Audi URs4, 20VT, AAN engine with non-factory assembled v3/9467 (purchased as mainboard and custom-assembled, non-standard in a couple of ways)
Installed camHALL inverter and [car finally started, without any problems] CONGRATS.. Thanx for cooperating and thematically sorting it. Have fun with your project (tuning, idle, power, gizmos, whatever...). If you happen to order another VEMS v3 ECU (econoseal, motronic55 or other type before 2016-07) (with your real name/addr we know from elsewhere), write in the order comment: "-150 EUR ... and refer to this URL."
Problem was that triggerlog showed no secondary trigger or spark events. As well as no RPM while cranking. Primary trigger seems fine.
Information from Peep (who was involved with assembly):
- v3/9467 device was assembled without CAM hall inverter (just a jumper)
- without cam-HALL inverter is nonstandard (since ages): since it is much more prone to any noise (from wheel, or ign, or other) that way. As it turned out, the standard cam-HALL inverter was REALLY needed for this install. So much sisiphos-work was caused by saving a small NPN (~0.02 EUR) as originally assembled (by a reseller) in a nonstandard way. Let's learn from that.
- also, without cam-HALL inverter, default AAN config needs ignout and triggertooth change to work
- instead of 62deg BTDC cyl1 crankhome pulse (reaches the processor with CAM hall inverter), the 134deg BTDC cyl5 pulse reaches the processor (without CAM hall inverter). See InputTrigger/AudiTrigger So ignout should be rotated so cyl5 is the topmost entry (assuming reference tooth 0 is also on top, as usual), and (assuming divider=3 and next trigger tooth=18) First trigger tooth=9''' (TDC after the trigger appr 58 deg, no change there)
- no +5v pullup
- that is quite non-standard (depending on the HALL sensor in use, it might not make a difference during operation but inconvenient during diagnostics: a pullup resistor is needed to pull up the HALL input signal to +5V when sensor is disconnected: it appears from below, that this was discovered, and 10k pullup applied).
- no P259 internal connection
- no big deal by itself
- sectrig VR LM1815 pin5 = +5v.
- should be OK
- sectrig DC voltage should be measured, Measured camHALL sensor 0-3,9V; mostly 3,9 and for a short period 0.
At the moment everything is stock, besides VEMS V3.6 and NTC type IAT sensor. Running firmware 1.2.2
Thanks for the [vemslog, config. and triggerlog], we reviewed. See below.
Hello Iwbs,
I have reviewed your config and triggerlog, indeed no secondary triggers present.
Lets start with the basics, how have you wired up the trigger part of the ecu (e.g. which sensor is connected where)?
My investigation shows this board (not assembled ecu but main board only) with serial v3_u009467 was sold to a reseller who must have manually assembled the ecu himself, i would advise you to contact the person you bought the ecu from, most likely there is some board setup error; in any case he should be able todo bench trigger testing for you to confirm good ecu operation.
Best regards, Dave
Just INFORMATION (in this chapter nothing to do): auditrigger HALLscope is a convenience feature of v3 when sold as [VEMS ECU]
Only available with auditrigger w/camhall inverter:
- a special p259-channel activated pulldown on camHall pin EC36/13
- equivalent of disconnecting HALL sensor and grounding the camHALL signal manually
- yes: it's really just that, a convenience feature (1 less step can be useful when expert's VT is connected to v3 over the internet: otherwise it's fairly easy to unplug camHALL-sensor connector)
- when sensor input is pulled down (either with the p259-channel or manually) => camHALL-inverter output is +5V and the crankhome-VR signal goes through
ECU (sectrig=auditrigger w/camhall inverter) originally ordered as assembled is set up that way (unless requested otherwise)
- v3/9467 was sold as
- it does NOT effect normal operation at all; just debugging: cam-HALL sensor needs to be unplugged (as a test, if no sectrig pulses in triggerlog, and new triggerlog captured), read below
IMPORTANT: Before sending device anywhere:
in case of any auditrigger sectrig issue, capture 2 triggerlogs : one with camHALL sensor disconnected; and one with camHALL sensor disconnected, and one with signal wired to GND:
- unplug cam-HALL sensor and capture triggerlog (due to internal pullup, which is hopefully installed, the EC36/13 camHALL signal will be high, appr 4V).
- With input=High: Secondary trigger pulses will show up in triggerlog if ECU is set up withOUT camHALL inverter
- assuming still no sectrig pulses seen: connect the camHALL signal to GND with a small wire, or low value resistor 0..270 Ohm and capture another triggerlog (cranking 8..9sec):
- With input=Low: Secondary trigger pulses will show up in triggerlog if ECU is set up with camHALL inverter
That is very easy, and either it confirms the ECU has camHALL inverter, or if it is without camHALL inverter (both will work for an AAN, unless the flywheel is alien, donated from a 7A engine which has some "ghost ferromagnetic anomaly").
optional (but useful): Verification of internal pullup (with DVM: 5 minutes)
- as a simple verification test, insert a 1.0 kOhm resistor between camHALL input and GND.
- Assuming Rpullup=2.7k internal pullup (to 5V) and 1.0 kOhm applied to GND externally during the test: 5/(Rpullup/1.0 + 1) = 1.35 V will be measured approximately.
- else you can calculate internal pullup resistance Rpullup from the measured value using this formula. Substitue the 1.0 with your actually applied pulldown resistance value (in kOhm: eg 0.27 if 270 Ohm pulldown was applied while measuring 0.13V ) : solving for Rpullup=(5/0.13 -1) * 0.27 = 10.1 kOhm
- Assuming Rpullup=2.7k internal pullup (to 5V) and 1.0 kOhm applied to GND externally during the test: 5/(Rpullup/1.0 + 1) = 1.35 V will be measured approximately.
If still no sectrig pulses in any triggerlog (captured with camHALL=hi and low), than crankhome-VR setup must be reviewed carefully.
- Eg. measure crankhome-VR sensor resistance: around 1100..1300 Ohm normally
- measure cranhome-VR signal DC voltage (+50..400mV when sensor is connected; ECU powered up but engine is off).
- internally: sectrig LM1815/pin5 can be left open (for adaptive hysteresis) or connected to +5V (fixed threshold). Possible to get it work either way, but it would be useful information to know which is applied.
- also a scopeshot (say, with an USB scope) could be helpful (perhaps the crankhome-VR signal was inverted twice ?).
- Taking note of wire colors and photo of the VR-connectors would be useful too
- it is possible (and perfectly valid) to connect PC speaker (or line) output to ECU sectrig crankhome-VR input (normally EC18/12, directly or through 270 Ohm resistor). GND must be also connected of course. With max volume, play some high amplitude wav c002 or m071 (from VT trigger play; or even some beep or Pavarotti; while capturing triggerlog). PC linout is min 2V peaktopeak (usually 2.8V so +-1.4V) and (if sectrig input is set up correctly inside the ECU) the pulses should show up in triggerlog if DC level is between 0 and 0.5V (obviously pulses not aligned to crankwheel :-)
•Thank you for your help!
Today I tested camHALL sensor itself with DVM and scope. Scope showed me no trace of it. Measured voltage between sensor signal and ground while cranking the engine and it stayed 0V. Absolutely no changing in voltage like there should be.
I will try another HALL sensor and see how it goes. Also I have camHALL sensor mounted right under the exhaust cam pulley, so there's no way it could've turned away from the crankhome position.
Power feed to the HALL sensor was 4,9V. That is normal.
• I tested again camHALL sensor, made some silly mistakes before testing it. As it turns out camHALL sensor is fine. Power feed stayed at 5V, ground is fine and signal changed from 0V to 5V. Tomorrow I will take a scopeshot to verify alignment.
- forgot to say, that it's also possible to verify camHALL sensor output with DVM (DC voltage measurement), eg. moving the car in gear 5, and noting the position of window (and relating that to TDC, or the crankhome-VR pulse).
- But of course scopeshot is the best: if captured with 2 channel scope, with crankhome-VR signal (alternatively with 1 channel scope 2 resistors, say 10kOhm each, can be used to mix the 2 signals).
● Engine won't start
Got solution for my trigger problems, (non-factory assembled) ECU was missing 10k pullup and that's fixed now, although it's still giving me sometimes errors less pr. trigger and tooth length bad even with injector and ignition fuses removed. There were no trigger errors when I removed spark plugs. I'm pretty sure this is not the issue anymore cause as long as I have reviewed other similar projects, engine should still start.
Tested my ignition order in vems test mode and it was 1-2-4-5-3 like it should. Also injectors were tested in test mode and they all open and close. Pulsewidth while cranking is about 14 to 15 and they are set to open individually. Req. fuel is 8,7. Used timing light to verify that 1-st cyl is firing at TDC. I'm not quite sure about my tooth reference table cause I couldnt find any logic in this. Also spark plugs are wet after crancking. I can hear some puffs in the exhaust while crancking, but it won't start. Here's my latest config: http://vems.hu/vemstune/bugreports/reports.php?cmd=view&key=zqwXvE#comments
Hello Iwbs,
I have reviewed your trigger, it seems you are still having some hardware issues, your spark events are causing noise on your secondary trigger. That said, does your board have the cam-hall inverter applied or not ?
- no cam-HALL inverter
Best regards, Dave