The effects of AFV on open loop areas

[mayor]

I know this has nothing to do with Harley's. GM with their vortec head design wasn't that designed to increases combustion chamber turbulence for better fuel air mixing and flame front  control? It sure made a difference in the old 350 engine.

yes, turbulent combustion chambers....but not turbulent intake tract.  Turbulence in the intake tract is a power robber, not power adder. 

[Steve Cole]

The Vortex head design was to CONTROL the turbulence in the combustion chamber which is a big difference than just having a turbulent chamber. The key to it all is having control over things not just tossing it together and hoping you hit upon a combination that works. Doing testing to prove out what you have is key, but just as key is not to misread the data being provided by the test equipment. You can only figure this out if the test you are performing has equipment being used that is capable of measuring what your looking for. This is a very big mistake people make, assuming that the reading they are getting is real when it's NOT.

[Sporty 48]

You know, I am not sure.
Have tried 4 different air filters, this latest made it lean over 3,000 rpm.
Stopped  intake/exhaust experimentation while trying to figure out tuning, still working on the tune.

Some ask if it is a turbo, I say no I just want lots of clean cool turbulent air.

are you sure you want turbulent air?     I think what you gain by using a longer throat on the intake tract is tuned airflow....which is likely the opposite of turbulent air flow.

[1FSTRK]

You know, I am not sure.
Have tried 4 different air filters, this latest made it lean over 3,000 rpm.
Stopped  intake/exhaust experimentation while trying to figure out tuning, still working on the tune.

You do bring up a good point about tuning. The pro tuners see a lot of combinations that work and a lot that don’t. When we tune our own bike we see only one bike with one combination so how does one tell when we have reached the point that hard parts need to be changed.
When the 1100 Sportster came out I had one and it had pipes and an air cleaner change and it ran ok but it never seemed smooth in the 2-3000 rpm range. Then the CV carbs came out and I swapped out the butterfly carb for the CV carb and a jet kit. I could not get it to run smooth at all, took it to the local tuner and he took one look and said “you can’t run that pipe with that air cleaner and a CV carb” He had been down that road before. He had an AC cover with a piece of clear plastic glued in the center so you could watch inside with the bike running. With no cover the bike ran fine, put the cover on and the carb slide would just bounce up and down at 2600-3000 rpm. I could run stock pipes with that air cleaner or my pipes with the stock big air box and it was smooth as silk. I tried different springs in the slide and finally gave up. Now fast forward to today and my very modified EFI twin cam and no slide to watch, what are the chances that if there were a slide in the TB that it would be bouncing all over right where I’m having trouble? Now how would the ECM and the AFV react to that when it closed loop. The thing is I think EFI is so much better that we can actual tune a bad combination that we would have otherwise given up on, not great but good enough that we just won’t give up and change a hard part.
Some folk are just learning about the pulse waves that we deal with in these things and others refuse to even admit they exist but there is a point in the tuning process where it is time to change one of the hard parts and see if it will tune.
Just my 2cents mix with some first hand experience.

[wurk_truk]

I'm not so sure the effect is only ever richer; I believe it goes both ways in our system.

you are incorrect in your assumptions, Glens.  Only goes richer in OL is a left over from the original base delphi programming, of which MOCO did not change when Delphi wrote the MOCO code.  The programmers felt going leaner than the map settings, using AFVs in OL, would be beyond stupid.  Now, the bike WILL lean back to original map settings as AFVs change.  They will never go below what is on the base map while in OL.

CL is a different animal completely.

[glens]

It'd be nice to assuredly know this sort of stuff...  Got a source you can name for the info?

[04FLHP]

Quote from
Steve Cole
Re: The sort of thing that gets my attention
« Reply #423 on: Wednesday, September 21, 2011. 05:30:18 PM.

In the HD Delphi code only, what is learned in closed loop is applied in open loop. This is not always the case in other applications. The most common approach I know of is to use what is learned provided it is richer and not use any enleanment in open loop.

I'm taking this as saying that the HD Delphi closed loop learned can and will drive open loop rich or lean.
Maybe Steve can correct me if that is wrong.

[wurk_truk]

FBRR, is whom I am quoting.

[wurk_truk]

Quote from
Steve Cole
Re: The sort of thing that gets my attention
« Reply #423 on: Wednesday, September 21, 2011. 05:30:18 PM.

In the HD Delphi code only, what is learned in closed loop is applied in open loop The most common approach I know of is to use what is learned provided it is richer and not use any enleanment in open loop.

THIS is what the english easy to understand side says...  Steve is quoting me quoting FBRR

[Steve Cole]

WURK_TRUK

Not me quoting you quoting FBRR. HD doesn't do it the regular way from what we have found.

[WVULTRA]

Quote from
Steve Cole
Re: The sort of thing that gets my attention
« Reply #423 on: Wednesday, September 21, 2011. 05:30:18 PM.

In the HD Delphi code only, what is learned in closed loop is applied in open loop. This is not always the case in other applications. The most common approach I know of is to use what is learned provided it is richer and not use any enleanment in open loop.

I'm taking this as saying that the HD Delphi closed loop learned can and will drive open loop rich or lean.
Maybe Steve can correct me if that is wrong.

I can't think of a reason anyone would run an OL cell/area leaner than a CL setting.........

[wurk_truk]

Say what you have found, then, Steve.

[glens]

I can't think of a reason anyone would run an OL cell/area leaner than a CL setting.........

I can't either, but I'd sure want an appropriate AFV to take things leaner any appropriate where in open-loop if that's what was determined needed doing any appropriate where in closed-loop.  If there's a reason, like I went from corn gas to just petroleum gas, or my fuel pressure regulator was putting out more pressure, etc., I'd like the right thing to be done open-loop as much as I want it done closed-loop.  Either direction.

[cts1950]

OK  this is confusing. Steve has repeatedly said that if the ecm was set in OL it dose not care what the o2s are reading. Is the afv active in OL or not and how dose it know how to adjust the fuel mix if it dose not use the o2s for feedback in OL.

[mayor]

OK  this is confusing. Steve has repeatedly said that if the ecm was set in OL it dose not care what the o2s are reading. Is the afv active in OL or not and how dose it know how to adjust the fuel mix if it dose not use the o2s for feedback in OL.

what is confusing to me is, if the bike is set to OL.....how could it be considered OL if the o2 sensors are still providing feedback for adjustments to the fuel supply (essentially CL)? 

[glens]

OK  this is confusing. Steve has repeatedly said that if the ecm was set in OL it dose not care what the o2s are reading. Is the afv active in OL or not and how dose it know how to adjust the fuel mix if it dose not use the o2s for feedback in OL.

Not confusing at all if you know what's being talked about.  Nobody's saying if you reset the AFVs and take away closed-loop operation across the board in your fuel tables ("unnatural" open-loop operation) that any AFVs are going to dick with what you've done (or ever save your butt .  What's being discussed here is open-loop operation which occurs "naturally".  That is, in the far right and the bottom of the fuel tables all the time and anywhere in the "14.6" area whenever transient conditions occur, such as when AE (for example) kicks in momentarily, in which case the ECM drops out of closed-loop for the duration (but you still reap the benefits of the pertinent AFVs all the while).

So when you're running a properly-set-up closed-loop setup, the AFVs developed during normal day-to-day operation will maintain your tune over time, both in the closed-loop areas of the fuel tables (both when running closed-loop and open-loop there), and in the open-loop areas.  Now this might not happen as well as designed if you try to get clever and only set a tight cruise area at highway speed to closed-loop or something; that would depend on how the AFV "tables" might be set up internally.

It guess it boils down to Steve saying "whenever it's in open-loop", not "whenever it's (entirely) set open-loop".

[add: Whenever it's in open-loop, for whatever reason, no AFVs will be getting developed, however if AFVs have gotten developed, they will always be put to use no matter the current *-loop situation. ]

[1FSTRK]

I believe that it works as follows:
Closed loop = O2s active and ECM will richen or lean real time to keep O2s centered
AFV = a saved calculated adjustment applied to one of the base fuel tables (VE or AFR) to better center the AFR starting point based on how far the original tables have to be adjusted to keep the O2s happy. (Say 2% across the board increase or decrease to make it quicker and easier for the ECM to adjust for proper O2 readings)
If this is how it works and because you have closed loop readings consistently off far enough to apply a 2% AFV as a new starting point to the base table then regardless of what caused it (fuel change) that 2% multiplier should be applied to the open loop cells as well.
The open loop does not have the O2s on line for feed back but the AFV of 2% can be applied to the base table because what ever change in the bike, fuel, or conditions will exist in both open and closed loop. It is kind of like the ECM can do an internal v-tune run to generate an AFV for small changes that occur.
Of coarse this is all just my take at this point in the discussion.

[Steve Cole]

Open loop operation = the fuel tables in the ECM +/- stored AFV"s

Closed Loop operation = the fuel tables in the ECM +/- actively changing AFV's

[1FSTRK]

Open loop operation = the fuel tables in the ECM +/- stored AFV"s

Closed Loop operation = the fuel tables in the ECM +/- actively changing AFV's

It doesn't get much easier to understand than that.
Thank you Steve

[wurk_truk]

No it is NOT simple.  I wish to know why Steve is saying the Delphi team is giving me incorrect info and I will want this to be backed up.  I'm NOT Andy, nor do I expect to be treated as if I am.  If OL on a closed loop tune can go leaner than the map settings, I want to know what testing you did?  As it stands, without your giving up the data, even in a phone call...  I will EITHER believe the guys that INVENTED the code, or I will know CL is a bunch of crap.

[mayor]

so is it safe to say that if Steve said it this way you would agree:

Open loop operation = the fuel tables in the ECM and stored AFV's that add to the fuel pulse wave, but not ones that subtract

Closed Loop operation = the fuel tables in the ECM +/- actively changing AFV's

[1FSTRK]

No it is NOT simple.  I wish to know why Steve is saying the Delphi team is giving me incorrect info and I will want this to be backed up.  I'm NOT Andy, nor do I expect to be treated as if I am.  If OL on a closed loop tune can go leaner than the map settings, I want to know what testing you did?  As it stands, without your giving up the data, even in a phone call...  I will EITHER believe the guys that INVENTED the code, or I will know CL is a bunch of crap.

Did your Delphi team work on the AFV code for the HD version or Delphi in general?

[glens]

I will EITHER believe the guys that INVENTED the code, or I will know CL is a bunch of crap.

Putting aside the question of whether or not writing software properly constitutes an invention, why would you take such an extreme stance?  Am I correctly understanding the last part of your statement, fleshed out, as "I will know Closed Loop is a bunch of crap"?  Why?  Why would you be so offended at the notion that fueling adaptation could either add or remove fuel when/where appropriate?

[cts1950]

OK
If a normal closed loop tune contains open loop cells such as idle and lets say 80 kpa and above will the afv correct the entire afr table or just the cells that are in closed loop.  I can see the ecm adjusting the closed loop cells, but the cells in open loop that the ecm is not collecting o2 data for, I would not I would not expect the afv to change these cells.

[glens]

There's not "the AFV".  There are many of them, in various locations, per cylinder.

They wouldn't be needed for closed-loop areas except for a starting point at new startup or to use whenever closed-loop is temporarily disabled.  They'll always be needed for areas which never themselves get active trim updates to their fueling.

If Delphi didn't write in ability to remove fuel as needed in the outlying areas of the fuel tables, and MoCo came along and added the functionality, I'm glad for it.  It's much more sensible that way.