What is the point of PE mode?

[pauly24]

I don't understand why they have a PE mode.

If you want it richer in "power mode" then you would set the AFRs in the VE table that are over say 70kpa to 12.5.

And lean mode is when you have low MAP values you can lean it up.

Why does it go to all the trouble of having a normal mode and a PE mode that needs all these parameters set to enter the mode, when you could get the same effect by having a VE table with different AFRs throughout the map?

[joecar]

The VE provides cylinder airmass (VE has units g*K/kPa).

The MAF table indirectly also provides cylinder airmass (MAF has units g/s).

Neither VE nor MAF provide AFR.

When in OL, the AFR is looked up from the richer of PE B3618 (if PE has triggered) or CFOL B3605.

So, the point of PE is this: when throttle is opened up, the load on the engine will now increase, so PE provides enrichement.

[joecar]

PCM roughly follows this sequence:
- compute airmass from VE and/or MAF;
- lookup AFR to command;
- compute fuelmass required to meet that AFR;
- apply CL trim to fuelmass;
- look up injector flow rate;
- compute injector pulsewdith required to deliver that fuelmass;
- look at O2 response to determine new CL trim;
- repeat.

[joecar]

When the VE table correctly models the engine's cylinder airmass, then in SD/VE mode, the measured AFR will equall the commanded AFR.

But instead of speaking of AFR, we should be more correctly speaking of Lambda or Equivalence Ratio (1/Lambda).

[pauly24]

Sorry yes I got confused, what I mean was, in the VE table, you could trick the ecu into thinking the engine was getting more or less air, therefor creating a rich or lean situation.
EG. what I mean was you can control the AFR through the VE table.
Even though this does not model the airflow correctly it would provide the same result would it not and you would not need the PE mode.

I think im just use to Nistune, where it has a MAP vs RPM and its a mixture coeffecient correction table, so you just punch in the AFR you want for each RPM/MAP point. This way its all controlled from one table, you know how much load its under (high map, low rpm = high load) so in that cell you would richen it up.

[joecar]

Yes, you could manipulate the VE table to achieve the same result.

The COS's (custom operating system) replace table B3605 (ECT x MAP) with table B3647 (RPM x MAP); table B3647 lets you set the AFR for all (RPM, MAP) operating points, this allows you to disable the PE table.

The advantage of keeping the PE table enabled is that it has ECT and IAT modifier tables if you wish to use these... for example, you could set the IAT PE modifier to give extra fuel when IAT is hotter (i.e. to promote a cooling effect on combustion chamber temps).

[redhardsupra]

you should sell these framed to hang in the garage of every ECU geek out there

PCM roughly follows this sequence:
- compute airmass from VE and/or MAF;
- lookup AFR to command;
- compute fuelmass required to meet that AFR;
- apply CL trim to fuelmass;
- look up injector flow rate;
- compute injector pulsewdith required to deliver that fuelmass;
- look at O2 response to determine new CL trim;
- repeat.

[ScarabEpic22]

you should sell these framed to hang in the garage of every ECU geek out there

Id take a t-shirt...

[redhardsupra]

joecar, i got this writeup i haven't posted yet on the fueling cycle from the point of view of the ECU. you wanna audit/edit it for me? you obviously have some proper insight into the process.

[joecar]

Marcin, pm sent.

[joecar]

you should sell these framed to hang in the garage of every ECU geek out thereI'll look into what's required to produce a laminated/framed card... :grin:

[swingtan]

I'll expand on Joecars points a little....

When the original tune was made for the car, GM/Holden had a certain set of parameters that they needed to work within, basically these were...


The engine must work well for a huge range of conditions.
The engine must keep working well over a very long time frame with no adjustments.
The engine must be tolerant of people using vastly different qualities of fuels and even contaminated fuels.
The engine must do all this under vastly different driving conditions.
For your car, this must be all done while complying to Euro-4 emissions standards.
The one tune must suit all cars for all of the conditions.


The next thing to note is that the AFR used for a given fuel gives different results in terms of economy and power. It centers around the "Stoichiometric" point of the fuel, or the point at which the amount of fuel in the intake charge is completely consumed in the combustion process. This is the point at which the least amount of hydrocarbons are emitted from the exhaust, which is why the car is set to run as close as possible to this point. If you run the car a little over Stoich ( leaner ) the fule burns to early and you can get nitros-oxide emissions, but you can also get better mileage. Run it a little under Stoich ( richer ) and you can get more power, but emit more hydrocarbons as the oxygen is consumed before the all fuel is.


<------power-------stoichiometric------economy------>
12.5:1----------14.7:1------------15.6:1

Representation of AFR's only - don't take these are actual rules

So to do all of this, the management system must use some sort of feed back to help the ECU know what the engine is running like, so it can help control everything it needs to. Part of this is the Closed Loop fueling, that maintains the actual AFR / Lambda values at the stoichiometric point of the fuel used. This helps give good economy and power when cruising on the highway and motoring around town, where probably 90% of the engine life is likely to be. So it all makes sense to use this method of control.

The problem is that when you want more power, for acceleration, overtaking or towing, the engine needs more fuel to burn. If you are running in CL mode, the O2 sensors will not allow the ECM to run richer. So we have "Power Enrichment - PE " mode to allow the engine to run richer when needed. When the ECM goes into PE mode, it ignores the current O2 sensor inputs and uses set tables and possibly long term fuel trims to "estimate" the required fuel quantities needed to achieve the commanded AFR. This then allows for a richer mixture to give the additional power required.

As for "messing" with the VE table..... The general view of this is that at the very best, it's a crude hack, most would say you should never, ever "fudge" the VE table as there are a number of other settings that this will effect. To be be "correct" you should make the VE table "correct" and then use other tables to command the fueling needed. I was talking to Jezza about this earlier this week and we came up with the following list of some of the tables that are based on the VE tables...

Firstly, the PID DYN_CYLAIR is taken directly from the VE table. This PID tells the ECM the amount of air entering the engine in terms of gm/Cyl, tables that use this PID are....


The main spark tables.
Most, if not all, of the calculate engine torque tables ( think torque and traction control )
If an auto, this may effect the TCM tune.


So you can see that having an incorrect VE table can effect more than just the fueling. By fudging one table, you then need to fudge a few more to try and fix the things messed up by the initial fudge....... It's a vicious circle that's better not started. Of course I'm not saying things won't work at all and sometimes a "fudge" might be used to "prove a point" or "test", but in the long run, getting the base tables correct has a lot less headaches for a daily driver.

Regards,

Simon.

[WeathermanShawn]

I'll expand on Joecars points a little....

When the original tune was made for the car, GM/Holden had a certain set of parameters that they needed to work within, basically these were...



The engine must work well for a huge range of conditions.
The engine must keep working well over a very long time frame with no adjustments.
The engine must be tolerant of people using vastly different qualities of fuels and even contaminated fuels.
The engine must do all this under vastly different driving conditions.
For your car, this must be all done while complying to Euro-4 emissions standards.
The one tune must suit all cars for all of the conditions.



The next thing to note is that the AFR used for a given fuel gives different results in terms of economy and power. It centers around the "Stoichiometric" point of the fuel, or the point at which the amount of fuel in the intake charge is completely consumed in the combustion process. This is the point at which the least amount of hydrocarbons are emitted from the exhaust, which is why the car is set to run as close as possible to this point. If you run the car a little over Stoich ( leaner ) the fule burns to early and you can get nitros-oxide emissions, but you can also get better mileage. Run it a little under Stoich ( richer ) and you can get more power, but emit more hydrocarbons as the oxygen is consumed before the all fuel is.


<------power-------stoichiometric------economy------>
12.5:1----------14.7:1------------15.6:1

Representation of AFR's only - don't take these are actual rules

So to do all of this, the management system must use some sort of feed back to help the ECU know what the engine is running like, so it can help control everything it needs to. Part of this is the Closed Loop fueling, that maintains the actual AFR / Lambda values at the stoichiometric point of the fuel used. This helps give good economy and power when cruising on the highway and motoring around town, where probably 90% of the engine life is likely to be. So it all makes sense to use this method of control.

The problem is that when you want more power, for acceleration, overtaking or towing, the engine needs more fuel to burn. If you are running in CL mode, the O2 sensors will not allow the ECM to run richer. So we have "Power Enrichment - PE " mode to allow the engine to run richer when needed. When the ECM goes into PE mode, it ignores the current O2 sensor inputs and uses set tables and possibly long term fuel trims to "estimate" the required fuel quantities needed to achieve the commanded AFR. This then allows for a richer mixture to give the additional power required.

As for "messing" with the VE table..... The general view of this is that at the very best, it's a crude hack, most would say you should never, ever "fudge" the VE table as there are a number of other settings that this will effect. To be be "correct" you should make the VE table "correct" and then use other tables to command the fueling needed. I was talking to Jezza about this earlier this week and we came up with the following list of some of the tables that are based on the VE tables...

Firstly, the PID DYN_CYLAIR is taken directly from the VE table. This PID tells the ECM the amount of air entering the engine in terms of gm/Cyl, tables that use this PID are....



The main spark tables.
Most, if not all, of the calculate engine torque tables ( think torque and traction control )
If an auto, this may effect the TCM tune.



So you can see that having an incorrect VE table can effect more than just the fueling. By fudging one table, you then need to fudge a few more to try and fix the things messed up by the initial fudge....... It's a vicious circle that's better not started. Of course I'm not saying things won't work at all and sometimes a "fudge" might be used to "prove a point" or "test", but in the long run, getting the base tables correct has a lot less headaches for a daily driver.

Regards,

Simon.

Word!http://forum.efilive.com/images/icons/icon14.gif

[gmh308]

The problem is that when you want more power, for acceleration, overtaking or towing, the engine needs more fuel to burn. If you are running in CL mode, the O2 sensors will not allow the ECM to run richer. So we have "Power Enrichment - PE " mode to allow the engine to run richer when needed. When the ECM goes into PE mode, it ignores the current O2 sensor inputs and uses set tables and possibly long term fuel trims to "estimate" the required fuel quantities needed to achieve the commanded AFR. This then allows for a richer mixture to give the additional power required.

As for "messing" with the VE table..... The general view of this is that at the very best, it's a crude hack, most would say you should never, ever "fudge" the VE table as there are a number of other settings that this will effect. To be be "correct" you should make the VE table "correct" and then use other tables to command the fueling needed. I was talking to Jezza about this earlier this week and we came up with the following list of some of the tables that are based on the VE tables...

Firstly, the PID DYN_CYLAIR is taken directly from the VE table. This PID tells the ECM the amount of air entering the engine in terms of gm/Cyl, tables that use this PID are....


The main spark tables.
Most, if not all, of the calculate engine torque tables ( think torque and traction control )
If an auto, this may effect the TCM tune.


So you can see that having an incorrect VE table can effect more than just the fueling. By fudging one table, you then need to fudge a few more to try and fix the things messed up by the initial fudge....... It's a vicious circle that's better not started. Of course I'm not saying things won't work at all and sometimes a "fudge" might be used to "prove a point" or "test", but in the long run, getting the base tables correct has a lot less headaches for a daily driver.

Regards,

Simon.

To Simon's point...even GM needs to "fudge" the VVE to get it right for all conditions. Depending on how they might be grouped and the engine type (DOD/AFM/VVT) there are between 10 and 20+ other tables to fine tune the 30 VVE zones for static/dynamic conditions to produce the right AFR's for starting at the very least and also to get the torque calculations right (for example) as these do affect auto trans shift parameters.

These aren't currently exposed in EFIlive but they can make considerable difference to fine tuning if they arent correct. (like difficulty starting even when running standard MAF/Closed Loop mode.)

And arguably as running MAFless in most cases has no advantage over a well tuned MAF tune (proved many times by Simon I would guess, which I would echo) in PE mode/open loop the main airflow calc is still taken from the MAF with the VVE & correction factors contributing to smoothing the MAF signal so yes it ends up being a "chase the tail" exercise if the VVE was fudged to provide PE type fueling.