NefMoto

Noob Zone => Noob Questions => Topic started by: DonkeyDub on October 10, 2019, 01:30:36 PM



Title: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 01:30:36 PM
2004 AWP 1.8t Jetta k03s 6Speed
Unitronic Stage 2 Flash
Full 3 inch exhaust W/ Cat and muff
Forge Inlet & CAI filter

3rd gear logs Engine is pulling 11+ CF and audible knock is briefly heard.
Pulled n75 electrically and car was still pulling 6CF under load.

Issues:

-Ignition timing goes negative :o

- The ECU is demanding 1.0 lambda under full boost/190% load- until it pulls timing or hits 4k rpm

-MAP values are low compared to  boost gauge. MAP values never show boost spiking close to 25psi or
over 2543millibar but the Gauge will.

I believe my issue is Lean AFR





Hope someone can take a look at this log and give me some feed back.
**During this log 4.5 degrees of ignition timing was removed via UNIsettings.**

Thanks,

DonkeyDub


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 10, 2019, 01:37:29 PM
You are asking for trouble with that boost request w/o the 5120 hack.

stick to no higher than 2450

And yes, you need a lot more fuel at or before peak torque.

If you still get zero timing at peak torque, i'd consider much lower boost between 3000-4000 rpm or run water meth, better gas, or lower compression.

GL, fighting with knock is a pita on these high compression motors.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 02:25:13 PM
Hello Nyet,

Is 2550 boost request normal for uni stage2?
Sensor seems to cap at 2543.

WaterMeth would be ideal, ECU seems fixated on 13 AFR.

I contacted Unitronic about the ECU requesting lean AFR's and they
replied that Under Full load, Engine should be demanding .78 lambda.

Their only comment to my ECU requesting 1.0 lambda at 190% load is I should bring it
into an 1.8t/20v specialist, very supportive.

Where is peak torque, 2800 rpm?

Another AFR oddball, Sometimes while cruising at 2000rpm, low engine load 12.5%.The engine will request
.75 lambda :S


- What is a good AFR while turbo is spooling up? 12.5?

Heres another log with -8 Ignition timing,+15 Main Fuel trim and 80SEL via UNIsettings

Thanks,

DonkeyDub


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 10, 2019, 02:33:54 PM
Is 2550 boost request normal for uni stage2?
No clue, don't care, doing adjustments on other people's tunes is asking for trouble, and having them adjust stuff for you never works out well. It is the reason I started DIY in the first place
Quote
Sensor seems to cap at 2543.

Yes. Running a PID near the end of the process measurement limit is a BAD IDEA

Like, terrible.

Quote
Where is peak torque, 2800 rpm?

Peak torque is generally around peak load.

Quote
- What is a good AFR while turbo is spooling up? 12.5?
As long as it gets to 12.5 ish by peak torque you're good. Make it too rich during spool and you may give up spool time

Quote
Heres another log with -8 Ignition timing,+15 Main Fuel trim and 80SEL via UNIsettings

At this point you are going to have to start from scratch with a stock file and DIY, or find another tuner. Tweaking with UNI settings isn't going to help.

The reason you are seeing weird fueling only sometimes is apparently UNItronic doesn't know the difference between BTS fueling and LAMFA fueling.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 03:02:40 PM
Quote
No clue, don't care, doing adjustments on other people's tunes is asking for trouble, and having them adjust stuff for you never works out well. It is the reason I started DIY in the first place

Yeah I think thats the route I will be making.
Silver Lining, i'll share my hind sight for others
save that Flash money for a custom tune or good water/meth kit And DYO Flash.

Quote
At this point you are going to have to start from scratch with a stock file and DIY, or find another tuner. Tweaking with UNI settings isn't going to help.

The reason you are seeing weird fueling only sometimes is apparently UNItronic doesn't know the difference between BTS fueling and LAMFA fueling.

What are my options for getting back to square one, stock tune. Is this something I can do myself or Will I need a Uni approved shop to flash the stage2melter off my ecu.

Time to hit the books!

Just a wild guess is BTS related to Engine load and LAMFA  part throttle/idle?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 03:19:21 PM
The issues were there long beforehand trying this.

But I didnt get the memo about Not Full Flash Reading Unitronic files,
Attempted that the other day, It was unsuccessful and Gave DTC P0602 which comes back
after erasing. I also got one P0606 Code but I believe that one was random as it has not returned.

Thanks again,

DonkeyDub


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 10, 2019, 03:38:15 PM
What are my options for getting back to square one, stock tune. Is this something I can do myself or Will I need a Uni approved shop to flash the stage2melter off my ecu.

Depends if there is a hardware riser on the ECU, or if UNI disabled flash write. If there is no riser you can bootmode it if normal obd flashing doesn't work.

Quote
Just a wild guess is BTS related to Engine load and LAMFA  part throttle/idle?

See my sig.

BTS is related to modeled EGT, LAMFA is related to pedal position.


Title: Re: 1.8t/20V KnockKnock
Post by: Blazius on October 10, 2019, 04:12:49 PM
BTS is used for fueling for longer travels and high EGT's but ofcourse you can use it all the time.  LAMFA uses pedal torque request, and this is a desired value, and since desired is always ahead of actual , it is usually used for acceleration enrcichment.

negative timing on high rpm/boost is real bad, also 11 cf too, you need to get your stuff sorted man, dont do pulls like this.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 04:18:35 PM
Flash Read didin't work, is Flash Write still possible?

Will a flash write affect IMMO?

Is Bootmode only accessed through a Galletto cable?

Wilco!

Much Appreciation

Thanks,

DonkeyDub


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 10, 2019, 04:20:52 PM
Flash Read didin't work, is Flash Write still possible?
I doubt it.

Quote
Will a flash write affect IMMO?

It should not

Quote
Is Bootmode only accessed through a Galletto cable?

Yes but the galletto tool can be hexedited to allow using any ftdi cable.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 04:29:23 PM
Hello Blazius,

Ahhhh and to that, I think both BTS and LAMFA maps are awful on my vehicle.
Fuel Enrichment by knock  ::)

Ignition was retarded after CF's of 11, but yes. No more boost until resolved.



Thanks,

DonkeyDub.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 04:42:53 PM
Oh wicked! nyet,


Just to summarize. Lean AFR's being requested is likely the cause of knock?

Because I still find it odd that
Boost data does not match the mechanical gauge.
It'll hit 25psi :o

Data usually shows hitting max boost at 3200rpm, Boost Gauge looks more around 2600

Was wondering if I just have a bad MAP sensor.

It'll also hit boost without requesting it  :S

Thanks,

DonkeyDub


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 10, 2019, 04:54:26 PM
Just to summarize. Lean AFR's being requested is likely the cause of knock?
Along with too much boost and too much timing request.


Quote
Boost data does not match the mechanical gauge.
It'll hit 25psi :o

Of course not. This is why requesting boost near the process variable limit is bad.

Quote
Data usually shows hitting max boost at 3200rpm, Boost Gauge looks more around 2600

Was wondering if I just have a bad MAP sensor.

I doubt it.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 10, 2019, 05:04:24 PM


Cheers and thanks again


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 11, 2019, 03:24:24 PM
Alright

Vehicle is now flashed back to stock
Here is a second gear log.

Still looks lean


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 12, 2019, 03:05:54 PM
Here is another Log, oem flash





Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 13, 2019, 04:32:11 PM
1) stock is almost always lambda 1
2) your idle rich condition is problematic, possibly due to your CAI. Good luck getting fueling right w/o the stock airbox. It can be tricky to get right.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 13, 2019, 05:10:18 PM
It is  pain off idle, power dogs off when engaging the clutch !!

Would resetting trims and unplugging the maf be a good test to rule out the CAI??

Odd how in the last log, CF was mainly cylinder 2 and this log it's number 3
Maybe ill check the timing and do a leak down test for peace of mind.
When plugs were last pulled, Cylinder 1 and 2's piston tops were pretty damn clean
while 3 and 4 had darker tops





Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 13, 2019, 05:37:00 PM
Would resetting trims and unplugging the maf be a good test to rule out the CAI??

No, not really, unfortunately. You can pressure test though to find leaks that might be contributing to the rich condition, though missing metered air during idle is pretty rare; more commonly an intake leak would lead to a lean condition (vac sucking in extra unmetered air).

Quote
Forgot to mention the vehicle has BF transmission and Dog Bone mount, and also
a Valeo Single mass flywheel.

Shouldn't affect fueling.

Quote
Odd how in the last log, CF was mainly cylinder 2 and this log it's number 3

I wouldn't read too much into this just yet.

Quote
Maybe ill check the timing and do a leak down test for peace of mind.
When plugs were last pulled, Cylinder 1 and 2's piston tops were pretty damn clean
while 3 and 4 had darker tops

Can't hurt but also likely won't help with fueling issues...


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 13, 2019, 05:59:47 PM
Yeah CF at this point is likely due to fueling.
Pressure tested a couple times and passed.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 14, 2019, 11:32:10 AM
on ignition tables
Looking at '' KFZW '' X and Y both appear to be RPM
and ''KFZWOP '' Y = RPM and X = Load?

I don't think I will but if the CAI is causing fuel issues,
To get it to run smoothly, would most corrections be made in the table ''KFKHFM''.


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 14, 2019, 01:26:39 PM
Looking at '' KFZW '' X and Y both appear to be RPM
That makes no sense.

Quote
and ''KFZWOP '' Y = RPM and X = Load?

This and more is covered in both the s4wiki and the FR, and extensively throughout this site.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 17, 2019, 03:30:55 PM
In OEM2 Log during Cruise and idle

Lambda Req hits  0.80

Is this normal?
 


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 17, 2019, 06:44:10 PM
In OEM2 Log during Cruise and idle

Lambda Req hits  0.80

Is this normal?
 

No.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 17, 2019, 07:30:52 PM

What would cause that?








Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 17, 2019, 07:51:56 PM
bts, lamfa, atr/egt, or kr enrich.

All are covered in the FR and the tuning wiki along with which variables to log to diagnose.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 17, 2019, 11:51:54 PM
 Variables added: tabgbts_w  ,   tikatm_w   ,  tkatm    ,  lambts_w 


Actual boost being higher than desired at idle and part throttle.
Does this reflect a bad sensor or some other issue?


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 18, 2019, 12:55:50 AM
You managed to miss kr and lamfa fueling


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 18, 2019, 02:10:11 PM
Looks like it could partly be BTS maps,

-It does req .80 at times without meeting BTS threshold

-Tries to maintain 760RPM Idle but it bounces between 720-760 RPM

-Added Supply voltage variable and it dips to 12.5 Volts



Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 18, 2019, 02:23:58 PM
Looks like it could partly be BTS maps,

-It does req .80 at times without meeting BTS threshold

-Tries to maintain 760RPM Idle but it bounces between 720-760 RPM

-Added Supply voltage variable and it dips to 12.5 Volts



You know if you actually logged bts and kr lambda req you could stop guessing.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 18, 2019, 05:12:58 PM
Like this?


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 18, 2019, 09:25:06 PM
Wow. Yes.

That is truly bizarre. I have no idea where your lambda req is going wrong.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 19, 2019, 09:33:48 AM
 
Before Lambda Desired ( lamsbg_w) is .80 ,

''tabgm_w'' ( BeforeCatEGT ) crosses ''tabgbts_w''. ( EGTPartProtection ).

but ''Lambts_w'' looks unactive , weird.

06A906032RN KFLBTS MAP .80 lambda starts at 4500RPM, My Issue is around 2000RPM.

Anything else look odd to you?
What about ignition timing and AFR on deceleration before it goes .80

Downshifting the other day, I got a little gun shot out of the exhaust.


Title: Re: 1.8t/20V KnockKnock
Post by: RBPE on October 20, 2019, 08:54:58 AM
Not able to check your logs but have you checked your lamfa map for being correct? Some of the 1.8T's are jumbled so may look right but rpm axis wrong sometimes ( & poss cross flashing prob  doing that if a commercial tune maybe?)


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on October 20, 2019, 11:23:56 AM
Not able to check your logs but have you checked your lamfa map for being correct? Some of the 1.8T's are jumbled so may look right but rpm axis wrong sometimes ( & poss cross flashing prob  doing that if a commercial tune maybe?)

His lamfa_w is 1.0 throughout.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on October 20, 2019, 02:18:17 PM
I want to log the same variables but instead use an .ori instead of a 06a906032RN.bin from the forums.

Neftmoto Ecu Flasher disconnects when trying to read info

Unplugging Instrument cluster fuse resolved this error but Full flash read gave Immobilizer not authenticated error.

- Edit - Turn key, let immobilizer authenticate and then pull fuse for instrument Cluster. It was #11 for my 2004 Jetta.
This will set an Airbag Fault and requires the Full version of VCDS to clear it.

Fuse 15 also powers the cluster and may allow connection without giving an airbag fault, I will try that next time.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on November 09, 2019, 02:47:18 PM
My ps_w is over actual boost.
It doesn't max out but is this from the CAI?

S4Wiki
Quote
Another good MLFHFM sanity check is to log ps_w vs actual boost at WOT - ps_w should generally be slightly below actual boost. If it is over, your MAF is probably overscaled

When I set MLOFS to - 0 , are MLFHFM values in tuner pro supposed to change?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on November 13, 2019, 05:37:43 PM
Here's a log showing ps_w going over,

To get AFR from Wide band 02 voltage, is it 2*Voltage+10= AFR?


Any advice on getting this to run better, will be appreciated.


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on November 13, 2019, 05:55:43 PM
How about starting with not overboosting?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on November 14, 2019, 01:41:53 PM
-


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on November 17, 2019, 06:19:30 PM
Lowering KFLDIMX doesn't seem effective or give any change to
how the wastegate DC acts.

Would KFLDRQ2 be next to change to try and dampen overshoot
or is it neccesary to change KFLDRL with an OEM k03s turbo.





Title: Re: 1.8t/20V KnockKnock
Post by: nyet on November 17, 2019, 06:27:24 PM
Impossible to tell until you log with useful variables, such as all the PID variables.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on December 17, 2019, 08:32:41 PM
Adjusted CLRSKA & CLRSHK, no more .80 lambda request under deceleration.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on February 24, 2020, 01:38:30 PM
Hey ,

How closely should ps_w Follow pvdkds_w?

My ps_w goes over pvkds_w when under boost, sometimes by .1 Bar

Out of boost ps_w is 300mbar while pvkds_w is 1000mbar .







Title: Re: 1.8t/20V KnockKnock
Post by: nyet on February 24, 2020, 02:37:45 PM
Hey ,

How closely should ps_w Follow pvdkds_w?

My ps_w goes over pvkds_w when under boost, sometimes by .1 Bar

Out of boost ps_w is 300mbar while pvkds_w is 1000mbar .

.1 isn't bad.

Out of boost that is expected. ps_w is modeled MAP, pvdkds is pressure upstream of the throttle plate. If it isn't clear why those are expected to be different, may want to do further reading before doing any more tuning ....


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 11, 2020, 05:24:28 PM
-With ps_w being over, do you suggest scaling back the Maf?

-WOT pulls Ignition isn't following KFZW and it goes low (& negative ) between 2000 - 3300 rpm  with Actual < Specified , Miist doesn't follow Mifa.
there isn't CR. I also get this ignition cut off idle.


Does this look like ARMD?

I know and and I will look in the FR but should Miist follow Mifa?




Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 12, 2020, 07:25:53 PM
wtf kind of log is this? There's barely any locations logged.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 15, 2020, 03:18:19 PM
Limited, get timeout if I go over 32 bites.
Wouldn't want to but could try with cluster unplugged.

Playing with Imx and and got adaptions to zero, still had too much overshoot.
Lowered Imx bit more for this log and it doesn't look as good and gives small corrections..

Going to go back and try Q2 to tame overshoot.

On the last KfldImx axis my input
from 1750-3500 I have 50
4000-6500 is 60.

In the log it shows Boost PidIMax higher, is this showing it with the corrections added?




Ignition drops low even with Actual < specified, from the wiki
Quote
If you are seeing timing intervention when PED is >60% you are likely seeing ARMD intervention

The defined file I use has KFDMDADP and KFDMDARO 0, KFDMDARO 1, KFDMDARO 2 , KFDMDARO 3 and
KFDMDAROS.

Which map should I adjust? Just KFDMDARO 0?

Also tempting the idea of Scaling back the Maf.


Thanks for the patience!
 


 


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 15, 2020, 10:11:43 PM
Without proper logs its going to be impossible to do anything, especially if you cant' even log req boost and wgdc


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 19, 2020, 05:42:09 PM
What should I log to diagnose this timing drop?

I don't see Requested boost in config.

What is Requested wgdc?


Drop Occurs with Actual Load < Request Load

and also my Actual Load>Requested load when Actual Boost< Desired Boost.







Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 19, 2020, 05:44:30 PM
Tested it with TMAR -143 and got the same results.


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 19, 2020, 06:23:00 PM
I'm having trouble getting the ECUxPlot filter working

Please log pedal position


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 19, 2020, 06:27:04 PM
I'm not seeing any timing pull


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 19, 2020, 06:29:56 PM
1) nothing wrong with your load
2) nothing wrong with your timing
3) fueling looks fine but you need to log lambda control



Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 19, 2020, 07:07:53 PM
EcuXplot gave JNI error and wouldn't open,
I'll look into it but am happy with Vlogger :)


Wilco.


Here is my KFZW and KFZW2 Map.

In that Screenshot my Zwout is 0 at 2544 RPM
168 rl_w and Load requested is 171 KFZW/KFZW2 at that Rpm is 3.75

@4000Rpm Zwout : 4.5
Zwsol: 3.75
 KFZW/2 at that Rpm is 8.25/9.00
and rl_w < rIsol_w

No KR, wondering where my timing is hiding.






Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 19, 2020, 07:09:01 PM
KFZW/KFZW2 map


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 19, 2020, 07:50:05 PM
there are a bunch more timing corrections in the ignition path.

just increase req.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 19, 2020, 09:14:53 PM
Thought that might've been it.

If actual Load is still below requested, why do I get corrections?
Checked Ignition adaptions and they were 0. Is this Pre-emptive before actual load goes over requested?



So I should raise KFIRML to get more requested load?

I want to keep it at 1.25bar? I thought a higher load request will request more boost?
And this is already capped by my HBN  set at 2.25. Even with a higher KFIRML, wouldn't
my load still be capped similarly because of HBN.

I know , noob status.

It seems like requested load isnt = to request boost.









Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 20, 2020, 12:53:22 PM
Is zwgru direct output or can dzwwl lower it?

Unfortunately dzwwl isn't in logger config


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 20, 2020, 01:04:49 PM
ahhhh,

dzwkg, dzwol and dzwoag are applied to KFZW before becoming zwgru.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 20, 2020, 01:26:15 PM
KFDZWKG map ( Correction of the ignition angle by shifting the knock limit ) outputs - zwkg

dzwol - Lambda ignition angle correction of the optimal ZW
                                   
dzwoag - Exhaust gas return rate correction of the optimal ZW

Are you suggesting to raise KFZW values to counter these corrections.

I'm curious how KFDZWKG works. Shifting the knock limit?


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 20, 2020, 07:52:00 PM
Yes. Increase KFZW. OR:

If your actual load is overshooting your request habitually, yes, you can retune KFMIOP/KFMIRL. But if they're already inverses of each other, you may just want to underscale MLHFM a bit and bring up KRKTE accordingly until actual load is always under req. I know some purists will object, but slight load underscaling never hurt anyone, plus it will automatically bring up your KFZW output for you as a side effect without ever having to touch the table itself.

Also compare ps_w to actual ps during wot to see if it is a MAF too high issue.

Do not touch any knock control maps. They should be fine stock.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on March 21, 2020, 03:22:26 PM
Stock KFZW, too much but It did of course bring up Zwgru.
Need to tweak them down.


What throws me off is Zwout = Zwsol when rl_w< risol_w.
I need to look more into FR and Torque monitoring. If I had mibas and mimax, it'd likely be very obvious.

I'd be more tempted to under scale MLHFM if the above wasn't happening. Example in screenshot.


Still unsure of KFMIRL/IOP.   
I've seen post about always having actual torque<requested.

In stock iop it seems actual = requested or actual>requested.

IOP:1240 RPMS @ 140 load = 70.59   KFMIRL: 1250 RPMS @ 70% Torque = 138.82

If the former is true, mine is incorrect and I should Increase KFMIRL.

Side question, how closely should msdk_w follow mshfm_w? is there an acceptable tolerance?







Title: Re: 1.8t/20V KnockKnock
Post by: nyet on March 21, 2020, 06:25:47 PM
You're overthinking things, especially now that you are hitting knock control

Stop worrying about torque intervention. It isn't a problem in your logs, you're making enough timing to hit knock limits as it is.

You didn't bother logging actual boost and ps_w, which is what you are really looking for if you are worried your MAF isn't scaled right.

I don't recall if you mentioned you had a stock MAF or not, but a CAI is going to fuck with your MAF readings.

Not enough for torque intervention but it might affect fuel trims.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on April 01, 2020, 12:07:38 PM
Stock maf.

Pid work and zwgru raise fixed it. 

A steady 18lbs with timing the clutch slipped between 2000 and 3200 rpm in high gears

Going for 1 bar now.


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on April 01, 2020, 01:07:27 PM
A steady 18lbs with timing the clutch slipped between 2000 and 3200 rpm in high gears

careful with this. once you get this happening, there is no going back. Things are going to get exponentially worse...


Title: Re: 1.8t/20V KnockKnock
Post by: Auriaka on April 01, 2020, 03:48:25 PM
Must has a valid point

I originally had clutch slip just at last 1000 Rpms. My response was to taper the boost out to bring down the power. Still slipped, was noce to it and kept out of the throttle to hard and by the end of 2 weeks I couldn’t get on it past 40% pedal before it would just free rev on gate pressure only (8psi)

Once clutch slips, it does not come back.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on April 02, 2020, 12:11:12 PM
Glazed clutch can recover.


With this issue,
Midway through a casual drive, I hit the clutch pedal. It shifted well but there was
A noticeable difference in pedal pressure. Valeo SMF kit in it is just over a year old.
All functions are fine besides pedal difference. (Besides the slip which may be from too much TQ for the clutch kit)


I'm unsure whether clutch kits wear in after so many miles and have a change of feel or if they should feel consistent from install until broken/worn.

The kit is rated at 300ft lbs and between 2000-3000rpm it is well near there at 18lbs.

If the clutch pedal should be consistent then there is likely something wrong with my pressure plate.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on April 07, 2020, 08:07:45 PM
1 bar is good!
Going to clean up the pid controller, get it dialed in and post a log.

Crazy how the Unitronic tune would hit 26 psi but 18lbs with a proper tune makes more power.
Proper is a given to more power but still suprising.

I'd like to have more  of that low end torque between 3000-50000rpm. But I'd rather steady boost instead of ramping up and down.


Nyet, for egts on a gasoline. What is max temps I should shoot for one k03s?



Title: Re: 1.8t/20V KnockKnock
Post by: nyet on April 07, 2020, 09:14:40 PM
900C

PRJ would probably say 850 though :)


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on April 30, 2020, 10:29:06 AM
Vehicle has a 3inch turbo back exhaust and No egt Gauge.

How accurate is tagbts_w? Does the Exhaust affect its accuracy?

If I aim to keep this under 900c during a fourth or fifth gear pull.

Is this overkill or desired?

I'll check FR as well.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 15, 2020, 11:46:16 PM
What other maps share RPM axis with KFLDHBN?
I'd like 3520 or 5000 to be 4500.

What Injector duty Cycle do you use as a maximum? 85?




Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 16, 2020, 11:06:56 AM
With Methanol injection. Have you experienced any damage to aluminum parts?
Worried about interactions with the 1.8t head.

Is methanols corrosive proprieties milder in vaporized form?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 16, 2020, 11:16:25 AM
Quote
How accurate is tagbts_w? Does the Exhaust affect its accuracy?

If I aim to keep this under 900c during a fourth or fifth gear pull.

Is this overkill or desired?     



Title: Re: 1.8t/20V KnockKnock
Post by: nyet on May 16, 2020, 04:32:30 PM
Why on earth would you post a screen shot? It can't be indexed by search engines, so any words in the picture are irrelevant.

Not only that, but if the page changes, the contents of the photo are out of date.

What you meant to do was this:

https://s4wiki.com/wiki/Exhaust_gas_temperature

Quote
VAG-COM blocks 034 and whatever else only display an internally calculated O2 sensor temperature (tabgbts), which is what most people think are real EGTs. That value may or may not have anything to do with the real EGTs or temperature at the primary O2 sensors; in fact, those temperature values are stored in 2 maps where the inputs are load and RPM.

Block 112 is the only block where a real EGT value (tats) is displayed, and only when the actual EGTs exceed 945°C. This is only due to the design of the sensors and their electronic modules, which makes that they can only "read" 945-1035°C and that's all. So basically if all you see is 945 all the time, it means your actual EGTs never exceed 945.

If you see block 034 EGTs hitting 750 or 800 or 850, those values have never been, are not and will never be actual EGTs. The only real EGTs are in block 112. If you want to know what they are when they're below 945, then you have to use RS6 sensors with adequate programming, so you get a range of -40 to 1100°C and then you will know what your real EGTs are.

Screen shots of text are worse than useless, IMO


Title: Re: 1.8t/20V KnockKnock
Post by: prj on May 17, 2020, 11:18:51 AM
950C max pre-turbo, 850C max post-turbo...

And you need an actual EGT sensor. The calculated stuff is totally wrong once you change anything.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 17, 2020, 06:18:45 PM
Alright, no more screenshots of text.

Would be a small resolution anyways, but Tats is not available. Displays constant 1110.
Yes, I do need an EGT gauge.

Quote
The calculated stuff is totally wrong once you change anything.
With the improved exhaust flow, would calculated be higher than expected?
Just to clarify anything, does that include parameters?


Looking for CWMDAPP to work with KFLDRL.
Compared against HN. Got 181BC but no conversion.
Does CMWDAPP max out at 8? If so , would it be 0.031373*X.

P/N:06A906032RN

 


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 17, 2020, 10:00:55 PM
Set CWMDAPP to Hex 08 and it did not limit WG DC to
Set KFLDRAPP of 10.

To Set CWMDAPP for KFLDRL purposes.
Should it be FF?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 19, 2020, 11:12:32 AM
Checked FR and searched threads.
What I found was mention of setting Hex 08 or 01.

Did that but it didint follow set KFLDRAPP.
Hmmm, 181BC may be the wrong address.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 19, 2020, 04:49:00 PM
Zeroed out LDIATA but Pid I Max (ldimx_w) is higher than what I've set.
E.g KFLDIMX is set to 70 but ldimx_w ( PID-I-Max) displays 76.
IMax corrections are negative, -0.15. Odd, I'm definitely missing something.

Haven't tried To FF CWMDAPP,  do not think that is the answer but am confused why 08 hex set didint work.


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 19, 2020, 10:19:28 PM
Ahhhhhh easy to check, should've looked harder at FR. Will do for reassurance.

Thought PID-I-Max follows KFLDIMX directly. Makes more sense that It would
Follow KFLDRL output from KFLDIMX input. Derp, alright.

Still muddled about CWMDAPP


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on May 21, 2020, 11:33:57 AM
Thought PID-I-Max follows KFLDIMX directly. Makes more sense that It would
Follow KFLDRL output from KFLDIMX input. Derp, alright.

I actually think it doesn't make any damn sense since imx should cap pre-DRL wgdc but who knows?


Title: Re: 1.8t/20V KnockKnock
Post by: DonkeyDub on May 21, 2020, 05:12:18 PM
Will look into it further but it does appear that

ldimx_w = KFLDRL output by KFLDIMX input (capped by TVLDMX?).

Correlates with logs.

Me7.1FR:868

Ignoring Corrections.
KFLDIMX and TVLDMX input a Min/Max table? Outputting
ldimx_w (PID-I-Max)


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on May 21, 2020, 05:20:24 PM
Will look into it further but it does appear that

ldimx_w = KFLDRL output by KFLDIMX input (capped by TVLDMX?).

Correlates with logs.

FR:868

Ignoring Corrections.
KFLDIMX and TVLDMX input a Min/Max table? Outputting
ldimx_w (PID-I-Max)

I haven't looked but how does the cap feed back into pre-lin wgdc then?


Title: Re: 1.8t/20V KnockKnock
Post by: nyet on May 21, 2020, 05:23:00 PM
I dont see this at all for ME7.1 anyway

Maybe it is different in ME7.5.

LDRL is only used to map pre-lin (ldtvr_w) wgdc to post lin wgdc (ldtv)