,is this normal? ** of course NOT! increase tps_low ---- '''wishes for next v3.x ECM''' *tachometer output (weryimportant for 5 coil config) * primary trigger: VR crank 60-2 (changeable to HALL with dip switch if possible, so only the endplate needs to be removed) secondary 1 window hall camsync * High voltage flyback (2+6diode) * All 8 ignition/injector channels * EGT and Knock * 4bar MAP sensor onboard * Programmable Shift light output (a p259 channel) * baro (altitude) compensation: MAP measurement at startup. Separate baro sensor is not justified when you have WBO2: when we tune at (say) 200m above sea level, we add +10% fuel in the VE table, so ego_correction=-10% (-25 on display) that is good to appr. 5000m, and normally the WBO2 based ego correction will take away as much of the extra fuel as necessary. '''under investigation:''' * WBO2 analog output (for DVM or other ECM), eg. ** lambda=0.6 .. 1.45 => 0.6V .. 1.45V ** 1.45V .. 1.499V not used ** O2%=15..21% => 1.5V .. 2.1V (or should it be max 1.96V so the DVM can be left in 2V mode? maybe configurable) ** below < 0.55V: measurement error (eg. WBO2 sensor not heated). Alternatively this condition could be mapped to >2.5V for better margin, but when WBO2 controller is unpowered, 0V is "easier" to maintain * Warning light output: activates when any measured data (that I list here) exceeeds it's programmable warning treshold: ** CLT: engine too hot ** EGT: exhaust too hot ** MAP: boost too high (BoostController misconfigured, or pneumatic problem) ** ... ---- '''Problem after engine was put back''': * 4 out of 10 startup is unsuccesful. LCD displays silly things and car runs like shit! Sometimes after succesfull starting LCD displays some kind of table data instead of realtime datas (as usually), but no keyboard is connected. ** you forgot to mention that this was after you pulled the engine and the ECM was also away for weeks. Check what could have been misconnected, or damaged. Check GND, GND5 and flyback too. Try to disconnect the LCD (maybe VCC occasionally shorts to GND in the LCD ?) * Check firmware version (a report is neglected without this information and impossible to help) * and verify firmware integrity * or reflash with a new firmware * check the config (publish mcd, mct: a report is neglected without this information and impossible to help) and compare with old known-good config * logs can also help to identify the problem ---- Questions,PLEASE ANSWER ALL OF THEM!!!!! *MAT retard??very useful and indispensable on high output turbocharged engines where MAT can vary. *For more programming room IAC settings can be simplified. If coolant bins are locked please extend its range from -20C to +120C *My coldstart problem : Tried to lower my -40Cranking PW to 10,no start,injection style All banks,than slowly increased to 1.5,and it catches.No start for firs ignition sequence tough,only for second.Now im trying to use priming which i really hate!5ms with 0% tempscaling. *After setting up idle to 0.98-1 (with warm engine) and riding 50-60kms,sometimes it richens up to 0.91.Whats happening there??? *Made some logs and with megalog viewer i can see some suspicious PW pikes at mindless points.(no change in tps) i will try to include the logfile [http://www.vems.hu/files/NanassyPeter/datalog200511161915_hazafele3.xls Logfile] Comments from Emil: I never found the pikes, its a pretty long file and i haven't got too much time now, Could you provide time: entrys from those pikes? I also suspect that you air temperature sensor catches heat from the engine bay when idling for a longer time, and after driving the sensor is reading a realistic temp (beginning of log 25degC, minimum while driving was 1.7degC), how is that sensor mounted? You can see the result of this at time: 3510 [http://www.vems.hu/files/MembersPage/MattiasSandgren/BmwSevenTwentyEight/bmw_728_iat.jpg This is a example of a pretty good budget sensor install], We had huge problems like this on Boströms BMW, IAT raised to 50+ degrees while driving slowly.. This caused a very lean situation when driving in dense traffic. Hi Emil, first of all thanks for your reply! My IAT sensor is mounted at Audis stock manifold,stock place. After throttle valve in intake manifold. I always tought that when driving slowly (so IC not working) IAT can be 50+ celsius.But i see the problem at 3510.... But anyway i think between 7 and 37C is only 7 percent fuel enrichment,i my problem is not this.... Gve Spikes are at: 2677 2695 My other problem is that tps acc is always 100% so no enrichment,why?in MT i set some numbers there ,are the too big or out of range? Thanks again,Peter I tried various Acceleration Enrichment settings,nothing worked,i monitored it with realtime display and even with the crazyest numbers and fastest throttle opening there was no affect.My mixture of course leans out and car hesitates if i accel hard. how can i find information about firmware upload?? ----- '''Config and tables''' - also generated mtt files From your config mtt files created for 1.0.19 and 1.0.23 firmware: * http://www.vems.hu/files/MembersPage/NanassyPeter/Nanassy_mtt_files_v3_1019-1023.zip ** '''Note that global.h variables in 1.0.19 and 1.0.23 files are identical, so config.mtt files are compatible - actually, exactly identical, bit-by-bit, and MegaTune for either should work for the other too''' (so you can try mt-r027 and mt-r028 with either) ** the tables.mtt is independent of firmware version, there's never been a change there (but in MegaTune, "12x12 default" must be selected). The included tables-1.0.19.mtt and tables-1.0.23.mtt are identical The config.txt didn't seem seriously damaged at brief review, variables seemed mostly sane. * alternate=14 # means all 5 injectors are fired simultaneously during cranking * cwl=D7 # therefore this cold pulsewidth is extremely long: might appear as 21.5 msec in MegaTune but in fact 200 * 0.1 + 15*0.5 msec=27.5 msec * cwh=0B # while at warm, 1.1msec sounds a bit low (but possible) Send the config.mtt and tables.mtt files with TerminalProgram ("send file" option, or even copypaste is possible) ---- '''Readback''' *config and tables read back from 1.0.23 ** http://www.vems.hu/files/MembersPage/NanassyPeter/r023_config/mcd.txt ** http://www.vems.hu/files/MembersPage/NanassyPeter/r023_config/mct.txt *config and tables read back from 1.0.19 ** http://www.vems.hu/files/MembersPage/NanassyPeter/1019_config_kiolvasott/mcd.txt ** http://www.vems.hu/files/MembersPage/NanassyPeter/1019_config_kiolvasott/mct.txt * config and tables read back from 1.0.14 ** http://www.vems.hu/files/MembersPage/NanassyPeter/1014_config_kiolvasott_WORKING/config.txt ** http://www.vems.hu/files/MembersPage/NanassyPeter/1014_config_kiolvasott_WORKING/tables.txt Looks like there ARE differences from the original!!! Are they serious differences ? ** likely just a few values probably changed from MegaTune ? ** or complete stupid values, maybe offsetted (values shifted to different variables). ? Should I make mtt files for the original? ---- '''Verification''' Verify 5..10 values (some from the beginning, middle and some from the end) manually. I verify with ''diff -U 1 etc/config.txt readfrom1.0.19/mcd.txt'' <code> -rpmk[1]=60 +rpmk[1]=1E tpsdot_kpadot_conf=00 @@ -250,6 +250,2 @@ tach_channel=FF -tach_divider=FF -shiftcut_conf=00 -shiftcut_time=01 -shiftcut_channel=FF - +tach_divider=FF </code> The rpmk[1]=1E instead of 60 is very strange, definitely bad. It results in a -2.5% (repeatable) error in RPM reading. The 3 extra variables from the end are probably left out in the read-back file. I verify with ''diff -U 1 etc/config.txt readfrom1.0.23/mcd.txt'' <code> --- etc/config.txt 2005-11-28 16:23:23.000000000 +0100 +++ readfrom1.0.23/mcd.txt 2005-11-28 17:48:31.000000000 +0100 @@ -42,3 +42,3 @@ rpmk[0]=09 -rpmk[1]=60 +rpmk[1]=1E tpsdot_kpadot_conf=00 @@ -252,4 +252,3 @@ shiftcut_conf=00 -shiftcut_time=01 shiftcut_channel=FF - +shiftcut_time=01 \ No newline at end of file </code> The rpmk[1] is the same, and there is no other difference (the "No newline at end of file" is harmless). '''verify readback tables''' diff -U 1 etc/tables.txt readfrom1.0.19/mct.txt shows only minor differences, nothing harmful. (eg. different boosttarget at very low-RPM, where it does not matter anyway) '''WARNING!''': ''diff -U 1 etc/tables.txt readfrom1.0.23/mct.txt'' showed big differences ! Almost everything differs, I just pick the most shocking RPM and KPA bins: <code> -k[0]=14 2D 3C 50 64 78 8C A0 B4 C8 E1 FF -r[0]=07 0C 0E 14 19 1E 25 2B 32 3C 43 4A +k[0]=8C A0 B4 C8 E1 FF 1B 1C 26 4F 67 7E +r[0]=25 2B 32 3C 43 4A 14 2D 3C 50 64 78 </code> Eg. RPM starts from 0x25=37=3700 RPM. Looking at the end of the readfrom1.0.23/mct.txt r[0], we see '''14 2D 3C 50 64 78''' which should be at the start of k[0] !!!''' This is 6 bytes offset . Similarly, 8C at the start of k[0] should be at position 6! '''What you noticed in MegaTune, is exactly the same: RPM starting way to high and vector-war screen, because kpa is not monotonously increasing''' (but decreasing from FF to 1B at the middle) Seems like it was left over from your ooold 1.0.15 (or 1.0.13?) >>>i think it was 1.0.14!! .Possible causes: * If you sent the tables.mtt via serial, maybe the data stream got through damaged ** possible solution: send 2..3 times, and verify the read back mcd and mct values (even if just manually, a few values here and there ) * maybe a bad MegaTune startup option was selected, not the "12x12 default" that matches the firmware. ** unlikely, right ? >>i always select the default 12x12 Uploaded the files you provided and upgraded to 1.0.19,looks like everything is fine now.I found some stupid numbers (nothing serious),maybe these caused my problems. EGO-correction seems to be automatically disabled sometimes during idle, resulting in rich 0.89 lambda instead of the desired 0.95 .. 1.02. (what is VBatt during this situation? How much lower than normally? can this be spotted in your logs ?) * WBO2 warmup absolute limit: try to increase from 160 to 180..220) * WBO2 PID ** Heater PID integral limit: increase (even up to maximum) ** PID Heater KI: increase slightly, but take care with this, increasing too much will result in divergent (oscillating) heater (and WBO2) behaviour ----- *i need an ALS wiring sematics ** since you have internal MAP, '''EC18pin6 is a free input. Use 10k pullup to 5V and switch to GND (EC36pin26) for activation. This does not require disassembling the box.''' ** note that current ALS code MembersPage/GaborRacz/NewAlsLaunchAndOthers uses fixed digital input, that is not available on assembled controllers; minor firmware mod is needed * i need injector setting recommendations for 14ohm high Z bosch injectors (420ccm,short pencil style ,from AUDI S3) ** injpwmt=FF # activate after 25.5 msec (actually 47.5 msec) => never. This alone would be enough enough ** injpwm=FF # PWM duty=100% .. for safety ** injocfuel=.. # 900 usec ... just a guess. Depends on flyback, relatively short with your powerflyback
Optional: Add document to category:
Wiki formatting: * is Bullet list ** Bullet list subentry ... '''Bold''', ---- is horizontal ruler, <code> preformatted text... </code> See wiki editing HELP for tables and other formatting tips and tricks.