Adamw

That sounds much like cam lope to me.

I initially tried 6KHz and wasnt getting reliable knock detection, so I done a frequency analysis and you can see why... Frequency is the Y axis on this spectrograph, notice at 6KHz you can hardly tell the difference between the area with no knock Vs the heavy knock. The best signal to noise ratio looks to be around 8-8.5KHz, you can see quite a clear difference in sound energy between no knock and heavy knock.

Yep, thats all a E-throttle is. Keep max current in mind, I think 4.4A continuous and short peaks to about 7A is what the driver is rated for. And you want a dual track TPS. It is pretty common in touring cars etc to use a maxon motor and planetary gearbox. Pics below of the last one I set up. This one was geared too low so was slower than I would have ideally liked. It looks like it says 23:1 on the gearbox but they didnt list that ratio, so im pretty sure it actually is 231:1. The customer was happy with the response so I didnt test or research any further. Beautifully made swiss gear boxes though.

That uses the std Bosch NTC calibration. ECU wise the idle control looks reasonable now. You could probably drop the dashpot offset so it is not so slow to come back to idle after you lift off. Cant see the misfire or reason for it in the log. These errors are common with a noisy power supply. So you have the lambda on its own relay connected direct to the battery, or is there other stuff sharing the same supply?

I would usually do a custom stream, you typically dont need a lot of data to a PDM, usually fan and fuel pump request, maybe e-throttle and then usually only a couple of button presses coming back the other way. Yes, but if you then turn the ignition on after that they do very weird stuff, all the voltages and currents read wrong, settings revert etc.

Correct our ECU's dont have a direct thermocouple input, you will need a 0-5V thermocouple amplifier or a CAN bus egt amp. If you only want a single TC, Texsense have a nice analog one, and a branch in Deutschland. https://www.sens4speed.com/en/products/59-thermocouple-amplifier/95-thk-200-400-1250degc-3.html

Can you attach your pdm and ecu configs. From memory the hardwire cant take compound messages so it probaby wont be able to interpret the generic dash stream. Also, if you havent already discovered yourself - make sure you have the ignition switch on before you plug the usb cable into those things, they do some fruity stuff if you have the usb cable connected before ign switch is on.

Ok, if they were both set to falling there would have been a clear edge clashing issue, but the rising edges are well away from each other so we can rule that one out. Is this something that has only started recently whereas it was previously ok, or is it a fresh install? Are the coils a recent addition? There is one spot in one of the triggerscopes that is potentially a capture of the issue happening. It looks like the CAS disc has suddenly bounced backwards as the waveform initially started to rise earlier, then suddenly drops all the way back to zero before rising again a millisecond later following the same shape as the other adjacent "teeth". Since this rises above the 1.5V arming threshold and drops back below the 1.0V low threshold, this little zigzag would be considered an extra tooth. Is there anything loose in the CAS or drive system? It gives me the impression it is more mechanical than electrical since voltage doesnt drop below or spike above the normal waveform voltage. The normally do have some backlash but I assume there must usually be enough drag to keep the backlash in check. Whats the trigger 1 filtering level set to? I dont think filtering will help but it might be worth a try.

Whats your trigger 1 edge and trigger 2 edge set to?

Before you waste too much time on it I would first get it tuned so its torque output is more optimised and has a better relationship to throttle. In your log the throttle is moving between about 6% and 26% just to maintain 50kmh. You should have a huge increase in torque between 6% and 25% throttle, so something seems very wrong there (the idle ign def wouldnt have been doing you any favors). A typical torque to throttle relationship curve shown below, the torque numbers arent relevant but the exponential curve is always there with a butterfly throttle. Looking at logs from our subaru at 50kmh cruise the throttle is between about 8-15%, ~1900RPM (not sure if 4th or 5th gear). At 105kmh it is much the same, around 10-16% ~3600RPM (suspect 5th gear). This is just a random log I had on my PC, 3mins of west auckland motorway driving so it is certainly not flat road. The largest error in speed vs set speed after the 8:40 mark where I bumped the set speed up is 0.6kmh across the whole 3mins. You could get speed control even better than this but fuel economy takes a hit with the throttle working harder. This log was from some early testing before we changed the way the feedfoward is applied at the initiation of cruise which is why you see the spike in proportional and a small speed error at the start of the cruise then a large integral for the rest (throttle used to initially drop as you lifted off the pedal). Initiation is now much nicer than this and no integral wind up.

No it will still only have 2 injector drives so sequential wont be possible. A few improvements in G4 that I can think of: Much nicer software to use - especially the logging tools and multiple pages/tabs. Closed loop boost control. Internal datalogging. Quicktune. Many more options for inputs and outputs - virtual auxes, 3D PMW, custom sensor calibrations, vast range of analog input functions etc. GP limiters for engine protection. VVT control also.

Not directly at present. In the near future we will be releasing a device that could do closed loop position control of two DC motors, but for your application I would be leaning more towards CO2 boost control. With that you are controlling only the wastegate dome pressure so the boost control is not affected by all the dynamics going on in the intake manifold with launch limiters etc. The vast majority of boosted drag cars use this method as it gives very stable and repeatable launch boost. Effectively you have a realtime user adjustable spring in the wastegate. An electronic wastegate for drag racing would typically be controlled by pre-throttle pressure which is less stable and also throws a big spanner in the works if the driver has to pedal the throttle at any time during the run. The E-throttle blow off valves that you see in the intake pipes are not usually controlling boost - they are just bleeding air so that there is enough air flow through the (very large) compressor to get it operating above its surge line. In the few high powered drag cars I have been involved with they used a conventional BOV with PWM CO2 to control it. Basically it was fully open on the launch limiter and progressively closed with a short timer after launch - so you didnt get a big boost spike as it closed.

In G4X if you want to do complex stuff with more control dimensions you can use a GP PWM (or more) to control an axis of your function. Example below that would achieve one part of the functionality that kaptainballistik asks for above, where the retard gets reduced above 600C. You could use the same "% aggression" on your cut table and boost table. So to explain the flow here, we have a GP PWM table to vary our "aggression" based on EGT and RPM. We have 100 where we want full effect and 0 where we want no effect, I have an RPM axis in this table because you usually have 0 cut and retard below a certain RPM and TP. We than have this % Aggression on the axis of our antilag retard table. At 100% aggression we are operating in the far right hand column so are going to get 50deg retard below 20%TP and if say our EGT creeps up to 595deg, then our aggression will drop to 50% which would reduce retard to 25deg.

Yeah you have a lot of trigger errors in there. Even sometimes on overrun when ignition energy is low so it is probably not ignition noise related. Can you do a few trigger scopes at around 2300/2400/2500RPM as most of your trigger errors happen around that RPM.

I have made some changes to the idle control settings, the main ignition table, accel enrichment and I have turned off the dual fuel table. I turned off fuel table 2 because there are some cells with -11% in them with zeros on either side, at idle it is bouncing between the -11 and zero. Did your old G4 not have the open loop AFR table enabled? It is very odd for that to be disabled with a GTR and that probably explains why the dual fuel table is so extreme. This should really be enabled before you re-tune, but it will likely through your fuel tune out quite a bit. Pav-Own idle changes.pclx

@Vaughan The only logic I could put to needing that ">=" instead of just "=" in there is aux 2 status going into fault condition which would be a higher enumerated value than active. Some fuel pump controllers have a clamping diode in them which prevents the aux voltage going high enough so it fails the "open load" test and reports a fault.

Im not just a pretty face you know.

Attach the tune and a short log of it idling.

One problem I see is you have idle control kicking in and out during cruise so that's pulling a lot of timing out. Set up the speed lockout should solve that. Secondly, Im not very familiar with the mini engine, but the ignition advance seems very low at low load cruise speeds, has the ignition been tuned to MBT? For example in our subaru at 2000RPM 40kpa I have 39deg advance, yours has 28. So those changes may help some, but I think you will struggle to achieve great cruise control at 1500RPM in anything except a big lazy V8. Im not sure off the top of my head how much effect the E-throttle target has on cruise control as Ive never really needed to think about it, I assume the PID loop controls the throttle plate set point directly, but there must also be some blend that goes on in the background between driver commanded and ECU commanded under some conditions as it is a smooth transition if you give it some pedal when cruise is active. My suspicion is if pedal is sitting at zero then the E-throttle target table probably has no effect. You could try adding say 5% to the e-throttle target when cruise is active to see if it has any influence. If the target table does actually have an influence then you could set up a math channel "Cruise set speed - driving wheel speed" to generate a cruise speed error value, then put this on the axis of your throttle target.

You said in your first post you were using Aux 1 for the engine fan. Its really up to you. It will be either GP output if you want on/off control or GP PWM if you want to ramp the 2nd pump in slowly. Of course you will need a fuel pump controller or SSD if you want to do speed control. I dont particularly like just on/off control of a 2nd pump as you get an initial spike and then a step up in pressure when you turn it on with makes tuning more difficult. Typically when I need two pumps I just run them both together all the time, but I dont do a lot of high power stuff so dont have issues with regulation etc.

Yeah thats what I suspected. Use Aux 2 for your 2nd pump.

Very rarely will you need anymore than about 3ms with those coils, unless very high boost and methanol and/or a lot of water injection. Use the minimum you need to run without misfire. The manufacturer specifies a max long term dwell of no more than 40%, you can go higher than that for shorts bursts for less than say 10 seconds with a long cooling period between bursts. They saturate at about 9ms and are pulling about 20A at that. However it would only be ~3000HP promod cars with lambda at something like 0.6 that will need to use them at that level. With a rotary sparking every 360deg you will run into the duty cycle limitation if you run too much dwell. Duty cycle = (dwell x 100)/(60000/RPM). So for example 4.0ms dwell would be 47% dutycycle at 7000RPM. If you have a misfire anywhere then only increase the dwell in the specific RPM or MAP area that needs it.

Yeah I have heard the PMU overheats very quickly when PWM'ing fuel pumps and fans. I use that same SSD for closed loop fuel pump control on my injector flow bench and it has worked well for that. The pump would only be ~10A most of the time though. Another couple that I have seen used for fans is the Dorman 902-310 (jeep/dodge etc) or RY330K (Chrysler). I have no idea how much current they handle but they are designed for fans originally and I suspect fairly big fans on the vehicles they are from.

Yeah strange for a factory ecu to not have more error checking. The ignition timing would have been bouncing around all over the place.

Get a MCP601 in DIP format - it will have a part number MCP601-I/P or -E/P. Probably RS components will have them in small qty's. You can solder it to a bit of strip board or just solder the wires direct to the pins. Pin 7 is the 5V supply (grab from TPS). Pin 4 connect to sensor ground (again TPS). Pin 3 is the TPS signal from the throttle body. Short pin 2 & 6 together and connect a wire to the short, this is your buffered TPS signal output to the AT controller.