Couple of things on WRX107X

[Rozsko]

Just got my new G4X yesterday, but obviously it is locked and since I ordered it from Australia, and it was Friday night there, I have to wait till Monday to get the unlock codes .

Anyhow, I didn't stand to plug the ECU in and check a few things.

First of all, I have to tell you that I love the new approach to the trigger scope. It is awesome that it is now integrated into the logging functionality instead of saving as an image.

One thing in the base map I realized that the PS switch active state is high, but based on the old G4+ map I had I think it should be low.

The other things I would need some help with are the following:

• ANV4 and 5 are TGV sensors, which are disabled for me, yet the ECU reports low voltage fault codes on these. Is there a way I can clear these so they never come back?
• I also have a fault code#9 "settings underflow error", but the help file doesn't have #9 fault code listed. From an older G4+ thread I guess it might have something to do with my CAN settings (maybe in the divider/multiplier), but I am not sure.
• I also had a similar fault, but the opposite. Settings overflow error. It is not active right now, but I assume it might be due to the same subject as previously.
• When I was cranking the engine there were a few things I spotted, but not sure if that is dues to the ecu being locked yet, so again if you could confirm that would be good. VVT inlet bank 1 signal was inactive (thought in the trigger scope I could see the signal); there was no fuel pressure (but the key on prime was there)
• I was not able to calibrate the VVT as there were no offset degrees displayed in the runtime values while cranking. Is this because of the ECU being locked or because the engine needs to run?

I think that's it for now, but I'm sure there will more things.

Thanks

Béla

 

[Adamw]

1 hour ago, Rozsko said:

First of all, I have to tell you that I love the new approach to the trigger scope. It is awesome that it is now integrated into the logging functionality instead of saving as an image.

Yeah, it also captures the DI's and an AN Volt for MAP sensor sync and a lot more diagnostics info too which will make our life easier.  Even in the G4+ you could save as log though.

 

1 hour ago, Rozsko said:

One thing in the base map I realized that the PS switch active state is high, but based on the old G4+ map I had I think it should be low.

Yep, I think you're probably right on that one.  I can see it has a pull-up added to AN Volt 6 on this ecu so I would say most likely that it switches to ground when PS is active.  Please confirm when you have it running and I will get the base map corrected if it is wrong.

 

1 hour ago, Rozsko said:

ANV4 and 5 are TGV sensors, which are disabled for me, yet the ECU reports low voltage fault codes on these. Is there a way I can clear these so they never come back?

Just go to the TGV menu (aux outputs) and set bank 1 and bank 2 source to none should fix that.  Set the PWM output to none while you are there.

 

1 hour ago, Rozsko said:

• I also have a fault code#9 "settings underflow error", but the help file doesn't have #9 fault code listed. From an older G4+ thread I guess it might have something to do with my CAN settings (maybe in the divider/multiplier), but I am not sure.
• I also had a similar fault, but the opposite. Settings overflow error. It is not active right now, but I assume it might be due to the same subject as previously.

This should be fixed if you update to the latest firmware.

 

1 hour ago, Rozsko said:

When I was cranking the engine there were a few things I spotted, but not sure if that is dues to the ecu being locked yet, so again if you could confirm that would be good. VVT inlet bank 1 signal was inactive (thought in the trigger scope I could see the signal); there was no fuel pressure (but the key on prime was there)

ECU being locked will prevent the triggers working (I think, havent actually tried in the g4X myself), but most other stuff should work.  I suspect the VVT signal inactive is because of the lock but none of the VVT will work until it is running anyway.  Fuel pressure will only work if you have fitted a fuel pressure sensor.

 

1 hour ago, Rozsko said:

I was not able to calibrate the VVT as there were no offset degrees displayed in the runtime values while cranking. Is this because of the ECU being locked or because the engine needs to run?

Engine will need to be running.  

 

[Rozsko]

Thanks a lot for your quick reply Adam! Really appreciate it!

3 hours ago, Adamw said:

Yeah, it also captures the DI's and an AN Volt for MAP sensor sync and a lot more diagnostics info too which will make our life easier.  Even in the G4+ you could save as log though.

Correct, you could save as log in the G4+, yet this is much better.

 

3 hours ago, Adamw said:

Please confirm when you have it running and I will get the base map corrected if it is wrong.

Sure thing.

3 hours ago, Adamw said:

Just go to the TGV menu (aux outputs) and set bank 1 and bank 2 source to none should fix that.  Set the PWM output to none while you are there.

First thought when I looked at your suggestion: Am I really that dumb that I did not do that??? Fortunately not. There is no TGV option for me in the ECU config and the ANV4-5 do not have any function assigned. Firmware update?

3 hours ago, Adamw said:

  Fuel pressure will only work if you have fitted a fuel pressure sensor.

I do have a sensor and the gauge itself was showing the 3bar on initial key on pump prime, and was showing 0.00 when cranking.

[Rozsko]

Just updated the firmware.

After clearing the fault codes, it seems the ANV4-5 low voltage errors did not come back (and now I have the TGV option in the menu) and the Underflow error is also gone.

However the overflow error (code#8) came back.

Any more suggestion on that?

Thanks

Béla

[Rozsko]

Played a little more the afternoon and finally connected the rotary switch I mounted earlier.

While doing that, I spotted a small glitch either in the software or in the help file. There is no Label field, but the help says there should be one.

Also wanted to ask about the wiring of the rotary switch. Is it ok if I used the sensor ground on the expansion connector to ground the switch? Or should I just use a chassis ground to do that?

 

When I had my G4+ previously, I once realized that the fuel temp value was -34 C, so I set it inactive/off. At that time I did not have the time to check if there is actually a sensor in the tank or not, but now I did and there is no sensor, and the pin on the ECU (B17) has no wire either.

I am not sure if this is an EDM specific thing or not, but if it is, I guess it would be nice if you could mention this somewhere in the help wiring or pinout functions, as I am sure there is a big difference in the calculated pulse width with -34 vs +20 C.

 

[Adamw]

17 hours ago, Rozsko said:

Just updated the firmware.

After clearing the fault codes, it seems the ANV4-5 low voltage errors did not come back (and now I have the TGV option in the menu) and the Underflow error is also gone.

However the overflow error (code#8) came back.

Any more suggestion on that?

I might call upon @Vaughan to reply on this one, he may know more than me.  I thought this should be fixed in 6.16.27 and I havent seen it on my car.

 

12 hours ago, Rozsko said:

Also wanted to ask about the wiring of the rotary switch. Is it ok if I used the sensor ground on the expansion connector to ground the switch? Or should I just use a chassis ground to do that?

Senor ground is suggested for an analog input reference.  

 

12 hours ago, Rozsko said:

Played a little more the afternoon and finally connected the rotary switch I mounted earlier.

While doing that, I spotted a small glitch either in the software or in the help file. There is no Label field, but the help says there should be one.

I might ask @Vaughan to get that on the fix list too.

[Rozsko]

48 minutes ago, Adamw said:

Senor ground is suggested for an analog input reference.  

Then it is good, as I used both the gnd and +5V from the expansion connector and it feeds ANV11.

Thanks

[Vaughan]

12 hours ago, Adamw said:

On 6/7/2020 at 3:03 AM, Rozsko said:

Just updated the firmware.

After clearing the fault codes, it seems the ANV4-5 low voltage errors did not come back (and now I have the TGV option in the menu) and the Underflow error is also gone.

However the overflow error (code#8) came back.

Any more suggestion on that?

I might call upon @Vaughan to reply on this one, he may know more than me.  I thought this should be fixed in 6.16.27 and I havent seen it on my car.

TGV was only added in a more recent version, make sure you are using the sample map from the newest release (can use the compare function to see changes) as there are some things like the 4D fuel table which is TGV based. Shouldn't be able to get Fault 8 or 9 in the newest release as the code to set it as a fault was removed so please clear it again and see if it comes back (could theoretically show up when loading older base maps as the fault codes are stored in them).

12 hours ago, Adamw said:

Played a little more the afternoon and finally connected the rotary switch I mounted earlier.

While doing that, I spotted a small glitch either in the software or in the help file. There is no Label field, but the help says there should be one.

just had a look and yup no labels in the release but they are in the newer stuff and so will be in next release sorry

[Rozsko]

On 6/7/2020 at 10:36 PM, Vaughan said:

TGV was only added in a more recent version, make sure you are using the sample map from the newest release (can use the compare function to see changes) as there are some things like the 4D fuel table which is TGV based. Shouldn't be able to get Fault 8 or 9 in the newest release as the code to set it as a fault was removed so please clear it again and see if it comes back (could theoretically show up when loading older base maps as the fault codes are stored in them).

just had a look and yup no labels in the release but they are in the newer stuff and so will be in next release sorry

Thanks a lot. The error code is gone now. Not sure what it was or why it came back the first time, but now it is gone.

[Rozsko]

On the other hand, I finally got the unlock code, so I fired up the car.

First time the idle was horrible. Actually didn't even want to idle and what was more weird is that the base position table had no effect on the idle speed. Neither in open or closed loop.

This one I figured out after a few minutes. Not sure how or when, but the actuator was set to None, which is strange as in the first map I created based on the G4+ map I had in the car previously has this set to Ethrottle. Anyhow, now it is set to properly, but the closed loop and ignition idle controls are still bad. The car idles very nicely in open loop, but in closed loop (no ignition control) the PDI values have only Integral value and even that is very small and 0 P and 0 D, while the in the ignition idle control it has huge fluctuation in P and D, but I is always 0. For the ignition control I tried to pull down the P and D values to 5 and 10 respectively (there seems to be no option to specify I and this pretty much is in line with how the logged values look like), but it did not have any effect. It seems the the ignition control is just jumping between the min and max clamps, without any smoothness. For the closed loop mode, I just realized that the integral value can be changed in the actuator setup, so I will try that later, but here again in line with how the PID values are logged, there seems to be no option to change the P and D portion of the control. Not sure if these are by design (and should play a bit more with the PID numbers) or a fault, but as is it stands right now the best option is to go open loop. I attached a log file (with some markers and section to help with what's going on) and my current map in case you can take a look.

What I also noticed is that the base position table requires very different values (with the same VE numbers) compared to the G4+.

I also attach an other log file where the map itself wasn't changed at all, yet the idle was hugely erratic and the overrun fuel cut was jumping in (I guess) to decrease RPM. Without any change, at the next startup this has gone.

Closed loop, open loop, ignition control: https://1drv.ms/u/s!Ale4oyMCOgLThMlGQbqbtgRpIkRdBA?e=B7Lgrx

Cyclic idle with overrun fuel cut: https://1drv.ms/u/s!Ale4oyMCOgLThMlH4xi9oIFymOmOaw?e=lXuKgU

 

An other issue I experienced is that (at least) some new features, like torque management has no help. Is this an issue with my install or these help sections are really not there yet?

I still need to confirm the PS switch active state.

Thanks,

Béla

 

Initital startup.pclx

[Rozsko]

Played a little more this afternoon with the idle PID, and if set any of these above 2, then the idle just gets cycling. I ended up using 0.5 P, 0.2 D and 0.1 I. With such low values it is probably better then simply open loop, but as these values are far afar away from the default map values, I am really not sure if this is good, or I am missing something very obvious.

As I was looking through the latest log, I ended up checking the accel enrichment and I think the help and the actual software layout are not matching.

Accel sens is not a single parameter any more, but an individual table. Whatever is written in the help makes sense, but for me it is confusing too, or I should rather say not clear enough. Is this table setup properly? I mean should this really be based on RPM and not MAP? And shouldn't the values in this table gradually decrease as RPM/MAP increases?

If I am talking too much newbie nonsense and this is self explanatory for others, let me know and I will shut my "mouth".

Thanks,

Béla

[Vaughan]

A few notes:

• In G4X set your ethrottle target table to 0% at 0% accel pedal and set the idle base positions to the appropriate tps targets for idle. (G4X uses a clever bit of code to combine ethrottle and idle targets to ensure on exit of idle that the position doesn't dip).
• Closed loop idle only uses Integral control, Ignition idle uses Proportional and Derivative control
• Torque management help hasn't been done yet and that code isn't officially finished yet, I plan to test it more (has already had a decent amount of testing) and write the help for it either this week or next week.
• Accel sens was initially a single value but it was requested that it be made into a table to reduce it's effect "at the top end". I would recommend turning on asynchronous injection (under fuel main) and then seeing if you actually need accel enrichment, I found with my car that asynchronous injection is faster and more precise than a more traditional accel fuel. I'll make sure this help is corrected if I haven't already done so.
• Have a play with that accel fuel axis to see what suits your car the best, any feedback on this is much appreciated.

I've attached a copy of the basemap from my 2GR-FE, It should be a half decent example of an ethrottle idle setup although I am not using closed loop idle, just open loop and ignition.

vaughans_mr2.pclx

[Adamw]

2 hours ago, Rozsko said:

Played a little more this afternoon with the idle PID, and if set any of these above 2, then the idle just gets cycling. I ended up using 0.5 P, 0.2 D and 0.1 I. With such low values it is probably better then simply open loop, but as these values are far afar away from the default map values, I am really not sure if this is good, or I am missing something very obvious.

No that is about right.  Typical P I have found around 1.0-2.0, D normally not needed (zero).  Intergral around 0.3-0.5.  I have asked the firmware engineer if these values can be set up as the defaults but Im not sure if it is possible in yet.

 

1 hour ago, Vaughan said:

Accel sens was initially a single value but it was requested that it be made into a table to reduce it's effect "at the top end". I would recommend turning on asynchronous injection (under fuel main) and then seeing if you actually need accel enrichment, I found with my car that asynchronous injection is faster and more precise than a more traditional accel fuel. I'll make sure this help is corrected if I haven't already done so.

I have found this too, on the two cars I have tried it on recently, enable Async, set min Async injection to 0.5ms, no or very little accel enrichment needed for normal driving.

[Rozsko]

Thank you guys.

Made a few changes to the ethrottle target (but I think I will need to undo some of it as at low APS% now the pedal is too gentle and I don't like that feel) and a little to the idle PID settings. I don't think it is perfect, but not sure if I can get it any better.

This is what I ended up with:

 

I also confirmed the PS switch active sate is as we talked about previously.

I did some cruise test and made some changes to fueling, but next week I will go on the dyno. This is going to be my first dyno session, so I am very excited about that.

Will certainly play with the async injection and accel sense.

Thanks again for all the help.

[Vaughan]

with the E-Throttle target keep 0% tps for 0%aps but make it more linear from there to make it more aggressive, I found with the big v6 I need to put in a decent curve to give me better throttle control down low

[Adamw]

1 hour ago, Rozsko said:

Made a few changes to the ethrottle target (but I think I will need to undo some of it as at low APS% now the pedal is too gentle and I don't like that feel) and a little to the idle PID settings. I don't think it is perfect, but not sure if I can get it any better.

This is what I ended up with:

 

I would have ignition target lower than that, typically around 10.  At 25 you are already sitting near MBT so there is less ability for the ECU to quickly increase torque when it is needed.  You may find the MAP lockout will need to be a bit higher so that it works reliably during coldstart - many engines when cold will sit above 50kpa after start for a while.  An actuator deadband around 40 is usually better for E-throttle to stop it dithering.

 

[Rozsko]

11 hours ago, Vaughan said:

with the E-Throttle target keep 0% tps for 0%aps but make it more linear from there to make it more aggressive, I found with the big v6 I need to put in a decent curve to give me better throttle control down low

That's exactly what I mean. Thanks again.

 

9 hours ago, Adamw said:

I would have ignition target lower than that, typically around 10.  At 25 you are already sitting near MBT so there is less ability for the ECU to quickly increase torque when it is needed.

Initially (I mean with the G4+ yet) I had  it at 15 degrees, but the engine wasn't happy, it didn't sounds good at all. Will give it an other try with the G4X.

 

9 hours ago, Adamw said:

You may find the MAP lockout will need to be a bit higher so that it works reliably during coldstart - many engines when cold will sit above 50kpa after start for a while.  An actuator deadband around 40 is usually better for E-throttle to stop it dithering.

Thanks for the recommendation, will try and see.

[Rozsko]

On 6/11/2020 at 12:59 AM, Adamw said:

You may find the MAP lockout will need to be a bit higher so that it works reliably during coldstart - many engines when cold will sit above 50kpa after start for a while.  An actuator deadband around 40 is usually better for E-throttle to stop it dithering.

You were absolutely right. Thanks for the advice.

[Rozsko]

A new thing I realized in the base map:

Bsed on teh G4+ settings, I guess the vertical axes should be TPS.

[Adamw]

You can set the axes to whatever you like. But this table should normally be turned off for modelled fuel equation.  Charge temp based air density calculation is already part of the equation.

[Rozsko]

 

11 minutes ago, Adamw said:

You can set the axes to whatever you like. But this table should normally be turned off for modelled fuel equation.  Charge temp based air density calculation is already part of the equation.

Yep. I already did change that to TPS (I just mentioned this as I think it would be nice to correct this in the base map, so people don't overlook at it, like me so far), but why should this be turned off for modelled fuel equation? This is ignition trim and not fuel.

thx

[Vaughan]

11 hours ago, Rozsko said:

A new thing I realized in the base map:

Bsed on teh G4+ settings, I guess the vertical axes should be TPS.

My bad, part of the risk of trying to convert basemaps using a custom program. New release will have it corrected to the G4+ axes.

[Rozsko]

Sorry for bringing this back again, but did not want to open a new thread for this.

Did not pay much attention to this yet, but to me it seems the gear status runtime value is 0 all the time. Checked the gear ratio table is populated with default values, so I am not sure what's the problem.

Any idea?

Thanks,

Béla

[Adamw]

You need to calibrate the gear ratios.  Instructions are in the help file.

[Rozsko]

19 hours ago, Adamw said:

You need to calibrate the gear ratios.  Instructions are in the help file.

ok, thanks. Will give it a try. With the G4+ I wasn't able to do that. for some reason I got an error during calibration. Will see if I am more successful with the G4X.