Audi 5 cyl 20vt AAN engine
Non-standard flywheel, crankhome-VR signal way lower than standard, apparently signal is lower than noise ?
- Stock sensor (but non-standard pin in the flywheel) only 130mV AC (10 times lower than factory).
- Modified sensor 225mV AC
- reason for the low signal: flywheel was replaced with a spec unit and the pin is not as bulky as the original pin on the original flywheel.
- does this mean +- 225mV ? Or 225mV peak to peak ?
- See test3.ods (made from test3.txt) in http://vems.hu/files/MembersPage/MorneAudi/morne/triggerlog/ but note this looks like primary trigger (135 tooth) and also a bit weird, distorted.
- Bad news the stronger pullup resistor does not work.
- I fitted a trimpot 80-100kohm and adjusted voltage from 130mv to 170mv in 5mv increments with no luck.
- Please find attached log files. where exactly ?
- It seems we have to go 60-2.
- there is one more thing that could be used to suppress the extreme noise (compared to the very low level signal), a capacitor between crankhome-VR signal and GND:
- maybe 100nF initially, good to test the concept
- one option: using a reasonably bulky screw (it does not matter if angular position is different, as long as safely within the HALL window: TDC-delay angle can be adjusted)
- other option: make the sectrig extremely sensitive (we did that, but noise does not become a problem).
After a long-long time and many iterations it turned out:
- Is it the sensor that causes the low voltage signal output for the trigger ?
- __The sensor is not an original OEM part__.
- Will the signal output be corrected with the original OEM part ?
Likely but hard to tell for sure.
- At least the controller passed same signal-level tests that always worked with the factory sensor (with large margin).
- We'll redo the min signal-level tests,
- and you'll have to specify your lower than standard signal levels.
28/06/12
Hi Marcell.
New trigger results with new vemscfg has been send to you via e-mail.
Please send my login password to download results to my members page.
Thank you.
28/06/12
Hi Marcell.
Unfortunately I am unable to test the sensor with scope.
But sensors has been replaced with brand new ones on Monday, so I doubt if the sensors are faulty.
Did test to see if pulse on sensors go up to wiring harness connector to ECU and it does.
I will only be able to send sensors out next week to be tested with scope if you do need the results.
Please find attached the results for trigger log, but still no sec trigger on the results send via e-mail.
Thank you.
Morne.
We did test the Hall sensor ignition on engine off
- Live to Hall 5.082volts
- Signal from sensor 4.123volt
- Signal from sensor switched 0volt
- camHALL window position relative to crankhome-VR: unknown
Can not find secondary trigger on trigger log (below)
This makes us suspect the crankhome-VR even more if the triggerlogging was done with cam-HALL grounded
Third trig - latest filenames and triggerlog suggest third trig disabled now, good
- Motorsport/Camshaft angle control/third trigger=disable
- burn
- reboot
- note: the third trigger is connected for diagnostic, and useful for verifying cam-HALL position relative to crankhome-VR
- but only the secondary and primary trigger must be enabled during normal operation, the third trigger should be disabled for normal operation in this case (as this is an audi with a cylinder head without camshaft-angle-control, no VVTI)
to see for 100% that crankhome-VR gets to the microcontroller
- disconnect HALL-sensor and pull cam-HALL input to GND (camHALL signal is motronic55/pin8 but easier with a wire at the HALL-sensor plug)
- on newer controllers (since 2012) this can also be done from software (eg. with [this config]) by activating p259/2 (internal output channel) via misc_output (by activating always on: from 0-25500 RPM, 0-510 kPa, 0-100% TPS, etc...) but we recommend the good old grounding wire
- capture triggerlog !
- if secondary trigger shows up in the triggerlog (captured with HALL-input=GND), we know the cam-HALL signal is constant high (4..5V) or the window is at bad position for some reason.
- measure cam-HALL signal DC voltage (ignition on, engine off). Is it usually 4V, and sometimes changes to 0V when rotating the engine manually ? (pushing car in gear 5).
If the motronic55/pin8 was grounded this makes it extremely likely that the crankhome-VR signal does not go through the LM1815 for some reason (maybe the crankhome-VR signal does not go sufficiently negative ?)
- the secondary trigger is an internally gated signal combined from the cam-HALL and the crankhome-VR, so it is likely that the filename is wrong and the cam-HALL was actually grounded
- Note that this is still not the .triggerlog file, just an image, but we assume that the secondary trigger didn't show up at all.
Can you make a scopeshot about the crankhome-VR while cranking ?
- or describe eg. "every 340 msec signal goes up to +1.2V than go down to -0.8V"
v3 serialnr 4622
- motronic55 box (2012-06-01)
- with cam-HALL inverter
- if cam-HALL signal is at suspect, pulling cam-HALL signal to GND (unplug from HALL-sensor and use a wire) will reveal this (either during triggerlog or startup attempt: engine will backfire every 2nd attempt on average, but at least we know and might allow limp-home)
- motronic55/pin8 (cam-HALL signal) connected to third trigger input inside (to allow scoping signal for future diagnostics only, not for running: third trigger should be normally disabled)
note: use FileArea to upload files (under http://www.vems.hu/files/MembersPage/MorneAudi/morne/ where the other related files are).
RPM=0 => Tools/capture triggerlog so we can see which trigger seem to be missing
- [vemscfg can be found in the report zipfile]
- [vemscfg file], also default AAN file and the "diff -U2" difference
- no serious problem in the configuration at first glance:
- prohibit seems good for 1.1.96
- configswitch is disabled
- some calibration values differ (good)
- vemslog: STILL not available
- click on "LOG" icon until it becomes green, attempt startup, publish the related file from VEMS_Files directory
- triggerlog: not available (important if RPM=0)
- Tools/Capture triggerlog, attempt startup, publish the related file from VEMS_Files directory
- we would need to check the RPM at the startup attempt !
- is RPM=0 at startup ?
- or RPM=normal (around 180) ? Or nonzero, but abnormal (if so, how) ?
Ignition outputs...
- Ignition outputs visual, switch to test mode
- important if RPM is good at startup attempt
- Did you see spark ? (use a sparkplug on the cylinderhead, make sure sparkplug-body is actually grounded to cylinder head). Do not touch !
- All 5 High-Voltage channels work correctly
- yes, sparking OK
Be careful !
- maybe it's something as simple as blown fuse ?
See [VT help] and tutorial videos.
Measurements
Note: triggerlog was not uploaded, but the images show something surprising:
- The "Camshaft exhaust" is surprising (1 tooth for every 90 primary trigger tooth is OK, it's 270 divby3), we're investigating why it is not "secondary trigger".
Copied from [report] (note: vemslog was not captured or not included in this report either)
log for audi s4 aan
ignition coils tested ok
measurement sensor
ohm between 44/30 ,728kohm
ohm between 45/30 2,69kohm
aan pin 10 0,4ohm
aan pin 19 1,3ohm
aan pin 24 1,2ohm
aan sensor pin 30 4,7mohm
- measurement trigger 49/48 1kohm
- ALSO MEASURE pin49 DC voltage (ignition on, engine off)
- measurement trigger 47/48 1kohm
- ALSO MEASURE pin47 DC voltage (ignition on, engine off)
- aan sensor triger 48 open
- What does this mean ? Sounds to contradict the above 2 measurements (if "open", then it would be infinite ohm, not 1 kohm
Symptoms
- no start
- 5 sparks tested OK
- no rpm signal
- Assuming RPM=0
- do not know which firmware has been used
Hi there, please find voltage measurement as requested.
Pin 49 0volt ignition off, ECU connected.
Pin 49 2.865volt ignition on, ECU connected.
Pin 47 0volt ignition off, ECU connected.
Pin 47 0volt ignition on, ECU connected.
Pin 48 on harness tested infinity to earth.
But test 1 kohm to pin 47 and 49 on G4 and G28 sensors.
New log file has been send on 26/06/2012.
Let me know if there was a problem again on the log file.