Boost control change testing (MS2 and MS3)

[sheek]

ok i will use normal so more duty =more boost.


this maybe the issue why my closed loop is erratic

[Barton]

I have the closed loop boost controller working in a eagle talon (4g63 1g) at 24psi but using the 1.2.3 code, and it is absolutely impressive, I got 24psi from a precision 6765 at 5500rpm, the closed loop made the turbo spool impressively fast, if I remember correctly that code already used the max duty to get the boost target faster, but maybe this way is even better, my only concern is there should be a specific value to let the code know when to start the control:

for example my case:
wasted spring: 15psi
boost target: 24psi
heavy car with low compression engine and big turbo (almost all the time the load is between 80kpa and 120kpa) to drive around

the wastegate won't open until at least 10psi (without using any control), but the code goes to 100% of duty from vacuum and the boost solenoid gets hot because it is always trying to raise the boost, but if I use a value to start that feature depending on my setup I wouldn't be worried about the boost solenoid.

[sheek]

i could be wrong but i think that is addressed by using the delta.

[Barton]

i could be wrong but i think that is addressed by using the delta.

the delta is the value to start the pid, not the feature to spool faster, I'm concern about that because the solenoid goes really hot.

[muythaibxr]

It is meant to spool faster and start PID.

[ol boy]

Can I get the MS2 version of this different boost control? Is it in the beta6 release?

[M30guy]

There is so much post to read through and I'm sorry that it has been answered before
but I'm using 3.2.4 ms2
and the boost control was very easy to tune, and boost is maybe 1/2psi off but just going 1/2psi higher
on the table is good enough and everything is rock solid, no oscillation.

What does this new code fix exactly? If I use it and retuned what does it do better?
The "creep" is tuned in and what I see is what I put into the kpa tables. I love it.

This website crops the picture so you can't see what the lines are. Here is the link to pic.
http://img.photobucket.com/albums/v163/ ... fa74fa.png

[muythaibxr]

The main thing it should do is spool more predictably and repeatably on engines or setups that are more difficult to tune.

[jsmcortina]

Can I get the MS2 version of this different boost control? Is it in the beta6 release?

It doesn't exist on MS2.

James

[muythaibxr]

Not entirely true. There is a version that has the initial value table. The change to make that standard PID + a bias table has not yet been made. I don't know if any of the betas have even the initial value table yet though.

[Zaphod]

Could we please get a "Sticky" with a tuning advice for the new boost control.
I am on the latest beta and I think I got the Initial value table reasonable well, but now i need to fine tune the PID and it's nice not to need to search through all the 12 pages of this thread (also because I think this thread is about two different stages of this changed boost control algorithm...)

Thanks in advance!

[tutuur]

So is there or is there not a version of ms2 with new boost control algorythm?

I would like to tune closed loop now as i have open working pretty well but as you all now it's pretty sensitive to temp changes. I have worked with the current closed loop but reading here the new/newer algorythm should work way better....

[muythaibxr]

MS2 has the same algorithm that's in MS3 1.3. It does not to my knowledge have the version that's in the MS3 1.4 beta.

[tutuur]

James and I have made the changes we have been discussing in the MS2 and MS3 sections to the boost control behavior:

MS2:
- Added delta setting and initial values table
- Made boost solenoid stay at 100% duty until (target - delta is reached) before jumping to the initial value duty and enabling PID.

MS3:
- Made boost solenoid stay at 100% duty until (target - delta is reached) before jumping to the initial value duty and enabling PID.

How to tune/test:
- Turn on closed loop boost control
- Turn on the initial values table
- Set your delta as low as you can (around 10-15 % load probably, make it larger if you need to in order to avoid overshoot)
- Make sure overboost protection is configured and working.
- Set control interval to about the same as valve frequency (so 19.5 Hz should be ~51ms control interval).
- Set tuning mode to "advanced"
- Set PID gains to all 0
- Tune the initial value table until you are spooling the way you like and reaching the target at various target values and RPMs
- Tune the I term to hold the target after PID is enabled.
- Tweak the initial values table to hit the target you want initially
- Add P and D term if needed to avoid overshoot and react to changes in throttle position after PID engages.

Please send me a PM if you want test code. Include the following:
- Your email address
- MS2 or MS3?
- screenshot or log of current boost behavior.

Keep in mind that this should be considered "test code." It has changes that we have only tested on the bench and have not tried in a car.

Discuss in this thread.

Ken

so that is this code then? could you e-mail me the ms2 version? i really like the idea of the initial value versus pid all the way.
ntw, does this still work on base of inverted control or normal as i know there has been a change?

[tutuur]

James and I have made the changes we have been discussing in the MS2 and MS3 sections to the boost control behavior:

MS2:
- Added delta setting and initial values table
- Made boost solenoid stay at 100% duty until (target - delta is reached) before jumping to the initial value duty and enabling PID.

MS3:
- Made boost solenoid stay at 100% duty until (target - delta is reached) before jumping to the initial value duty and enabling PID.

How to tune/test:
- Turn on closed loop boost control
- Turn on the initial values table
- Set your delta as low as you can (around 10-15 % load probably, make it larger if you need to in order to avoid overshoot)
- Make sure overboost protection is configured and working.
- Set control interval to about the same as valve frequency (so 19.5 Hz should be ~51ms control interval).
- Set tuning mode to "advanced"
- Set PID gains to all 0
- Tune the initial value table until you are spooling the way you like and reaching the target at various target values and RPMs
- Tune the I term to hold the target after PID is enabled.
- Tweak the initial values table to hit the target you want initially
- Add P and D term if needed to avoid overshoot and react to changes in throttle position after PID engages.

Please send me a PM if you want test code. Include the following:
- Your email address
- MS2 or MS3?
- screenshot or log of current boost behavior.

Keep in mind that this should be considered "test code." It has changes that we have only tested on the bench and have not tried in a car.

Discuss in this thread.

Ken

so that is this code then? could you e-mail me the ms2 version? i really like the idea of the initial value versus pid all the way

[jsmcortina]

could you e-mail me the ms2 version? i really like the idea of the initial value versus pid all the way

It hasn't been written yet, so no.

James

[tutuur]

just upgraded to 3.3.2 beta 4 and noticed the initial value table is in there

[yellow-fever]

Hello all.
Had open loop working pretty good, so I figured why not try 3.3.2 closed loop on MS2extra. Running an extremely small turbo that spools fast and huge open tial wastegate with 5lb spring with DIY valve set up so more duty is more boost. Starts to creep open early due to the peanut of a turbo and 5lb spring and I have found it beneficial to hold 95% duty cycle at higher rpm to hold it shut as long as I can before the turbo falls off the map and gate creeps open. I think the only reason this setup is still alive with the tiny turbo on a 2.7 straight 6 is because the huge external wastegate that creeps open at higher rpm even if its set to 95% duty "wonder what would happen with a 4 port wastegate valve taking pressure away from the bottom of the diaphragm at the same time its adding it on top, I probably don't want to know? LOL! Crossed over my open loop duty table to the initial values table. Set boost limit at 200kpa with 10kpa hysteresis. Closed duty 5 and open duty 95. Target table 190kpa all bins, lower delta set to 10kpa. Boost tuning mode set to basic with the slider at 200. It is working pretty well and I know I can tweak the gains in advanced mode to hold my target more closely after pid is engauged, but I'm not sure if what I am seeing is normal operation for X reason or code that could improve closed loop operation? What is happening is duty cycle is at my max 95% for fast spool, when delta is hit at 180 the pid instantly drops down to min 5% and then up to max 95% before it drops back down to around 40% under the pid control. After the initial oscillation, I know I can tweak the gains with advanced control, but the initial max, min, max, then finally pid is bugging me and you can see it in the data log at the beginning of every pull in nearly every gear. I was under the impression that as soon as target delta was hit "180 in this case with delta set to 10kpa with 190 kpa target" it should go to where the initial duty table is set to and control the pid from there? This drop down to min duty of 5% and back up to max 95% before the pid settles down near where it should be around 40% causes a flat spot in the boost at 5% and then when it jumps back up to 95% causes a spike before it settles back around to near where it should be around the initial duty table. Should the drop to min duty and back to max duty when pid engages after delta is hit, normal and why, and would it not work better if after delta was hit it went instantly to where it should be on the initial duty table and control with pid from there vs jumping low and then high and then to initial duty like all my data logs show? I have been through my logs and have not found many instances where it needs less than a 40% duty so I am going to set my min duty to 40% vs the 5% to try to get rid of the flat spot, and change delta from 10 to 15kpa to see if I can calm down the slight overboost just after 95% duty cycle after the flat spot when it should be 40ish if the flat spot was gone. When delta is hit, should it go to min duty to max duty and then to initial duty table setting? Does this need to happen for some reason for the pid to work correct after pid engagement at delta. Sorry for the long winded post, but as you can see the 95% to 5% to 95% to initial table then pid control is kind of strange operation that happens every time I hit delta at 180kpa on the way to 190kpa target like clockwork. If it stops raining I will make the changes mentioned above and log a few good pulls. It should help greatly if I understand this correct? Sorry for the long winded post. Why didn't anyone warn me that boost and megasquirt was worse than crack! LOL! Discussion?

-Dan

[yellow-fever]

image.jpeg

[pit_celica]

Please post the actual log instead of a picture. If the log is too big, try to zip it.

I think that it goes back to 95% because the MAP goes under the 180kPa (under the delta). Try to increase your delta. 10kPa before the target isn't that much. When you reach this 180kPa, the code sees that the boost is near the target (because of the low delta) and that the boost is still rising fast, so it freaks out and command a 5% duty. This cause a drop in MAP under the 180kPa and this is where it goes back to 95%.

An improvement on this could be a kind of hysteresis when the target-delta MAP load is attained.

Again, try to increase delta. On a fast spooling turbo like yours, I'm pretty sure you won't notice the difference in spool and you will gain control.

Sam