6 cyl 2jz Toyota with ETC - Norway
Log with the strange behaviour would be of utmost importance. (preferrably with annotations like at ... sec happens ...)
The .vemslog file contains an immense amount of information (especially with 60-2 or 36-2 trigger, even per cylinder behavior). We cannot analyze the engine behavior from hearsay.
Because of the ETC, it is really important to sort this, and overview of the full install/setup is needed.
oid11617 (repair /upgrade).
Project URL provided: no (this was created later to facilitate progress).
On short paper, he requests to test inj and ign (logiclevel) outputs (every input and output passes tests).
At higher output engine does not work OK (maybe leans out ?)
ECU in/out seem to work OK.
HW Testing
We tested the v3/10001 multiple times, and a bit puzzled.
It passes output and input functional tests.
- Although these tests methods are primarily made for manufacturing (component, solder) defects, not for possibly external induced device damage, it usually shows any such problems clearly.
So the whole setup needs review.
We see a config inside, abstract:
- 6 cyl ( 6cyl 2jz Toyota acc. to orig oid)
- 6 logiclevel ign outputs
- 36-2 with camsync
- ETC enabled (seems intentional, orig oid had bridgeoutputdriver !)
- SD card logging is disabled. There is a 19 sec log, leftover from manufacturing. Unfortunately no log to review.
We will review the wiring setup carefully (ground, flyback, ... thickness and wire colors would be useful in the same table), and compare with the config we have (or any config or log provided on same thematical page). Log with the strange behaviour would be of utmost importance.
Reviewing ecu vemscfg (not much use without vemslog, or any setup details)
Some ETC misconfiguration present. See [VT ETC help]
- ETC integral window curve bad setup
- error threshold 5v (disabled), tps1/tps2 calibration curve not filed (prolly not connected); pps1 has narrow range,
- most likely using pps2 as pps1 (also bad), pps2 calibration curve confirms (filled 0.6-4.08 range)
- fuel wise, the deadtime table is filled with quite high values, seems to have forgotten to substract closing time;
- req fuel is tad high but compensated by scaled down (bad resolution) ve table (otherwise could be okay-ish)
- Dave says: if he provides a vemslog i can do a more thorough review
Leaning out on high power can be so many other things:
- like alternator low capacity (dropping voltage at high rpm)
- fuel pump to small, pressure dropping
- regulator problem
- ...
v3/10001 Orig specs
1 x ECU VEMS V3
- 1-wire interface: no
- EGT input: 1
- Flyback: HighZ-only 30V flyback
- ignition driver: 2+6
- Knock input channels: 0
- LCD connection: no
- MAP connection: 400 kPa simplenipple
- mounting-style: flanged case
- primary_trigger: EC36/27 VR
- secondary_trigger: EC36/13VR
- Service: normal
- withSDcard: removable
1 x MAT sensor
5 x Bluetooth-RS232 adapter
- With wire: Wired-in 12V Dual USB Power Supp
Orig order request confirms ETC application:
We will probably use all analog inputs here so:
- ETC = 3 without pullup + tps input.
- ALS on/off with pullup
- Oil-pressure 3pin = without pullup
- oiletemp = with pulluo
- launch on/off = with pullup
- EGT 1x
Please configure BridgeOutputDriver with 2 pushpull + 4 highside drivers.
Please equip VEMS ECU with flying loom connector for BridgeOutputDriver with 4 highside drivers, and min 3+1 analog inputs on EC18
Log file from problem -> http://www.vems.hu/files/Reidar/ECU_10001/v3.3_u010001-2015.07.21-18.06.24.vemslog
See 1.54.281 sec in log.
Log file from Dyno -> http://www.vems.hu/files/Reidar/ECU_10001/DYNO%20LOGv3.3_u010001-2015.04.24-21.46.24%20.vemslog
I have not had the chance to troubleshoot this myself. This provided information is from talking to customer:
Fuelpump output not working => Car stalls (0 RPM)
- from orig text it appeared more like: Car stalls (0 RPM) => Fuelpump output not working
- clarified, guessing from the "fuelpump-relay swapping" action below
Bypassed relay with jumper cable in "relay connection", car runs fine.
- resistance of relay coil ? ... Ohm
- What type of relay ?
- does the coil have diode or resistance flyback protection inside ?
- If not, is there a flyback diode installed at the relay socket, or in the harness ?
- isn't there a FET, IGBT or stepper output available to drive the relay ?
- was this with a certain (but for some reason unnamed) output for fuelpump ? Or even if some other type of output was selected for fuelpump ?
- did he not specify which output he thinks "not working" ? Explicite is always better than to "assume", or to leave others no other chance, than to assume.
- Are we talking about EC36 pin 15=P259/ch5 (fuelpump_channel=57) ?
- or same behavior with a different output selected ?
- The ECU fw or HW actuates the fuelpump similarly at all RPM-s (except 0 RPM: fuelpump deactivated after configured time, this is number1 criteria to comply, for desired behavior at any accident)
- before sending the ECU anywhere, it would have been reasonable to take some notes and req help, and assign miscout to that questioned output, and see if fuelpump stays on that way
- and connect fuelpump relay to another output, best as recommended by the provided paper (relay page of [Rescuekit.pdf]): injector, ignition IGBT, stepper recommended if available, and as a last chance: the low current output
- we need to find out if it's related to the ETC setup deviating from [VT ETC help], a bit unusual
- note: the ETC "off-target" condition would primarily manifest as injector fuelcut (similar to TPS< IAC tps_threshold RPM overrun)
- we've seen some installs (although rare, and more easily happens when modifying some factory harness), where some load was powered through some surprising path: eg. not directly from +12V (well, via a relay and fuse, not literally directly from node 30 of course), but through some OTHER load (eg. PWM IAC valve, or boost valve or other relay). This is tricky to notice, because the +12V measurement can be (false) reassuring.
He also claimed having problems with cyl2 ignition output
- also unnamed EC36/...
- note: but (according to config extracted from ECU) certainly a logiclevel i259 output like the other 5
- how was this conclusion arrived at ?
- scope or other measurement ?
- by swapping 2 ignoutputs in wiring ? (and the problem moved ?)
- by learning that removing injector from cyl2 makes no difference in engine power, but finding that cyl2 inj is good ?
- are there any perceived/suspected conditions under which "cyl2 ignition has problems"
Since repeated instrumental measurements show that all inputs and outputs work, it would be of high importance to explicitely name which actual (EC36/...) output(s) are suspected, and what exact measurements lead to the conclusion. If we need to speculate/guess (because report lacks important info), that makes it very difficult to help.
Is it possible to check grounding in harness ?
- eg if all 5 grounds are connected ?
- thick, min 1 (or better min 1.23) mm2 wires ?
- maybe some receptacles slided back-out from the plastic connector housing ?
- was the econoseal backout plate installed ? If not, is it possible to insert before engine start attempted ?
- any chance that some screw came loose ?
- experience suggests, that if something "ECU does not work well" happens at "high engine load", it is related to GND wiring (rarely +12V)
- '''if our assumptions are good, and "EC36 pin 15=P259/ch5" is the "misbehaving output", together with "happens at high engine load" => fingers point at grounding to be doublechecked
- in some cases of insufficient GND (or sufficiently wired for normal load, but somehow overloaded), the P259 chip dies (often protecting other, more valuable parts of the device).
- In this case P259 didn't permanently die => which actually made the situation less obvious.
- there is a traditional quick test for GND pins and connections: +12V, through a 25W brakelight bulb, that is 2A current. Any individual GND receptacle contact and related harness wire (to the GND junction point) should drop max 10mV voltage.
- the good 0.005 Ohm resistance is so small, that DVM resistance or diode mode cannot indicate because the DVM probe's contact resistance dominates, obscuring the real var to be measured. By measuring voltage, the resistance of the DVM's probe contact becomes largely irrelevant.
- Easiest to measure if there is an unpopulated EC36 male connector at hand, but some EC10 male tabs (pins) do perfectly, or even SSC male pins. If the EC36 is populated, the current is divided between contacts, and acceptable max=2mV/2A ... but less precise this way and might not show if 1 contact is poor.
The ECU / fw happily and precisely drives outputs at 16000 RPM,
- and ECU otherwise does not mind if MAP input voltage relates to 414 kPa (for the controller, they are just numbers, and certainly no numeric overflow or similar, even at 100% injector duty)...
- but if GND5 (power GND) jumps to 0.7V above GND because of peak current, or current spikes, that results in P259 permanent damage, or transient hazards.
Toyota 2JZ ETC is a bit special with only one TPS sensor, and a clutch inside the throttle body, used instead of the safety relay. This is the standard 2jz vvti ETC. ETC works without problems, even though PPS1 and PPS2 are swapped.
- TPS2 should be connected. Normally separate TPS signal (with different slope polarity) should be available for any proper electronic throttle valve, and preferrably routed separate from TPS... but :
- If separate TPS and TPS2 signal not available for some reason, than TPS2 input connected to TPS is still better than not connected at all. In this case really route the wire on alternate path (and in separate insulation tube: less chance for any mechanical hazard to cut both simultaneously)
Injector latency values are supplied with injectors.
FIC1100cc`s. compensation works good, and tested/compared to voltage drops / disabled alternator and 500W fan.
- OK, good to have that tested.