Tuning Boost Control With Bias Table

[E30_Driv3r]

My findings.

-went to release code, that didnt work too well for my blended fuel tables so no testing done
-changed frequency = no change
-played with slider = no change
-even went back to open loop to see if it wasnt anything mechanical = confirmed working
-reading nathaninwa's post i tried to copy his setup in its working state. ONLY and only when I turn off table blending, use boost by gear, in setup mode does the code use my bias table. basic mode = overboost, advanced = overboost.

I'm happy to have a somewhat functional boost control, but this new code needs work i think. I would like the table blending to function as I have different targets for pump 93 vs e85.

I have my e85 base tune pretty much wrapped up and then my car goes into the body shop so hopefully when its out again there will be some updates i would be willing to try out.

Thanks guys James and Ken for giving us these awesome features! Excited for when its all dialed in

[nathaninwa]

Well I'm glad you got it working. I wonder what's going on? I'm able to use basic mode with some slider, and get corrective boost control without overshooting

[E30_Driv3r]

Well I'm glad you got it working. I wonder what's going on? I'm able to use basic mode with some slider, and get corrective boost control without overshooting

What frequency are you using? I believe we are using the same 3port Mac valve that the AEM tru-boost uses. I was running 26hz before but everything is a toss-up at the moment!

[nathaninwa]

Yep, 3 port Aem valve at 26hz

[muythaibxr]

Tried 19.5 Hz?

[aidandj]

So I went to the dyno, and am retargeting boost at 190-195kpa. Set up my bias table. Got it working ok, but still major spike at spool, and then it would drop down to target. Then this morning I was driving and it was a bit colder out. Instant overboost. Like spooled straight to overboost.

[muythaibxr]

Have you tried 19.5 Hz?

Ken

[aidandj]

Yes. And it made tuning better. But still weird. Like P doesn't affect the overshoot at all. Anywhere from 0-200. Its almost like the bias table is too heavily weighted in the control equation.

[aidandj]

Of course as I type that answer things start to work a little better. I really lowered the bias table, and I stopped getting overshoot. It still doesn't fully spool to my goal but I will try and tune that some more. Here is another log and tune.

graph5-4-15.png

graph5-4-15-2.png

[jsmcortina]

Please attach the logfile.

James

[aidandj]

Please attach the logfile.

James

See last post. Oops, sorry it didn't attach.

[aidandj]

It was too big, so the attachment failed. Here's a link to it: https://drive.google.com/file/d/0B7v8e7 ... sp=sharing

I will try again in the morning when its cold and see if it overboosts again.

[jsmcortina]

See also: http://www.msextra.com/forums/viewtopic ... 11#p439411

James

[aidandj]

I'm not using blended tables but it may have just been tuning errors.

[aidandj]

Went out this morning, it was pretty cold. Car warmed up and I did a 3rd gear pull. Straight to overboost. I tried to log it but for some reason the log file is empty. I will try again tomorrow morning.

[nathaninwa]

There is a couple threads now and hard to keep them straight without rereading, so I'll ask. What happens if you turn boost control off and run Wastegate pressure? Does it climb and maintain or does it overboost some and settle?

What if you use the open loop table and target a dc near 200kpa? Let's say it's 50% dc. Just make the open loop table 50 and see what happens?

I'm just wondering if it's a mechanical deal causing the wg to not function fast enough, like say a pinched line or boost reference hose that is to small?

I have also found that tuning on the dyno and tuning on the street are way different for boost dc numbers. My car is heavier than the dyno rollers so to make boost I have to add about 8% to the table. But have to make sure I pull it back out or I overboost! Same with fueling. I come off the rollers lean a bit

[aidandj]

It worked yesterday just fine. It happens every morning. I will try wastegate tomorrow too.

[nathaninwa]

Hmmmm. Air density then. Would be interesting to see a Wastegate pull and then an open loop pull to see if a spike is still present

[aidandj]

Yeah but isn't the whole reason for using closed loop to make that not an issue?

[E30_Driv3r]

I also see 100% overboost with every mode outside of setup mode. Setup mode I stay around my targets +/- 1 psi using only the bias table which should be close enough for pid to take over the rest. Pulling up logs the code is still using extremely high numbers not from my Bias table, and yes boost blending is off. Will attach a log later.