Dano's Boost Tuning 101 + Cool Stuff

[fjames]

I dunno, you were talking about WGDC and killing the BD table, so I thought you were talking about boost control? Don't see what fueling has to do with boost targeting? I was just pointing out that whatever you do, you're still targeting a table that is sensitive to throttle/pedal position. Since the ots mapping is very aggressive there, and no one changes it, I just thought I'd remind folks it could be relevant.

[Dano]

haha

I was typing faster than I was thinking...n/m my post

[Frequentflyer]

Regarding the Boost Comp:gear tables; I'm having trouble figuring out how these work. There is RPM in both the X and Y axis. WTF is the idea behind this? It seems there really only needs to be one row and not 14 rows.

Then of course there are the Boost Comp A and B tables, which I don't get either. So many tables that seem to do the same thing.

[Dano]

hahah

I haven't got around to those yet...went with APP instead.

I don't think it matters that its a 3d table but you can put lower values in the a table and see what happens.

IDK why they have different tables but I would guess that A is for normal operation and B is for DSC active but then again does DSC even work above 3rd gear??

who knows

dive in and see what happens and report back.

I am having great success with APP translation. I need to switch to my MBC and see how it will react but feel it will indeed limit power.

did the BD table value help your boost dropoff?

[Frequentflyer]

Well I made my 1st and 2nd gear tables lower (see below), so we'll see what happens tomorrow after I flash it. I'm tempted to lower it to like .50 all over to see if it cuts boost in half.

I can't tell if the BD did anything yet. It may have a little bit, but I really think the stock turbo just can't keep up after 5k RPM to keep producing 19psi. In the summer, I usually run out of DC, but that's not the case. I think I'm running out of turbo. I did a 5th gear log for as long as I could yesterday and I was pretty happy with it. I ran out of road though.

[fjames]

Did you ever log 19 psi at 6K+ with the stock dp? If you could hold it then, but not now, it makes sense that with the lower restriction of the new dp, the pour little stocker can't keep up.

But still, you have wg to spare ... and your throttle was opening a little, so it's trying, but why not use the WG?

You could switch to load tuning, put some big numbers up high and see what boost it results in. Fine tune that, then you'd know what the system was capable of when running how it wants to. Hopefully you could then transfer the boost numbers over to the toggle.

[acidbbg]

So I have been following this thread and it seemed like people are having good luck tuning via boost targeting toggle in the ATR. I have been struggling with load based tuning trying to meet targets. I was constantly over or under shooting.

I followed Dano's example but I wanted to start slow and see if I could hit my targets psi of 16.5 psi with just 20 in the WGDC...

Low and behold I am hitting my target psi of 16.5 with only 20's in the wgdc.

I did notice a couple of spikes.

I have attached screenshots of my boost targets, wgdc & wgdc iat.

I want to turn up to wick and set it to ~18 psi but won't until we figure out how to better control boost targeting in 1st & 2nd gear.

Let me know what you guys think.

P.S. I also notice the throttle feels softer when compared with load based toggle/tuning.

[super_pablo_]

subbed for interest to read later... lol

[Dano]

Originally Posted by fjames Did you ever log 19 psi at 6K+ with the stock dp? If you could hold it then, but not now, it makes sense that with the lower restriction of the new dp, the pour little stocker can't keep up. But still, you have wg to spare ... and your throttle was opening a little, so it's trying, but why not use the WG? You could switch to load tuning, put some big numbers up high and see what boost it results in. Fine tune that, then you'd know what the system was capable of when running how it wants to. Hopefully you could then transfer the boost numbers over to the toggle.

in my testing with BD zeroed out and WGDC table set to say all 20 I see logged DC start to drop around 5400RPMs and target isn't met yet..

I have multiple logs that show this and have attached one.

That has me thinking there is another limit table somewhere that knows the K04 can't go much beyone 5K.

I also have the reverse upon spool up. you can see that DC will not hit the target of 49 until after 3K. IDK what is going on.

thoughts?

[fjames]

That looks like a load target log - it's holding load nicely with boost building with rpm, and WG diminishing to accommodate it. BTs must be fun, seeing less WGDC as boost builds at high rpm lol.

But you're saying it's a boost target map?

[Dano]

yessir! Boost toggle enabled and my CMD load values are higher than what is being logged. Of course I have WGDC lower than needed so I can walk it up to target.

I agree it looks like Load or Load Dynamics table is coming into play somewhere.

I have moved DC up and it appears that a % is being applied early and late in the RPM band. That is to say that my 22 DC at 2.5 K may yield 48, instead of 53, but if I move it to say 30 it Will give me 50 which should be 69. So it is possible to get what I want it just doesn't fit the formula exactly.

Here are my current values and log so you can see what I mean.

35.00	35.00	35.00	35.00	35.00	30.00	27.00	30.00	30.00	30.00	30.00	28.00	28.00	28.00	20.00

I have one more rev of DC modification and then I'll but my BD table values back in and see what happens.

BTW zeroing out BD is the absolute best way to get your WGDC tuned. Get within 2 to 3PSI of target without BD making changes then turn BD on with the lower values Christian has suggested.

works like a champ as long as you make all your changes at the same AMB temp of course.

[Dano]

Originally Posted by acidbbg So I have been following this thread and it seemed like people are having good luck tuning via boost targeting toggle in the ATR. I have been struggling with load based tuning trying to meet targets. I was constantly over or under shooting. I followed Dano's example but I wanted to start slow and see if I could hit my targets psi of 16.5 psi with just 20 in the WGDC... Low and behold I am hitting my target psi of 16.5 with only 20's in the wgdc. I did notice a couple of spikes. I have attached screenshots of my boost targets, wgdc & wgdc iat. I want to turn up to wick and set it to ~18 psi but won't until we figure out how to better control boost targeting in 1st & 2nd gear. Let me know what you guys think. P.S. I also notice the throttle feels softer when compared with load based toggle/tuning.

that second log just looks like a spool up spike and isn't much to worry about. The last two look good.

Are you running OEM boost control?

The best way to get accurate logs and a good picture of what's going on is to do 4th gear pulls starting below spool so like 2.5K to RL. Get at least 3 logs per revision to rule out anomalies.

lowere your DC table starting at 2.5K down to like 18 and get some 4th gear logs.

[acidbbg]

Originally Posted by Dano2010 that second log just looks like a spool up spike and isn't much to worry about. The last two look good. Are you running OEM boost control? The best way to get accurate logs and a good picture of what's going on is to do 4th gear pulls starting below spool so like 2.5K to RL. Get at least 3 logs per revision to rule out anomalies. lowere your DC table starting at 2.5K down to like 18 and get some 4th gear logs.

Yes to running OEM boost control.

I try to tune in 4th but it's very difficult to do with out getting a speeding ticket.

I created a map targeting 17 psi up top.
Also, I decreased wgdc to 18 at 2.5 & 3k rpm.

It's supposed to snow tomorrow but if it's nice out I will do some more runs and post some logs.

[Dano]

hahahah

yes dyno time is less expensive than a ticket ...

or you could get invest in an Escort 9500xi...best detector on the market IMO..GPS enabled so almost NO false positives...has saved my ass at least 5 times since i got one this summer.

it will remember auto opening doors like ones at retail stores and never beep again...if this mutherphucker beeps its the popo...

and before someone toutes a Valentin...fuck your POS....it's 15 year old tech...GPS is the future.

[acidbbg]

Originally Posted by Dano2010 hahahah yes dyno time is less expensive than a ticket ... or you could get invest in an Escort 9500xi...best detector on the market IMO..GPS enabled so almost NO false positives...has saved my ass at least 5 times since i got one this summer. it will remember auto opening doors like ones at retail stores and never beep again...if this mutherphucker beeps its the popo...

I have the escort passport 8500 X50..Basically purchased when it first came out.
That thing has save my a$$ so many times it's not funny.

[Dano]

Originally Posted by Frequentflyer Well I made my 1st and 2nd gear tables lower (see below), so we'll see what happens tomorrow after I flash it. I'm tempted to lower it to like .50 all over to see if it cuts boost in half.

in my experience, it is best to put in exaggerated values in to see if they even work then fine tune for the results you want.

cut them in 1/2 to start out then bring them back up to what you want...LOL

I have cut APP in 1/2 to get good results in the 40* temps here for this time of year.

I let them ramp up until they plane out then flatline to a total of 50% APP which works great for 3rd gear but I need less for 1-2nd.

stay tuned for my table updates.

[Frequentflyer]

Originally Posted by fjames Did you ever log 19 psi at 6K+ with the stock dp? If you could hold it then, but not now, it makes sense that with the lower restriction of the new dp, the pour little stocker can't keep up. But still, you have wg to spare ... and your throttle was opening a little, so it's trying, but why not use the WG? You could switch to load tuning, put some big numbers up high and see what boost it results in. Fine tune that, then you'd know what the system was capable of when running how it wants to. Hopefully you could then transfer the boost numbers over to the toggle.

Looks like boost tapered off on its own before in my older logs pre-downpipe, but WGDC's were also significantly higher without the downpipe. So it's most likely the turbo running out of steam.

[Dano]

Well I think I just about have the boost portion of my tune buttoned up. I have just a tad of overrun to tweak out.

It looks like at about 1 PSI over boost the BD table doesn't have authority to remove DC so I'll add a -0.01 to the 1.0 over run value there.

I may also need to lower WGDC values in the RPMs I am over boosting, IDK yet.

It's interesting that running with the BD table zeroed out, WGDC began to not follow the 2x+9 formula once you started getting close. It would run less than what the table called for so I would put even more DC into the table. When I repopulated BD with non zero values DC shot up to above table value so I basically overshot what was necessary.

Another thing to note is with the BD table zeroed out my boost curve is very smooth and predictable which you can see in the 010711g15 log. As soon as I enabled BD things got a little dicey which you can see in my latest log 010711g16. Hopefully the -0.01 value will correct those overruns

It would be nice if this ECU would follow its own rules consistantly...oh well.

Here are my current values and latest logs. I will go out later today with my BD table mods and hopefully get rid of my 21+PSI overruns. Car is mad fast at 21-22PSI tho


This is Christians' EBC BD table unaltered - This is where I'll add the -0.01 value to the 1.00 over run value

[amoosenamedhank]

So at first you mentioned that you believe the most accurate way to dial in your WGDC is with you BD table zeroed. But then you said after repopulating the BD your DC was way too much.

So it sounds like dialing the DC in with the BD table zeroed out isn't all that helpful if you everything is skewed afterwards.

[Dano]

hahah

Actually I still believe my method works great. All I had to do was reduce my BD table values on the left side to keep it from over targeting boost. It was a 15 minute fix and a few logs to confirm.

I guess you could inch up DC until you start to see it not logging what is commanded then enable the BD table value and see how far off you are. DC will follow comanded to a certain point right on the money [WGDC table value X2 + 9] then it stops following commanded and you get a % less than commanded. There is some logic or a table in the ECU that causes this I suspect.


I'll post up final results and thoughts as soon as I am finished. Snow here now so no more logging for a few days.

[Frequentflyer]

Originally Posted by Dano2010 hahah Actually I still believe my method works great. All I had to do was reduce my BD table values on the left side to keep it from over targeting boost. It was a 15 minute fix and a few logs to confirm. I guess you could inch up DC until you start to see it not logging what is commanded then enable the BD table value and see how far off you are. DC will follow comanded to a certain point right on the money [WGDC table value X2 + 9] then it stops following commanded and you get a % less than commanded. There is some logic or a table in the ECU that causes this I suspect. I'll post up final results and thoughts as soon as I am finished. Snow here now so no more logging for a few days.

What do you have in your Boost Comp RPM tables? I zero'd mine out, but the stock A table looks like this while B is zero'd out:

[amoosenamedhank]

Question, when I'm datalogging my logged max psi doesn't match the max PSI reported by the AP using the 'live data function'.

Example...on my last log my boost max out at 18.2 psi. While monitoring on the AP, the majority of the pull I saw 18psi. When I took my eyes away from the AP to watch the tach to shift, I looked back down and the AP is reporting a max 22psi... but the log does not reflect this. This is common occurrence , so it wasn't some freak anomaly.

[evidence]

Originally Posted by amoosenamedhank Question, when I'm datalogging my logged max psi doesn't match the max PSI reported by the AP using the 'live data function'. Example...on my last log my boost max out at 18.2 psi. While monitoring on the AP, the majority of the pull I saw 18psi. When I took my eyes away from the AP to watch the tach to shift, I looked back down and the AP is reporting a max 22psi... but the log does not reflect this. This is common occurrence , so it wasn't some freak anomaly.

Wondering if it's because your logging too many variables and only getting 1/2 polls a second in your logs, so it's not picking up the boost spike that the live data display is.

[amoosenamedhank]

Well here is one of my previous logs (live data said max 21psi) and you can see the psi logged got no where near that level.

Edit: Also when this is happening, I'm watching the live data as long as I can and I never see psi numbers this high, but when I look back down the max is way to high.

[Dano]

can't really tell anything from your logs but the live dispay does round up so if you were running say 16.5psi the display would read 17. From your logs it doesn't even look close though.

I would trust the logged data over the live data...as evidence said maybe try to log only PSI and the sample rate will go up so that perhaps you could see some sort of spike but I doubt you would. You are getting ~3 samples/sec which is about what I get.

[amoosenamedhank]

Originally Posted by evidence Wondering if it's because your logging too many variables and only getting 1/2 polls a second in your logs, so it's not picking up the boost spike that the live data display is.

Originally Posted by Dano2010 can't really tell anything from your logs but the live dispay does round up so if you were running say 16.5psi the display would read 17. From your logs it doesn't even look close though. I would trust the logged data over the live data...as evidence said maybe try to log only PSI and the sample rate will go up so that perhaps you could see some sort of spike but I doubt you would. You are getting ~3 samples/sec which is about what I get.

Thanks guys... I'll try just logging psi. I really want to just trust the logged numbers but I always get nervous when I see that 22psi. This brings up another concern of using the 'live data' as an accurate source. Because the more I think about it, this has happened to me before with my AF ratio as well. It reports a max rich of 10:1 on the AP but the logged data only richened to 11.36....

[Dano]

Ok I have logs from my Saturday runs with the changes I spoke of trying a few posts back. Namely adding the -0.01 value in my BD table for the 1.00 over run cell. Well after logging, I determined it wasn't necessary to add more authority for over run protection but rather to lower BD values in the under run area so less WGDC was added to the table value.

Once I did this I had much less over run and in the first tab of the attached logs the run was perfect! What you will still see in the the other tabs is that I do have an over run of ~20PSI around 3.7k and BD kicks in and lowers DC and I lose boost until around 4.4K.

What my next iteration will include is lower WGDC table value in the 3K range which will hopefully prevent the over run at 3.7 and keep BD from removing too much boost.

One thing to keep in mind when WGDC tuning, is that you need to think ~300 RPMs in the past when looking at your boost targets and setting WGDC. Meaning that if your target at 4K = 20 PSI what got you there was basically WGDC values at 3.5K. It would really be the interpolated value between 3.5k and 4K but you get the jist.

I have learned that it is better to have to little WGDC values set and have BD add DC than to have too much WGDC values set and have BD remove DC. When BD removes WGDC, too much boost is lost, at least with the faster reacting EBC setup.

Well enough of my ramblings, here are the map values that go along with the attached log.

Keep in mind my values are for an EBCS control system and will not work very well with OEM BC

Another note on my WGDC setup: I think my ATP setup still has a tiny bit of "controllable" boost creep and this is why you see that my DC values actually go down as RPMs rise. a typical setup would have DC increase throughout the RPM range.

[cld12pk2go]

Dano,

Have you considered putting a pill (~0.025-0.050'') on the boost source line to your EBC solenoid in intercept mode?

This should allow you to run larger values in your BD table and gain resolution for better control.

[Dano]

Originally Posted by Frequentflyer What do you have in your Boost Comp RPM tables? I zero'd mine out, but the stock A table looks like this while B is zero'd out:

man I have looked at them briefly but haven't made any changes....APP appears to be a very easy way to limit power per gear and will do the same if you are running EBC or MBC and I have both available.

IIRC table A is all 1.00 which would mean 100% of what was caculated then table B has the lesser values....I would think we need to modify table A as perhaps table B is for DSC active or steering angle sensor.

Originally Posted by cld12pk2go Dano, Have you considered putting a pill (~0.025-0.050'') on the boost source line to your EBC solenoid in intercept mode? This should allow you to run larger values in your BD table and gain resolution for better control.

I have not tried that as I was under the impression that EBCs were designed to work without any pills.

I can see where that would make the EBC less reactive and thus higher values in the BD table would make less of a change.

Good idea and something to try out.

[Dano]

Originally Posted by amoosenamedhank This brings up another concern of using the 'live data' as an accurate source. Because the more I think about it, this has happened to me before with my AF ratio as well. It reports a max rich of 10:1 on the AP but the logged data only richened to 11.36....

man you have some wierd chit going on with your AP... Are you compairing Actual AFR to Equiv Ratio maybe? They will be different depending on how your MAF is calibrated.

Actual AFR logged and Actual AFR dispaly should be the same.

[cld12pk2go]

On second thought...

If you are in intercept mode and the WGDC signal represents the % of the time the solenoid is open (passing boost), then wouldn't you need the left side of your BD table to be negative and the right side positive to obtain the desired response?

Example: If you are under your boost target you would want lower WGDC to open the solenoid less of the time to pass less pressure to the Wastegate to keep from over coming the spring to open the flapper?

This would also mean that your WGDC % table would be inverted as well (50 = min boost, 0 = max boost). Your log appears to confirm this is the case as your WGDC % decreases as RPM's increase when controlling to your boost target.

This should be the exact opposite of when in bleed mode where higher WGDC = more bleed = less pressure to WGA = most boost.

[Dano]

Hey 12pk,

do you have any idea what the axis values actually are in the Boost Dynamics table? They can't be actual PSI over/under...right?

When I first got my EBC I was thinking the exact same thing...things would operate in reverse to what the table values would dictate...IDK

I'll have to look back at the actual hose routing but IIRC IT is what puts it back in the correct order. Meaning % time the EBC is driven is % time it is open but routed to the TIP

I have it very close to working now either way.

[cld12pk2go]

The axis should be the boost error in PSI.

So once the ECU determines it's boost error it uses the BD table to find the correction factor to feed into the control loop to adjust the WGDC to get to target.

To me it looks like you have neutered your BD table and are driving almost exclusively off the WGDC table.

The only way this could be different is if your solenoid is open all the time at 0% WGDC instead of 100%, but I thought the industry standard was open when driven.

You can test it easily, just plug in something like the following into your BD and see if it controls:

-0.08 -0.05 -0.03 -0.02 -0.01 0.00 0.02 0.04 0.06 0.10 0.20

If that seems interesting, then try:

-0.15 -0.08 -0.05 -0.03 -0.01 0.00 0.03 0.06 0.09 0.15 0.30

Both of those are relatively mild compared to the stock BD table.

[Dano]

Originally Posted by cld12pk2go They should be the boost error in PSI. So once the ECU determines it's boost error it uses the BD table to find the correction factor to feed into the control loop to adjust the WGDC to get to target. To me it looks like you have neutered your BD table and are driving almost exclusively off the WGDC table.

ok lets think about the BD table for a minute.


BTW I never changed the right hand side values from Christians "neutered" values and when I over shoot target, DC is cut so I don't think the EBC does any reversing of boost control logic.

Back to my neutered BD.. you would think so but it ain't so.

With BD zeroed...WGDC values in the 20's would yield logged values using the below formula

table value X 2 +9

Christian has confirmed he sees the same thing... in his example 0 WGDC table value would yield 9% actual DC

This was on the money up until around 2PSI of under reaching boost targets as defined in the BT table. Meaning that as I inched DC values up the formula held and actual matched the formula until I was close to 2 PSI under target. At that point, the actual value was LESS than commanded in the table and it didn't follow the formula nor achieve boost target.

So I started to get a random % of Actual DC vs. commanded and just increased DC until I reached .5 PSI of target.

Then I re-enabled BD with Christian's values and immediately overshot boost targets. it seems that something else is going on somewhere.

[cld12pk2go]

Originally Posted by Dano2010 ok lets think about the BD table for a minute. The read only axis values are -9 -7 -5 all the way to +9. in the middle you only have 1PSI of boost error yet I can get within .5 or less. IDK it seems the ECU would never need to correct for 7-9 psi of over or under boost. Anyway I know the values look like they should be PSI but I am not sure. with zero values up to -5 in the table I saw no change in actual boost and I was within 2PSI of target. When I changed the -7 value my boost went up and met target. BTW I never changed the right hand side values from Christians "neutered" values and when I over shoot target, DC is cut so I don't think the EBC does any reversing of boost control logic. Back to my neutered BD.. you would think so but it ain't so. With BD zeroed...WGDC values in the 20's would yield logged values using the below formula table value X 2 +9 This was on the money up until around 2PSI of under reaching boost targets as defined in the BT table. Meaning that as I inched DC values up the formula held and actual matched the formula until I was close to 2 PSI under target. At that point, the actual value was LESS than commanded in the table and it didn't follow the formula nor achieve boost target. So I started to get a random % of Actual DC vs. commanded and just increased DC until I reached .5 PSI of target. Then I re-enabled BD with Christian's values and immediately overshot boost targets. it seems that something else is going on somewhere.

Actually, now that I am reading this:

Perrin Performance - Boost control explained

It looks like "on" could mean not allowing pressure to pass, which would make sense...

Still seems odd that your WGDC decreases with RPM though...


Oh well, my GS EBC solenoid is on the way, so I get to figure it out here in a few days...

[Dano]

yes exactly...% driven is still on but driven to the TIP or VTA port not the WGA port.

That's what I meant by the hose routing making the difference.

Christian's BD table philosophy is that OEM gives way too much authority to the ECU to alter DC and this is what I chose to follow. So my DC will need to be much closer but there will be less error correction which with an EBC was causing my boost to oscillate too much.

with that said, your idea of running closer to OEM BD and a pill should yield very similar results.

[Dano]

Originally Posted by cld12pk2go Still seems odd that your WGDC decreases with RPM though...

I think is due to the little bit of ATP boost creep....LOL IIRC on MBC it was around 1PSI by RL

Originally Posted by cld12pk2go Oh well, my GS EBC solenoid is on the way, so I get to figure it out here in a few days...

Please post back here or link here to your EBC tuning results. I would be very interested to hear your thoughts.

[NoLeftTurn]

Originally Posted by cld12pk2go Actually, now that I am reading this: Perrin Performance - Boost control explained It looks like "on" could mean not allowing pressure to pass, which would make sense... Still seems odd that your WGDC decreases with RPM though... Oh well, my GS EBC solenoid is on the way, so I get to figure it out here in a few days...

As I understand it, a 0% WGDC the solenoid is closed therefore not allowing any of the boost pressure to bleed off (lower boost). The waistegate would open at whatever spring rating it has internal to it. A WGDC of 100% would mean the solenoid is bleeding as much as it can (raising boost).

Maybe the reason WGDC decreases at a higher RPM is that more air is being forced through the turbo. To compensate for this the solenoid reduces the cycle to reduce the exhaust energy being sent to the turbo (lowering the boost). If I didn't take this precaution the car could go into overboost potentially. Does that sound feasible to anyone else?

[Ziggo]

Originally Posted by 3speedy As I understand it, a 0% WGDC the solenoid is closed therefore not allowing any of the boost pressure to bleed off. The waistegate would open at whatever spring rating it has internal to it. A WGDC of 100% would mean the solenoid is bleeding as much as it can. Therefore, as RPM's increase the WGDC would have to increase (bleed off more boost) to maintaining the same boost given the engine load and throttle postion were constant. Then again I could be completely wrong.....it has happened before many of times.

Half right. WGDC does go up as the RPM goes up since the turbo needs more energy from the exhaust to spool faster.

It's purpose is not to bleed boost, but to lower the boost pressure the WGA sees so it will stay closed at boost pressures above where it would usually open if only the spring was controlling it.

[NoLeftTurn]

Originally Posted by Ziggo Half right. WGDC does go up as the RPM goes up since the turbo needs more energy from the exhaust to spool faster. It's purpose is not to bleed boost, but to lower the boost pressure the WGA sees so it will stay closed at boost pressures above where it would usually open if only the spring was controlling it.

You prematurely quoted me. I saw the error in my ways and edited the post. Is my statement right now???