Adamw

Ok it all looks happy in that log. But things might be different when the engine is running and there is boost etc. So my next suggestion is to set up ECU logging like below and next time it plays up, down load the log out of the ecu as soon as you can. The ECU will have enough memory to hold about the last 1.5hrs of running set up like this. This may give us more clues. Set up like below except for aux 8 status - replace that with whichever aux controls your DBW relay.

Can you try changing the transmit rate to 20Hz and turn the OBD2 off and see if that makes any difference.

Were you driving when you done that triggerscope? There is no signal showing on DI3. I want you to do the triggerscope while driving so we can try to capture what is wrong with the speed signal.

Can you tell me what firmware version you are using. >help>ecu information. It looks like there was a change to MX5 triggering in 6.17.8 so if your firmware is older than that then that would be the first thing to try.

I made many changes as it had some pretty odd settings in your tune. Can you do a couple more logs using the tune below. Since ive changed quite a bit it may still not be right so dont freak out if it is worse not better initially. We should be closer after the next log. Idle tweaks .pclr

I havent seen the odd blank screen like you show there before. But yeah, make sure the power supply is goo is the first thing I would look at. If you attach your ecu tune an the dash config I will take a quick look at your software setup too.

The log is not much use as only some basic parameters are logged and there is no lambda etc, but from what is there it mostly looks ok. I had a quick look over the tune and there are a few aspects that I dont like. The first one is the lambda target table only spans to 0Kpa boost (i.e it suits an NA engine). So above 0kpa boost it is still targeting 0.885 lambda which is very lean for a boosted engine. For all we know the tuner could have completely ignored this table and tuned it to run richer but there is no way to know (most tuners that know how the system works would generally tune to the value in the target table). Similarly knock control is only spanned to 100kpa MAP (0 boost), so may or maynot work above that. The ignition table doesnt really look tuned or optimised either. My suggestion is it should probably go to a tuner for a check and tickle up. If it drives ok it probably is not going to need much time spent on it to confirm the tune is ok, but I think there are enough questionable aspects that it would be prudent to get at least a power run with a lambda and some knock ears connected to confirm.

Trigger 2 looks fine now. That is fine/correct. It will work fine, but just FYI, the built-in 1K would have been fine to - or even marginally better. It is the same sensor Link sell.

Do you get spark if you use the ignition test function? Can you confirm in ignition setup>ignition main, that spark edge is set to falling. Do you have an oscilloscope or LED test light available?

Yeah pretty much, G4X looks a little different, but similar enough I think you will work it out on your own.

This may be an ECU related problem - there are some specific models where we have seen something similar. What causes the problem is at key on the ABS ecu expects an acknowledge very quickly from the engine ECU - but the Link ECU is slower respond than the factory ecu at boot up time when it is doing lots of high priority tasks. Because the ABS doesnt receive the acknowledge instantly it throws an error. I will PM you some instructions with a simple wiring change that is needed to solve the issue.

Yes that should be fine. It will be a little tight on inputs/outputs but it is capable. Use Aux 5 or 6 for the Vtec solenoid and Aux 4 for the VVT solenoid. Connect inlet cam sensor to DI1 (B7). Connect crank sensor to pin B16 and exhaust cam sensor to pin B12.

Can you give us a PC log with some heavy throttle/foot movements - engine doesnt need to be running. The error 73 maybe a tune issue rather than wiring issue - if the ECU shuts off the E-throttle relay due to errors/safety checks then this error can appear.

The aux out fault wont prevent it from running - that just indicates the voltage on that pin is not reaching the expected level. Try unplugging the AC clutch relay for now to see if that error goes away.

Is the MAP sensor hose connected? It is showing 102kpa all the time - it doesnt even flinch with the engine running so it appears not? Have you checked base timing?

Eth content and boost switch added. To calibrate the fuel level, go to the live measures screen and note down the mV voltage from your fuel level sensor at various known points. You can either do two point calibration just full and empty or you can get more accurate if you want and note down mV every say 10L added. Then you need to go to the custom sensors screen and fill out the calibration like below. Eric 03.zconfig EricsFinaltune New CAN.pclx

Most of the warm-up idle looks ok in that log. Can you try to get a log of this near stall effect.

Ok, try these. Note in the dash the fuel consumption channel I have assigned units as "ml/s" since AIM dont have cc/min, but I didnot scale it to read in ml/s, to me I dont think that will mean much to most people - So it will still read in cc/min. New CAN.pclr LINK_Heags86_@20200920_091942_007219.xc1

I would lean more towards it being a intermittent connection or possibly failing sensor. The triggerscope in the G4X now captures the DI's 1-4 as well as the triggers so we may learn something from that. Can you get a few triggerscope captures when it is playing up and we''l see if there are any clues in there. You might have to do 4 or 5 since it only captures a short time period and may miss of of these spikes. You could also try changing the GP Speed input active edge to raising to see if that is any cleaner.

Yes the BA sensor works fine.

You can make it work by fudging the firing order in the software but that is asking for trouble if you ask me - I would fix the wiring. If you fudge the firing order and say the tuner finds for instance cyl #1 is more prone to knock and wants to retard it a little, if you adjust cyl #1 in trhe software it would actually be adjusting cyl #4. It will be fine if you are never going to do any individual cylinder trims, but I dont know if I like doing it that way.

Just go to the CAN protocol page in RS3 put a check beside yours custom stream and hit export.

Ok, It looks like my changes to base position may not have accounted for the AC step and fan step. Can you do another log of some idle and blips with AC and fans off.

Sorry, I must have missed your original post. I dont know of anyone that has ever used the EL sensor so it would probably be better to fit a wideband sensor from the get go so as you dont waste hours on it only to find it tops out at moderate RPM. First things first though - does your ECU actually have knock control? Only the Xtremes did in the G4 range as far as I remember.

Im no coder, so some of this is in laymans terms but my understanding is it is basically not possible with the G4+. The whole G4X code has been written from scratch with functions such as this in mind. It stores values as 32bit floating point numbers which allows both a huge range of numbers and high resolution. In the G4+, most values are stored as 8 or 16bit integers, so you have either a decent range or decent resolution, but not both together. So this makes user math much more difficult as you would have to scale every part of the equation so that every step of that equation and any possible result of that equation didnt lose too much resolution and will still fit in the 8 or 16 bits of space that it had to fit in. You could possibly use a GP PWM table to calculate delta temp provided you have a PWM capable aux and frequency cable DI spare.