Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by cj

  1. cj

    Repeating Fault Code 75

    Those arent parameters you would normally worry about once you have a "happy" ethrottle. As far as I understand it the logging is limited to a certain number of MB, and the more parameters, the quicker you fill it up, so you just reduce the amount of history/time you can hold in the ECU.
  2. cj

    Subaru Version 8 G4+

    The inital gauge sweep in a lot of the newer subaru's is controlled by the gauge cluster and it doesnt "accept" rpm signal input until a couple seconds after startup. Do all the gauges sweep at the same time? Which model ECU have you got, and was this car originally e-throttle or cable throttle? (it changes the pinout). Sounds like some of the pins are connected/assigned wrong as your idle rpm doesnt seem to relate to your idle valve DC% at all.
  3. cj

    Repeating Fault Code 75

    Unfortunately that log doesnt contain values for TP Error Accumulator or TP/Target Error Accumulator. If we cant fix it here, add these 2 into the logged parameters for next time so we have better info next time it faults. For now, we can see here that the accel pedal has no issues and rule that out - both main and sub track nicely together and show no AP errors Comparing TP main & Sub we see the 2 of them track together quite nicely so its unlikely to be a TPS error. These 2 values seem to be within 0.2 - 0.3 at all times which is fine. comparing e-throttle target against TP main & Sub postitions however, we see that every time you get off the throttle, the TPS postition almost "bounces" back up a couple %. Usually this is caused by PID setting issues in the ethrottle control. in your case it looks like they might be a bit too aggressive, and its causing some pretty bad overshoot (and for it to not quite oscillate, but still jump back in the wrong direction once it overshoots). This screenshot shows the worst example I could find of both the "bounce" when you go from full throttle to 0%, and for general poor tracking of target -> actual TPS angle. Note the blue TPS line overshooting by 10% ish repeatedly between 10:04 and 10:08. There is another good example at 13:40 +/- 5 seconds. I'd suggest setting your ethrottle to "always run", then with the engine off, tune your PID settings a bit while logging these same values, until you get TPS to track target reliably thorugh both slow and fast throttle transitions
  4. cj

    Repeating Fault Code 75

    Can you please post the actual log? there are 3 or 4 error counters around main vs sub correlation, AP target vs actual correlation etc, and we need to see which one is tripping.
  5. The jumpy MAP value and low-ish voltage suggest it is cranking, although the description isnt very clear on this. It looks like you have no trigger reading happening at all. Can you grab a trigger scope while cranking it? If it worked yesterday, there is a chance that one of your trigger sensors has had the plug wiggle off or rub through a wire.
  6. cj

    Base timing discrepancy

    When you have it in set base timing mode, you garauntee the timing will be at 15*. Sounds like at this time, you do have stable timing. This means your trigger1 offset is correct. When at idle, you are NOT garaunteed to be at 15* timing. If you are idling at 18* timing, you will be 3* ahead of your timing mark, exactly as you are seeing. You havent posted an idle log so I cant be sure, but looking at your ignition table, if you are idling above 1500rpm, or above ~50kpa MAP, then your timing will start to advance so 18* seems possible. without locking the timing, it will move around a bit at idle based on RPM and MAP.
  7. cj

    R33 GTR Speedo

    I dont know the answer to your question about speedo needle drive, but I've got one more problem to add to your list. The VR signal from the gearbox increases voltage (and frequency) as speed increases. Assuming you work around the signal direction issues, you likely need some kind of voltage attentuator on the wire before you feed this into a DI on the link. Factory manual says the signal voltage is approx +/- 7v @ 40kph and +/- 10v @ 60 kph . Extrapolating this out, at 200kph its going to be well over the 12v you'd normally feed to a DI and might damage something if fed in un-filtered. Is all this work just so you can do speedo correction or is there another reason?
  8. cj

    bench test ICV

    with the engine off but key on, look at the current coolant temp, ensure idle control is open loop, then change the idle base table value for your current ECT. The valve should move to the requested position.
  9. kph value / 60 = km per min (a) inst fuel (in cc/min) / 1000 = Litres per min (b) with these 2 values now both in X per min, you can remove the "minutes" from the equation and use the 2 values as Kilometers per Litre so (a/b) To calculate it as km per (whole) Litre, multiply a by 1/b to change it from km per litre, to litre per 100km, you take 100 divide by the km-L value Example: 50kph, 150cc/min 50kph / 60 = 0.8333 km/min 150cc-min / 1000 = 0.15 L/min now that "minutes" is constant to both numbers you can simply remove it to give 0.8333km per 0.15L mulitplication factor is 1/0.2L = 6.666 so 0.83333km x6.666 = 5.55 km/L 100/5.55 = 18.01 L per 100km
  10. you need the vehicle speed too, but as long as you have this, yes you can calculate it, on a spreadsheet or with a calculator, or a script on your pc. Can you do it in the ECU to have it displayed on a dash etc... I dont think so.
  11. How did you test this? backprobe with a multimeter? most DBW pedals are wanting to see 5v input, not 12. Pedal wiring should normally be 2x feeds from the 5v ref output, 2x feeds from sensor ground, and 2x feeds into ANVolt inputs unless you have a rather weird pedal. The aux9/10 supply error has nothing to do with your pedal however. It means the ECU is not seeing 12-14v on the red wire top center of plug B. This pin needs to be supplied with 12v, which is then used to drive the TPS motors. Have you got 12v supplied to this pin? Regarding your trigger scope, were you cranking the engine when you captured this? Also what voltage are you using for the power feed to the sensor? The trigger scope shows a constant ~7.5v but no teeth etc.
  12. What the chance of getting a junkyard ABS pump? Like I mentioned before, if you want a direct feed from the ABS sensors to the ECU you will have to cut off some teeth, and this will break your ABS functionality. You could also buy/build a frequency divider to drop the input frequency from a few thousand Hz to a few hundred Hz. You can buy these for roughly $100 from a few places online or build your own for about $10 of parts at jaycar or whatever electronics supplier is local to you. Once you get the frequency down to <500Hz @ 200-250kph (whatever max speed is), you can feed it straight to a DI. You would first need to confirm that the output is a square wave - check the manual or scope it yourself.
  13. This sounds like the way to do it. Logically you cut the wire from abs -> factory ECU, and put the link in the middle of that cut wire. (DI as input, aux out as output). You define a DI as LR wheel speed, and set the calibration number. Adjust this number until the speed shown in link matches the gps on your phone. Grab an aux out and configure it as "speedo out" - I think this has to be one of aux 1-4 but will depend on the expected frequencies. Not all aux outputs can PWM signals at a high enough frequency. Once you set this, another option will pop up under the aux outputs "folder" named Speedo Out. Click on this and you can set the source DI you are "copying" as well as multiplier + offset values to adjust the frequency. Note that these multiplier values dont take any notice of the calibration for the DI - its just frequency in -> frequnecy out (with multiplier)
  14. Which car are you working on? A lot of the earlier japanese stuff ran gearbox -> speedo then speedo out -> all other systems that want a VSS signal. you'll need the factory wiring diagram to figure out where the signal goes once it leaves the gearbox (assuming its not canbus - at which point it comes from ABS). If its old enough to be a mechanical speedo drive its another story. In theory you want to change the wiring so the link is the first point of contact with the speed signal, do any modification, then have another output on the link to output the new speed signal and have this continue on through the rest of the factory "path" so other computers/gauges all get the modified speed signal
  15. So to clarify this - do you have working ABS? If the ABS is gone, you can grind off teeth on the tone rings until you get down to a more reasonable number - say 5 or 6 - just need to confirm the number so you can figure out an even spacing between teeth. If you have ABS working (assuming its the 2006 pump + controller), you can run wires to the CANBUS pins and just have a very short canbus only from ECU to ABS.
  16. On a thunder yes, on the lower ones not directly as the frequency is too high for a DI. The 48 or 50 teeth they use gives you a max speed reading of about 10-20kph before you break 500Hz. You can read it from the canbus directly though if you are tied into this - on a mid 2000's legacy I have some details on here this was CANID 0x512 bytes 2 & 3 - read as a 2byte number then divide by 16 to get vehicle average wheel speed in kph. You can get individual wheel speeds from 0x513 as well.
  17. cj

    Injectors Not firing

    you can take the ignition table and most of the triggering etc from the supra map - ignition isnt affected by the fuel table config (except for AFR's impact on knock), but take the fuel table from any of the VE/modelled based maps. you'll notice that the majority of the VE fuel tables are relatively similar, because within reason most (stock) modern engines are in the same ballpark when it comes to fuel efficiency ranges. That should at least get it to run enough to start tuning it. If you have already tuned a bunch of the WOT cells you could copy those into your "new" fuel table and then interpolate from the tuned cells down to the "copied from another map" cells so you are even more in the ballpark for starting figures.
  18. I think you frequnecy -> speed calibration is a bit wrong. you're not maxing out the frequency input and I dont see anything that looks like it goes "flat" at any point. If you know your actual trap speed you can pretty simply calculate what that 260 value should be to get your speedo to match trap speed. The log shows a max of 190kph
  19. Couple of things you can change to get it a lot better, then grab another log if its still not right 1) Disable VVT around your idle RPM by either upping the lockout RPM on your VVT to say 2k or, change the VVT target table so it has 0 target at 2k below 30% or add in another column at ~1800 and set this to 0 below 30%tps. This is because you VVT is advancing by 10deg or so as soon as you get above idle rpm and throwing things off 2) as suggested by RobW, change idle control to open loop with the engine dead cold, then let it warm up and as it hits the center of each cell in the idle base position table. In general, your table values are about 45 and closed loop is running the idle valve at about 55 to keep the idle rpm you want. If your open loop table values were much closer to what your engine actually wants it will be depending less on closed loop control to keep stable. 3) its going very lean (1.2 lambda) at a couple points around idle. setting the fuel table value for 1500 @ -60 MGP and the same row @ 1000 rpm about 5-10 points higher might help as well. Once you have it approximately stable, try to smooth out the ignition table around idle rpm, and once its about right, turn on idle ignition control which will help keep idle stable - you just shouldnt use it to fix idle that is otherwise all over the show.
  20. sounds like its not going to be a fuel problem then. A few things you could do to rule it out but i'm leaning more towards triggering issues from what you've said so far.
  21. filters, sizes of the pipe between the pump and the tank if they are external, everything that will be under suction instead of pressure when fuel is pumping. Do you have a surge tank and/or feed pump for this?
  22. Those pumps are kind of overkill and might even be a problem if you just have a single wire feeding both of them. 8gauge wire should be rated for ~73amps draw though so unless there are some weak points/small wires/fuses/etc further up the wiring this should be the problem. You could get a friend to hold a multimeter across the pump terminals and verify that you still see ~14v when it starts to fail just to rule it out. [edit] what do your fuel pickups look like? if you are sucking a large amount of gas out of the tank at once and through small pickups you might be causing cavitation/air bubbles in the fuel. there are 2x back of the envelope calculations you can you to estimate fuel pump size 1) 4x 1000cc (per sec) injectors x 60 seconds = 240LPM max flow through your injectors @ rated pressure @ 100% injector duty cycle. your injectors and pump(s) will have a different flow rating at different pressures, typically the rated flow is at 43.5psi (3bar) and drops a bit at higher flows, but some pumps actually increase their flow at high pressures, and most injectors flow more so you really need the spec sheet to confirm this. 2) horsepower / 2 = LPM requirement. Assuming you have 400 BHP, you need ~200LPM (on petrol) for example the walbrough 465 pump spec states: 465 LPH @ 40PSI & 13.5 Volts, 360LPH @ 120PSI & 13.5V. The trigger scope part is that these engines can get "noisy" at higher rpm which can throw off the RPM triggering which then causes all kinds of thing to go wrong around spark and fuel numbers. The trigger scope feature in the ecu is like an oscilliscope that shows you the trigger voltages and pattern the ECU is seeing. With the engine running, click ecu controls -> trigger scope, and then save the log file it creates. This only take a couple seconds. Please grab one log at idle, and another with the engine at 5500rpm
  23. I dont see any limits/fuel cuts/etc kicking in. Your logs dont show the lambda values or fuel pressure though so its hard to get the full picture. Without this I can really only guess Do you have a screenshot of the lambda graph from the dyno from either of these pulls? specifically I want to know if gradually crept up once you were on full boost, if it suddenly just jumped at some point, etc. Is your fuel regulator boost reference working correctly? max calculated fuel flow in those logs is around 2050cc/sec - so around 120LPM which even after losses in the lines etc, you should be able to hit with any decent aftermarket pump (assuming your lines are big enough and the pump has big enough wiring to supply current) Were you feeling misfires as it got above 5k on the 3rd gear pull? your RPM jumps around a little bit which could be misfire or it could be trigger issues. If all else fails, grab a trigger scope at 5400-ish rpm.
  24. cj

    3sgte goes lean

    ah, you didnt say it had a misfire. that combined with the sudden start of this problem means i'd check the plugs for damage, strange wear, etc. The last log your fuel number looks pretty good for near idle @ ~0.95 so I doubt you have rich/lean misfire happening now so check plugs, coils, HT leads, etc - whatever ignition system you have on the engine.
  25. cj

    3sgte goes lean

    Doesnt look fuel pump related. Its actually flowing less fuel at that ~1400rpm 1% TPS range than you are at idle. If you run that same check but at say 20% tps (so 3-4k rpm) does it show the same behaviour? Looks to me like your fuel map isnt granular enough at idle map values (30-35kpa) and so its bouncing around a bit as it goes too high/rich -> too low/lean and cycles back and forth between the 2 states. Your idle seems ok but its at ~0.87 lambda whereas that 1% tps rpm is between 1 and 1.15 which is quite lean. try increasing the fuel table values in this region by a few points. you may need to add in a line at 30kpa as well and let it interpolate as a starting point. The other thing that might be at play here is if your injector dead times are wrong. Are the numbers you have in there from a reliable source and/or tested? You also have no short pulse adder numbers which can help this kind of behaviour too as at ~1 ms you might be on the edge of non-linear injector flow. you've also got launch control permanently enabled when the car is under 15kph (VA1 config). This isnt directly impacting it but its stopping closed loop lambda kicking in and probably some other things. If none of the above help, disable ignition idle control and swap idle, lambda correction etc all back to open loop/disabled so you can rule out these systems "over-reacting" which can cause cyclic rpm issues like this.
  • Create New...