Boost control tuning

Tuning, troubleshooting and the nitty gritty of using rusEFI to make your engine run nicely!
Post Reply
blundar
contributor
contributor
Posts: 141
Joined: Tue Jan 07, 2014 4:38 am
Location: Cincinnati, Ohio
Github Username: blundar
Slack: Dave B.
Contact:

Boost control tuning

Post by blundar »

Jimmy asked me about how to go about tuning boost control for his twincharged merc.

This is the general procedure I would follow:
1. Figure out the LEAST amount of boost you want the turbocharger to run. Put this wastegate spring in the WG.
2. Unplug electrically any boost control solenoids. You want the wastegate connected directly to a pressure source.
3. Drive the car at WOT. There are several things you're interested in:
a. What RPM does the turbo start to build any boost in a low gear (1st, 2nd)? What RPM does it stabilize?
b. What RPM does the turbo start to build any boost in a high gear (i.e. 4th)? What RPM does it stabilize?
c. Once the turbo spools up to a steady boost figure, what does the shape of the boost curve look like in a moderate-higher gear (3rd, 4th) where you're not spinning the tires? Possible answers: flat as a board, drops off close to redline, starts increasing before redline and is higher than the initial steady-state value

At this point, you can start deciding how to use boost control. You need to think about GOALS before putting in values. Are you looking to spin the tires as soon as you can as much as you can? Are you looking to make the car accelerate as much as possible? Can the car maintain traction at maximum boost in 4th gear? 2nd gear? 1st gear?

Let's say you are at lea level (atmospheric = 100kpa) and your car has an external wastegate and starts to make positive pressure at 2500 RPM and is at 165kpa by 3900 RPM. It then holds 165 kpa to redline of 7000 RPM. Let's also suppose that the rods can probably deal with about 250kpa of boost before having issues and the driver wants the car to have all the jam, all the time.
What I would recommend:
1. Set the boost cut to a really low value (i.e. 110kpa) that you will definitely be able to hit. Hit it on purpose. MAKE SURE THE ENGINE IS STOPPED BY THE BOOST CUT.
2. Once you have verified that it works, set the boost cut 10-15kpa higher than your target boost level
3. Disable closed loop boost duty adjustment by setting gains to zero.
4. Make an educated guess about how much duty you'll need. A rule of thumb is that you will be able to make 3x your wastegate with an external gate and 2x with an internal gate. I.e. wastegate boost = 165kpa = 100kpa atmosphere at sea level + 65kpa boost, 65kpa * 3 = 195kpa so 100kpa + 195kpa boost = 295kpa absolute as an estimate. THIS IS A GUESS. We're going to go for 230kpa target. 230kpa = 100 atm + 130kpa boost. 130kpa boost / 65kpa spring = 2.0 2.0 / 3.0 max(estimated) = 0.67 Drop it down a little bit for safety and call our guess 65% Set the boost target table to 65% everywhere.
5. get the car to 4500-5000 RPM, steady state cruising in 3rd or 4th gear, as you feel comfortable. ROLL into the throttle and see if you either hit boost cut or manage to accelerate with the boost stable. Note how much boost you have and how closely you are to target.
6. If you hit boost cut, drop target. If you stabilize at a lower target, bump duty up a bit. Go back to step 5 until you are reasonably close (5-10kpa) from your targets.
7. Repeat steps 5-6 but choose a higher RPM. Map out duty cycle values to achieve your target at the alternate RPM. Pro tip: If you find that you need to add a lot of duty to maintain the same boost level at higher RPMs, you're probably running out of compressor. If you find that you need LESS duty at higher RPMs ("boost creep") you probably need a larger wastegate.
8. Repeat steps at a lower boost target. Halfway between your wastegate (165 kpa) and high boost target (250kpa) is probably a good value. Map out the duty cycle required to achieve your target
9. linearly interpolate where WG = 165kpa = 0% and your "medium boost" level according to load (vertically I think??). Interpolate between your measured "medium boost" and measured high boost. Maybe do a little smoothing.
10. Turn on closed loop boost control.
11. Choose targets that suit you
12. Hopefully have your car hit your targets.

To maximize spool, you want to hold the boost control solenoid completely closed as long as possible. Many boost control systems have a "quick spool" option that holds the solenoid closed until a certain pressure. This tends to help the turbo spool faster. It's not unreasonable to see a 500-1000 RPM increase in spool time, especially with small low pressure wastegate springs. This can also be accomplished / hinted at by setting the open loop wastegate duty table to 100% for the entire area under normal spool RPM. In our example, we had positive at 2500 and WG at 3900. If you set the open loop table to 100% from say 2000 RPM -> 3800 RPM for all boost pressures, this is a way to poke the ECU into spooling the turbo faster. You will need to play with the exact RPM because setting it to the same as where your turbo spools without help will tend to overshoot boost targets. MAKE SURE YOU HAVE A BOOST CUT SET AND WORKING BEFORE PLAYING WITH TRYING TO MAKE A TURBO SPOOL FAST!

At this point, you will have a car "set to kill." In most cases, you're not going to want "all the jam" all the time - you won't be able to keep traction in lower gears. Once you have a boost control system where you can get what you command, you can start figuring out what to ask for. I don't think that RE has a lot of this as of the time of writing, but things like this have been really helpful on other systems:
TPS vs RPM = table to scale boost targets (so you will drop to a lower boost by taking your foot off the gas)
gear x rpm = boost table for boost targets specific to each gear
gear ratio targets used to scale target boost by difference in mechanical advantage vs. 4th gear
gear ratio vs. multiplicative factor table used to adjust overall boost targets by multiplying by a different factor for each gear

More later..
mx5newbie
Posts: 6
Joined: Sun Mar 20, 2022 4:14 am

Re: Boost control tuning

Post by mx5newbie »

Can you tell me what values to start with for tuning closed loop boost control? I have my open loop pretty much dialed in like I want it.
1990 Miata w/Mamba Td04-13T turbo, FIC 775cc injectors
mck1117
running engine in first post
running engine in first post
Posts: 1337
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Boost control tuning

Post by mck1117 »

mx5newbie wrote:
Thu Jun 30, 2022 2:50 am
Can you tell me what values to start with for tuning closed loop boost control? I have my open loop pretty much dialed in like I want it.
What values where?

Here's an example of a tune with pretty good boost control (I've endurance raced on it with good control!): https://rusefi.com/online/view.php?msq=969
mx5newbie
Posts: 6
Joined: Sun Mar 20, 2022 4:14 am

Re: Boost control tuning

Post by mx5newbie »

mck1117 wrote:
Thu Jun 30, 2022 3:45 am
mx5newbie wrote:
Thu Jun 30, 2022 2:50 am
Can you tell me what values to start with for tuning closed loop boost control? I have my open loop pretty much dialed in like I want it.
What values where?

Here's an example of a tune with pretty good boost control (I've endurance raced on it with good control!): https://rusefi.com/online/view.php?msq=969

The PID values. They seem to be a little different when compared to MSQ users I see online.
1990 Miata w/Mamba Td04-13T turbo, FIC 775cc injectors
User avatar
AndreyB
Site Admin
Posts: 13000
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Boost control tuning

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
mck1117
running engine in first post
running engine in first post
Posts: 1337
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Boost control tuning

Post by mck1117 »

mx5newbie wrote:
Thu Jun 30, 2022 10:53 am
The PID values. They seem to be a little different when compared to MSQ users I see online.
MSQ users of what? rusEFI?
mx5newbie
Posts: 6
Joined: Sun Mar 20, 2022 4:14 am

Re: Boost control tuning

Post by mx5newbie »

mck1117 wrote:
Thu Jun 30, 2022 5:53 pm
mx5newbie wrote:
Thu Jun 30, 2022 10:53 am
The PID values. They seem to be a little different when compared to MSQ users I see online.
MSQ users of what? rusEFI?
Was speaking from a Megasquirt perspective. As I’ve read, they suggest starting with a high number in P, like 185. I’m trying to figure what is considered a high number in rusEFI? My unit came loaded with the values as shown below. Hope I’m making some sense here.

P = 0.500
I = 0.300
D = 0.000

Min adjustment = -20
Max adjustment = 20
1990 Miata w/Mamba Td04-13T turbo, FIC 775cc injectors
mck1117
running engine in first post
running engine in first post
Posts: 1337
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Boost control tuning

Post by mck1117 »

I have no idea how the megasquirt's PID boost control works.

The rusEFI values are all in understandable units:

P = % per kPa error
I = % per kPa-seconds
D = % per kPa/sec

So:
If the boost error is 5kpa, and your P is set to 2, it will make a correction of 10%.
If the boost error has been 5kPa for 2 second (10 kPa-seconds) and I is set to 0.1, it will make a correction of 1%
If the boost is changing at 10kPa/s and D is set to 0.5, it will make a correction of 5%.

Those three terms are then added together, and limited by the PID min/max settings.

That is then added to the value from the open loop table, and sent to PWM the valve.
mx5newbie
Posts: 6
Joined: Sun Mar 20, 2022 4:14 am

Re: Boost control tuning

Post by mx5newbie »

mck1117 wrote:
Thu Jun 30, 2022 8:09 pm
I have no idea how the megasquirt's PID boost control works.

The rusEFI values are all in understandable to units:

P = % per kPa error
I = % per kPa-seconds
D = % per kPa/sec

So:
If the boost error is 5kpa, and your P is set to 2, it will make a correction of 10%.
If the boost error has been 5kPa for 2 second (10 kPa-seconds) and I is set to 0.1, it will make a correction of 1%
If the boost is changing at 10kPa/s and D is set to 0.5, it will make a correction of 5%.

Those three terms are then added together, and limited by the PID min/max settings.

That is then added to the value from the open loop table, and sent to PWM the valve.
Awesome. That’s what I needed. Loving the ECU.
1990 Miata w/Mamba Td04-13T turbo, FIC 775cc injectors
User avatar
MHTSOS
Posts: 57
Joined: Sun Jun 02, 2019 6:46 pm
Github Username: MHTSOSVOODOO
Slack: Dimitri C

Re: Boost control tuning

Post by MHTSOS »

This explanation of PID has to be in the Wiki if it's not already there.

Στάλθηκε από το VOG-L29 μου χρησιμοποιώντας Tapatalk

Post Reply