Over run fuel cut clarification

[b3tuning]

When Over run fuel cut  (decel fuel cut) is active, is the fuel pulse reduced by a certain percentage or clamped by the minimum pulse width?  Or is injection completely disabled or ramped down with ignition to 0 duty cycle?  Or is the logger not calculating duty cycle and injector pulse width properly during over run?

Have noticed via logs or while driving with my laptop riding shotgun that when over run fuel cut is active that fuel is still being injected, Injector duty cycle only drops to ~3% on average.  Wideband does show going lean (at least to 16:1 afr, 1.10 lamba petrol)

Curious if I'm missing out on some engine braking and fuel economy, or a small bug in the logger or modeled fueling?

Will post my current cal and log if needed.  G4+ Xtreme (red single CAN)

[Davidv]

Instead of duty cycle have a look at the log of Injector Actual Pulsewidth

When the injector cuts you'll see a delay in the reading changing which I believe is for two reasons: 

1. The amount of gas coming past the wideband sensor under these conditions is very low so its slower to respond

2. There is still the fuel film build up on the port walls which is coming into the cylinder, although the amount of fuel is low the amount of air is very low too when the throttle is shut.

But I see a delay in the readings here, and look how long it takes to register again once the injectors turn back on.

 

Edited September 20, 2017 by Davidv

[b3tuning]

Instead of duty cycle have a look at the log of Injector Actual Pulsewidth

When the injector cuts you'll see a delay in the reading changing for two reasons: 

1. The amount of gas coming past the wideband sensor under these conditions is very low so its slower to respond

2. There is still the fuel film build up on the port walls which is coming into the cylinder, although the amount of fuel is low the amount of air is very low too when the throttle is shut.

But I see a delay in the readings here, and look how long it takes to register again once the injectors turn back on.

 

I get the increased manifold vacuum pulling the fuel film off the walls, but at least in my case, the injectors are still pulsing during over run and there is a measurable fuel mass... I've included a screen shot showing that over run is active, and actual injector pw is NOT 0, and duty cycle is ~3%.  This is ALWAYS the case for me during over run.

 

[Davidv]

Weird! 

Can you post up the fuel cut decel setup page and the overrun deactivation table? 

But does seem to look like a bug. 

I'm running a red G4+ xtreme as well, latest firmware version - You know what version you are on?

Edited September 20, 2017 by Davidv

[b3tuning]

Weird! 

Can you post up the fuel cut decel setup page and the overrun deactivation table? 

Thanks David for chiming in... I've included screen shots of the tables along with my target idle table to show that there is no conflict...

I agree this is weird, and hopefully easily resolved. 

The only thing that I can possibly think of that can be a contributor to this bug is that I do have an Xtreme affected by the cold start hardware bug.  As I am in the U.S. and this is my daily, I have not been able to send my Xtreme in for the update...  surprised I am the only one to notice this... Can Adam or Simon chime in that this could be related to the cold start hardware issue?

[b3tuning]

I'm running a red G4+ xtreme as well, latest firmware version - You know what version you are on?

Sorry didn't see your edit.

I'm actually on release 5.6.4.3229

I had planned on updating to the latest firmware when it was released 5.6.5.3338, but when it was rolled out there were some posts about a few bugs so I held off..   3338b was released shortly after but still didn't see anything in the notes that my setup benefited from.

I'm on PCB version 1.4 which only has 1 CAN bus, and is affected by the cold start bug.

[Steve]

Edit... Never mind

[Adamw]

I just tested this on a simulator this morning.  There does seem to be a bug that effects the displayed actual and effective PW during overrun fuel cut.  My actual goes close to zero but not exactly zero.  My effective holds on to what ever value was last used when overrun is activated.

The good news is the overrun fuel cut does work as expected - injection is competely cut when overrun is active.

 

I'm on PCB version 1.4 which only has 1 CAN bus, and is affected by the cold start bug.

Can you expand on this some more - no one here at Link remembers any hardware issues in G4+ V1.4?

[iliasfyntanidis]

I am facing a problem with overrun fuel cut. If I set it to 2000(by default is set at 2000 rpm) rpm when engine is in warm up to temp everything works as it should. The problem is that the car is jerking when the engine speed hits 2000 rpm. So I have to press the clutch. Plus I like to have it lower than 2000 rpm, for example 1400.There isn't any conflict with idle target rpm which is set at 1200 rpm. When I set to 1700~1500 rpm for example the following strange thing happens. The car either keeps the idle at 1500~1600 rpm or the idle goes lean 21 AFR and finally dies. Any suggestions? Does the rpm lock out on idle speed control (closed loop control) has any effect? 

 

Basically it does not like at all overrun fuel cut lower than 2000 rpm. 

 

[Adamw]

48 minutes ago, iliasfyntanidis said:

I am facing a problem with overrun fuel cut. If I set it to 2000(by default is set at 2000 rpm) rpm when engine is in warm up to temp everything works as it should. The problem is that the car is jerking when the engine speed hits 2000 rpm. So I have to press the clutch. Plus I like to have it lower than 2000 rpm, for example 1400.There isn't any conflict with idle target rpm which is set at 1200 rpm. When I set to 1700~1500 rpm for example the following strange thing happens. The car either keeps the idle at 1500~1600 rpm or the idle goes lean 21 AFR and finally dies. Any suggestions? Does the rpm lock out on idle speed control (closed loop control) has any effect? 

 

Basically it does not like at all overrun fuel cut lower than 2000 rpm. 

 

Log and map please.

[iliasfyntanidis]

This my latest and best map.The log is too big to upload.About 1.7 mb.I can send it via email.

I have done some improvements.Have adjusted idle from Open Loop,then proceed to Closed Loop etc etc...Its rock solid.Never understood why someone has to go through Open Loop and then to Closed Loop,probably the enginners now better....

I want to ask the following because i think that many things happen in the backround like idle speed control where the user first has to set the Open Loop and later the Closed Loop.

How does rpm lockout on Closed Loop idle speed control,max and min clamp,idle target RPM Table and Overrun Deactivation Table are related?

 

Regards

 

Mazda Miata Vi-PEC i88 Final Map - Link Ignition Map.pclr

[Adamw]

2 hours ago, iliasfyntanidis said:

The log is too big to upload.About 1.7 mb.I can send it via email.

Share it using google drive, dropbox, onedrive etc.

[iliasfyntanidis]

https://drive.google.com/file/d/1CIIB3g6LcraDnki_bBmCrd309C6IHgFW/view?usp=drivesdk

Despite the overrun which I might have fixed at the very end of the file on parking manoeuvres the engine stalls and dies. This happens many times and I am breaking my head to figure out why. 

[Steve]

Ilias. You seem to have some groundissues affecting your sensors. Look at the ECT as an example

Also your 4 thoot triggerwheel doesnt seem to play nice. Also see the warmup enrichment going active again because of the sensor readings.

 

Are your fuelpressuresensor hooked up? The sensor is static.

 

Your lambdasensor seems to of line now and then. Like where the marker is in the next pic.

 

Also idle target is set to 1200rpm and your idle rpm lockout is set to 800rpm. Which ads up to 2000 rpm

[iliasfyntanidis]

What would be your suggestion about idle rpm lockout Steve? 

The fuel pressure sensor is not working by the way. 

 

[Steve]

Try lowering the lockout rpm to say 400 and see if it changes anything.

But the sensor readings would be high on my prioritylist to fix

[iliasfyntanidis]

Steve to be honest I don't see the situation/problem you are describing. The values look good to me. 

By the way the RPM lockout was set previously at 400 rpm. By default. The problem is the same. 

 

[Steve]

Look at the coolant temp line in one my screenshots. Thats not normal for coolant to vary that much, that fast. In the last one at the bottom you can see the warmup enrichment come on line giving up to 10% extra fuel because of this

[cj]

1 hour ago, Steve said:

Look at the coolant temp line in one my screenshots. Thats not normal for coolant to vary that much, that fast. In the last one at the bottom you can see the warmup enrichment come on line giving up to 10% extra fuel because of this

In this particular circumstance I have seen this happen before. The car is warm, sits powered off for some time, and while the water in the radiator cools down quickly (especially if you have the key on and therfore probably the fan running), the water in the engine absorbs the heat from the hot engine however. If your sensor is on the way "out" of the engine, what you see is immedaitely on cranking, the temp increases 4 or 5 deg as the hot water in the engine goes out past the sensor, then it drop by 10 or 20 as the cool water from the radiator makes it way through the engine to the ECT sensor. 

There are however some weird things going on the the confguration of the sensors...

There are 2x coolant temp sensors defined, and both seem to be working (they are within a couple deg of each other but sometimes they diverge, so they are clearly not 2x signals from the same sensor) The Temp3 values align with ECT so it looks like temp4 is just being ignored, but having 2x sensors defined as the same things is asking for trouble. Same thing is configured for IAT - Temp1 is being used, and temp2 is being ignored. If these extras are just for logging you should define them as GP Temp and just label them, rather than telling the ECU you have 2x IAT's and 2x ECT sensors.

There is something odd going on with the oil pressure sensor. It reads ~115 kpa with the engine off. Is this an absolute pressure sensor rather than a gauge sensor? if so, you should offset the calibration by 100kpa so it reads ~0 when the engine is off. Otherwise any safety's you set up will be wrong by ~100kpa.

8 hours ago, iliasfyntanidis said:

The fuel pressure sensor is not working by the way. 

Is it the same type of sensor as the oil pressure, and its wired up but not installed in the fuel line? I think you have the same gauge vs absolute thing going on here too. 120kpa is a bit co-incidental of a value for it to stick on for no reason. If so, you need to do the same thing with offsetting the calibration.

The reason you need to set up open loop before closed loop for most systems is that the closed loop logic is something like "use the open loop value, +/- a small correction based on feedback from sensors". If the open loop values are completely off, the "start" point for any closed loop control is way off and it may not be allowed to go that far away from baseline, and it will take some time to learn this "error" each time you start up (or return to idle for idle control) so it will run badly for a bit until it figures out what realistic values should be for the various control systems.

The various "lockout" values only apply to the system they are set for. Idle lockouts control when the idle systems kick in. They have no bearing on other systems (ie overrun fuel cut). If you add the value "overrun fuel cut status" to a log view you can see when this kicks in. In your log about -5:00 you can see if coming on and off at the upper end of the idle range, but it is giving you the fuel back below ~1400rpm. Any stalling behaviour is caused by soemthing else as fuel isnt being cut below 1400rpm.

I'm suspicsou you might have an intake leak. Your MAP is at ~95kpa at -2:30 in your log. I'm guess you have ITB's or huge cams but thats still very low vacuum. If you set it to open loop idle does it run OK? It seems off that the idle solenoid can be reporting 60% DC and yet you still only have 600rpm and bascially atmospheric MAP. Is the throttle screw on the TPS set correctly so you have a slightly low but passable idle with the idle solenoid disconnected?

 

 

[Steve]

Looks like you are right. Just seemed odd for ECT to drop 15 degrees over 10 seconds. But i dont see abrupt changes when he cranks it (which usually shows ground issues like a sore thumb), So it seems to be alright after all. My bad.

[iliasfyntanidis]

Cj 

I have 2 ECT and 2 IAT involved. They 4 read the same but only the 2 of them take part in the ECU.The others are for future use with my Rotrex. 

The camshafts are big about 60° of overlap thus I have my load axis set to TPS. 

The sensors are PX3AN2BS100PAAAX sealed gage both of them. 

The idle stop screw is set correctly. 

Basically I just got back from a small ride and noticed that overrun issues are solved. The problem that I have now is at tight maneuver and light throttle openings where the RPM drop and the engine stalls. 

[Steve]

You could most probably benefit from better resolution at small throttleopenings in your fuelmap. Instead of 0 ,5 ,10 and so on you should space them closer down low. Like 0,  2,  4, 6, 10, 15 and so on. There are BIG changes in the airflow at small openings requiering finer adjustments of the fueling

[cj]

7 hours ago, iliasfyntanidis said:

The sensors are PX3AN2BS100PAAAX sealed gage both of them. 

In that case you should set the top/B value in the calibration to 100psi rather than 114.7 as it is now. That will be skewing the entire range. I'd guess you set them from the honeywell spec sheet that shows the sealed gauge pressure graphs tops out at 114.7? the X axis on the graph is Absolute pressure, so 114.7 abs is 100gauge - ie its the same range as the 100psi absolute sensor which is 0-100psi absolute, but with the entire scale moved 14.7psi to the right to make it relative(gauge), so you still read it as 0-100.

Rather than us all making educated guesses, can you try things in a bit of a planned order

1) turn off fuel overrun cutout. I dont think this is active at the time you have a problem but you say it was the last this you enabled. Do you still have a problem?

2) unplug the connector for the ISC solenoid and check that your idle is still ok. (reason for this is that at one point you have ~700rpm idle and the ISC reports as fully open)

What injectors do you have and where did you get the deadtimes/short pulse adders for them? at idle you are running ~1ms effective pulse which might well be into the non-linear range. Your AFR is still showing at 0.9 which suggests fuelling is ok, but somehow you've got an idle valve maxed out, seemingly good AFR's, and still only 600rpm idle. Something isnt lining up.

[iliasfyntanidis]

@cj

Thank you very much for the answers and usefull questions.

Please tell me how would you setup the calibration table depending the following image.Would you set 0.5volt as 0 psi and 4.5volt as 100psi?My error is that i have set 4.5 volt as 114.7 psi correct?

Honeywell PX3 by Ηλίας Φυντανίδης, on Flickr

I am using Siemens Deka 550cc fuel injectors.The dead times are the following

  6              7          8            9         10            11       12          13          14        15

4.999    2.024    1.33    1.032    0.883    0.685    0.586    0.487    0.387    0.338

I will do the rest as suggested.By the way are you a tuner?Because i see you are always spot on.....

 

Thank you for you time and effort.
 

 

[cj]

Your picture hasn't come thorough, but assuming its the same as the px3 datasheet here. Set 0.5v = 0psi, 4.5v = 100psi. Because you bough a gauge sensor it does the atmospheric offset for you.

Essentially if that graph had an X axis of "pressure - gauge", the absolute line + numbers would be correct, and the 14.7psi atmospheric pressure offset would be invisible.

Those dead times don't match what is configured in your map. Changing these will require a bit of subsequent changing of the fuel table numbers, especially in the lower flow areas but should make a reasonable difference to your idle assuming they are correct.