MEGAsquirt A place to collectively sort out this megasquirt gizmo

Trying to sort out AC idle up and down RPM oscillations

Thread Tools
 
Search this Thread
 
Old 01-16-2020, 08:26 PM
  #21  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 857
Default

@Silver NB Having now looked at your info, I would start with taking away the "D" in the idle PID.

You could also try going open loop to see how the car responds to that.

A little dip or flare can occur under certain conditions, but the oscillation, you should be able to tame.

When you do get it like you want it, you may wish to allow the A/C to stay on above 40 MPH. Lastly, when you put the turbo on, you will have to expand your VE table. May as well do it now, as you have the spark table. AND, set your boost cut before you take the car out on even the first run.

Other guys.EDIT: I don't think MS will allow Lambda. The AFR table won't allow numbers below "1". See Post #22 below. However, I see no benefit. There is the same learning curve to what is 0.8 vs 1.0 as 12 to 14.7; and you end up with tables that folks here cannot easily read.

If you are heart set on Lambda, just make Stoichiometric 10.0 (lambda * 10). Recalibrate the AFR input from the controller to the lambda * 10 equivalents. Then you only have to move the decimal in your head.

Now, if you could read 1/Lambda, then to me that would be an improvement. FAR rather than AFR. Then bigger numbers would mean too much fuel, rich; and smaller numbers would mean too little fuel, lean. That I could see value in. But again, one would be speaking Greek in an English world.

irodd, given

DNM

Last edited by DNMakinson; 01-17-2020 at 10:10 AM. Reason: Correct my erroneous conjecture.
DNMakinson is offline  
Old 01-17-2020, 09:30 AM
  #22  
Junior Member
iTrader: (4)
 
irodd's Avatar
 
Join Date: Apr 2014
Location: Toronto, Canada
Posts: 173
Total Cats: 16
Default

To switch from AFR to Lambda in Project properties change Wideband to Lambda.
Same for Fahrenheits to Celsius.

irodd is offline  
Old 01-17-2020, 11:53 AM
  #23  
Junior Member
Thread Starter
iTrader: (1)
 
Silver NB's Avatar
 
Join Date: Feb 2019
Location: Commerce, TX
Posts: 46
Total Cats: 2
Default

Originally Posted by DNMakinson
@Silver NB Having now looked at your info, I would start with taking away the "D" in the idle PID.

You could also try going open loop to see how the car responds to that.

A little dip or flare can occur under certain conditions, but the oscillation, you should be able to tame.
I'll give that a try and see what happens. I thought that the D term was what was used to tame overshoot, but like I said before, I may be misunderstanding how the PID terms operate here. If that doesn't help, would raising I be the next step? My previous experience with PID controllers was with tuning racing drones, so the experience may not directly translate.

Originally Posted by DNMakinson
Lastly, when you put the turbo on, you will have to expand your VE table. May as well do it now, as you have the spark table. AND, set your boost cut before you take the car out on even the first run.
Where would you recommend putting the boost cut at? I don't know what spring my actuator has, but I think it's supposed to be for 8 psi. If I did my math right, 8 psi should be right around 150 kpa. Maybe set it for 160 kpa or would you start lower?
Silver NB is offline  
Old 01-17-2020, 03:36 PM
  #24  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 857
Default

Set it for just above where you are tuning and then increase along. In your case, the 160 would likely be safe. But see, if you go real lean at 140, that could still be a problem.

more important with EBC, when the sky is the limit.
DNMakinson is offline  
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
andym
MSPNP
4
11-20-2018 04:22 PM
thebigtuna
MEGAsquirt
0
08-23-2017 04:08 PM
joe morreale
MEGAsquirt
13
06-20-2017 07:23 PM
Nicolas L
MEGAsquirt
16
05-18-2017 05:56 PM
dale
MSPNP
8
01-04-2016 11:32 AM



Quick Reply: Trying to sort out AC idle up and down RPM oscillations



All times are GMT -4. The time now is 03:52 PM.