Boost control change testing (MS2 and MS3)

Tuning concepts, methods, tips etc.

Moderators: jsmcortina, muythaibxr

nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Boost control change testing (MS2 and MS3)

Post by nathaninwa »

I was thinking table switching too. But just like you see Eric, the target is table 2. We compared .msq's last night and nothing within boost seems funny. Maybe a possible log with bias table 1 set to table 2 would help. Hhmmmm.

edit, I'm doing that gear blend switching, but have been tuning both tables the same, maybe at lunch I'll change table 2 and see what happens.
Last edited by nathaninwa on Wed Mar 19, 2014 6:04 am, edited 1 time in total.
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
elaw
Super MS/Extra'er
Posts: 2926
Joined: Fri Oct 16, 2009 6:20 am
Location: Wilmington, MA

Re: Boost control change testing (MS2 and MS3)

Post by elaw »

I probably should have worded my post better - I meant there possibly might be a problem in the code?

It looks like tableswitching has caused the MS to select boost target table #2 (based on the logged boost target), but is still using bias table #1 (based on the fact the valve duty goes to 0 rather than a higher number when PID engages).
Eric Law
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking

Be alert! America needs more lerts.
nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Boost control change testing (MS2 and MS3)

Post by nathaninwa »

I was thinking what you were last night. Him and I were emailing logs and stuff to try and figure it out, and we/he couldnt get it to stop cycling 0/100/0/100. I didn't see anything in the Msq, and sent him mine for him to verify. We couldn't find anything out of place.

I edited my last post.
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
Jdmfreak
Helpful MS/Extra'er
Posts: 71
Joined: Thu Dec 12, 2013 11:32 am

Re: Boost control change testing (MS2 and MS3)

Post by Jdmfreak »

ive verifi ed in open loop that higher DC = higher boost

I can set my bias table 1 when I get off work tonight and see if that changes anything (have to rock the work van today)
1990 supra 7mgte ms3x v3.0
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Re: Boost control change testing (MS2 and MS3)

Post by muythaibxr »

I will have a look at the bias table code to see if the table switch stuff works.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Boost control change testing (MS2 and MS3)

Post by nathaninwa »

As luck would have it, my laptop didn't sleep last night, just enough battery to change bias. set bias Table 1 to 0 at the 260 range, then died while getting a log started. But I can confirm, with gear blended switching, the boost targets change, but its using bias Table 1 the whole time. I got 10psi oscioscillation with bias at 0. My current pid is all zeros.

edit I should note, IvI've been making both tables the same while tuning.
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
jsmcortina
Site Admin
Posts: 39587
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK
Contact:

Re: Boost control change testing (MS2 and MS3)

Post by jsmcortina »

whittlebeast wrote:Is there a way to kill all power to the boost solenoid when in a totally non performance part of the map? It does not seem right to constantly cycle a mechanical valve when there is no advantage to do so.
That should already be in place.

Off-boost the duty should be 0% and the solenoid is off.
As you get above baro, the solenoid should go to 100% duty for fastest spool.
Then when you are within "delta" of the target it will switch the initial values/bias table, then engage closed loop and aim for your target table.

James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Boost control change testing (MS2 and MS3)

Post by nathaninwa »

I noticed the "off" duty cycle is the limits set on the boost page for open/close %'s. Once I set mine up to 0/100 vs the 40/80 I had, then I had no valve dc till I wanted boost.
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Re: Boost control change testing (MS2 and MS3)

Post by muythaibxr »

OK, I was able to see in the code why the second table wasn't working for bias.

Some of the logic in that area of the code could use a longer look than I have time to look at it tonight (interaction between the various things that can set target and which bias table should be associated). I should have time to fix it this weekend, so hopefully we'll have test code out by the end of the weekend.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Boost control change testing (MS2 and MS3)

Post by nathaninwa »

Thankyou Ken.
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
Jdmfreak
Helpful MS/Extra'er
Posts: 71
Joined: Thu Dec 12, 2013 11:32 am

Re: Boost control change testing (MS2 and MS3)

Post by Jdmfreak »

Awesome thanks :-)



(to both of you)
1990 supra 7mgte ms3x v3.0
tutuur
Experienced MS/Extra'er
Posts: 288
Joined: Tue Nov 06, 2012 5:27 am

Re: Boost control change testing (MS2 and MS3)

Post by tutuur »

I haven't had time or oportunity to test this yet but all i can say is: keep up the good work, it's massively appreciated!
duxthe1
Helpful MS/Extra'er
Posts: 137
Joined: Fri May 25, 2012 8:32 pm

Re: Boost control change testing (MS2 and MS3)

Post by duxthe1 »

All this talk about boost control and I decided to finally attack it. I'm running MS2E pre3.3.2 beta3. I've got the control circuit installed and working. I'm using the pierburg boost valve found on newer volvos at 26Hz. My turbo is quite a bit undersized for the engine. I've got some boost control functionality but I caught a pretty severe oscillation in a datalog. I'm using the basic tuning mode with the slider at 238. Should I try to move the slider around or is it likely I'll have to tune PID in advanced mode? Any other advice?
90 Mercedes 300 TE
3.0L Inline 6, AWD
T-3 turbo @ 11 PSI W/G, 18 PSI MBC, A2W I/C
MS2Extra V3.0, 3.3.2 beta7
60-2 with direct coil control
40 lb/hr high imp, AEM wideband
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Boost control change testing (MS2 and MS3)

Post by muythaibxr »

Did you tune the initial values table first? Please read the first few posts on this thread. Basically you should go to advanced mode and turn your gains to 0, and tune the initial vals table first.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Re: Boost control change testing (MS2 and MS3)

Post by muythaibxr »

Nathan, do you have a blended msq I could use to test my fixes? I'm testing right now using my own settings, but it'd be nice to have settings from a running engine.

Ken
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Re: Boost control change testing (MS2 and MS3)

Post by muythaibxr »

My fixes seem to work on my own settings. I can send test code if you want.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
duxthe1
Helpful MS/Extra'er
Posts: 137
Joined: Fri May 25, 2012 8:32 pm

Re: Boost control change testing (MS2 and MS3)

Post by duxthe1 »

muythaibxr wrote:Did you tune the initial values table first? Please read the first few posts on this thread. Basically you should go to advanced mode and turn your gains to 0, and tune the initial vals table first.

I've been working at it but dont seem to be having much luck. I'm starting to think this control strategy wont work with my setup. The turbo is small for the engine on both the hot and cold sides. Nonetheless it will spool hard and can hit my overboost quickly with the EBC set to do so. I have been working on the initial Vals table and they seem to be dialed in pretty well. In the attached PNG the initial val setting got the boost to 1.4KPA over target. The problem illustrated quite well in the screenshot is how fast boost falls off. When the initial vals takes control of the wastegate the turbo loses its head of steam and just cant recover despite falling below the delta and reverting back to 100% of the valve's duty. I have the gain's 0'ed except an I of 10 but as you can see the boost falls so quickly PID doesn't have a chance. The trough of the dip following the peak is 8 PSI below target and pretty much wastegate boost.

Any help or suggestions are welcome at this point.
90 Mercedes 300 TE
3.0L Inline 6, AWD
T-3 turbo @ 11 PSI W/G, 18 PSI MBC, A2W I/C
MS2Extra V3.0, 3.3.2 beta7
60-2 with direct coil control
40 lb/hr high imp, AEM wideband
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Boost control change testing (MS2 and MS3)

Post by muythaibxr »

I think you answered your own question. If it gets to 100% duty and it still doesn't get the boost to increase, you need to do something about that tiny turbo!

That said, you might try adjusting the PID numbers higher to see if it can compensate.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
elaw
Super MS/Extra'er
Posts: 2926
Joined: Fri Oct 16, 2009 6:20 am
Location: Wilmington, MA

Re: Boost control change testing (MS2 and MS3)

Post by elaw »

duxthe1 wrote:I've been working at it but dont seem to be having much luck.
Can you post your current .msq file?

Honestly I know you say this isn't the case, but it looks like your initial values table is way off. When PID kicks in and your boost duty drops to around 27, presumably that 27 number is coming from the table. And if the number is right, the boost would start to level off at a value near the target. The fact that it drops like crazy would point to the initial value being too low.

But other than that, Ken makes a good point too. You'll notice that when PID kicks in and the boost drops, the valve duty doesn't increase significantly. That means the PID really isn't doing anything, due to the gains being set low.
Eric Law
1990 Audi 80 quattro with AAN turbo engine: happily running on MS3+MS3X
2012 Audi A4 quattro, desperately in need of tweaking

Be alert! America needs more lerts.
zmanco
Experienced MS/Extra'er
Posts: 329
Joined: Mon May 28, 2007 2:16 pm

Re: Boost control change testing (MS2 and MS3)

Post by zmanco »

duxthe1, I have a similar setup with the same issues: 2.8L Datsun 6 cylinder with T3/T4 hybrid turbo. A/R=.63 which I now realize is too small. I see the same thing you do and finally concluded that the issue is due to too much backpressure in the turbine. A WOT pull in 2nd gear where the boost builds quickly will result in overshoot followed by a drop in boost pressure - without any BC from MS, just letting the wastegate work by itself 100% mechanically. I believe that the backpressure is large enough that it's preventing the actuator from quickly closing the WG and hence the boost drops too much. I've found that if I do a slow application of throttle in 4 gear such that the WG doesn't open more than needed to hit the steady state boost target that this doesn't happen, leading me to further believe this is the phenomenon at work.

It would be interesting to see a pull of how your WG performs without any MS control.

BTW, in my case, about the only time I'm running full boost for more than a second or two is on the track, and when I look at the datalogs I see that I almost never go to full throttle suddenly, as we do when we're tuning BC. I'm not prepared to change the turbo so I'm in the process of tuning using closed loop BC under conditions where the boost builds gradually, which reflect how I actually drive the car most of the time.
Daniel
'73 240Z
Castle Rock, CO USA
Post Reply