TUNER HELP NEEDED: Boost Flatline at high RPM with throttle closure. Load to Torque Limit??

[all4bspinnin]

Good morning all,

Im tuning my first upgraded turbo car and im running into a small issue. Around 6000 RPM boost goes flat and the throttle progressively closes to keep it at 15.1 PSI.

As you can see, I've pulled a ton of wastegate base and airflow out thinking it was just an over boost. Problem is, if I lower wgdc base/airflow the problem just follows it. I could set wgdc to 0% and it would still do the same thing.

This leads me to believe I've hit some sort of limiter. Datalog below:

http://datazap.me/u/all4bspinnin/log...og=0&data=3-24

Last night I raised the bottom rows of load to torque 1 through 3 and i'll post another log from that in a few hours.

Car is 09 335i with MMP Stage 1 turbos, Stg 2 fuel pump, Hotside inlets, and FBO, 93 oct


Thanks for your help

[all4bspinnin]

48 views.... anyone have ideas?

[RSL]

If you think there's a problem with L2T or you're making adjustments to it, at least log torque output and torque limiter active. Are you making adjustments to L2T blindly? Actual load and boost exceed load requested and target. What's more, boost mean continues to rise through that entire area, so you've got ~14.5psi target, actual boost slightly higher and 18+psi boost mean there. It's exceeding targets and the DME is doing exactly what's it's supposed to: closing the throttle to try to keep the overshoot in check. You need to increase target or reduce actual...adding some room between them wouldn't hurt either.

[CarAbuser]

Maybe I'm missing something. But it looks to me like your throttle is closing because you are going over target boost.

If you want to target higher boost then raise the load target.

Also increasing the values in the load to torque table will mean the car just reports higher torque. The Load to Torque table is just a lookup table for the car to workout how much torque it's producing based on load and RPM.

[all4bspinnin]

If you think there's a problem with L2T or you're making adjustments to it, at least log torque output and torque limiter active. Are you making adjustments to L2T blindly? Actual load and boost exceed load requested and target. What's more, boost mean continues to rise through that entire area, so you've got ~14.5psi target, actual boost slightly higher and 18+psi boost mean there. It's exceeding targets and the DME is doing exactly what's it's supposed to: closing the throttle to try to keep the overshoot in check. You need to increase target or reduce actual...adding some room between them wouldn't hurt either.

The L2T 1 - 3 tables are based on a cobb stage 2 aggressive. After I read this comment, I went and reworked the L2T tables to match the output of the car in torque nm based on an earlier log. I'll flash it later and see how it works.

Im going to ask a stupid question here, but how would you suggest to drop load actual? I'd prefer to keep the boost around 14 to 15 psi until I get this resolved.

Thanks for the reply @RSL

[all4bspinnin]

Here's a log with torque included.

http://datazap.me/u/all4bspinnin/log...21-106&mark=84

[bradsm87]

The one that's used for boost control is "Boost Mean". You need to drop your WGDC more as it's way above target and causing the throttle closure.

[all4bspinnin]

The one that's used for boost control is "Boost Mean". You need to drop your WGDC more as it's way above target and causing the throttle closure.

Brad, I don't have any more room to pull wastegate. I've lowered the WGDC base number to damn near 0 in those cells... I think they're like 1 to 2. That's why im wondering about this.

Am I supposed to raise the values? I was of the understanding that to correct overboost we had to open the wastegate, thereby lowering the values in cells at x.xx boost setpoint and xxx load actual

[bradsm87]

Are your wastegates adjusted too tightly? That's very low WGDC across the board.

[V8Bait]

There's a few ways to fix this. Lower wgdc in the cells over 390 MAF at 2.3 setpoint and below. Lower MAF adder in any cell over 390. Lower WGDC minimum. Raise load target (will also raise boost a little though). Lower boost request offset %. Raise PID P factor for negative values. Raise D factor for positive values. These must be MMP good luck they are boosty little guys I had to rewrite everything for them.

[all4bspinnin]

There's a few ways to fix this. Lower wgdc in the cells over 390 MAF at 2.3 setpoint and below. Lower MAF adder in any cell over 390. Lower WGDC minimum. Raise load target (will also raise boost a little though). Lower boost request offset %. Raise PID P factor for negative values. Raise D factor for positive values. These must be MMP good luck they are boosty little guys I had to rewrite everything for them.

Thanks for the reply V8bait. I've been lowering the wgdc cells over 380 MAF on the base table at 2.3 setpoint. I've also reduced the adder to 1.25 above 350 MAF. I raised the P factor and d factor cells a slight bit in the cells at those sites. ......glad to see im on the right track here. whats funny is that youre exactly right... these are MMP stage 1s.

What does boost request offset do again? i'll be reading up on that one again.

I turned the load target up from 170 to 185 so we'll see how the car reacts tonight. If I ever needed to learn how to tune a upgraded turbo car, this is good practice lol

[all4bspinnin]

Im tempted to turn this thing on...

[noorj]

Do it! Let us know how tuning it goes, it would be super easy to do if MHD would start doing histograms in the logger

[hydra]

Im tempted to turn this thing on...

Sure makes boost control a lot easier and more intuitive if you're used to tuning boost on other platforms.

[all4bspinnin]

So here's the load at 185 dyno run. If you compare this against last run you can see where WGDC base crosses WGDC PID around 6000. Here its narrowed the gap and moved to 6500.

I also increased p factor a slight bit above 6k. I'll probably bump that up a bit as well. No changes made to WGDC Base or Adder

It had a bit of correction on 1 and 3. So we'll probably do 190 next so I can make sure its not correcting. Im thinking if I raise it to 195 it should fix the problem.


http://datazap.me/u/all4bspinnin/mar...26-101&mark=88

[V8Bait]

It's better to sort out the wgdc so the PID doesn't have to react negative, but that's much better. Why so little boost to begin with though? Bad gas or something? I'd be running 18+ even with pump on any upgraded twin, they do better there than stockers.

[all4bspinnin]

It's better to sort out the wgdc so the PID doesn't have to react negative, but that's much better. Why so little boost to begin with though? Bad gas or something? I'd be running 18+ even with pump on any upgraded twin, they do better there than stockers.

Being my first upgraded turbo car, I wasn't sure where to start load-wise/boost-wise and not knowing how healthy his car was I was probably a little over-protective. Its a 93 octane car as well so I wasn't sure of the quality of the gas.

Either way, youre right, 18 psi is probably a good spot for these turbos.... just playing it safe. I think 18psi will be a good pump gas target. I plan on turning it up a bit tonight. Thanks again for your help. I'll post more after tonight.

[hydra]

if 18psi is good, 21psi is better

[all4bspinnin]

Im tempted to turn this thing on...

So, i dont understand why everyone with mhd isnt running this. Made things so much easier.

One thing that im still wondering is WGDC base in the logs should show a true base now before PID correct? @jyamona@motiv @V8Bait @RSL @CarAbuser @bradsm87

The wgdc base in the logs doesnt reflect the number in the boost setpoint x rpm table. It was my understanding that it removed the WGDC postion, adder airflow, ceiling, etc. What else comes into play here to produce the WGDC base number in the logs using this feature? Obviously PID kicks in on top for the WGDC after PID

Its my understanding that base is calculated by base + adder > that number is then used as x axis in WGDC to WGDC position > stores that value > uses that value to lookup in WGDC position to WGDC > takes that lookup value + base = final wgdc base.

So if we removed all except base, what would still add to it?

[bradsm87]

So, i dont understand why everyone with mhd isnt running this. Made things so much easier.

There is one advantage to the stock tables and that's closing the wastegate HARD when it's known to be closed anyway to stop rattle and wear. That's what the WGDG to WG position table is for. Once it's deemed to be closed, lots of WGDC is added in the WG position back to WGDC table.

[RSL]

That new logic is only on I8A0S and IJE0S, so I've never used it and haven't paid much attention to how it functions. Is the base table even still used with it?

[all4bspinnin]

There is one advantage to the stock tables and that's closing the wastegate HARD when it's known to be closed anyway to stop rattle and wear. That's what the WGDG to WG position table is for. Once it's deemed to be closed, lots of WGDC is added in the WG position back to WGDC table.

Brad, very good point.

That new logic is only on I8A0S and IJE0S, so I've never used it and haven't paid much attention to how it functions. Is the base table even still used with it?

At this point im not sure. The WGDC base didnt follow the number in the table. I was able to stop the throttle closures up top by raising the load target tables.

@jyamona@motiv

[RSL]

I think it's just the WGDC tables watered down to a single cumulative table that is looked up by set point/RPM rather than set point/MAF, but are still just starting points being acted as usual. If the new logic/table includes base, I have no idea what base might show up as in logs separately then, but guessing probably some % of the total. Hopefully, Jake will pop in and explain it, but as long as base is close to or slightly under "after PID" numbers, the only thing you really need to do is adjust that table to keep boost close to target without overshoots to make less work for the PID.