How does one avoid a low-load overboost error and limp mode?

[Jeffman]

Questions Presented:
How does one avoid a low-load overboost error and limp mode?
Which of the Overboost - Adder TunerPro tables should be adjusted to avoid triggering a simple overboost error giving rise to 3100 error limp mode?

Background:
My latest custom 93 Oct ECU bin runs great with xHP Stage 3 OTS TCU bin. But while cruising on the highway at 70 mph in 6th gear (D6) and then accelerating with ~30% peddle request to 80 mph while staying in D6 I've gotten half-engine light, reduced power, and 3100 and 30FE (overboost) codes. Hit clear codes twice to clear and continue - didn't even need to stop the car.

Possible Solutions:
I suspect an easy fix is to increase the overboost pressure amount that triggers the code, "Pressure Deviation to Trigger Overboost" (mine is currently set to 1.450) located in the XDF for IJEOS under the "Overboost - Load Adder" tab. But maybe it's not the magnitude of the overboost in pressure delta that's my problem, but rather the duration. So the other tables like Time Delay at Overboost Start" and "Overboost Duration (total)" may be the right ones to adjust.

Request:
Before tweaking and experimenting and reinventing the wheel, I thought I'd reach out to the community to see if anyone has dealt with, and solved, this minor irritating issue.
Any suggestions?

Thanks in advance.

 

[RSL]

Those tables apply to 335is and 1M cars only unless they changed the logic or unlocked it in non-N54B30TO DMEs. They are parameters of the factory overboost load offset feature on those DMEs in the stock bins, not tweaks to adjust where overboost code is thrown. You've more or less answered your own question though: 30% pedal is obviously requesting more output than target there. I'd suggest looking at the torque request, torque request (driver) or even PID tables. You either want more target, less output, bigger brake on boost overshoots or some combo of all.

 

[Jeffman]

Thanks for the suggestion.
Btw- The latest XDF for IJEOS (Dec. 2016, IIRC) has these tables.

 

[RSL]

Sure thing. I know they are, but just because they're defined doesn't mean the N54B30 DMEs have logic to look at them or use their values in any way unless Martial or Jake added or uncorked it.

 

[Jeffman]

Ok. That's good to know. I had assumed that if the tables were found they were their for a reason. I guess it's possible the developers may have left them in but unused in the logic.

 

[Jeffman]

Follow-up with a log:
I was able to replicate the error while logging. Check out my datazap, zoomed into the section where I put my foot down half-way on the highway to accelerate from ~70 to 80 mph in D6. The mark on the right is just before when the DME threw the 30FE overboost error then 3100 limp mode.
http://www.datazap.me/u/jeffman/30f...1-17-18-30&solo=2-5-6-7&zoom=417-512&mark=479

Comments about the log:
I'm in spool mode in D6 while accelerating from ca. 70-80 (ca. 2200 to 2600 rpm).
Just before the error and limp mode:

Boost target pressure just before the error is about 6 psi.
Boost pressure (approx. pressure in the intake manifold) is about 6.3 psi.
Boost mean pressure (approx. pressure in the charge pipe) is about 14 psi.
Boost Error P-Factor is about -5.6 psi.
Throttle plate is at 21.5%
WGDC bank 1 is about 55%
​

Comments about my bin:

This is a scaled bin to support the N20 TMAP with MBoost option to measure boost pressures greater than 22 psi.
I'm one of the relatively few who uses the simple WGDC (RPM x Boost SetPoint) logic. See attached table below.
My spool mode max rpm is 3000 rpm.​

What would be a reasonable way to adjust the bin to prevent the large Boost Error?

1. Get the hell out of spool mode much earlier in the rpm, e.g., reduce spool mode max rpm to 2500 rpm (i.e., the stock value)?
2. Adjust the WGDC (RPM x Boost SetPoint) table at around (RPM = 2600; BoostSetPoint = 1.42) to make sure the WGDC values are not so high to cause the overboost.
3. There's something funky about my scaled tables - maybe a table wasn't scaled correctly?
​

I think the answer is 1. Reduce Spool Mode Max RPM.

Am I thinking this through correctly?

 

[RSL]

I was always under the impression that boost mean was calculated and boost psi was measured. MHD doc says the opposite, but whichever is correct, one is calculated/filtered and one is measured and they are obviously miles apart.

It would take about 5 minutes to determine if your option 1 has any impact or not, so try it and see. If I had to guess though, nowhere near enough PID with WGDC still climbing to 70% on increasing -P error (and throttle nanny active, which might be amplifying your problem if mean is, in fact, actual measured in the charge pipe). Like I said above, one of the things is probably needing more brakes (a lot more it seems) on overboost and I'd still take a close look at the other tables too

 

[doublespaces]

Single Turbos are particularly susceptible to this, I'm told its because in partial throttle situations, there may be targeting 8 PSI but you'll never be able to do that when you run a 1 bar spring. I'm curious about this subject, hopefully more people can chime in.

 

[Jeffman]

I edited my comments and provided a third reason - I didn't scale a table properly since I'm using the N20 TMAP.
But in the meantime I'll go try Option 1.

 

[WOT808]

Looks like your WGDC base is higher than WGDC after pid in that area.
Do you want to get actual boost to match requested?
Or, do you want to stop it from going into limp?
I think there's an overboost error toggle in our bins. My SPX base toggle was set to off.

 

[Jake@MHD]

I suspect an easy fix is to increase the overboost pressure amount that triggers the code, "Pressure Deviation to Trigger Overboost" (mine is currently set to 1.450) located in the XDF for IJEOS under the "Overboost - Load Adder" tab

This table is not what you think. All tables in that folder control the factory 1M / 335is style "overboost", which will temporarily increase your load target for X amount of time. Those tables allow you to use it on any N54 (those without it from the factory just had the timer set to 0).

To control part throttle overboost situations, you really want to get your boost target up higher so that is doesn't overboost, or lower WGDC so that it doesn't overboost. For ST guys who must run a minimum spring pressure, you either have to raise boost target in the effected areas or let the throttle close some to limit pressure. The later can cause compressor surge on larger spring gates though, and might not be the best option.

 

[doublespaces]

To control part throttle overboost situations, you really want to get your boost target up higher so that is doesn't overboost, or lower WGDC so that it doesn't overboost. For ST guys who must run a minimum spring pressure, you either have to raise boost target in the effected areas or let the throttle close some to limit pressure. The later can cause compressor surge on larger spring gates though, and might not be the best option.

Great info for the uninitiated

 

[Jeffman]

Oh I'm initiated alright - Trial by Fire!

 

[Jeffman]

Looks like your WGDC base is higher than WGDC after pid in that area.
Do you want to get actual boost to match requested?
Or, do you want to stop it from going into limp?
I think there's an overboost error toggle in our bins. My SPX base toggle was set to off.

I want to avoid limp. But I imagine someday one of my waste gates may get stuck at 99% (does this happen???) so I'd want the safety to shut it down.

 

[Jeffman]

This table is not what you think. All tables in that folder control the factory 1M / 335is style "overboost", which will temporarily increase your load target for X amount of time. Those tables allow you to use it on any N54 (those without it from the factory just had the timer set to 0).

To control part throttle overboost situations, you really want to get your boost target up higher so that is doesn't overboost, or lower WGDC so that it doesn't overboost. For ST guys who must run a minimum spring pressure, you either have to raise boost target in the effected areas or let the throttle close some to limit pressure. The later can cause compressor surge on larger spring gates though, and might not be the best option.

Thanks, Boss. I should have connected the dots better with the 1M's "Overboost" function. I got thrown off by seeing those tables in the latest IJEOS XDF.

Edit: Thanks also for the informative comments, Jake. As you have instilled upon us (I think it was you), the MSD80/81 logic doesn't target Boost but rather Load. So looking at my log linked above I see I am indeed hitting my Load targets. But isn't this mainly the case during main fueling mode 2 and not during spool mode 20? My case of pressing the accelerator to ~40% while cruising at 70mph to pass a car at 80mph while staying in D6 all happened when the engine was in spool mode 20 because my spool mode max rpm was set to 3,000 - spool mode is designed mainly to spool the turbos to build pressure in the intake system, which is measured by my N20 TMAP in the charge pipe. So I think in my situation while my WGDC was set high for the conditions, I should be able to avoid this over spooling / overboost by staying out of spool mode while at highway cruising speeds. I remapped this to 2,500 rpm and will report back tomorrow.