Braineack

iTrader: (62)

i can get on board with chi's timing table.

Chiburbian

iTrader: (1)

One thing I see right off the bat is your "Lower Limit Delta" of 80kpa.

Tunerstudio documentation says this:



What you are saying in effect is, start trying to use closed loop boost control until I get within 80kpa of my target. If you are at 100% throttle and your target is 200, it is going to be doing math once it crosses 120kpa. The way I understand it, is you should keep the delta as low as possible so it keeps your wastegate closed as long as possible. Try putting 30kpa in that and see how it goes. If it overshoots, raise it. If it is slower to build boost but doesn't overshoot, lower it.

Mech5700

I read that in the reference manual, however one part I don't quite get with that statement is valve held open (0% DC I assume?) = pressure open to WGA, not open valve = closed wastegate. But either way, say if the process starts at 120kpa, it should still be holding the valve closed till closer to the target, would you agree? It knows what the target it, why would it open the valve so far from it?

I'm definately gonna give it a shot tho. I can see that the DC is 100% till 123kpa, and may target was 203 kpa... 80kpa away. Good lookin out man

Chiburbian

iTrader: (1)

I pulled up your log in excel which while it doesn't let me see pretty graphs (by default) it does show me some interesting things.

From what I am seeing, the Megasquirt is pulling your boost duty down to 35% at 4100rpm because it's initiating the closed loop process and starting at 35 per your initial duty table.

You have two things working against you. First, you have a HUGE discrepancy between your target and your actual boost level. Additionally, your PID loop is starting with 35 and making adjustments from there - except it doesn't seem to be doing anything... If anything it is lowering your boost duty? Maybe?

I am a big fan of closed loop, but I still don't really understand how to tune it properly. If you did a fresh install of tunerstudio, what are the default values? Should you adjust the sensitivity slider before or after adjusting PIDs? What should the initial value be for the sensitivity slider?

Start with changing your delta and do another run. (I bet it will scare you)

ETA: Where did you get that initial duty map from? Did you do an open loop tune to see what duty cycles create what boost? It seems like your low rpm initial duties are a bit low to say the least.

Per Curly's open loop boost tune of Blaen99's car:

Mech5700

I don't think the 35%DC is from the initial duty table. At 4100 RPM and 125 kpa, the ID table is at about 20%

I set the delta to 27.5 kpa (4psi) but have not done a test run yet. Gave the car enough of a beating today...

And for tuning PID properly? I've read set to 100, 0, 0 and lower the P from there till you reach target. Then add I (consequently adjust P again) to maintain boost across the RPM range, then add D to dampen oscillations. Slider? not sure about that... It looks like I can do slider only or slider and PID combined.

And I think we got the ID table from roughly what I had set in OL. Is curly suggesting 100% from top to bottom from 0 to 2500, and 80% from top to bottom 3250 and up?

Chiburbian

iTrader: (1)

Also, something cool to get your settings a little more accurate...

Switch to open loop mode. Make sure to copy down your existing open loop numbers.

Then, put all cells to zero. Record what your max boost achieved is.

Progressively work your way up from zero. Maybe go by 5's initially until you go over your wastegate pressure. Once you break wastegate pressure, back down by 1 at a time until you once again settle on wastegate pressure. That is your "zero" for "minimum duty percentage".

I am not sure how to compute the maximum off the top of my head.

Chiburbian

iTrader: (1)

At 4,100rpm your TARGET at 100% throttle is 203. Your initial duty table lookup at 203kpa and 4100 rpm is 35%.

BTW: I was unclear as to how the target duty table worked until looking at your log by the way. Now it all makes sense

Chiburbian

iTrader: (1)

I can't say. That is what it took for open loop to reach his boost targets.

Maybe invite Curly into this thread? (send PM?)

Mech5700

at all 0's, it wouldn't reach 10 psi, 0 is full open. Do you mean set to 100%/close the solenoid off? Overboost set to maybe 16#?

Mech5700

There are so many things going on at once, I wish I knew what kind of order they are in.

Initial Duty table
Boost Target Table
Boost by gear/TPS/VSS calculations
PID
Slider

It seems like it should be simpler. Stab the throttle, MS does whatever it needs to in order to reach target. Or am I just overthinking it?

Chiburbian

iTrader: (1)

No, what I am saying is that at 0% duty, the boost signal is traveling through the fully open EBC solenoid into the wastegate. 100% of the boost signal MAY still hit the wastegate at 5% EBC duty... Maybe at 10%...

I can't think of a good way to put it at the moment.

curly

iTrader: (25)

Hi.

Yes, from top to bottom (because TPS was acting funky) was 100% until 2500rpm, and 80% at 3250.

Since I didn't have any experience with EBC, I went with what my very experienced tuner was used to, which was 0% = 0% boost (wide open) and 100%=100% boost, or completely closed.

There's a option in your boost control to switch it from inverted or normal (not typing on a TS computer, so those menu names are off the top of my head), make sure you know which one you have selected, and what each one means in terms of 0% or 100%.

Also, I used open loop boost control. I'd suggest you do the same until you find out what percentage equals what boost level, what boost level you actually want for your engine/turbo/clutch/transmission setup, what boost you tune to at the dyno, and then enter those desired parameters to tune your closed loop until it starts controlling boost properly.

Mech5700

Ah I see what you are saying. This is to figure out at what % does the solenoid actually do something/give you anything over WG pressure. I get you now. Then with that % being the minimum, It will never allow straight WG pressure.

curly

iTrader: (25)

Something like that. Sometimes you want it to be straight wastegate, but if that's at 50%, there's no reason to be below ~45%. Otherwise it'll be hunting forever to get back up 50%, and then it'll finally start increasing boost.

Mech5700

Hi curly, thanks for joining in.

I guess I should've said it wouldn't allow anything LESS than WG pressure.

And yes, mine is set and verified to be 0%=open=minimum boost, and 100%=closed=max boost.

You are probably right... maybe I'm just not ready for CL quite yet.

Mech5700

Can't get my laptop to connect to the web at work but here's a shot of the log.



I set the delta to 27.5 psi before target, and it holds the valve 100% DC till just about 10 psi, but it still doesn't reach full boost till 5700 rpms in both 2nd and 3rd gear shown here.

Chiburbian

iTrader: (1)

Ok cool. I checked my settings this morning.

My delta is 14kpa and my control interval is 10ms. Although - that interval may be too short. The manual recommends 20ms to start.

I hate to just cut/paste timing maps *****-nilly but I'd be interested to see how your engine reacts to my timing map. It is based on one that is pretty well respected around here (I can't remember who's map it is). Are you running 91 octane or 93? I have had no problems with this map on 93...

Maybe lower your target boost (at 100% throttle) to 190kpa for safety's sake and load my timing map and see how it pulls up to 190 to start?

Mech5700

I run 93 as well so it's worth a shot to give it a try.

Braineack

iTrader: (62)

what happens during spool with the vacuum line to the wastegate pulled?

looking at that log it almost appears like the function is inverted, where 100% is wastegate.

just seems like such slow spool... what turbo is it? unless it's ungodly in size, most turbos should be hitting target boot levels between 3.5K and 4.5K.

Chiburbian

iTrader: (1)

Seems like the safer thing to do is to just turn boost control off, no?

It's a 2860 Brain.