COS5 Turbo Questions

[ferocity02]

I am finally getting started with COS5. I think I have the baseline tune setup, if someone could take a look at it and let me know if they see anything strange that would be great! This is just my 12212156 baseline turbo tune flashed into COS5 with the out-of-range values fixed using the COS tutorial. There was little to no tuning done with the turbo, it is still the conservative baseline tune.

Will this work with a 1bar MAP sensor and the MAF (for now)? I'm looking to go 2bar MAP and MAFless in the next few days.

How does B3647 work? Will it use this table for fueling while in open loop, and then go to the MAF once in closed loop?

Can I use the VE tables exclusively (w/o PE mode) for fueling in vacuum and boost? If so, how is this setup?

Would B3647 be better to control fueling? If so, would I need to enable PE mode to add fuel above 100 kPa?

Long story short, if running MAFless with 2bar MAP, what is the preferred method to control fueling in both vacuum and boost?

Vehicle is a '01 Silverado 5.3L, turbo up to maybe 15psi, meth/water injection above ~8psi. I read through the entire COS5 sticky and joecar's notes and the COS tutorial.

[joecar]

Yes, ok with 1-bar MAF'd for now (avoid running engine at load) until you go 2-bar MAF-less.

[joecar]

B3647 is commanded fuel in OL, with an added feature: SOL (semi-open loop), this does STFT trimming on stoich cells in B3647.

Do not confuse OL/CL with MAF'd/MAF-less, they are othorgonal (i.e. there are 4 modes: OLSD, OLMAF, CLSD, CLMAF).

So, B3647 provides commanded fueling on OL regardless of MAF or MAF-less.


[ also note: running in CL (CL/LTFT enabled), when you go to WOT you're still in a CL mode (not OL), so B3647 does not become active (so commanded fueling comes from PE B3618) ]

[joecar]

You can use the VE table with or without PE (I would recommend PE as a safety net) for vacuum and boost.

You force VE-only mode by failing the MAF (cause a MAF DTC to become present).

[joecar]

Use B3647 to command fueling for vacuum, and use B3618 to command fueling for boost (see B3613 PE MAP Threshold, set this to 105 kPa)...

this gives you the most fueling control.


You will have to edit A0008 to control warming-up fueling.

[ferocity02]

Yes, ok with 1-bar MAF'd for now (avoid running engine at load) until you go 2-bar MAF-less.

Awesome!



B3647 is commanded fuel in OL, with an added feature: SOL (semi-open loop), this does STFT trimming on stoich cells in B3647.

Do not confuse OL/CL with MAF'd/MAF-less, they are othorgonal (i.e. there are 4 modes: OLSD, OLMAF, CLSD, CLMAF).

So, B3647 provides commanded fueling on OL regardless of MAF or MAF-less.

[ also note: running in CL (CL/LTFT enabled), when you go to WOT you're still in a CL mode (not OL), so B3647 does not become active (so commanded fueling comes from PE B3618) ]

Can the STFT trimming feature be disabled? Or will it always be active as long as there are EQ Ratio=1 cells in B3647?
So if I'm in closed loop, B3647 is not used at all?
Looks like I have the option between OLSD and CLSD... is one preferred over the other?



You can use the VE table with or without PE (I would recommend PE as a safety net) for vacuum and boost.

You force VE-only mode by failing the MAF (cause a MAF DTC to become present).

Does going MAFless not necessarily force VE-only mode? I assumed that MAFless was equivalent to running on the VE table only.



Use B3647 to command fueling for vacuum, and use B3618 to command fueling for boost (see B3613 PE MAP Threshold, set this to 105 kPa)...

this gives you the most fueling control.

Will setting it up like this mean that that VE tables aren't used? Or does it use the VE tables to determine the amount of air entering the engine, looks up how much fuel you are commanding from B3647 and B3618, and computes the amount of fuel to deliver?

[joecar]

Can the STFT trimming feature be disabled? Or will it always be active as long as there are EQ Ratio=1 cells in B3647?

SOL (OL STFT trimming) can be disabled by changing any EQR 1.00 cells to something other than stoich (e.g. say EQR 0.99 or EQR 1.01).



So if I'm in closed loop, B3647 is not used at all?

Correct, in CL B3647 is never used (even when you go WOT).


Looks like I have the option between OLSD and CLSD... is one preferred over the other?

CLSD probably gives you a little better MPG; OLSD means you never have to replace NBO2 sensors.



Does going MAFless not necessarily force VE-only mode? I assumed that MAFless was equivalent to running on the VE table only.

The PCM will only failover to the VE table if it sees a MAF DTC (if there's no MAF DTC then it continues using the MAF table, even if the MAF is physically absent);

so if you remove the MAF you have to make sure that a MAF DTC triggers immediately on key on.

Yes, MAF-less (aka SD) is running from the VE table only (i.e. the PCM computes cylinder airmass from VE, and ignores the MAF).

BTW: if you use VE units g*K/kPa you will see that the VE table B0101 is really a "pressure/temperature-normalized" airmass table.




Will setting it up like this mean that that VE tables aren't used? Or does it use the VE tables to determine the amount of air entering the engine, looks up how much fuel you are commanding from B3647 and B3618, and computes the amount of fuel to deliver?
Yes, running SD/MAF-less (verified by immediate trigger/presence of a MAF DTC) the PCM computes cylinder airmass from the VE table only (and ignores the MAF);

then, it determines commanded fueling from the richest of the enabled fueling tables (in OL: B3647, and B3618 if PE enables; in CL: trimming, and B3618 when PE enables)

[ see post #4 here: Summary-Notes (http://forum.efilive.com/showthread.php?14188-Summary-Notes) ]

and then it applies any commanded fuel EQR modifier tables...

and then it applies B3601 to compute fuelmass (from airmass and commanded fuel EQR)...

and then it uses the injector tables to arrive at a pulsewdith;

then it sprays the injectors.


Then, if CL, it reads the NBO2's and determines the trim to apply to the computed fuelmass to correct fueling to stoich.

[ferocity02]

Thanks for the help! The puzzle is slowly coming together in my head.

So OLSD means the fueling is always determined by B3647 plus modifiers from PE mode and others (such as A0008). And the VE tables are used to estimate airflow. In B3647 cells that are EQ 1.0, "semi" open loop is used which makes corrections using the NBO2's and STFT's to achieve stoich.

And CLSD is the same as OLSD until you enter closed-loop mode, and B3647 is ignored and the LTFT's are used to achieve stoich everywhere, except when PE mode is active. And again the VE tables are used to estimate airflow.

Sounds like OLSD might be the way to go, seems like there is more fueling control this way. And you could potentially do lean cruise with B3647 correct? Or could you still use the lean cruise calibrations in OLSD?

[joecar]

OLSD can be broken up as follows:

SD:
cylinder airmass is directly computed from VE only;
the VE table provides cylinder airmass normalized for pressure and temperature (VE units are g*K/kPa);
the estimated airflow is for information only;

OL:
commanded fuel is determined from the richer of B3647 and B3618 (when PE enables), and modified by A0008 during warmup;
B3647 also allows OL STFT trimming on stoich cells (SOL);

CLSD can be broken up as:

SD:
same as SD above;

CL:
runs in OL until CL enables;
uses CL LTFT trimming to stoich;
when it leaves CL, it enters a partial CL mode where it adds the last LTFT (rounded up to zero if negative), and it does not use B3647;
when PE mode enables, it commands fuel from B3618 and it adds the last LTFT (rounded up to zero if negative), and it does not use B3647.

[joecar]

Yes, OLSD gives you very good control over fueling.

Lean cruise fueling is more controllable using OLSD rather than LC mode.

[5.7ute]

If you use lean cruisemode with a COS you need to match your PE map trigger to the enrichened cells in B3647. For example, if you command stoich up to 80kpa, then progressively enrichen the cells up to 105kpa, you need PE the map enabler to trigger from 80kpa. Also the TP position enabler will need to be set just above your usual cruise position. This is the only work around I have found to prevent a "rich cruise mode" Which will occur if the cruise timer is hit during an uphill climb.

[ferocity02]

If you use lean cruisemode with a COS you need to match your PE map trigger to the enrichened cells in B3647. For example, if you command stoich up to 80kpa, then progressively enrichen the cells up to 105kpa, you need PE the map enabler to trigger from 80kpa. Also the TP position enabler will need to be set just above your usual cruise position. This is the only work around I have found to prevent a "rich cruise mode" Which will occur if the cruise timer is hit during an uphill climb.

Using your example of enriching after 80kPa in B3647, what about just setting PE MAP trigger to 105 kPa?

[5.7ute]

Using your example of enriching after 80kPa in B3647, what about just setting PE MAP trigger to 105 kPa?
In the example I stated, that is what is necessary if using the lean cruise patch. If you are not using this patch, enrichen B3647 as you would for an N/A application, trigger PE at 105 kpa & place your boost enrichment values in the commanded fuel in PE table. You can lean out the cruise areas of B3647 for an easy lean cruise setup.
PS sorry for the late reply but we are swamped here at the moment.

[ferocity02]

I have my B3647 set to 0.99 everywhere except for the higher MAP columns. But once it gets into idle mode at zero VSS and zero TP the GM.EQUIVRATIO goes to 1.00. Can this be adjusted? If so where?

If EQR is 1.00 at idle but 0.99 in B3647, will it go into semi-open loop?

[joecar]

If wideband AFR is tightly oscillating around stoich then SOL is active, see below pic AFR_WB waveform:

http://img13.imageshack.us/img13/9413/solf.png

[ferocity02]

Hmm, mine doesn't look nice like that, but it's definitely oscillating.

Why is the idle EQR 1.00 when B3647 is commanding 0.99?

Here's a log from today.

[darcy]

Why is the idle EQR 1.00 when B3647 is commanding 0.99?


Try setting B4101 to B4104 to 0, to stop it entering closed loop idle.

[ferocity02]

Try setting B4101 to B4104 to 0, to stop it entering closed loop idle.

Thanks, I'll give that a shot.