Cold Accel Setting with Flex Fuel

[DanielCPG]

Hello. I am finding ,on all my flex fuel tunes, that the cold correction value for e85 and for 98 is very different. Ive heard that there is going to be an update to add fuel content to the cold correction table. Wondering if this is true and how far away it is?

Also i am having some trouble with closed loop fuel and boost control . I was wondering what the update rate is for Closed loop boost control? 

With closed loop fueling, i see the update rate max is 10. Is there a case for increasng this to 20 or 30 and having lower PID numbers. I always find that on a full throttle run, the CLL is too slow to reach the target, increasing the tClosed loop gain then results in a Overshoot/Undershoot in lower throttle positions/ engine speed. Perhaps theres a  case for a custom axis on the gain control table?

My current numbers in this car are in the closed loop gain table are  

1 2 4 8 16 32

[Adamw]

7 hours ago, DanielCPG said:

Hello. I am finding ,on all my flex fuel tunes, that the cold correction value for e85 and for 98 is very different. Ive heard that there is going to be an update to add fuel content to the cold correction table. Wondering if this is true and how far away it is?

Have you tried the mass of fuel option?  That potentially still wont take care of the lack of fuel vapor but it should at least increase the accel dose in proportion to the stoich ratio and density change.  

 

7 hours ago, DanielCPG said:

Also i am having some trouble with closed loop fuel and boost control . I was wondering what the update rate is for Closed loop boost control?

From a quick test I just done it looks like the boost control loop runs at 50Hz.   I havent seen any signs that is too slow, especially considering the valve is typically only running at 25Hz its going to take at least a full period before it can change position.  If you attach a log and tune I may be able to make some suggestions.

The CLL loop does only run at 10Hz.  There is a task to investigate increasing that in the future, but this was mostly only so we could run it slower at idle as 1Hz is the slowest you can update at the moment and with many cars the system response is longer than that at idle.  I dont think I have ever needed or wanted higher than 10Hz for car engines, possibly high RPM sports bikes or RC engines etc may need a faster update. 

7 hours ago, DanielCPG said:

I always find that on a full throttle run, the CLL is too slow to reach the target, increasing the tClosed loop gain then results in a Overshoot/Undershoot in lower throttle positions/ engine speed. Perhaps theres a  case for a custom axis on the gain control table?

My current numbers in this car are in the closed loop gain table are  

1 2 4 8 16 32

No enough gain, especially the left hand end, and update rate too high.  You ideally dont want to update any quicker than it takes for the sensor to respond to a change in inj PW.  You can test this with CLL off by holding it at steady RPM/load, then change the fuel table by say 10% and watch how long it takes for the lambda to reach about 90% of the way to that 10% change.  The update rate table is usually ok with RPM on its axis but I prefer something more related to mass flow, if using modelled fuel equation then Air mass estimated is a good option, if using traditional the inj DC is a good indicator.  

The update rate and gain tables from my car below as examples, these values seem to be about where I end up for most road cars I tune.  As a rough guide, a value of 60 in the gain table will mean it will correct the whole lambda error in one update (there is a little more to it such as it depends how far off stoich the target is etc but that is a good way to visualise the magnitude of the effect). Injector characterisation needs to be reasonable to have the gain at 60 and the lambda controller needs to be fast responding with the sensor as close as possible to the engine for the update rates at this level. 

   

[Adamw]

For your interest, CLL doing a decent job here keeping the lambda flat on target on both sides of a V10 race engine at 10.5K while I was playing with injector timing.    

[DanielCPG]

Thank you for the response 

I will have to test the closed loop on the next car i do and get some logging.

I will try the fuel mass option for Accel, but in either case is it possible to make the tradition cold Accel correction table axis 3d

[DanielCPG]

@Adamw also. How do I change the update rate axis. It's locked as rpm from me

[Adamw]

Yes that is a bit of a hidden trick, you need to set the trim limit tables setting to 3D and this will also change the update rate table into a configurable 3D table.  Since we are using something relative to mass flow we dont really need the other axis though.  

[mapper]

Agree, the 2D table should be changed to mass flow. 

What I also would like to see is an additional Integral part for CLL. 

[DanielCPG]

I have adopted the changes mentioned in this post in regards to both Accel and closed loop and have seen some good improvements.

I would still like a custom axis on cold correction as although it has improved, I am finding myself making changes cold on X fuel, it isn't feasible to then drain the tank to go back to Y fuel, check the Accel cold on y fuel,, then drain and go back to X. It would be alot easier if I could set my Accel cold on X fuel, and make changes to my Accel cold Y fuel without affecting my X fuel. 

 

[DanielCPG]

@Adamw any advice on the option on added a cold Accel axis. I think there's a pretty case for it to be available

[Adamw]

I think the ability to enable a further axis on the accel sens and cold correction table is a valid request and something I will add to the request list.  I have never personally needed it for flex fuel but I have always been careful with injector selection in that application and it doesnt get particularly cold here either so my experience may not reflect the typical.  But I do think there will be other use cases for 3D sens especially like staged DI/port injection or inboard/outboard staged injection where it could be beneficial or even necessary.

[mapper]

I defenetively agree for the need of correction for flexfuel. Especially when cold, ethanol needs much more enrichment. 

 

[DanielCPG]

15 hours ago, Adamw said:

I think the ability to enable a further axis on the accel sens and cold correction table is a valid request and something I will add to the request list.  I have never personally needed it for flex fuel but I have always been careful with injector selection in that application and it doesnt get particularly cold here either so my experience may not reflect the typical.  But I do think there will be other use cases for 3D sens especially like staged DI/port injection or inboard/outboard staged injection where it could be beneficial or even necessary.

I think 3d sense is already available.

I normally use rpm and tps delta..I find that I usually have large numbers at low delta and small numbers at low delta. 

Curious do you normally run just one axis on the sense table 

 

[mapper]

9 hours ago, DanielCPG said:

I think 3d sense is already available.

I normally use rpm and tps delta..I find that I usually have large numbers at low delta and small numbers at low delta. 

Curious do you normally run just one axis on the sense table 

 

thats exactly how I setup the sense table.
I my eyes the colds start correction table should be 3d, do adjust for ethanol. i don't think sense table should need any adjustment. Just the overall mass needs to be higher. 
I also tried the fuel mass accel mode on a RB26. However, even on lowest setting, enrichment was to much. 
And BTW, something I'm really still missing a lot on Link is deaccel enleanment.