Dome control closed loop misfires, open loop works fine..

[nvr2fst]

Hi everyone,

I am hoping someone can help me solve an issue I am having with closed loop dome control on my ms3 ultimate factory 1.5 firmware.

When I enable closed loop dome control and set my boost target value above 235kpa in my 95% and 100% throttle cells the car runs like total crap. Lean pops at even part throttle and will not go into boost.

While it is poping and lean I can change either the boost target lower or change it to open loop and the car runs instantly runs perfect. Wth is going on? I have turned off boost tolerance and over boost protection and no change. Any thoughts?


-Josh

[jsmcortina]

[nvr2fst]

Sorry, I was working from my phone last night.
Here are the files:

2017-06-21_06.07.28.msq

Dropbox link to the log files from yesterday.
https://www.dropbox.com/sh/08acqkxaz8xe ... Ehbsa?dl=0

[jsmcortina]

Where in the logs (what time) are you seeing a problem? (Not sure what I'm looking for.)

James

[nvr2fst]

Where in the logs (what time) are you seeing a problem? (Not sure what I'm looking for.)

James

Hi James,

For easy conversation take a look at WOT TPS. I renamed and updated the log file in the download link to show "good pull" vs "bad pull" I will start the car and specifically take a log with Open loop vs closed driveability issues for better comparison later today.

In the "Good Pull", around Record 2600, you see all AFR's and egt's are normal. I have individual EGT and wideband sensors so look at AFR 2 and EGT2 in bank one and AFR 3 and EGT3 in bank 2 compared to any other cylinder in the matching odd or even bank.
For some reason, in the "bad pull" those two cylinders "drop out" in some error state if I set my boost target higher with closed loop boost control on. Also notice how ALL the cylinders AFR is going nuts in the section where Cylinder 2 and 3 drop out. Before and after all is well.

In the "good pull" vs Bad pull the change that made the difference was to enter Open loop boost control. If I enter closed loop cylinder 2 and 3 always drop out causing the car to stall.

I have almost 45 dyno pulls trying to solve the issue. I have tired, adding/removing timing, adding/removing fuel, moved around and tired different plugs, moved around and tried different plug wires, moved around coilpacks and moved around injectors.

No positive change except for turning on/off closed loop boost control or lowering the target boost.
This car will make 1100hp all day long at 200kpa and never hiccup. If I set the target table at 260 It will misfire until MAP pressure goes over 180ish kpa. After I cross the 180ish kpa threshold it will make a beautiful pull and make 1300+ at 240kpa.
Look here at the attached photo:

msextra.png

It will do this misfire exactly on repeat over and over until I lower target or goto open loop boost control.

[nvr2fst]

My suspicion at this point is, if Boost Target VS Current MAP get more than 100kpa apart they have to come within 80kpa to "clear the error state"

I guess that is an easy enough theory to test but I am really tired of putting stress on the car and setup to troubleshoot. I think I have the stress equivalent of around 10 track days in dyno time to try and figure this out.

*edit- I looked back through previous logs and this is not the case. It is not an exact KPA difference when the issue occurs.

[jsmcortina]

I've dug into this some more on the bench and I agree that something is wrong.
Two issues I've identified so far.
1. The "Lower limit" setting is preventing any boost from happening. Set it to the max for the moment.
2. Dome response seems very slow even with the maximum sensitivity and gains turned up.

James

[nvr2fst]

Thanks James! The fact you found somewhere to start makes my day! I have been going crazy with this!
Let me know if you need anything specific and I will see if I can get it for you.

[jsmcortina]

I've done some more work on this and believe I've improved the issues that I was seeing and fixed your misfire.

The code was overwriting an output port in error when in closed-loop Dome mode only. (Null pointer.)

James

[nvr2fst]

I've done some more work on this and believe I've improved the issues that I was seeing and fixed your misfire.

The code was overwriting an output port in error when in closed-loop Dome mode only. (Null pointer.)

James

Excellent! When or How can I get a copy of this?

[jsmcortina]

When or How can I get a copy of this?

Email sent.

James

[jsmcortina]

With the new firmware, reduce your "lowerlimit" setting to perhaps 20kPa. This will use the bias table value as the dome target until MAP is within 20kPa of the boost target. Within that 20kPa it will then use PID to match it. This of course requires a reasonably well tuned bias table - use "setup" mode for that.

James

[nvr2fst]

With the new firmware, reduce your "lowerlimit" setting to perhaps 20kPa. This will use the bias table value as the dome target until MAP is within 20kPa of the boost target. Within that 20kPa it will then use PID to match it. This of course requires a reasonably well tuned bias table - use "setup" mode for that.

James

Thanks James! I will give it a try tonight when I get home.

[jsmcortina]

Any updates?

James

[nvr2fst]

Any updates?

James

Hi James, I am afraid not. It was pooring rain until I left for Vacation. As soon as I am back I will get everything updated and put it back on the dyno so I can test and report back.

-Josh