Dano's Boost Tuning 101 + Cool Stuff

[djuosnteisn]

Originally Posted by Dano but doesn't checking the BT toggle eliminate the load cap? Isn't that what dustin does...or does he do that and use an MBC? it certainly doesn't eliminate load targeting...sooooo maybe try the toggle and set boost targets out of reach an load tune as normal. zero BEC

This is what i was eluding to... using the BT toggle simply to eliminate some of the load based logic (the portion that is likely causing the reduction in WGDC and throttle closing). Keep all things the same, except for enable the BT toggle.

@cld12pk2go

[Enki]

With any luck, this won't be an issue anymore in the near future.

[cld12pk2go]

Originally Posted by djuosnteisn This is what i was eluding to... using the BT toggle simply to eliminate some of the load based logic (the portion that is likely causing the reduction in WGDC and throttle closing). Keep all things the same, except for enable the BT toggle. @cld12pk2go

I am now on my second BT map revision starting at 16 PSI to make sure I am stable on the control loop.

What would you recommend on the LD/BEC/LEC tables for a starting point?

Tks

[djuosnteisn]

Originally Posted by cld12pk2go I am now on my second BT map revision starting at 16 PSI to make sure I am stable on the control loop. What would you recommend on the LD/BEC/LEC tables for a starting point? Tks

Theoretically... according to the diagrams David has posted, the toggle alone will neuter the offending load table and leave LEC as the sole error input for the load based portion. And the BEC is primary error correction logic until the "ramp" drops below RPM Comp A + B... then the LEC becomes the sole source for error correction over all.

Unfortunately we don't have access to the "ramp" value. It's not something we can log.


But for giggles, i'd try something like setting the RPM Comp A & B tables to 50 & 50... so that A+B = 100... and you can pretty much guarantee the ramp will be below it. That should eliminate the BEC influence.

At that point.... LEC is the only input term left... and the car should behave like traditional load based.


Summary:
- BT toggle enabled (eliminates 2 of the 3 load error influences, leaving LEC)
- RPM Comp A & B set to +50... making "ramp" always under A + B and removing BEC error influence.
- LEC and LD populated with OEM values... and start out with safe TRL load targets.

You should technically be able to set the boost targets to whatever you want, and they shouldn't cause issue... but you may wanna set them either lower than desired, or at desired boost pressure.

[Fobio]

Originally Posted by djuosnteisn - LEC and LD populated with OEM values... and start out with safe TRL load targets.

^^^This is basically what I've been doing since the V200 maps.

Use Boost Tune toggle.
Set your boost target.
Use math then to correct boost (thru WGDC) and AFR's (thru MAF).
Adjust TRL tables until all targets are met.
Fine-tune as necessary.
Done-ski.

[Dano]

Originally Posted by djuosnteisn - LEC and LD populated with OEM values... and start out with safe TRL load targets.

just to be clear OEM or Cobb OTS?

I assume you mean OTS but you know what they say about AssUMe

[atvfreek]

Originally Posted by Fobio ^^^This is basically what I've been doing since the V200 maps. Use Boost Tune toggle. Set your boost target. Use math then to correct boost (thru WGDC) and AFR's (thru MAF). Adjust TRL tables until all targets are met. Fine-tune as necessary. Done-ski.

I do the same, except with slightly higher values in the BEC table.

Tappin

[cld12pk2go]

Originally Posted by djuosnteisn Theoretically... according to the diagrams David has posted, the toggle alone will neuter the offending load table and leave LEC as the sole error input for the load based portion. And the BEC is primary error correction logic until the "ramp" drops below RPM Comp A + B... then the LEC becomes the sole source for error correction over all. Unfortunately we don't have access to the "ramp" value. It's not something we can log. But for giggles, i'd try something like setting the RPM Comp A & B tables to 50 & 50... so that A+B = 100... and you can pretty much guarantee the ramp will be below it. That should eliminate the BEC influence. At that point.... LEC is the only input term left... and the car should behave like traditional load based. Summary: - BT toggle enabled (eliminates 2 of the 3 load error influences, leaving LEC) - RPM Comp A & B set to +50... making "ramp" always under A + B and removing BEC error influence. - LEC and LD populated with OEM values... and start out with safe TRL load targets. You should technically be able to set the boost targets to whatever you want, and they shouldn't cause issue... but you may wanna set them either lower than desired, or at desired boost pressure.

That sounds like a solid plan. If I can avoid the current hidden load limits I have stumbled upon and still have load targeting capability, then that would be awesome. Thanks!

I do have a point about which I would like further clarification: What exactly is the definition of the "ramp"? It sounds like it is a rate of change of the WGDC loop driven by the PID feedback that is happening in the background (correct?).

[rfinkle2]

Originally Posted by cld12pk2go That sounds like a solid plan. If I can avoid the current hidden load limits I have stumbled upon and still have load targeting capability, then that would be awesome. Thanks! I do have a point about which I would like further clarification: What exactly is the definition of the "ramp"? It sounds like it is a rate of change of the WGDC loop driven by the PID feedback that is happening in the background (correct?).

It is the max allowable deviation from the base wgdc table (a static value, not a % of the wgdc table). (already took a jab on this one for testing it on my own car for the thread we've been questioning David in, but I digress).

It does seem that the further the ramp comps are away from eachother, the faster the summation and "reloading" of that value into the base wgdc table. (@ least to me)

[djuosnteisn]

Originally Posted by Fobio ^^^This is basically what I've been doing since the V200 maps. Use Boost Tune toggle. Set your boost target. Use math then to correct boost (thru WGDC) and AFR's (thru MAF). Adjust TRL tables until all targets are met. Fine-tune as necessary. Done-ski.

Right... the only thing your not doing is adjust thing RPM Comp A & B tables... to force the "ramp" to be below them at all times, and soley use LEC. So your still running with both BEC and LEC in conjunction... which is fine, cause both those targets are setup correctly.

Originally Posted by Dano just to be clear OEM or Cobb OTS? I assume you mean OTS but you know what they say about AssUMe

I prefer OEM for LEC and LD. And honestly, i don't think LEC and LD ever changed from OEM to OTS. I think BEC was the only one that got reduced drastically.

Originally Posted by cld12pk2go That sounds like a solid plan. If I can avoid the current hidden load limits I have stumbled upon and still have load targeting capability, then that would be awesome. Thanks! I do have a point about which I would like further clarification: What exactly is the definition of the "ramp"? It sounds like it is a rate of change of the WGDC loop driven by the PID feedback that is happening in the background (correct?).

Yup, it's most likely the current result from the PID weighted terms. Not something we can log.

Originally Posted by rfinkle2 It is the max allowable deviation from the base wgdc table (a static value, not a % of the wgdc table). (already took a jab on this one for testing it on my own car for the thread we've been questioning David in, but I digress). It does seem that the further the ramp comps are away from eachother, the faster the summation and "reloading" of that value into the base wgdc table. (@ least to me)

Your think of the "ramp limits", not "ramp" itself. The "ramp" is the current PID output... and the limits simply set how far the WGDC is allowed to deviate form the WGDC table.

[Dano]

so any ideas what is causing the "shark fin" effect.

I sent out another map to creep with 100 WGDC in it from 6K + and again it runs its normal DC ramping up according to BEC values [~70ish] then all a sudden it pegs 99 then ramps down according to BEC overrun until its close to target then jumps back to 99...rinse/repeat.

The good news is with populated LEC [LD still zero] the BEC table values are actually being added to DC now. I just need to add a bit more BEC which should get us on track.


I didn't want to create an overrun condition with LEC having my TRLS set to 3.0 so this is very conservative but it seemed to do the trick. just a lil push from LEC to get BEC to wake up

LEC values:

0.50	0.50	0.50	0.50	0.00	-1.25	-1.25	-1.25	-1.25	-1.25	-1.25

I should have more logs tonight to confirm this is now working. Still need to cut WGDC values down to avoid the "shark fin" effect.

[djuosnteisn]

Yeah, the only way i was able to rationalize that phenomenon was this:

He was over boost, IIRC, and the "ramp" was going negative. WGDC table had high values, like 100's, in higher rpm's. And also had higher load targets, like 3.0.

When over boost, "ramp" goes negative... dropping WGDC... and finally "ramp" drops below the RPM Comp A + B (-10)... then suddenly switched to LEC only... which was targeting 3.0... so WGDC increased quickly. Then once "ramp" was higher than A+B... it switched back to BEC... and started dropping again.


That's the best i got right now lol.

[Dano]

well kinda


the first "fin" occurs while .5PSI UNDER target...[Target is 19 in this map] DC is ramping up to attempt to reach target to around 80 then jumps to 99 then boost is overshot and the reduction starts for the second time.

but I have never seen DC "jump" to commanded target no matter how far off commanded was from actual....never...

hell OTS maps command 120% DC in that RPM range...lol

[djuosnteisn]

Originally Posted by Dano well kinda the first "fin" occurs while .5PSI UNDER target...[Target is 19 in this map] DC is ramping up to attempt to reach target to around 80 then jumps to 99 then boost is overshot and the reduction starts for the second time. but I have never seen DC "jump" to commanded target no matter how far off commanded was from actual....never... hell OTS maps command 120% DC in that RPM range...lol

I know what you mean. But it's hard to say what exactly was going on with the "ramp" variable at that time. If it's the resultant from a PID loop, then you have to take lots of things into consideration, like was it over boost for a period of time before the first fin, the current error, and how fast is the error changing.

I honestly have no idea. One of my main fears at the moment is that the RPM Comp A + B logic may not be as simple as advertised.... but who knows. David's the one bathing in the source code, so i trust what he says. I guess the lil voice in the back of my mind is just worried that the logic may be different, and perhaps the A table represents negative "ramp" switch limits, and the B table represents positive "ramp" switch limits.... or something more along those lines.


Once @cld12pk2go gets some feed back on the A+B = 100... it'll at least shed a lil more light on the subject.





And i agree dude...... i have never seen any of the recent funkiness before. Seems like something changed, either via colder weather... or a recent ATR update. Dunno.

[rfinkle2]

I think a 3 port is too sensitive to OEM values, but what do I know.

[Dano]

lol

I run higher than OTS BEC values on my GS to work correctly so there you have it.

this current tune is OE boost control hardware.

[rfinkle2]

Originally Posted by Dano lol I run higher than OTS BEC values on my GS to work correctly so there you have it. this current tune is OE boost control hardware.

On a pure boost tune correct? LOL

I'm stating that I think on a load tune the oem values are too aggressive.

Aren't we talking about cld12pk2go's tune?

[Lex]

So by setting the Boost Toggle and using LEC, are you finding that the mystery load table limiting load is completely eliminated and we no longer have to overshoot the targets in the TRL tables?

[djuosnteisn]

Originally Posted by Lex So by setting the Boost Toggle and using LEC, are you finding that the mystery load table limiting load is completely eliminated and we no longer have to overshoot the targets in the TRL tables?

TBD... cld12pk2go is giving it a try.

[Lex]

Originally Posted by djuosnteisn TBD... cld12pk2go is giving it a try.

... Because getting what you ask for from a complex control system seems to make sense

[Dano]

Originally Posted by rfinkle2 On a pure boost tune correct? LOL I'm stating that I think on a load tune the oem values are too aggressive. Aren't we talking about cld12pk2go's tune?

lol...talking about 3 tunes


"shark fin" = boost_creeps tune [funny that is SN is boost creep yet we can't hold target to RL hahahah]
my values are on a pure boost tune
I haven't commented on 12pks tune

and yes when I was load tuning with my GS my LD [pre LEC lol] was extremely reduced compared to OTS/OEM.

Originally Posted by Lex ... Because getting what you ask for from a complex control system seems to make sense

come on Lex...that's way too much to ask for.

[MattJackson86]

Hoping to get some input on my logs. 2 issues.

1) Shortly after going WOT I hit my target AFR, then I go lean for a couple hundred RPMs (it looked longer on my dashhawk but according to AP logs its not for long, but still).

2) My WGDC are not being affected by BEC what so ever. My boost just followed WG Duty Cycles exactly, even with huge values in BEC. This seems to have been the topic of the last few pages, and I will read through them.

Main issue is #1. Any other input would be great. Not had someone look at this tune for a while so input please!

4th.csv is a full 4th gear log. datalog3.csv is just a short 3rd gear log showing the lean issue.

[rfinkle2]

I've seen that a couple of times now in different logs (the lean issue), and more importantly, @ about 3.9V.

It just dawned on me after having had to manually adjust wot afr's (3.9 volts and the lean condition being a pattern).

Anyone else lean @ 3.9 volts?

What do your ramp high limits (boost and load) and rpm comp A & B tables look like?

[Lex]

3.9V is probably a common point where the turbo spools and open loop is entered.

[MattJackson86]

Originally Posted by rfinkle2 What do your ramp high limits (boost and load) and rpm comp A & B tables look like?

Boost Ramp High 15%
Boost Ramp Low -100%

Load Ramp High and Low 0%

RPM Comp A -20
RPM Comp B 10

Originally Posted by Lex 3.9V is probably a common point where the turbo spools and open loop is entered.

In my log I was jumping to OL around 3.5v. Load hit 1.25 and I entered OL, AFR dropped to target. It was around 3.85v and 2.0 load that the lean condition occurred.

For now I manually adjusted my WOT AFR in that range and will test out the map tomorrow. Just to confirm, tuning a WOT AFR, do I multiply by actual/target (12.0/11.8 i.e. 1.017) or the other way around? Want to double check my late night thinking before I make it go leaner by accident.

[Lex]

The "lean" bump as the turbo spools is very common ... it has been speculated that it can even be from O2 sensor latency.

[rfinkle2]

Originally Posted by MattJackson86 Boost Ramp High 15% Boost Ramp Low -100% Load Ramp High and Low 0% RPM Comp A -20 RPM Comp B 10 In my log I was jumping to OL around 3.5v. Load hit 1.25 and I entered OL, AFR dropped to target. It was around 3.85v and 2.0 load that the lean condition occurred. For now I manually adjusted my WOT AFR in that range and will test out the map tomorrow. Just to confirm, tuning a WOT AFR, do I multiply by actual/target (12.0/11.8 i.e. 1.017) or the other way around? Want to double check my late night thinking before I make it go leaner by accident.

I've tried various combinations of these, and until all of the relationships are 100% understood, I would leave those populated.

You are 100% correct as far as wot afr correction goes.

I remember it this way "erectile dysfunction" , LOL, or E/D = exhaust reading / "dic"tated.

That ought to stick in your mind!

[Dano]

I see this from time to time on different cars and unless the car is knocking I don't worry about it nor alter my commanded AF.

I typically command 11.8 on e-tunes and it can drift up to 12.0-12.2. Its nothing to worry about. I run 12.2 on my car even without meth or E and it never knocked.

its much to do about nothing.

with that said there is another lean condition where the car remains lean for an entire run and I've seen 13.x AF. That does need to be addressed and the clutch in/out fueling tables appear to resolve that.

[silvapain]

Just to piggyback on what Dano said, I also ran 12.2 WOT AFRs on pump gas with no knock.


Tapadatass

[Nataphen]

^As do I. 12.2 on 93 octane is no problem, even with extra timing.

[m4tic]

eff you guys.. 13.5 is where its at

[Dano]

lol your car is a freak!

I assume you are using the new super cool maps? With the new clutch in/out settings?

[MattJackson86]

Yea I wouldn't be so concerned, but I only run 91 and I seem to be knock prone. People always comment on my timing and how it seems low, but anytime i adjust the slightest amount it starts to knock. Just want to resolve it if at all possible. Adjusted my map and will see how it goes. Then on to the boost issues.

[Todd98SE]

Originally Posted by MattJackson86 Yea I wouldn't be so concerned, but I only run 91 and I seem to be knock prone. People always comment on my timing and how it seems low, but anytime i adjust the slightest amount it starts to knock. Just want to resolve it if at all possible. Adjusted my map and will see how it goes. Then on to the boost issues.

Matt you need to find you some E85 dude. I'm running about 2-3 gallons per tank and that shit is like magic. I just got started with the E but I'm currently running 11.6 AFR's and 19-19.5 psi with 18 degrees of timing at redline and it is completely knock free and runs like a beast.

Before the E I had to target 11.2 AFR's and the max timing I could use was around 9 degrees and even then I had occasional bouts of KR.

It's amazing what 1-2 gallons of E will allow you to do. Give it a try.

[Boosted Beluga]

I guess my motor is a boarder line freak then. I'm on 91 11.8 afr and running 16.5° of timing at redline with like 1.06 kr so I added some fuel up top.

[silvapain]

California 91 octane gas is complete crap. I'm not surprised you Cali guys can't run lean AFRs or advanced timing.

At least you can tell yourself that you have plenty of time to look at all that beautiful scenery when you're going slow.

[MattJackson86]

Originally Posted by Todd98SE Before the E I had to target 11.2 AFR's and the max timing I could use was around 9 degrees and even then I had occasional bouts of KR.

Glad its not just me. I am managing 11.8, but yea timing is 10 max with the odd degree or two of KR. I will definitely look into some E85.

[Nataphen]

One of the guys that I help tune is in CA, and it kills me to tell him that's all that I can do for him, LOL. I'll send him a similar map to mine that's completely KR free on TX 93 octane, but drop AFRs from 12.2 to 11.4, lower my timing about 4*, and he's usually just about maxed out there. Sometimes, I have to pull even more timing at a few points. He's probably taking at least a 20-30hp hit all through the power band.

I would highly recommend trying 2 gallons of e85 per tank, you usually won't have to adjust your MAF cal for 2 gallons or less.

[silvapain]

Originally Posted by Nataphen I would highly recommend getting the fuck out of California.

Fixed.

[MattJackson86]

So CA 91 + 2 gallons on E85 and all should be good? A MAF cal is easy if thats all it takes, just know nothing of E85 and its side effects (if any).