Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by cj

  1. I think there is something physically wrong with your fuelling system. with 550cc injectors on a 2.0 engine you should have an effective pulse width somewhere between 0.7-1ms. Yours is reporting as 4-5ms but your lamba is only reporting slightly rich. I'd guess either the injectors are blocked up, not actually 550's, or you have stuff all fuel pressure. Or you have something really quirky in your wiring like several injectors wired to a single output? This isnt a tune thing as it doesnt really matter how you get there with the numbers in the tune, that size injector at anything resembling a normal fuel pressure (300-400kpa) must be running around 1ms at warm idle. Once you find the issue, cut your master fuel value at least in half, maybe down to 1/3 to get the pulse rate something more reasonable. Once you have it idling ok, the numbers in the fuel table will behave as expected and you can start tuning it. Can you also please describe how you have your tps main sensor wired up? I think there is some electrical noise getting in to it somehow (and maybe other parts of the ECU control as well). those logs show that its jumping between 5% & 30% 5-10 times a second almost constantly. Maybe that's true because of the way you are testing/driving it, but it seems unlikely. If more important controls are also seeing significant noise it could explain many things
  2. I vote for gear shift control. a) its the only one with a -30* ign cut b) its set to trigger "powered up shift" above 4500rpm when DI 4 is active.... which is configured as ignition switch so its always active? Like Adam says, its really hard to be certain from that log as there are not many parameters logged. separately... Theres something wrong with your oil pressure sensor. No way do you have identical pressure at 4500rpm and 600rpm
  3. the install path shouldnt be relevant to driver issues as these all get installed in various system folders anyway (and yes they overwrite each other - the latest driver is usually fine though for old versions of the pclink software). right click the device with the yellow mark, go to properties, and see what the status/error is. If it just says something like no driver loaded, re-run the driver installer from your install path - something like C:\Link G4-567\Link USB Drivers (the click the 32/64 bit installer depending on your laptop). Is this win7 or 10? there was an issue with the 566 or 567 drivers not installing on win7 so you had to install those from 565 first. the 566/567 install would then just not touch the drivers and use the old ones, so this shouldnt have broken from an update.
  4. Getting into the more picky things now as its looking ok for a "about to go to the tuner soon anyway" state. 1) your igntion table is quite conservative especially at lighter loads and just off idle. This will make it feel a bit flat and may make it run a bit hotter. I dont think what you've got is dangerous at all though but it could be better. 2) your ethrottle response is still a bit slow. It is recording target match errors some of the time but only in low quantities so again, you could make it feel a lot better to drive than it currently is by making the PID settings for ethrottle more aggressive, but its not dangerous. There are a few posts on these forums on how to tune this better. 3) you're about 250rpm high on your idle rpm vs target. your igntion timing is only at 14* which is factory-ish, or possibly a couple * lower than factory so if it were mine I wouldnt drop idle rpm via lower ign angles, and there isnt much wiggle room in your base idle target table as its all small numbers. I'd drop the ethrottle target by 0.5 in that top row, then bump up all your base idle numbers by 0.5 to compensate, then look at reducing the numbers in the base idle table by 0.2 to 0.3 at temperatures where neccessary to get down to the ~800 rpm target it looks like you want. Best way to do this is from dead cold, fire it up, touch nothing in the car, and as your temp hits the middle of each cell in teh base idle table, bump it up/down until your idle rpm is what you want, then wait until it warms up 10 deg more & repeat. 4) hard to tell from your log, but there are still RH inlet cam errors being counted. The value doesnt go up at all in the log, but it is 200+ right from the start. Your VVT target tables are both all zero's so we cant see cam performance when advancing away from 0. Put somethign like factory numbers in the target tables, or just say 10* above 2krpm on inlet, and -10 on exhaust from the same point, then run another log while revving it up and down a bit so we can see the cams move.
  5. cj

    Batt volts reading low !

    The difference in the current draw between a "low power" coil and a high power one is about 5amp max vs 8amp max. The output Joules seem somewhat unrelated to max input current. There was a really good video on a bunch of the common australian coil swaps floating around youtube where they tested input current & joules of output to the spark plug. Assuming COP's and sequential ignition, the factory wiring can almost certainly supply the <10amps needed to power higher end coils. CDI's are another story but I dont think that's what you are asking here. Running wasted spark doubles the current draw through the harness (assuming shared ground and/or positive feed) so then you *might* want to look at wiring - you probably need at least 12GA wire to support ~20 amps but a lot of factory wiring will be this size already. Peak & hold injectors can draw more current than saturated but again its well under 10 amps per injector (usually more like 4-5) so assuming sequential injection, its probably fine on factory wiring.
  6. what are you currently using as your duty cycle? maybe try adding or removing 10-20% from it and see if that helps. You could also try putting in a 5ohm resistor rather than 10ohm as I seem to recall there was a lot of discussion on the correct value to use with some people even jumpering it with no resistor (but I wouldnt reccomend that). Its been a while since I modded a toyota tach and it was a couple years newer than yours but they certainly *can* function well from a link with just that resistor mod.
  7. first, what sort of inaccuracy are you seeing? there is a way to calibrate before you try before jumping into alternate PWM tables. This assume tacho movement is linear, but just not correct. * multiplier - if the tacho needle only moves 800rpm for every 1000rpm of engine speed, set the multiplier to 1.2 (or 0.8 to correct it the other way) If you have something non-linear in the tacho (correct up to 4k then only moves 500rpm for every 1000rpm or something), then heres how to do generic PWM output instead of a tacho signal. Set an aux output to GP PWM (aux 1-4 will do up to 500hz, 5-8 only 300hz). Note that 500hz only drives a 4 cyl tacho to about 8k. This is a limitation of the GP PWM tables - a normal tacho output can do higher frequency. Set the condition to "only cond 1" and something generic like "rpm > 400rpm" Set PWM frequency to "table" Set the PWM DC table for this aux to all 50% (axis options dont really matter, ive set it to 2x cells of RPM but not required). This value may need to be tweaked for your car Set the PWM frequency table to have only 1x axis of RPM Set the numbers in the PWM frequency table to be the frequency of pulse sent to the tacho at that rpm. [freq = rpm / 60 * number of cylinders .... plus your fixup number]
  8. im guessing the orignal tacho didnt run from an ECU at all as it probbaly didnt have one, but probably from one of the diodes on the alternator instead. You could pop the tacho out and see if it looks similar to the other 90's-ish toyota tacho's where you can just bypass an input resistor to make it accept a 12v sqaure wave. I know a lot of the 4wd forums talk about sourcing tachos from the newer hilux/surfs or from petrol models and swapping the unit over. Google is probably your friend to work out the exact input pulse the tacho expects. The link can only output square wave PWM signals so you can either a) find a tacho that accepts this and swap it in b) modify some resistors etc on your tacho so it accepts this type of signal (tacho dependent) c) get a tacho booster from a few online parts stores (never tried it but apparently you can wire up the coil side only from a relay to get the same high voltage pulse) d) open up your alternator and run a wire from one of the output rectifying diodes to the tacho to get the factory design working again. This will only read correctly if the crank -> alternator diameter ratio is the same as factory. Whether C works depends on the signal it is expecting. I'm not sure if a diesel alternator signal is similar to a coil negative pulse like used on older petrol tach's
  9. logically, *something* has to go to trigger2 (except in a few very special cases, not this one). If your engine had a non-vvt enabled exhaust cam to wire to trigger 2, then the 2x intake cams could be DI1 & DI2. If you dont have a fixed exhaust cam, and so your only 2x cam sensors are the 2x VVT inlets, then 1x of them has to go to trigger2. As far as I know, subaru didnt start putting exhaust cam sensors in until they released the quad AVCS models. Therefore, you only have the 2x inlet cams to work with, so 1x of them has to go to trigger 2. I did notice a typo in the help file where the EJ25 settings are correct, but the early EJ20 AVCS instructions say to use DI1+2 and set trigger2 VVT to none. a) you have an EJ25, and b) the EJ20 settings dont seem to make sense so i'd start with the EJ25 settings anyway even on an EJ20, but then check things before starting it.
  10. You have triggering errors counting up while cranking and erratic RPM signal. Something is wrong with your triggering. Can you most your pclr file as well, and a trigger scope capture? - to do this crank the engine, then while still cranking hit the scope capture button, then post the .llg file here please. trigger1 should be wired to crank trigger2 should be wired to LH cam sensor any DI should be RH cam sensor. what year & model were the old and new engine? maybe you have moved from an older to newer model crank wheel - these changed early 2000's.
  11. cj

    Injector duty cycle??

    duty cycle at idle is pretty misleading because it has so much time between injection events, that the "% on time" is always going to be low. normally i'd look more at "effective pulse width (ms)" as a measure of "normal". On a 1.6 with 550cc injectors its probably in the 0.7-0.8ms range.
  12. cj

    ECU for 1ZZ

    this is the 01-07 model? its essentially a corolla and the diagrams I can find for those years show the dash is still driven by 5-10 separate inputs - speed, tacho, various warning lights. celica's in the same year range did have canbus/bean (aka multiplex) but even then it looks like its mostly for the auto trans, and a bunch of stuff from the body module - park brake, door ajar lights etc. Without specific year/model, i'd say you'd get most of it working in a complete rip and replace job.
  13. So the next question then is, with the old intake (that works with an OEM ECU), does the link still work? I'm guessing not given you test with power straight into the ECU, but you will need to fix that issue on the new intake before you connect up a fixed Link or you might break that too. At a guess, you didnt have the ground from the left of the chassis (by the fuel lines) connected. On the older intakes it bolts through one of the intercooler mounts and if you dont have it connected, you get that dash staying on issue as its some important part of grounding the chassis.There is one on the right behind/above the air box which can either go to an intake bolt or to the top of the heads, depending on who last installed your engine. Other thing to consider, is are you using your original wiring and connecting it to the new intake, or did your new intake come with all the wiring attached? That wiring loom contains 95% of the engine->ECU connections, and i'd imagine the pinouts are different between years.
  14. cj

    Rpm signal required

    Yep, find any free aux output & set the config to "tacho". This will output a square wave signal that will drive most modern (post ~1990) tachos
  15. On subaru's the dash lights staying on after the key is turned off is usually caused by bad grounds, and it means a bunch of systems are grounding out through places they shouldnt be. The real question is why did your fuse blow? if you overloaded a circuit with too big of a stereo thats one thing, but if you have a short somewhere still in the car all kinds of things could be happening. do you see the power LED on the link light up when you turn the key on? have you checked power & ground to the ECU is still ok?
  16. cj


    deleted from your laptop or the ECU? If its your laptop, try Recuva - its free and pretty good at getting back recently deleted files. You will need to know where you had the file saved.
  17. Your post start enrichment table is all -20's which would explain the ~5 seconds of bad fuelling as that's about how long that lasts. Someone else had exactly the same thing a couple days ago so I wonder if one of the base maps is like this, or maybe there's a bug in a firmware update somewhere. It should look more like this As for your cam problem... its still showing no signal present on RH inlet cam. You could run the same checks on another cam sensor wire that is working, make sure you see the same voltages etc, ideally one that also goes to a DI so you know the pullup on the ECU is the same on the signal wire. Specifically that 5v power wire... I thought these ran on 12v? Its less likely on a low current circuit like this, but you can have really bad wiring that will show voltage etc on a multimeter but will not pass useful current. To test for this you need to put some load through the circuit and check for voltage drop or use a 1/2 amp test light (make one out of a 5w dash bulb, some wire and some electrical tape if you dont have one). Unplug both the ecu connector & the sensor, connect the ground wire to battery, the "other" wire to 12v+, then put a test light across the pins at the other end of the circuit (doesnt matter if you put the voltage at the ecu end or the sensor end but id probably put it at the sensor end to ensure you have the right pins). Then move the 12v input and the test light to the second pin to be tested. Both should light up brightly (same as if you put the test light directly to battery). If its really dim, you have marginal wiring. The other thing you could try (assuming you dont have a scope) is de-pinning the trigger2 wire and DI2 signal wires at the ECU connector, swap them over, disable fuelling, then try running a trigger scope while cranking. This will let us see what if any signal is being received from the RH cam sensor by the ECU. Might give us a direction to look at. Remember to swap the wires back after doing this too...
  18. Map Sensor - this is internal to the ECU. The config will show it connected to AN V3, but physically you need to run a tee-d off boost/vacuum line to the ECU. If you want to use an external MAP sensor, change the anv3 config to none, and wire in the signal line from the external MAP sensor to any free AN channel/reuse the AFM channel like you suggested. (and obviously set this channel to MAP sensor + set the calibration) canbus can certainly be used to receive the O2 sensor signals, and this is actually preferred over an Analog input as it prevents ground offset plus it gives you access to temp + error codes if the lambda controller sends them. The fact you previously had both modules on the same canbus means they should connect at the same bus speed. You will need to either configure your lambda controller to output message in "Link CAN-Lambda" format, or find/have a spec sheet of the messages it sends so you can configure the CAN receive channel in the Link to read the data. the flex fuel sensor goes to any available DI, yes. FYI the story on SR20 rocker arm failures seems to be that ignition cut limiters are a big cause of this. Either configure your limiters to kick in very gently, and/or use fuel cut as your rpm limiter method..
  19. your post start enrichment table shouldnt be negative. Yours is all -20 for some reason. Try something more like this Also, what temperature is it there? Your Air temp is showing -4C and your coolant temp shows 8C. Do these sound reasonable?
  20. You could verify the pinout by checking continuity between the crank sensor pins (unplug the crank sensor) and pin 17 & 43 or whatever it should be at the ecu connector end, and then do the same to the sensor ground & signal wires from the cam sensor (ECU unplugged as well - test at the connector). To test the 12v signal, plug in the ECU but unplug the cam sensor, then with key on, check for 12v on the pin you didnt find as cam or sensor ground. The 12v to the cam sensor may have been verified by your OEM ecu swap test, but its worth checking excplicty as some time OEMs can swap over to batch fire+waste spark mode if there is no cam signal and you wouldnt really know.
  21. 1) your fuel table is all "100". It *might* start like this but its not going to run well. As a starting point, copy the fuel table from any of the base maps running modelled mode - I think the monsoon base map has one you can use to start. 2) you could drop the ignition table 1 values for 100kpa 0rpm to 5 or 10, instead of 20. It should start as is, but its easier on the starter motor if you have a very low number in the cranking ignition cell. 3) you probably need to give it some gas pedal to fire up. I dont see any idle control systems configured so it will need some throttle to start, at least until you get it dialled in well. 4) not your starting issue, but you have 2x Check engine light outputs configured? 5) Youve got it configured for ~1000cc injectors. Is that correct? If your injector vendor supplies short pulse data you should enter this too. When you say "all timing is correct", does that mean you have run the trigger 1 calibration process and the timing mark is stable and at the right timing? Or are you referring to physical distributor timing?
  22. that scope shows a whole lot of nothing. Were you cranking the car already when you clicked the capture button? Adam is probably right about pinout differences but at least pin 16+17 which are trigger1 & 2 look to be the same pins, so you should see a trigger signal even if there are other issues.
  23. run a trigger scope capture while cranking it. Lets see what (if any) signal is coming in from the crank & cam sensors.
  24. the simple version of how to test this is what you are doing now - check voltage at a few points along the circuit and see when you see a significant drop. Remember to always use the same ground point because low voltage can also be caused by bad grounds. The more accurate way (that rules out ground side issues too) is to load the circuit up (remove the ECU, connect a head light bulb (so around 4-5amp load, which is safe on that circuit) between the ground and 12v ign wires, then turn on the key. Now set your meter to voltage and measure *along* the wires - eg first measurement will be battery positive to the screw down point in the main fuse box where that battery/alternator line comes in. On a healthy wire you will see something like 0.1-0.2v (or the same negative amount if you have the meter backwards, but it doesnt matter). If you see 1v or so on your multimeter it means you lost 1v between those test points and should go looking for a bad contact. Then repeat for each individual wire/pair of available test points. ie you could test both in & out sides of the EFI relay if you lift up the fuse box and pop the bottom off. Also test the grounds using the same process - ive worked on an sw20 before that had ECU voltage issues when cranking because the ground straps to the engine had poor contacts. You dont need the car running for this as the drop will be nearly the same at 12v vs 14v.
  • Create New...