______ 
| ___ \
| |_/ /
|    / 
| |\ \ 
\_| \_|
       
       

         
         
 ##  ##  
 ##  ##  
 ##  ##  
 ##  ##  
  #####  
         

 _______ 
(  ____ )
| (    )|
| (____)|
|     __)
| (\ (   
| ) \ \__
|/   \__/
         

 _______ 
(  ___  )
| (   ) |
| |   | |
| |   | |
| | /\| |
| (_\ \ |
(____\/_)
         

         
         
 #####   
 ##  ##  
 ##  ##  
 #####   
 ##      
 ##      

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:
Audi A4 B5 (VW)VR6 12V 2.8 engine. Stock VR 60-2 trigger wheel in the crankcase. camsync=disabled "bosch motorsport" ignition coil continental flexfuel sensor connected to Whspeed1st. 4.9WBO 6pcs low-z injector, 8ch Active flyback. removable SD card, Serious Primary trigger VR wheel "runout" (systematic amplitude variations within 1 rotation) suspected, at 2300 RPM less ptrig error. See InputTrigger/RunOut . Not this v3/14333 ECU, but a very similar standard VEMS ECU (VR/HALL but sectrig disabled) was tested with a runout primtrig signal: * 54..100% (and 125% tooth0 amplitude) ** worked perfectly with the factory R37=1M * 220nF (C38), well upto 10kRPM ** Marcell expected it to show trigger error at some RPM. ** Verified with scope that signal did NOT rail (amplitude really changed appr 2x fold each crankperiod ). ---- '''Russian roulette install''' * home-assembled (not benchtested before the field) ** this alone should not be a big problem (although adds to the uncertainty for sure),''' later benchtests showed that trigger input worked well (tested to 10k RPM)''' ** even with unusually high missing tooth amplitue [http://vems.hu/files/sensors/Trigger/60-2_aka_m582/ m582 signal wav] * installing far from home (on another continent, in USA) * GND problem => p259 damage * no scope * not even soundcard recording of VR signal * no wiki page at the time of install (this page was created later) The last 2..3 makes it extremely risky and waste a lot of time, basically impossible to support. In case of any install issue, '''without scopeshots we are showing darts in the dark''', (sometimes making things even worse, and not knowing). ---- The battery (+) disconnected and shorted to GND before first cranking and startup. The HIFI audio system has had a big capacitor so it was discharged and sparked * => p259 damaged. No 5V on the lm1815 Pin8, RPM=0 ** New p259 chip: RPM good, engine started up. TDC_delay set up with strobe light, '''"Less tr. error" flag and misfire at 1800-2000rpm. ''' VR sensor polarity was good, but tried to swap without result. Originaly r30=27K but after tried, 1K, 10K, 22K, 33K, and removed resistor(without pullup) without result, no different. Everytime got "tr error" around 2krpm. '''R181=150K originally, I changed it to 270K, This was the only improvement: The "tr error" came around 2300rpm''' * I kept the r181=270K and If i changed the r30 from 1k to 33K again. Without different. I tried to increase the R181 but the "tr error" came at lower RPM The '''best setup was: r30=22K, r180=270K'''. Originally lm1815 pin5=open (measured 4.95V, interestingly), i jumpered pin5=5V and changed R30 from 1k to 33k again, No result. * also tried 1k pulldown as adviced ? ---- '''Simple input threshold without noise suppression''' HALL input (as a last experiment). I tried to solder the ECU from VR to HALL as Marcel advised, R30= 1K. Ec18/27 voltage = 2.40V, no rpm signal read when cranking. This path was abandoned without trying to adjust input threshold to trigger (so engine did not start at all). Another possible path (this was not attempted) is to use an '''NPN inverter''' (external or internal) * possibly from the '''onboard sectrig circuit''', easy enough especially that sectrig is not used at all here: ** VR => sectrig and internally connect the NPN inverter output to primtrig (or as an experiment configure "sectrig used as primtrig" before doing the HW mod.): ** with some input bias (eg 900 Ohm sensor and appr 6..7kOhm pullup to +5V) it is possible to set appr +0.55V bias for the NPN inverter, so cranking triggers the HALL input This is similar to the simple triggerdetection with no noise suppression that simpler circuits use. Usually inferior, but might be the way to go if toothamplitude changes are extreme, but noise is not a problem. ---- '''Improvised bench-setup''' I tried the "Play trigger" tool * but i saw trigger error and wrong RPM reading all the time. * same on VR and HALL jumpered ECU Hw settings. * tried to play wav from audacity ? Perhaps some 3D sound special effect interfered ? Engine, body, battery and engine harness GND points checked and made to stronger. I tried to change the GND of the engine harness location from body to cylinder head, but no result again. Hard to check the trigger wheel runout because it's located in very bad position on the engine. This engine was running fine with another standalone ECU this before. Tried to lower the ignition dwell time from 3.01ms to 1ms maybe the coils makes electronic noises.... no result. I disconnected WBO sensor but no result. I made new shielded cable separated from the harness, GNDed(shield, VR(-) ) everywhere, no different VR sensor resistante=~900ohm it's OK, I changed the sensors, tried 3pcs, without result. I didn't have scope. I would have made scopeshots if i would have had. triggerlogs: http://www.vems.hu/files/Norbitron/v3.3_u014333-2018-01-26-09.39.36.triggerlog http://www.vems.hu/files/Norbitron/v3.3_u014333-2018-01-26-09.41.23.triggerlog Norbi From what I see in your logs and from my own experience I can say that trigger wheel in that engine is fucked up, I mean one tooth is bent or something, maybe the crank was dropped on hard surface and it's tooth 46 I believe. Valdas Tooth 45 time is slightly off indeed, but nothing extreme. * Wheel is most likely eccentric, ** which cannot be seen from triggerlog, unfortunately, but scopeshot or soundcard recording would show clearly, immediately. Marcell
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.