AlphaN problem with 1.1.8x and 1.1.9x and vemstune bug?

[z0tya]

Hello !
I have a test inline 4 engine with ITB-s, with connected map signal, but with pure alphaN strategy. And I don't understand how it works, or how need to work.
This is my logs:
cosworth.hu/misc/alphan/_0kpa_v3.3_n003420-2011.07.20-19.53.59.vemslog
cosworth.hu/misc/alphan/_110kpa_v3.3_n003420-2011.07.20-19.52.03.vemslog
Vemstune tuning gauge group seems confusing:


How does ecu calculate "VE"?
What about the row headers? (0-16-26-....-404-506)
Is it an interpolation error at the edge of the table?
I have tried with 1% tps in the lowest row, at first.
Vemstune help says with alphaN need change to after throttle - I see on injector calibration that alpha blending is 100% with this, but I don't know which kpa or tps lookup is good 0 or 100.

[rob@vems.co.uk]

The engine calculates VE based on the position of the throttle (TPS) (the numbers at the side of the table) and the RPM - TPS,RPM
For the TPS values I was recommended to use: 0, 2, 5, 12, 23, 35, 53, 75, 115, 165, 225, 255
Have you set the AlphaN blending correctly?

Rob

[z0tya]

I would not want blending, just pure tps.
I have tried 110kpa and 0 kpa in kpa or tps lookup in alpha n menu, but no difference. The map connection set to "after throttle"
I don't understand why show ve current arond 80 when I at 0 tps where 0s the table values.
I don't understand what meens the vertical row values on the Ve table on the screenshot. Why the pointer stay between 52-72 when tps 0, and when you open the Ve table from tuning menu there stay the pointer at the lowest (0) row.

[gunni]

The engine calculates VE based on the position of the throttle (TPS) (the numbers at the side of the table) and the RPM - TPS,RPM
For the TPS values I was recommended to use: 0, 2, 5, 12, 23, 35, 53, 75, 115, 165, 225, 255
Have you set the AlphaN blending correctly?

Rob

TPS values go from 0-100% now.

That RPM/MAP table is not the right one to modify.

[z0tya]

Any idea?

[Jim_W]

Just an oversight by the VT developers.  It appears that two tables are kept in VT.  When you open the VE table from the menu, the software looks at the firmware settings to find out which table (RPM/MAP or RPM/TPS) it should open.  When you open the predefined tuning views, this check is not performed and VT displays the RPM/MAP table.  The workaround is to simply open the VE table from the menu and do not tune using the RPM/MAP table.

[mattias]

Menu?
For each gauge group where the wrong table is used : Enter edit mode, right click table, select the correct table (RPM/TPS), enter view mode again, done.

Simply put, with Alpha-N use your eyes and check the  table outline -  if it says RPM/MAP, edit the gauge group and be done with it.

A work around to fix this was discussed long ago, it may come some day.

[Jim_W]

Alternatively, choose the correct table manually.  Open the view you want to change, click the Edit Mode button, right click the table, click Change table to, then select the correct table.

[z0tya]

It's okay but the bigger problem is the VE lookup and calculation:
See 110kpa log,  0 tps rpm around 800 and ve current is 100 at the first half of the log when kpa/tps lookup 110kpa.
Why?

[mattias]

Some things I noted:
MAP corr table should begin at 100 kPa, not at 0. Haven't tried to see what that gives..
Some features are turned on that don't make sense with your setup.
Ign+Lambda Y-axis should be MAP, or putting the MAP sensor hose after the throttle doesn't make sense.
You should use four ignition outputs, just repat the same output four times (single coil, distributor), this is why the recommended ref tooth table has "0,18,0,18" as the recommended setting.
Turn off all idle control features, not sure you're set up to use an idle air valve with Alpha-N at this point, nor idle ignition advance control.

[z0tya]

I think after / before throttle means only map correction.
Doc says:
after: map lower than barometric pressire will be ignored in fuel calc
before: map will be included in calc
(I think coders see the throttle from opposite direction )
I need alpha-n on ign and lambda too, so I changed axis to tps.

But I will try map corr axis hint.

[z0tya]

More log:
with 1.1.75 (this is older alpha-n strategy)
cosworth.hu/misc/alphan/alphan_test_1.1.75.v3.3_n003420-2011.08.06-12.02.13.vemslog
with 1.1.96
cosworth.hu/misc/alphan/alphan_test_1.1.96.v3.3_n003420-2011.08.06-12.18.12.vemslog
cosworth.hu/misc/alphan/alphan_test_acc_1.1.96.v3.3_n003420-2011.08.06-12.19.18.vemslog

It seems something very wrong with the VE interpolation, like getting table data from one row above as tps bin says:


I have no idea :/

[GintsK]

Doesn't your problem comes from Alpha-N IAC compensation? IMHO it is designed to add some IAC dependant VE value.

Gints

[Ignitec]

I did that this way

closed iac
open throttle enough for 900 rpm, remembered tps bin value from tps vs rpm
opened iac, and wrote tps bin to alpha-n iac compensation, and it worked.
i think trick is explaining to vems how much would tps be opened if there isn't iac, because you dont have map to measure anything...

you can change whatever you want in table while in idle, it wont change until you change ve in position with actual rpm and alpha-n iac compensation. when out of idle, everything works

[rob@vems.co.uk]

Good point, I had this with my SR20DE engine on throttle bodies, messing around with and IACV showed me that there needed to be some extra fuel control related to IACV airflow.