Jump to content

Adamw

Moderators
  • Posts

    19,853
  • Joined

  • Last visited

  • Days Won

    1,266

Posts posted by Adamw

  1. From a couple of references I have I would agree with your wiring - red being signal and white being the ground.  

    It is pretty odd for the cam signal to cause a trigger error with this type of missing tooth trigger, so this shouldnt be an issue.  Can you attach a copy of your tune, as well as a log of it cranking, one with the cam sensor connected and one without.  

    Is it using a Link loom or a modified factory loom or something?

  2. 20369 is from back in may and I suspect that was for the original mod we done to make the CAN 1 port work.  

    To give you a bit of background in case it hasnt been explained to you properly;

    There is a design oversight on the current JZX100 board, the CAN expansion socket that is available for connecting user accessories to is connected to CAN2, and CAN2 is also connected to another micro-controller on board that controls the temp gauge and AC.  This microcontroller for the temp gauge uses a bitrate of 1mbit, so that means anything else connected to CAN2 must also run at 1Mbit.  For 99% of users this creates no issue as most aftermarket devices either run at 1Mbit or can be set to run at that speed.  AEM is a bit unique in that they use 500Kbit and it is fixed/not configurable.  

    In most of our plug-in ECUs we have CAN2 doing the chassis stuff and not available to the end user, and CAN1 is the one that is available to the user in the CAN expansion socket to do anything they like with it.  

    So this is what we done in your ticket 20369, we disconnected CAN2 from the expansion socket and connected CAN1 to it instead so you can then use your AEM device at 500K plugged into the CAN expansion socket.  

    It now seems like someone has given you some bad advice and replaced the board which puts you back to square one, with no CAN1 available, and only CAN2 which is not going to work with your AEM device.  So far I have only found communication of the original modification and I havent found anything relating to why the board was replaced etc.  

     

  3. You have no accel enrichment working or async fuel, so you will have some issues with transients without these but I think you have something else going on too.  The reported lambda doesnt seem to follow the trend that injector PW shows - ie when PW increases you would expect lambda to get richer.  Obviously with a misfire the lambda sensor isnt useful so we cant put too much weight on lambda alone.   But the lambda doesnt look like it is showing the truth to me.  ie it is idling with the lambda pinned at 1.23lambda which is the leanest the x-series can read, there are not many engines that would be able to idle that lean.  Assuming it is not misfiring at idle then I would suspect maybe the lambda probe is dying or there is an air leak or something. 

    I suggest the changes in orange and red below, as a better starting point fr transients, but there is some other non-tune issue by the looks so this isnt going to solve the issue.

    5fTsNap.png

  4. Settings below from my evo 7.  Copy these in (except leave your existing base position table), after storing, do a power cycle (turn ign off for 5 sec then back on) and do another log, I will suggest some tweaks from there.  Your base position numbers are quite high already which may suggest the BISS screw needs to be opened a little more but give us a log with the new settings first since the step count setting will have an influence too.  

    3aTAjv5.png

  5. The V7 engine would have originally had a cam sensor at the front of one head behind the sprocket connected to trigger 2 then 2 further sensors at the back connected to DI's.  

    You will now no longer have or need the front sensor, the LH intake cam sensor at the back will now need to be connected to trig 2 instead, the other 3 cam sensors will need to be connected to 3 DIs between DI 1-4.  DI 1 & 2 will already be there at the back of the heads so you will just need to run 1 further DI from ECU to cam sensor - probably DI4 is easiest (was IC spray switch in some models).  You will need to run power to the cam sensors too.  

    The new extra 2 AVCS solenoids need to be connected to any aux between Aux 1-8.  Probably the easiest is Aux 8 (was o2 sensor heater) and Aux 5 (CE light), the CE light could be moved to any other spare output.

  6. 11 minutes ago, kirchoff said:

    hi resistor between pin 7 and 9 jumps a little bit but it is arround 0,6 MOhms

    Yep that would suggest there is something wrong with wiring or sensor there.  Check the resistance directly at the sensor pins also - if you get the same then sensor is open circuit.  If something more typical then investigate wiring.  Typically a VR sensor like this would be more like 500ohm.  Well at least less than 1000ohm.  

  7. Most of the older cars like this are happy to share ABS sensors in my experience.  I suggest do a real crude test to confirm, it doesnt need to be flash, just grab a couple of pieces of wire, backprobe into a sensor connector or at the ABS plug, connect to the ecu and see what happens.  If you get an ABS light, then swap polarity and try again.  

  8. I suspect it would have been pretty difficult to find that specific post using the search, I found it easy since I knew the symptoms and name of the damaged file.  I just linked to it to save me typing the same replay again - forums are good for that.   

  9. Can you give me a bit more info on why the bottom board was changed?  It looks to me like you or someone else must have originally added wires or links to make the CAN1 port functional since the AEM device needs 500Kbit whereas the JZX bus on CAN2 runs at 1Mbit so you cant have them both connected to the same bus.  

    But now you have replaced the board which means you are back to square 1 with no CAN1 access.  

  10. Your idle control is not working because your APS is never at zero, it never goes below 1.4%.  Not sure if that is because you havent calibrated it, or if you have your foot resting on the pedal.  You also have no integral gain on the idle actuator so that means you are effectively running in open loop.  Set it to 0.05 is usually a good starting point for e-throttle.

  11. The early AEM widebands like 30-4100 and 30-4110 were especially poor accuracy on the analog outputs, and to make matters worse they didnt have a separate ground so even if you done a fudged cal to correct the poor analog accuracyit, you could see the PWM heater circuit noise in the AFR signal as well as the afr would change when for example a fan turns on and creates a larger ground offset between chassis and ecu ground.  The X-series at least have a separate analog ground which should eliminate some of these issues, but a digital communication such as CAN is always the best option.  

  12. Yeah we had a discussion yesterday.  Andrew thinks the output drive of the trig 2 circuit when set to hall would be a little less sensative to noise but his feeling was he would still expect to see some indication on trig 2 if it was via the ground.

    He did bring up something I never considered though - he said he has seen sensor vibration cause an effect that looks like noise.  Your bracket looks pretty robust in your earlier pic, but you cant really tell how thick that mounting plate is or the length of the upstand.  Can you flex it at at the sensor end at all with your hands? 

  13.  

    15 hours ago, LEIGH211 said:

    Has anyone got any clues on how to drive it? Looks like it's just two coils with 4 connections?.

    I imagine there would be a driver chip on the pcb somewhere.  

    What did you try for a booster device?  I have never had much trouble getting high level tachos to work.  Did the tacho used to connect to the negative or positive side of the coil originally?  

     

  14. APS sub and main should folllow each other very closely, usually when you look at the together on a log they will be right on top of each other.  A section of your log where it failed is below, at the left hand end where I have placed the purple box you can see APS sub and main are pretty close to each other (although not as good as they should be).  However near the middle of the graph where I have marked with a red box you can see the green trace (APS sub) has started to become erratic and no longer matches APS main at all.  You can see the APS error accumulator at the bottom start to count up which means the ECU is not happy about the amount of error between them - however it then comes right after you lift off.  Further to the right you can see they no longer match at all and the accumulator rockets up to 100% so safety mode is initiated.  

    I would say the sensor is failing or possibly a loose connection somewhere.  Set up your laptop on the drivers seat or something so you can look at aps sub & main on the screen, then get under the dash and wiggle/shake/yank the wiring as much as you can to see if that shows any connection issues, otherwise I would replace the sensor.  

     

    2hH8bFl.png

×
×
  • Create New...