Boost PID

[le mans]

Does anyone know what the is the mathematical relationship is of the PID numbers in M232 to the base WGDC?

For example, in a Link ECU this is how the math works (cut and paste from their help file):

Proportional Gain

This adjustment sets the proportional gain of the PID algorithm. For a Proportional Gain of 1.0 and an error value of 1 kPa the output duty cycle will be adjusted by 0.1 %.

Example

Set this to a value to control the aggression or attack rate of the boost pressure. A typical value for this would be 3.0, but may vary largely depending on the turbo/wastegate/engine combination.



Example

With a base duty cycle of 50%, a proportional gain of 5.0 and a current error of 10kPa, the base duty will be trimmed to 55% = (50 + (5.0 * -10 * 0.1%)



Hint

Increase this value to decrease spool time; increase it too far and the boost will overshoot the target. This overshoot may be counteracted by increasing the Derivative Gain.



Integral Gain

This adjustment sets the integral gain of the PID algorithm. For an Integral Gain value of 1.00 and an error value of 1 kPa the rate of increase in output duty cycle will be 5.0 % /sec.

Example

Set this to a value to control the target offset correction of the boost pressure. A typical value for this would be 0.3, but may vary largely depending on the turbo/wastegate/engine combination.

Hint

Increase this value for better offset correction; increase it too far and the boost pressure will oscillate.



Derivative Gain

This adjustment sets the derivative gain of the PID algorithm. For a Derivative gain value of 10.0 and a rate of change in error of 1.0kPa/sec the output duty cycle will be adjusted by 0.5 %.

Example

Set this to a value to control the target overshoot of the boost pressure. A typical value for this would be 3.0, but may vary largely depending on the turbo/wastegate/engine combination.

Hint

Increase this value for better overshoot protection; increase it too far and spool time will decrease.

[le mans]

Anyone? I've also read about the D term 'flipping' in M232 - curious about that one too.

[jonn48]

Does anyone know what the is the mathematical relationship is of the PID numbers in M232 to the base WGDC?

For example, in a Link ECU this is how the math works (cut and paste from their help file):

Proportional Gain

This adjustment sets the proportional gain of the PID algorithm. For a Proportional Gain of 1.0 and an error value of 1 kPa the output duty cycle will be adjusted by 0.1 %.

Example

Set this to a value to control the aggression or attack rate of the boost pressure. A typical value for this would be 3.0, but may vary largely depending on the turbo/wastegate/engine combination.



Example

With a base duty cycle of 50%, a proportional gain of 5.0 and a current error of 10kPa, the base duty will be trimmed to 55% = (50 + (5.0 * -10 * 0.1%)



Hint

Increase this value to decrease spool time; increase it too far and the boost will overshoot the target. This overshoot may be counteracted by increasing the Derivative Gain.



Integral Gain

This adjustment sets the integral gain of the PID algorithm. For an Integral Gain value of 1.00 and an error value of 1 kPa the rate of increase in output duty cycle will be 5.0 % /sec.

Example

Set this to a value to control the target offset correction of the boost pressure. A typical value for this would be 0.3, but may vary largely depending on the turbo/wastegate/engine combination.

Hint

Increase this value for better offset correction; increase it too far and the boost pressure will oscillate.



Derivative Gain

This adjustment sets the derivative gain of the PID algorithm. For a Derivative gain value of 10.0 and a rate of change in error of 1.0kPa/sec the output duty cycle will be adjusted by 0.5 %.

Example

Set this to a value to control the target overshoot of the boost pressure. A typical value for this would be 3.0, but may vary largely depending on the turbo/wastegate/engine combination.

Hint

Increase this value for better overshoot protection; increase it too far and spool time will decrease.

Probably only prj knows this, he made the xdf file. Perhaps there are differences from the usual PID settings, since sometimes there are problems with the stability of the boost and an incomprehensible reaction to changing the PID parameters.

[prj]

There has been full damos leaked for this, and then I also saw a full FR.
So technically it is possible to answer all the questions about this ECU. Should it be done though? I don't know.

But I think the paper damos was posted here so you can check the addresses and verify what is what. All my stuff was reverse engineered without having any data at all. I am sure the "P term" is correctly labelled, and I tested that many many times, I am not as sure about I and D.