Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   MEGAsquirt (https://www.miataturbo.net/megasquirt-18/)
-   -   Closed Loop Boost Control Issues (https://www.miataturbo.net/megasquirt-18/closed-loop-boost-control-issues-109418/)

tfbmiata May 18, 2024 04:15 PM

Closed Loop Boost Control Issues
 
6 Attachment(s)
Hi All - I have an nb2 w/ a goal of 170kpa. I've been trying to tweak out a closed loop setup based on the information I've found here. I'm attaching my recent tune/logs. The issue I'm having is the 'basic' setup is still overshooting by quite a bit. My aim is 170kpa and my boost cut is 185kpa. I moved the slider bar all the way down to 150 for sensitivity, but I decided to stop as I've seen others post their setup of 250+ and they said theirs gets up to PSI and just stays without overshoot. Is it as easy as continuing to reduce the sensitivity slider?

When I go back to open loop it seems that depending on the heat soak of the engine, a 36% duty cycle will get my 170kpa or sometimes need a 40% duty cycle for the same. That is why I'm trying to get closed loop working. So anyhow, hopefully, that's enough info. I'm attaching my log and setup if anyone can help me figure out why I'm still overshooting by so much.

EDIT: Anyone ever thought about putting the boost controller in the area between engine bay and firewall? I have it tucked up under the driver's headlight and its still getting to 170F. I noticed the only "cold spot" is that place where I stick my ratchet between outside of the engine bay before it gets to the windshield. Hopefully I'm describing that location right? It was only like 95F after about an hour of street tuning.

Watterson02 May 20, 2024 12:58 PM

I couldn't get the basic control to work without fluctuations. Maybe it would work if you keep reducing the gain slider, but I still don't think the results would be optimal. It is my understanding that basic just sets the PID gain values to 100/100/100.

tfbmiata May 20, 2024 01:05 PM

2 Attachment(s)

Originally Posted by Watterson02 (Post 1650696)
I couldn't get the basic control to work without fluctuations. Maybe it would work if you keep reducing the gain slider, but I still don't think the results would be optimal. It is my understanding that basic just sets the PID gain values to 100/100/100.

I meant to update my thread over the weekend. I did eventually get basic to work. I had to reduce my bias duty table quite a bit lower than my open loop table and also ended up changing my boost target table (which I'm sure was also wrong) and reduced sensitivity quite a bit. I essentially have closed loop turn on close enough. My target is 170kpa and I would see it get as high as 175 to 180 and then drop to about 165 and then zero in. I'm sure I could tweak that out further. I here's my last tune and log for anyone curious. I'd give it a B+ or A-. I will lastly admit that while the car and oil were nice and warm, the boost controller itself (EBC Mac) didn't see nearly the temperatures it usually does when I'm done tuning (usually 175F but only was about 120F). This might have contributed to why I needed to reduce the boost bias duty table so much, but again once I did everything worked pretty grand. I'm going to be relocating to the EBC to the gap between the firewall and the cable below the windshield this week to see if I can keep the temperature range on the EBC much narrower.

tfbmiata May 22, 2024 10:17 AM


Originally Posted by Watterson02 (Post 1650696)
I couldn't get the basic control to work without fluctuations. Maybe it would work if you keep reducing the gain slider, but I still don't think the results would be optimal. It is my understanding that basic just sets the PID gain values to 100/100/100.

So right now on basic,I'm at about 80 on the sensitivity. My boost target is 170kpa and the closed loop turns on at 25kpa range. I noticed that the initial spool up is fast and then I over shoot to about 175 to 180kpa and then it comes back down to about 165 until it evens out. Does that sound reasonable?

redursidae May 22, 2024 10:37 AM

Yes, that's reasonable. You could try a higher delta, 30 or 35 kPa, and see if it gets you closer.

tfbmiata May 22, 2024 11:18 AM


Originally Posted by redursidae (Post 1650801)
Yes, that's reasonable. You could try a higher delta, 30 or 35 kPa, and see if it gets you closer.

Is it fundamentally possible to have it quickly come up to boost and then flat line at your limit or would I be chasing a pipe dream?

redursidae May 22, 2024 11:27 AM

Yes, in theory. But in practice, you’ll get it perfect for one day, at a certain altitude and temperature, and it’ll behave differently another day at different temperature and altitude. That’s why closed-loop is a thing. If it’s close and it’s consistent, I would say leave it alone.

Watterson02 May 22, 2024 11:27 AM

One might argue that you could get closer to what you're describing by actually tuning the PID gain values, but if you want the valve to stay completely shut until it is right at the target pressure for theoretically perfect spool then yeah I believe yeah that might be a pipe dream. Maybe if you had an actual electronically controlled wastegate, but I doubt it with a pneumatic one. It should already be better than the Open Loop though, so at least there is that. If you do go and decide to tune the PID terms, make sure to back up your current tune.


All times are GMT -4. The time now is 07:38 AM.


© 2026 MH Sub I, LLC dba Internet Brands