Issues with v1.5, Engine States and AEMnet CANbus - bug?

David_G · Post by **David_G** » Mon Mar 20, 2017 12:22 pm

Hi,

I've got a few issues with my MS3x since the update to v1.5.0

My problem is that the transition from overrun to CL Idle is not working as good as it did with v1.4.1 because there is only 1 MAP threshold available anymore.
I have set the acceleration thresholds (what is the fast accel/deccel value used for by the way?) and CL Idle is recognized correctly.

I used to have both of the MAP thresholds in v1.4.1 set to 28.0 (CL Idle) and 27.9 (overrun).
That way CL Idle and overrun could never ever activate at the same time.

Bildschirmfoto 2017-03-20 um 18.32.43.jpg

Now with v1.5.0 there is only one MAP setting left and I thought that this is simply the switch point for CL Idle and overrun. So overrun only being active below that MAP value and CL Idle only being active above that MAP value.
But obviously that is'nt the case and CL Idle can also activate below that MAP value as long as RPMdot is between the slow accel and deccel values, so RPM being more or less constant.

This gets problematic when being in overrun in a high gear and going only slightly downhill which can lead to constant RPMs (so RPMdot being between the slow accel/deccel values) while being in overrun.
I had it a few times now, that CL Idle activates in such a case even though overrun is still active. That of course led to closing my Idle valve which keeps the engine in overrun much longer because MAP is suddenly a lot lower and that made my engine die once i slowed down further and pressed the clutch.

That brings me to the next problem:
The "Highest and Lowest RPM to Re-engage Fuel" fields in the overrun Fuel Cut windows don't seem to have any effect anymore. During overrun, RPMs fall as long as MAP is below the "overrun MAP threshold" value in the Engine States window. If those would work properly, the above problem might not even appear, I don't know.

And a third thing unrelated to the above:
I have a AEM X-series wideband and I am using the new AEMnet CANbus protocol with that. From time to time when being in overrun, the display in TS does show the exact opposite from what is actually going on.
Instead of showing full lean (AFR 25.5), like the AEM gauge, TS is displaying full rich (AFR 4.0). This only happens sometime. In general, the displayed values in TS match the ones from the AEM gauge.

IMG_7442.jpg

I would like to have the same good CL Idle/overrun transition as in v1.4.1 but with the current v1.5.0 this seems impossible. Or am I doing something wrong and the engine states are supposed to be tuned totally different compared to the previous firmware?

Thanks, David

David_G · Post by **David_G** » Tue Mar 21, 2017 2:41 pm

Ok I tracked it down to 3 problems:

1)
"Highest and Lowest RPM to Re-engage Fuel" fields in the overrun Fuel Cut window do not have any effect at all. Overrun only stops when MAP is above the MAP threshold in the engine states window.

2)
CL Idle does not deactivate even after a "slow acceleration" is detected for example when engaging the clutch without pressing the throttle.
The engine state changes from "Idle" to "slow acceleration", but CL Idle is still staying active, no matter what the engine state indicators are showing.
Once CL idle is active, the only way it deactivates again is by pressing the throttle so TPS is above the TPS threshold in the engine states window

3)
AFR gauge in TS is freaking out when the AEM X-series gauge (connected via CANbus) is showing full lean in a overrun condition.
Either the TS gauge is showing full rich instead of full lean (see pic above) or it is rapidly flickering between 1.2 and 23.1 (see picture below)

Bildschirmfoto 2017-03-21 um 22.36.19.jpg

To me that 3 issues seem like a bug in the v1.5.0 firmware because I can't imagine how this can be right...?
As written above, points 1) and 2) were working flawlessly in v 1.4.1

dr.mike · Post by **dr.mike** » Tue Mar 21, 2017 9:47 pm

AEMNet uses a 16-bit unsigned value for lambda. 0.0000 to 6.5535
Lambda values over 3.2767( super lean ) might be misinterpreted if they are being read as a signed value ?

Post by **jsmcortina** » Wed Mar 22, 2017 2:55 am

David_G wrote:And a third thing unrelated to the above:
I have a AEM X-series wideband and I am using the new AEMnet CANbus protocol with that. From time to time when being in overrun, the display in TS does show the exact opposite from what is actually going on.

Please start a new topic with MSQ/MSL.

James

David_G · Post by **David_G** » Wed Mar 22, 2017 2:57 am

Will do

Any inputs on the 2 other, more important issues?

Post by **muythaibxr** » Wed Mar 22, 2017 3:44 am

The behavior listed in your number 2 item is fixed in the next version.

David_G · Post by **David_G** » Wed Mar 22, 2017 8:19 am

Ok great! Is there a known date when the next version will be published?

Issue 1) has at least the same importance to me because without that, RPMs drop far too much during overrun which makes the engine stall.
To overcome this for the moment, I set the MAP threshold relatively low, but this again lets CL-Idle activate too early which makes the Idle valve close. This lets MAP drop again and re-activates overrun fuel cut (CL-Idle still active). If I then press the clutch, the engine just dies straight away.

Please let me know if I should test any specific settings or behaviours to confirm that this is a firmware problem and not one with my settings.

Post by **muythaibxr** » Thu Mar 23, 2017 3:31 am

I will look into #1.

David_G · Post by **David_G** » Thu Mar 23, 2017 12:46 pm

Thanks, I appreciate it!

Post by **muythaibxr** » Sun Mar 26, 2017 7:36 am

Looking into it today.

Post by **muythaibxr** » Sun Mar 26, 2017 8:05 am

I think I see the problem. When I integrated the engine states code into the overrun fuel cut code, I didn't allow the code to break out of overrun fuel cut if the engine state was overrun, even if the RPM was below where it should be to break it out. That is fixed now on my laptop. Just need to test and check in.

It'll be in our next code drop. I'm not sure when that will be though.

Ken

David_G · Post by **David_G** » Sun Mar 26, 2017 10:54 am

Okay, Great, Thanks for looking into it!

Why is it that overrun- and Idle-engine state can both be active at the same time? I thought that was one of the main advantages of introducing the engine states, no?

And why is it that the MAP threshold in the engine states windows is used for Idle AND overrun instead of only for Idle?
Don't the 2 RPM values in the overrun fuel cut window make the MAP threshold for overrun fuel cut unnecessary?
What are the thoughts behind that?

Post by **muythaibxr** » Mon Mar 27, 2017 3:34 am

Overrun state and Idle state can't be active at the same time.

Overrun fuel cut and overrun engine state are not the same thing.

The MAP threshold is there to keep idle from engaging in certain situations where overrun is engaged.

David_G · Post by **David_G** » Mon Mar 27, 2017 12:12 pm

True, sorry, my mistake!

But overrun fuel cut and Cl-Idle still can be active at the same time (see screenshot)
That's what I wanted to say.

This happens if MAP falls below the MAP threshold again after CL-Idle has been engaged, which is closing the Idle valve because RPMs are of course too high for Idle.
This makes overrun fuel cut re-activate even though CL-Idle does not deactivate if the throttle has'nt been touched.
It just surprised me that this is even possible because I thought, that this would be gone completely with the engine states code.

Could you implement something like "no overrun fuel cut, if CL-Idle is active" or something similar?
In v1.4.1 there was the "cut fuel when greater than ... (rpm)" field which worked more or less the same.

montelibre · Post by **montelibre** » Mon Apr 03, 2017 3:03 pm

Hi

I have just downloaded the 1.5.0 firmware for my MS3 X and I have the same problem ( nº 1).

On the other hand, my engine stall sometimes when it is overrunning.
I have a Jeep V8 with automatic transmission and the closed loop stays always active when it is in light downhill, so the stepper motor goes to full closed position.

I have tried with setting use Vss, but if you start downhill without throttle, the closed loop is always on and the engine stalls later. (I think it would be easy to solve if closed loop is turned off when the Vss gets larger than the threshold speed).

Post by **jsmcortina** » Mon Apr 10, 2017 9:23 am

Please try out pre-1.5.1 beta 1.

James

David_G · Post by **David_G** » Wed Apr 12, 2017 10:21 am

Awesome, Thanks!

I just tried it out and it seems that #1 and #2 are solved, Thanks!

Unfortunately there is a new issue concerning the Idle Valve transition from Cranking to Run Mode (Crank to Run Taper [s] ). Before the update I had 2s in there and the transition was fast but smooth.
Now, after the update the PWM Idle Duty is rapidly switching to the run position no matter what value I put into that field, which is almost stalling the engine. I tried 1, 10, 20, 50 and 100 seconds. No matter which value, there is no transition happening but only a sudden switch. -> see screenshot

PS:
The "STOP Engine" and "Reset ECU" buttons are absolutely awesome, great idea!

David_G · Post by **David_G** » Sun Apr 16, 2017 12:01 am

I posted the Idle valve issue in the 1.5.1-beta1 development thread as this is basically a new topic unrelated to the above.
Thanks a lot for your help, very happy this could be sorted out!!

ridethecliche · Post by **ridethecliche** » Mon Apr 17, 2017 9:27 am

So the 1.5 beta fixed your engine state issue? I'm trying to figure out something similar right now where things apparently revert to CL idle anytime I let off the throttle.
I hope I didn't post this in too many places, but nothing I post seems to show up including a thread I tried to post yesterday.

David_G · Post by **David_G** » Tue Apr 18, 2017 2:44 am

Yes, #1 and #2 got fixed by the beta.
Your problems looks more like a tune error to me instead of a firmware issue.

1)
Your Initial closed-loop idle duty is far to high because after letting off the throttle, RPMs rise up to 3500rpm and CL-Idle has to bring the RPM all back down to 850.

2)
The idle valve opening directly after letting off the throttle is caused by the "idle valve close delay" setting. You let your idle valve close completely when pressing the throttle, so when letting off the throttle it has to open again with the values of the "Closed loop initial values" table.

Tune your closed-loop initial values to stop the rapid RPM increase after letting off the throttle.

Megasquirt Support Forum (MSEXTRA)