Afternoon All,

Currently mid mapping my car, my tuner is a remote mapper so we've been doing it in stages. So far I've had no issues of overboosting or any issue other then plumbing up the lines for the boost controller incorrectly which was a brain fart moment as i did the same with my last build.

For specs I have the following.
Kraken Top mount manifiold with a 3" downpipe
A Garrett T25 W/ Hybrid core W/ 10 psi spring mounted
540CC bosch EV14 injectors
Audi R8 COPs
Genuine Mac 3 port boost controller
ME221 Gen 2 ECU
AEM X series wideband

Basically in the last 2 maps in anything over 3rd gear at 100% throttle its building to high revs and then just does a massive overboost for a split second and then comes back and goes again.
I reverted to one of the previous maps and I can get through all the gears fine but theres not as much boost. My tuner has said he things its an actuator or something similar, I have checked the actuator and its not sticking or getting stuck.
I'm unsure what else to check.

Thanks for reading and thanks for any assistance or suggestions.


edit - attached no overboost and overboost logs - 11:43 is map 4 with no overboost - map 6 is with overboost 11:16

 

Post a log of it overboosting and the tune where it over boosts.

Bonus points if you post the old tune where it doesn't overboost

 

I have all of that, but they are too large to upload, the calibration file is 1Mb which is fine but both the non overboosting and overboosting files are over 20mb

 

If you compress it into a zip file it might fit, or upload them on separate posts. You could likely trim the logs down to the important sections

 

https://drive.google.com/file/d/13HP...w?usp=drivesdk

try this.

to add, my tuner has said it’s not boost creeping it’s overshooting.

 

Yeah, I'm not surprised. The boost control is setup so the boost controller doesn't start working until 160 KPA. In one of those logs there's about 0.2 seconds between 160 and 205 KPA when you hit boost cut. You can see the map start to drop ar





Personally I don't have experience on ME ECU's but it also looks like the initial duty cycle table for the CL boost control is completely untuned, but the setting is checked to use that table. I'd kinda expect the Y axis to be target MAP and not TPS, but again I'm not sure how ME does these tables compared to others.






IMO, enabled boost control earlier and get that initial value table tuned. Instead of spooling at 50 (looks like that was lowered to 47 after that log) maybe try 40. I doubt that will be perfect but it should at least give the closed loop time to figure out what's going on.

*I am not a pro tuner, so take it with a grain of salt

 

Seems to me that the Inactive Duty should be 0. Your controller is being held at 50% duty all the time off control, then a drop to 47% duty above 160kPa. Not only will that make it over-spool with your other settings as SimBa pointed out, but your valve is probably clicking away even at idle and won’t have a long life.

 

I can see a non-zero inactive to help spool up, but yeah, no point around idle/ cruise. I enable my boost controller at 50 kPa which is probably a bit lower than I need, but I like to get the turbo spooled up pretty aggressively.

I didn't capture it in that screenshot but in second gear it dropped from 50% duty to around 43% duty, fell well below the boost target and slowly started ramping duty in, so I'd say it's just poorly tuned base settings and likely poorly or untuned PID as well.

 

Hey guys, thanks for the comments.

So bit of convo with my tuner later, I redid the vac lines and it got worse.

my tuner is under the impression that the actuator has an issue and the wastegate isn’t opening fast enough.
so I’m continuing to look into the mechanical side of this.

 

I would be interested to see a log of the boost curve with the controller disabled. It did seem like 3rd gear built boost pretty rapidly in that log.

It would also be interesting to see what open loop around 30% duty cycle. On my turbo for boost levels closer to wastegate pressure I basically pull all duty cycle from the controller as I near redline because it creeps like crazy.

 

I see weird PID settings...P = 1, I = 700? Looks like Boost Cut is 205 kPa and Boost delta is 20 kPa. You could try setting Delta higher but if your Duty Cycle table is all set to 47, you need to focus on finding what DC works at each RPM to hold your target and/or set lower targets lower TPS inputs (X in the DC chart?).

Let me repeat the knowledge I gained reading another long discussion thread here about tuning closed loop EBC. First you should get a baseline data log of what your turbo setup is capable of. Set your boost target and disable PID controls either using open loop or with TunerStudio, I selected closed loop in setup mode. If open loop, set duty cycle to 100 (if your vacuum lines are connected properly, 100 should be closed, 0 open). so the valve is fully closed. Now find a desolate road or long on ramp with little traffic. Get moving and into your 1:1 ratio gear (5th with the 6-speed, might be 4th or 5th on the 5-speed?) at ~2000 RPM, maybe 2500 RPM and go WOT recording data up to your target boost. Then let off and stay out of boost until you can pull over and stop the log and/or reset the DC table to a more sane number. I'd recommend setting it at 10, then 20 and seeing what happens. With your 10 psi actuator you'll quickly find the minimum DC required to raise above 10 psi wastegate. That becomes your PWM minimum duty.

Read that data log, or compare 2-3 pulls and average the results. Use that data to populate your target table. Set boost targets and RPM columns based on how much boost and when the log shows you hit those levels. This gives the PID loop (if using closed loop) real world data to work on and avoids the I term ramping up if your delta setting is higher. You can find yourself in higher gears where the turbo cannot realistically achieve a high boost target you might have fed the loop in the target table. You want real world targets and accurate DC settings across the table.

Now change to open loop or CL setup mode and set your boost target to 10 psi or slightly higher, add that row into your duty cycle table and change DC in each RPM column (can set the entire table to the same in this first step) until you hit target. You want DC that minimizes overshooting target, even staying slightly below target rather than slightly over. You edit the DC table for your targets raising DC slowly at your next boost level until you have covered up to your final target boost level. CL PID uses that DC table to make adjustments, if it's off (mine was 2-3% too high after changing to the new, higher flowing FM manifold and outlet pipe) CL EBC will struggle to hit and hold target. Mine was spiking and falling below target as the PID loop quickly removed DC after spiking 1-2 psi over target. Then it wouldn't recover, or if it did boost climbed slowly and redline often came faster than it could recover to target. My 18 psi target was actually spiking to 19-20, then dropping to 16.

Yesterday I retuned my CL EBC settings. After dialing in the DC table for my newer FM parts it only over-shoots by .5 psi, moves just under target, then back and holds to redline. DC in the table wasn't even across the rev range for upper boost target rows, it needed to be lower at 4500, just back to wherever it was just before 4500 RPM, and ramp up higher towards redline. My setup is also on the edge of what works with a 7 psi actuator (GT2560R) running 18 psi max. That's slightly over the 2 - 2.5x ratio reported to be the maximum range using a common MAC valve (3 port I believe). You need to find the correct DC and realistic target settings. I'm currently running basic mode CL EBC, will now switch back to advanced and tune the P, the I, and finally the D terms separately to see if I can't get it working even better than the defualt (100, 100, 100). I only switched back to setup mode and then basic to dial it in and test, I ran advanced mode for more than a year before upgrading the FM parts and deciding I wanted CL EBC to be more accurate and reliable.

Hopefully this helps, it's a deeper dive into EBC. Make sure you communicate with your tuner, some will work alongside you, other's might feel like you're stepping on their toes and messing up their work or work flow. Just be clear if you change anything and maybe send your current tune with the matching data logs each time you submit for review and tweaking (IF they are willing to allow you to make adjustments before they are finished). That way you don't have to copy your changes into their latest revisions.