Adamw

That's not one I have seen before. Did it get to the stage where it tells you to do a power cycle? Are you doing it on the bench or is it still connected to a car/loom? If it is still attached to a loom I would remove it from the car and try again with just power and ground pins connected only. Does it still show in windows device manager?

Slower cranking will cause the lower voltage to be generated by the crank sensor but coupled with the noise I think the low voltage in this case isn't just because of the slow cranking.

The main power supplies are meant to be 49 & 59, pin 45 is usually just the "ign switch" signal that controls the ECCS relay. Im not sure if your harness even has an ECCS relay but it looks like you have something connected to pin 16? It should work the way you have it as 49&59 are linked internally anyway. What voltage do you have on pin 59? Note you will also want 12V on pin 58 (doesnt matter if constant or switched) as this is used for the idle valve flyback. This is not related to your low battery voltage issue, just something you will need to fix for the idle to work properly.

Yeah the trigger waveform looks terrible in that scope. The voltage drops very low in some patches so the arming threshold is not met and there are spikes of noise all over it. Its difficult to give solid advice on any specific cause as there is a huge range of possibilities, but my thoughts are most likely a poor connection/high resistance somewhere along the path between ECU and crank sensor. The cam signal is a lot cleaner so I think it is more specific to the crank sensor rather than general engine grounding or ignition noise etc. It could also potentially be the crank sensor failing internally but that would be less likely.

You can use an AN Volt input if you fit an external pull-up resistor.

Although this doesnt change my sentiment that the boost control set up is poor, with the newly added info that "it has worked for years like that" then I would now add this thought: If it now needs 20% more DC to reach target than it has needed in the past then I would be looking for a hardware issue. A big leak etc. Why has the fuel tune changed so much if its been running for years?

I cant really understand what you have? Does the engine loom have the superseal connectors for a Haltech wire-in ECUU and you have made an adapter loom to connect it to a Link plug-in?

I would say the gain was either very small or zero when that log was captured. Start with something like this:

I think you have clicked capture before cranking in that scope, make sure you only click capture when the engine is cranking.

The stall issue is mostly because the APS and TPS hasnt been calibrated so it is never enabling idle control. Some of your idle control settings are a bit odd also, so I would copy the settings out of our WRX11 base map as a better starting point - especially the dashpot settings, integral gain, map lockout and other offsets. Your idle base position table will then need tuning. Fuel is also quite rich so this will need to be closer before idle and drivability will be good. The re-start issue looks like it is not happy with something with the trigger waveform but that wasn't captured in the scope, you might have to try a couple more scope captures next time the hard start issue occurs to see if we can capture the problem. There is some odd ignition noise in that scope that may be the issue, but it wasn't bad enough to cause an issue when that scope was captured. What coils does it have? Are spark plugs resistor type?

The crank sensor is wired incorrect polarity, the cam sensor looks ok. Swap the +/- wires at the crank sensor plug.

Ignition test is permanently set to Ign 1 in your map, the triggers are ignored in test mode. I didnt even know that setting could be stored permanently on like that, but I think you will find your trigger will work correctly once you turn the ign test off.

So is the MAP reading correctly now? It sounds like you still have an issue if you need a master fuel of 7ms for 650cc injectors, a master of 7ms would be more typical for ~ 1200-1600cc injectors.

There is no aux 9 pin available on that ecu, Probably aux 6 or 7 would be the options for sig 2, you could move the FP relay or CE light to one of the auxes on the expansion.

The boost control set up is not even close to correct, you are probably better to put it back to open loop mode until you can get someone experienced to tune it. The smallest value in the target boost table should be no smaller than the lowest boost the system can make. Having a target of zero in there means you are already above target as soon as boost control enables, you will get no stage 1, and often no stage 2. The wastegate DC table is a long way off. For example in the first pull you needed 51%DC to get close to target boost at 5500RPM, but your wastegate DC table only has 36% DC in it for 5500RPM. At 6500RPM it needed 42% to get close to target, yet DC table only has 24%. Not relevant to boost control, but be aware your lambda at high load is very lean, you will likely damage the engine if you continue to drive it like that.

Your master fuel will need to be up around 20-30ms for injectors that small.

It sounds like you are connected to the wrong sensor. The sensor for the ecu should have a white or blue plastic housing and it uses the standard bosch NTC calibration (2.5Kohm @ 20°C). The black one is for the dash temp gauge and that is 5Kohm @ 25°C, this uses the Hella 6PT calibration.

There is only a single knock pin available on that ecu, pin 3O. You will only be able to run 1 knock sensor, connect one side to the factory knock sensor wire, the other side to ground, the factory sensor used chassis ground so that will likely be fine for the bosch too.

Attach a log of a start attempt and a copy of your tune. What injectors do you have?

Attached below are DBC files for the Link Generic Dash stream and also for the built-in Dash2Pro stream. Set the CAN ID to 1000 in the ecu setup for either. I tried loading both into your dyno and it looks like they have a bug which doesnt import the generic dash one properly so this wont be useful for you immediately but if you pass it on to Jostein he may be able to fix the dbc importer or manually hard code it. The Generic dash stream uses what is called a compound message which many DBC parsers dont seem to interpret correctly. The Dash2Pro stream is a more basic sequential format. Dash2Pro probably has most of the basics for dyno use like RPM, Lambda, MGP, temps & Ignition angle, Generic dash is a bit more comprehensive with cam angles, individual wheel speeds etc. Link Dash2Pro.dbc Link Generic Dash.dbc

For the MAP sensor you just need to set AN Volt 5 to MAP which is the only option. It should read about the same as BAP with the engine off and if you suck on the hose as hard as you can you should be able to see a change of 10kpa or so. There is no triggerscope in the G4+ Atom or Monsoon, you really need a standalone scope to diagnose a trigger issue if there is one. I would start by reducing the trig 1 arming threshold to 2V at 1000RPM.

I would expect the S15 map would do that.

Im suspicious of the MAP reading, there is no change in MAP when the engine starts, the throttle is open quite a bit but I thought we would still see some vacuum. In your first post it says it is wired to AN Volt 1, but in your map it is set to AN Volt 6, was this changed?.

It is overrun fuel cut as your TPS is less than 0.8% and your RPM is about the deactivation RPM.

I will likely have something suitable. Yourdyno, Perek or something else?