NefMoto

I have started this thread to split this issue out of my med9 tuning thread. I hope someone can help. I am having very inconsistent boost and I am not sure where to start looking. following is a wot throttle run in 3rd gear with the n75 in place but not connected to the wiring loom, from this it looks like my wastegate is working 100% and only cracking at 1.4bar and the boost stays linear. (stock file)

(http://i226.photobucket.com/albums/dd202/airtite/gti_n75_unplugged_zps53d2121c.png)

I then reconnected the n75 and this was the result. huge boost spike and then boost very lumpy.

(http://i226.photobucket.com/albums/dd202/airtite/gti_n75_changed_zpsf5e9b0d9.png)

I then set CWMDAPP = 8 set all cells in KFLDRAPP = 0 to try and start doing a pid calibration and my boost is all over the place, I am missing something surely the n75 should act the same at 0% duty cycle as it would if it was electrically unplugged? instead I am getting these results.

(http://i226.photobucket.com/albums/dd202/airtite/gti_wgdc_locked_0_run1_zps75ebc1c1.png)

(http://i226.photobucket.com/albums/dd202/airtite/gti_wgdc_locked_0_run2_zpsa0c4c439.png)

(http://i226.photobucket.com/albums/dd202/airtite/gti_wgdc_locked_0_run3_zpsee3eaa4e.png)

N75 shot?

Did you try swapping it for another one?

yup n75 changed, same result.

this from another forum

"I test myself and have found that with n75 unplugged the throttle plate closed. Set Cwmdapp to 1 and set KFLRAPP to 0%"

do you have it plumbed the right way?

both n75s were plumbed the same way on both turbos, if it was plumbed wrong surely it wouldnt have worked when disconnected electrically?

Yes, it would have worked if it was disconnected electrically, if you swapped pressure and wastegate lines.

pretty sure its routed correctly. bottom hose to compressor housing, middle hose to wastegate, top hose to turbo intake

(http://i226.photobucket.com/albums/dd202/airtite/IMG00357-20130228-0903_zps37c11b5a.jpg)

tried setting Cwmdapp to 1 but same result except worse I couldnt even get a decent log done because the pinging was so bad and I had to come of the throttle!

(http://i226.photobucket.com/albums/dd202/airtite/gti_wgdc_locked_0_new_run1_zps11636805.png)

bit
CWMDAPP
7 6 5 4 3 2 1 0
+---------------+
|x x x x x x x x|
+---------------+
ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ
| | | | | | | +--B_zwappl
| | | | | | +----B_fpwdkap
| | | | | +------B_wdksap
| | | | +--------*)
| | | +----------B_ldsafw
| | +------------B_plsolap
| +--------------
+----------------

B_zwappl = Zundwinkelapplikation torque condition without intervention
B_fpwdkap = DK control directly via the accelerator pedal
B_wdksap = DK control over fixed value, bit 1 has priority
B_ldsafw = LDR runs on control with fixed value LDRAPP
B_plsolap = Changeover: target boost pressure setpoint boost pressure on application

so am I not meant to have CWMDAPP set to 5? B_ldsafw = LDR runs on control with fixed value LDRAPP

B_ldsapp != B_ldsafw

Read the FR some more, there is a difference.
One setting uses a single value for N75 DC, the other one uses KFLDRAPP.

Oh, and of course not 5!
Either 8 or 16. This is binary representation!

*) B_ldsapp = B_zwappl or B_fpwdkap or B_wdksap or CWMDAPP.Bit3

IMO stop dicking with CWMDAPP. Put it back to stock.

Something about VCDS isn't logging the right WGDC with CWMDAPP enabled.

then where to from here, how do you suggest I sort this issue out................

anyone  ::)

so with the help of Rarak, i decided to do what I wanted to do initially and that was to copy across the PID maps from the ed30 GTI which is K04 in stock form and this has smoothed out the boost considerably. I forgot to enable turbo mode in vcds so the resolution isnt so high but you can see the difference. I will do more logging this evening to verify thats its better but the car feels smoother.

(http://i226.photobucket.com/albums/dd202/airtite/LOG-01-029-54C-1148-115C-2303_20130103_run1_zps526bbefb.png)

isn't that what you wanted to do initially??

exactly!!

In my experience, requesting far more boost (low revs) than the turbo can produce will sometimes mess up the boost control somewhat. Try to request sane levels of boost. You are requesting 1.1 bars of boost@2.100 rpms, but (if the log is somewhat close to normal behaviour) this is a boost level that isn't reached in reality until a thousand rpm's later.

Perhaps the high boost at low revs can be forced to appear during a static dyno pull with long duration before taking the measurement, but on the street it doesn't seem like it. I would try to adjust "requested" to better correlate with what is possible as it might improve the behavious even further.  :) 

http://s4wiki.com/wiki/Tuning#Positive_deviation

Not only that... Even if you disable boost deviation monitoring, the PID controller will be unneccesarily stressed when working way off target.

I and P are both maxed during spool anyway...

The only issue is I limit trims... which you can't fix by lowering req boost.

You have to move the RPM ranges around a bit for very laggy turbos so the trimming doesn't happen at low RPMs.

sure, but this was on the stock file and spool on the k04 shouldnt be far off the k03 so not sure if it was a boost request issue as much as what it was a boost control issue either way the car is much better now. I am however having a surging issue in the higher gears in the upper rpm under load especially uphill I will try and get some more logging in place now but at least I am moving forward.

the deviation code I was getting was positive deviation but not underboost it was an overboost condition mainly caused by actual exceeding requested in the upper rpm even though wgdc was low

Punctuation might help you.

better  ::) thanks for your valuable input

just to dig this up again, I wanted to make 100% sure that my wastegate actuator was correctly adjusted so I ran a vacuum line straight between the turbo outlet and the actuator and left the n75 electrically connected so the ecu would think it was still connected and this was the result, cracking pressure is 6.8psi and peddle position and throttle plate angle both at 100%.. so actuator adjustment is 100% and this is purely a PID issue, still need to work out why using the PID calibration method as described in here doesnt work but I am sure I will eventually figure it out.

(http://i226.photobucket.com/albums/dd202/airtite/LOG-01-021-54C-1148-115C-wastegatepressure_run1_zps8be20d29.png)

That doesn't explain this, unless VCDS is logging the wrong variable with CWMDAPP = 8 and KFLDRAPP = 0 AND you have the definition for CWMDAPP/KFLDRAPP wrong

(http://i226.photobucket.com/albums/dd202/airtite/gti_wgdc_locked_0_new_run1_zps11636805.png)

Or your N75 is bad, or it is plumbed wrong.

I will double check my definitions.


I double checked plumbing and its doing the same thing with 2 n75s both of them cant be bad.

None of the above, my definitions are correct. N75 swapped and triple checked plumbing. Anything else you can think of?

Did you ever manage to get this figured out?  I'm having the exact same issue on my B7 A4 with a S3 K04.  Only difference is my wastegate cracks at 10 psi vs your 7 psi.

since I couldnt tune the PID from scratch, I ended up just copying across most of the PID maps from the ed30 stock file which seems to have smoothed out boost control somewhat but its still not 100%