MicroSquirt.com read only forum (Jan.15,2018)

I'm trying to understand the relationship between GEGO, AFR target tables, and auto-tune and how they work together.

I gathered this much, that auto-tune uses 02 correction to correct rich or lean spots of the VE table. But what does it use as it's base value? Does it try to make everything 14.7:1? Now I'm assuming this is where AFR targets come into play. If you have AFR targets enabled, it will ignore it's arbitrary 14.7:1 preset and use what you've entered into the tables, which is to prevent you from running a stoich mixture during boost. Am I correct in assuming this is how it works?

If so, what happens to the VE table when you've got AFR targets turned on? Is it ignored? I know you can set it to use AFR targets only after a certain RPM/MAP reading, but you could theoretically run off the AFR table all the time right? Which leads me to another problem, if you're using AFR tables, and for some reason your WB02 sensor craps out, what does it do?

Also, lastly how does GEGO fall into the picture when you're running AFR tables?
Thanks.

The AFR table is used for closed-loop fueling without changing the VE table. EGO correction is the difference between the desired AFR vs the actual AFR from the VE table, that AutoTune use to correct the VE until there is no EGO correction under stable condition. The original MS-I firmware only has 1 switchpoint for closed-loop fueling, MSnS-E and MS-II have an AFR table in addition to the global switchpoint (which I believe is only used when the AFR table is disabled)

So the EGO sensor switchpoint is what auto-tune aims for when AFR tables are disabled? What I'd still like to know, is the relationship between AFR tables and auto-tune.

I believe that's the case - no AFR table => EGO correction will use the switchpoint. AFR table is what Auto-Tune eventually targets by correcting VE table to eliminate EGO correction

KGB wrote:I believe that's the case - no AFR table => EGO correction will use the switchpoint. AFR table is what Auto-Tune eventually targets by correcting VE table to eliminate EGO correction

Does this mean that in order to use autotune, I need to have EGO correction turned on? I had it disabled while I tuned. Do I need to activate it again?

Reason I ask is it seems that autotune isn't altering my tables any.

I believe you have to - as long as it's generating EGO correction value, auto-tune can use it

I to have the same problems. I have everything set as far as the auto tune settings go. I can have a air/fuel table selected as well and it seems like auto tune is always outside of window or something. I have been playing with ms for quite some time now and I have yet to ever see my auto tune do something. Is there a check list I can go by to make sure everything I have done is correct ? Thanks guys

-Chris

I am currenly puzzled by the same thing.  First time round Auto-tune did its thing no problem.  Now I've changed the AFR targets (and who knows what else...) and now it seems to refuse to Auto-tune even though EGO is sitting way off 100%, AFR is way off target as per AFR map and I am so certain that I'm in the AT zone having NARROWED the auto-tune params for the first time round.  Ive even relaxed the lumpiness to 50% as this seemes to be the only parameter that could possibly still prevent autotune from doing its thing.
 
I'm a newby but let me have a go at this hoping an expert will confirm/add:
 
1 ) Set the auto-tune parameters in custom.ini in the relevant directory:
#elif MS_II
   table = veTable1Map
      allowAutoTune          = on
      corrector              = egoCorrection1
      xLimits                = 1500, 4000 ; rpm within which AT should be active
      yLimits                =   60,   90 ; MAP values within which AT should be active
      zLimits                =   10,  200 ; VE values within which AT should be active
      xRadius                =  200 ; rpms above or below a cell engine should be for AT to adjust that cell
      yRadius                =    7 ; MAP above or below a cell engine should be for AT to adjust that cell
      initialStartupInterval =  1.0 ; seconds AT has to be within the window before it starts adjusting values
      updateInterval         =  1.0 ; times per second AT adjusts the cell
      proportionalGain       =  0.5 ; how much each change is allowed to be
      lumpiness              =    5 ; by how much the value in a cell is allowed to differ from any adjacent cell 
                                        ; (prevents AT from "running away" with a value)
 
2) Make sure AT is on
3) make sure the engine is in the window (I find using the 2D map by pressing "z" is better for making sure I'm in the AT window 
 
Kind regards
Philip
MS-2/V3/2.35/MT 2.25/Innovate LC-1/LSU4.2
Land Rover 4.6 std, 4-2-1 exhaust headers, IAC
'80 Jaguar XJS V12 5.3 (10:1 CR) is waiting for its turn


Posted by email.

Philip Lochner wrote:I am currenly puzzled by the same thing.  First time round Auto-tune did its thing no problem.  Now I've changed the AFR targets (and who knows what else...) and now it seems to refuse to Auto-tune even though EGO is sitting way off 100%, AFR is way off target as per AFR map and I am so certain that I'm in the AT zone having NARROWED the auto-tune params for the first time round. 

1 ) Set the auto-tune parameters in custom.ini in the relevant directory:
#elif MS_II
   table = veTable1Map
      allowAutoTune          = on
      corrector              = egoCorrection1
      xLimits                = 1500, 4000 ; rpm within which AT should be active
      yLimits                =   60,   90 ; MAP values within which AT should be active
      zLimits                =   10,  200 ; VE values within which AT should be active
      xRadius                =  200 ; rpms above or below a cell engine should be for AT to adjust that cell
      yRadius                =    7 ; MAP above or below a cell engine should be for AT to adjust that cell
      initialStartupInterval =  1.0 ; seconds AT has to be within the window before it starts adjusting values
      updateInterval         =  1.0 ; times per second AT adjusts the cell
      proportionalGain       =  0.5 ; how much each change is allowed to be
      lumpiness              =    5 ; by how much the value in a cell is allowed to differ from any adjacent cell 
                                        ; (prevents AT from "running away" with a value)
 
2) Make sure AT is on
3) make sure the engine is in the window (I find using the 2D map by pressing "z" is better for making sure I'm in the AT window 

I have since spotted another requirement for autotune to be active:
4) NONE of the enrichment parameters must be active.

The reason my autotune did not "work" the second time around is that I had my percentages too small

How is it workin for ya ? One of these days ill get it working as well. I still have no luck with auto tune. To this day it seems that I am always "out of window"..

-Chris