I have read the thread on EBC but it was not clear what changes were made for MS3. Ken said in one of the posts that he changed the code for MS3 and the theory for tuning the EBC was different.

Can someone please explain what PID does and give me a starting point? I am looking for 18psi.

Thank you!

 

You have found me. I am 18psi

You're welcome

 

Boost Control

 

oh wow that's perfect.

+props

That's what I was trying to explain to you yesterday all written out and with pictures

 

Applies for latest MS2 firmware as well (what I used when I tuned my stuff up).

 

yeah, The other **** people were linking to on DIY website was the old MS2 stuff and I was afraid to use it. Thanks Aaron! That should be exactly what I needed!

 

Closed loop fuel requires 3 things:

- a 12x12 or so table of AFR targets vs RPM and MAP
- an open loop 12x12 table of injector on-times vs RPM and MAP (the VE table)
- and the O2 sensor and PID to provide feedback to trim the AFR

Why doesn't boost control do the same thing?:

- an 8x8 table of boost target vs RPM and TPS
- an 8x8 table of open loop table of boost solenoid duty cycle vs. RPM and target boost
- and PID to provide feedback to trim the boost solenoid duty


And then just like fuel auto-tune changes the VE table, boost autotune should do the same for the open-loop duty table.

 

Ooooh, ooooh, ooooh. I'd love that!

 

And then just like fuel auto-tune changes the VE table, boost autotune should do the same for the open-loop duty table.

 

I generally welcome and enjoy your contributions Jason, but this one is making me wonder if you've ever actually used or tuned a MS3?

MS3 has several options for boost control:
-Standard open loop
-Closed loop with 8x8 boost target table
-Feed forward closed loop with 8x8 initial duty % table and 8x8 boost target table

Other algorithms exist to augment/supplement boost control as well, such as boost vs time after launch, boost vs VSS, boost vs gear, and a couple traction control strategies -- but I think these can be ignored for this conversation.

I think if you tried one of our closed loop strategies, you will find that the MS3 does an excellent job of closed loop control, and the "autotune" feature has potential to be a very dangerous solution to a problem that doesn't exist.

 

I'd be curious to see datalogs of results, duty cycle, boost and boost target... flooring it at different RPMs, from different starting values of MAP.

 

O.K. Jason, here you go.

Nr. 1
stomp on the throttle keeping it at 100%
4. Gear



Nr. 2
lighter throttle,
6. gear



Pretty darn good, I would say...

 

I've looked and looked again and was unable to find the feed forward CL option. Is it in a newer FW release? I'm running one of the 1.2.x versions.

 

I thought that was still in the alphas.

 

Zaphod: tune your enrichment and look under the file menu for "save graph as PNG". the world will be a better place!

I'm kinda sad I never got to play with the new method of boost.

 

The initial duty option is in the 1.3 chain, which is still considered alpha.
Downloadable here:
Megasquirt firmware downloads

This firmware also includes alternator control.

 

And I thought I'd do some work on the house this weekend. Looks like I'll be installing FW and re-tuning instead :-)

 

Zaphod,

Thanks. Do you have a log where you went WOT at high RPM, such as a WOT upshift?

 

I'll have a look.

 

Here are mine:

Varying TPS ergo varying boost target. Note how it doesn't overshoot much and it hits the changing boost target quickly:



WOT'ing at different RPMs. Note how it hits full boost quickly with no overshoots even at different RPMs: