NefMoto

Hey guys,

I've been running this tune for a while and it seems to be running pretty well. I've just had a suspicion that it is not quite perfect. When driving next to a wall or something I can hear the engine better, and during that time I hear the turbos spool up on hard accel. When the boost gets up towards its max requested, I can hear a sort of choppy kinda whistle, and I think it's the DVs not quite holding the max boost. Car is 00 A6 2.7t. DVs are adjustable Forge Hybrid Splitters running in re-circulation mode. I can't find the max boost spec ANYWHERE for these valves. I have done a pressurized intake boost leak test and it seemed to pass that. Unfortunately, it's harder to test the actual DVs separate for the max boost they will hold. Below is a graph showing a 2nd gear pull. I am running essentially the Notorious Tune, with very few changes. No changes to boost, fueling, etc.

Thanks!

Anybody have any ideas? Need to see any more logs?

That is not a log shane, it's a picture of boost which means next to nothing.

lol.

Post the full CSV of a 3rd gear 2500-redline log, along with your mods.

Sorry, didn't know what info you wanted to see. Here you go.

WGDC is 95% all the way to redline.

I'd say you have a boost leak or your DV's aren't holding, like you said.

Any particular reason you didn't include WGDC in your picture? :P

Also, when logging, please hold your accel pedal at WOT, or ecuxplot will filter that data if you have filter enabled.

Thanks. No particular reason...still learning. What does a normal/good duty cycle look like?

Depends entirely on the application... you will probably want to

1) read some materials on turbo/efi tuning
2) log a lot of cars
3) both of the above

ok...I mean, do you rarely see a healthy 2.7t with a WGDC above 50% for example?

When turbos are spooling WGDC should be high. When they are spooled it should decrease. This is all dependent on how hard they're working to meet requested boost.

Thanks,
makes sense. Do you have a graph showing boost and WGDC from your car that I could see?

(http://i41.tinypic.com/9rndkh.png)

Perfect. Thank-you. Picture definitely lets me know what to look for.

Depends on the WG and boost levels.

stock k04, stock wg will need 80%+ WGDC to maintain > 20psi at redline

k04 knockoffs with tight wastegates will do 23-24psi with 75%

another example - k04 knockoffs with somewhat tight wastegates.

Ok, thank you. So I just did another boost leak test, and it's coming up clean. I replaced a few vacuum lines, although none of them appeared to have any cracks in them. I will be investigating the DVs in the next few days. But, assuming that I find no issues with them, could there be an issue with the waste gate fully closing?

Also, I have new valve cover gaskets, and cam tensioner gaskets to put on (just been putting it off). The are leaking pretty bad. Can that have a negative result on boost being produced?

No. If the plug wells fill with oil you'll get misfires, but that'll be very obvious. Do all necessary maintenance, and log the car.

Okay, how about N75 valve? Could that be going bad and causing the WGDC to seem so high, when it's not causing the wastegate to close enough?

Disconnect the line from N75 to WG, and do a log. CAREFULLY! YOU MAY OVER BOOST.

If you want to be 100% safe, just try a spare N75.

Okay...I'll swap one from my other 2.7

Also, you can remove it, cap the long end, and blow into one of the short legs.

You should NOT be able to blow any air through the disconnected N75

(http://s4wiki.com/images/d/d1/N75.jpg)

Ideally, you should know EXACTLY why this is the case before proceeding anyway :)

ETA: oops scratch that. you'll need a battery or VCDS output test to test that :)

So. You shouldn't be able to blow any air through it when it's energized then?

yea.

if you energize it with a little battery (like a 9v) DO NOT LEAVE IT ON for a long time. you'll burn it out.

just long enough to close the valve...

do not energize with a full size battery.

Do you know the VCDS procedure to accomplish the same?

output test in Engine...

Great! A bad N75...well, at least I think I found the problem. I took it off and tested the waste-gate control lines for vacuum and they passed. Then I removed the solenoid, and blew through it, then blew through it while applying voltage...

The valve seems to be working fine, there was a click and the amount of air allowed to pass lessened a bit, but it still let a bit through.

These valves are bad if they let ANY air through right?

Yep. Any leakage is bad.

I hope I don't have this backwards but if the WGDC is at 75% the actual waste gate valve should be open roughly 3/4 of the way?
I have a knockoff k04 with what seems to be a 10 - 12 lb+ spring tension on the WG. I haven't checked it with a gauge but it was way harder to open by hand than the k03 with a 5lb spring setting. I thought about lessoning the tension but the manufacture put that green colored marking compound on the threads to see if the tension was changed from their factory setting. I'm still breaking in the clutch but I get mad overboost with barely any throttle. I mean needle bouncing off the 25psi mark on the boost gauge and it only goes to 25psi...kinda overboost at 2700+rpm. I don't have any logs yet as I am breaking in the clutch and dialing some ev14 550's using 2.7bar on an ECS tuning adjustable FPR and Geneses fuel pump by looking at the ST & LT fuel trims. Just been looking at them live after like 20miles of driving with VCDSPro.

Tighter wastegates=recalibrate the PID.

No. First, you have it backwards. The higher the WGDC, the more the WG is held closed.

Second, the "amount open" has as much to do with manifold pressure as WGDC...


You shouldn't have done any real driving w/o at least doing a little bit of WG tuning.

Okay, I've done a little more investigating and actually opened up the N75. It does not operate as previously explained. When the valve is not energized, it routes ALL boost pressure into the Wastegate actuators causing them to fully open. When the valve is energized, it opens a plunger inside that allows air to pass into the PCV spider hose system AND into the Wastegate actuators simultaneously. So, when testing...when the valve is energized and you blow into the bottom port, you will find that air is normally allowed to come out both other ports if the valve is good.

So, after I found that, (no real problems with the N75), I went and disconnected the tube to the Wastegate actuators completely, and cautiously did a short run to verify that boost would be there if Wastegates remained fully closed, and sure enough it was building much more boost, much more quickly. So, this doesn't make any sense.

What I'm thinking now is either the PCV valve is not working correctly and not properly venting the extra pressure from the crankcase and N75, or for some reason the crankcase pressure is too overbearing and when the N75 opens "to atmosphere", it can't properly went the extra boost because of pressure equalization.

So, I guess my next test will be to run it without the long port on the N75 connected.

Yep, this is why I said to cap the long port :)

BTW it has to open this way to relieve any pressure built up in the wg lines...

Yes, if you cap the long port, and energize it, air will still be allowed to come out the wastegate control port....

Anyway...if I end up having to change the PID tuning...would I begin by increasing KFLDIMX in the 850 and 1000 columns?

And, say I am running at approx 95% WGDC in the current tune. Is that referring to 95% of the physical capabilities of the valve, or 95% of the max duty cycle for this tune?

In other words, would changes to KFLDIMX accomplish anything if it is at 95% currently?

Thank-you

95% WGDC is N75 fully open, ECU requesting max boost

Ok, that's what is was thinking...if I disconnect the lines to the WG from N75, it will boost...what could be causing this then?

Disconnecting the lines from the N75 to the WG should cause uncontrollable overboost well in excess of the MAP.

What does "it will boost mean"?

It means I only got on it for about between 3500 and 4500 while logging once to see if I had a turbo or waste problem.

It built boost right up to 22 psi no problem, so I got what I was looking for. So, it's obviously not a massive boost leak.

No, it should not.

The N75 is a solenoid. 100% DC would burn it out. 95% means all manifold pressure is being diverted away from the wastegate, and is equivalent to disconnecting the wg line from the n75.

If you get different results (between 95% and wg line disconnected), your N75 is bad (or harness is bad).

Thank you for clarifying.
 As far as WG tuning, I pretty much combined elrey's kfldrl map to the first 3 columns of a stock Audi TT 225 kfldrl. and ran with it. But since ddilenger said that on the previous page about the heavier spring in the WG requiring the PID to be recalibrated then I guess it doesn't matter. It's just gonna overboost no matter what I guess. Maybe I'll run the map with like 30% DC across the board and see what happens before I delve into stuff I'm not too sure about yet(completely recalibrating the boost PID). Plus I'm still breaking in a clutch so I shouldn't be driving it hard anyway right now.

Yes it will. Have you disassembled one?

Here is a diagram of exactly how the valve section of the N75 works.

The plunger is held down by default by a small spring. When the solenoid is energized it pulls up and away from the tube that supplies the manifold pressure. As you can see, there is a direct link between the input and output to wastegates at all times. When you energize, it merely gives that pressure another route to travel.

Err. Sorry. You are correct. I assumed that only wg pressure could escape back into the intake/PCV. I was assuming you dont want manifold pressure to leak that way, but I guess it does.

.

Okay, so I believe my N75 is good and so why doesn't 95% duty cycle give me as much boost as without the valve at all?

WoW nice first post dude!! LOL

Perhaps your PCV is blocked? Try installing the N75, but vent the long leg to atmosphere?

because your N75 is still bad probably.
 You can test it with an ohm meter, if it is between 27 and 30 ohms it's good, otherwise throw it away.. this is per the Bentley diagnostic instructions for the N75s equipped on 1.8t and 1.9t Golfs and Jettas Yours may be a different part number(last digit) but I think the check is the same for all of them. If I were you I would maybe Google it or something. They are pulse width modulated and are set to have a certain amount of resistance for the ECU to "use" the N75 "properly".

Oh ya, you want to check it while its warm only.

I tried that, didn't really change anything. Today I also removed, checked and cleaned the PCV spider hose assembly while I was doing the cam tensioner and valve cover gaskets. Everything seems okay. The one thing that had me wondering was a green check valve immediately after the f-hose. It has an arrow on it (i believe to indicate airflow) pointing towards the f-hose, meaning that it should only allow flow towards the f-hose? (which would contain boost) But, when I blew on it, it allowed air to flow in the reverse direction of the arrow, and when I pulled vacuum on it, it would snap shut, not allowing flow. Doesn't this seem backwards? Are the arrows on these green valves supposed to indicate airflow direction?

I have't logged again after cleaning the PCV system and doing valve cover gaskets, but I don't think it's quite right yet.

Besides that check valve, the only other thing I'm thinking to try is bypassing N249?

I checked the original one from the car when I removed it. It was pretty hot/warm when I did, and it read 35 ohms. I read somewhere else that was at the top end of the normal range for those valves. 25-35 ohms I believe. I swapped it with another valve from my other 2.7t and it's still running the same (although I didn't check the resistance on that one yet, so it could be "bad" as well."

Good luck! I hope you get it figured out soon.

Thanks, me too!

Update...I found a bad green check valve just off the f-hose that must have been letting some boost go back into the low pressure side, so I replaced that with a good one. Still not getting full boost.
I put clamps on the vacuum/boost hoses that control the DVs, as I have not previously had them clamped. Thought that might be a source of a problem. At the same time, I replaced N75 with a used one that I got for a good price. Seems in better physical condition than the one I had currently. Resistance checked out ok, valve functionality checked out ok. After that I was running had to go right to work, so I didn't have a chance to log it yet. But, on my way to work, I got on it a bit, and sometimes on WOT it will stumble very bad, and the ABS and ESP lights come on simultaneously, but only for a few seconds, then they shut off. No wheel slip though, as this was in the higher gears.

Like I said I haven't logged it yet, and will tonight. Just curious if anyone knows what the ESP and ABS light coming on for a few seconds means?

Log your MAF signal.

Okay, will do. Does this indicate "limp mode"?

ABS/Check engine can be caused by a bad MAF signal. If you get logs, the problem will become apparent.

So, it literally won't do that anymore...which I guess is a good thing...

It must have had something going on with the fact that I was driving on the first start-up after changing the N75? I don't know...but I was doing MUCH more boost leak testing today and found some tiiinny leaks in a few places. I will log again.

The ecu was expecting the same "performance" from your changed out N75, once it realized something was different it recalibrated itself to use the new part. If you reset your adaption values on top of clearing all the DTCs( if there are any) you may get even better results than what you are currently experiencing.
About the small boost leaks, they add up to be a big boost leak so its good you fixed them even though the ECU can adapt to small "leaks" in boost/vacuum.

OP, are you still running K03's?

Yes.

Oh. In that case, your boost isn't surprising for those wgdc values.

Update...I at a point where I'm going to start repeating things that I've done, because I'm at a loss and can't figure out what's wrong.

I've tried 3 different N75 valves.
I just replaced the valve cover/cam tensioner/cam seals.
Just re checked and re regapped plugs.
Today, charged the system up to 20psi for boost leak test and followed from y-pipe all the way to the TB with soapy water spray. Found a few small leaks...like brake booster hose/map sensor/TBB joint. I believe I took care of all those.

Tune is essentially Corrado's V3 file.

I will post two logs that I just took. The first is a two - three gear pull, and you will see that my boost is peaking slightly higher than I was able to get before. I would get around 17 psi max before, I can get around 19 now sometimes, but it quickly dwindles down.  The boost always seems to peak higher in third gear...not sure why that is. You will also notice that something is happening around 6k or so where the WGDC goes to zero and I lose most of my boost. It will remain like this until I restart the engine.

The second log is just some city driving after the previous run, before restarting. Limp mode or something?


I really appreciate it guys! Please let me know if you can pinpoint what you think is the HW issue. I'll check anything you direct me to.

So, are you saying that's the max that k03's can produce?

Yep. Depending on dp/exhaust setup, it isn't unreasonable.

You're chasing a non-issue. I'm actually surprised you're holding 16+ to redline. Post LDRXN from your file.

Hrpmh i really should dig up and post a k03 compressor map at this point :/

meh...

Yeah exactly as I thought then.

What are you trying to achieve running 1.5 bar boost on your K03's?
Besides from trying to blow up your turbos I am not sure what you are trying to do.

And running that much boost to redline you are for sure exceeding 100% injector duty cycle, and eventually going to melt your engine with the high EGT.
You have no problems with your car, but you are running far too much boost. Your timing is also like 2 degrees in the midrange.

Case closed then...

Okay, I was under the impression that people were successfully running this tune with K03s. Thanks for clearing that up. I suppose I'll have to tune it down until I get some K04s. Curious though, was it limp mode that was causing the car to not boost after a run...and was it constant negative deviation that caused it to do that?

Also, if someone could just let me know basically what I would have to alter to lower max requested boost in this file, I would appreciate it. I'm on the wiki right now trying to figure it out.

Can I just alter KFMIRL and/or LDRXN to slightly lower maximum requested boost?

Well, I gave that a whirl. Could someone check this out and see if it would accomplish about 17 psi max?
And let me know if there's anything else I would have to change?

Thanks

Get a wideband and make sure your fueling is sufficient.

Why do you say that?

Because you're out of fuel with those boost levels. Do you want to melt things? I think not.

I'm understanding that, is corrado running k03s or k04s and stock injectors? Anyone know?

The calculated AFR was still looking on the rich side (if the calculated is even reliable), although I know it's really pushing it on the higher rpms.

Could you take a look at the file I posted above, to see if it should be safe to try and log, and if it looks like it would accomplish what I want?

This is a k03 file. It will not hold boost to redline, k03's must taper. What problems are you having? I'm not understanding your issues I guess.

Well, I thought that corrado's file was a k03 file...no?

It isn't, in the least. It isn't doing much more than back calculating what the ECU already does (MAF,RPM->injector on time)

I had a feeling...so you really need a wideband to know for sure at higher boost.

If you're serious about tuning, get a wideband.

That, and start reading. Everything you ever wanted to know (at this stage) has been covered.

Otherwise, run the file in blissful ignorance as so many have done in the past.

Okay, can you see if I made the changes to the tables in the file in post 71 correctly?

I'd say closer to 18, but it's an estimate until you log it.

K thanks, but, I'm on the right track with the changes?

Read the s4 tuning wiki. If you've read it, read it again.

http://s4wiki.com/wiki/Tuning

I have, and I am currently. I guess I just can't seem to find information on the technique used to figure out the values for the tables. Do I basically just try to keep the same kind of ratios/curves between the values as I increase/decrease them..then log?

Me7 doesn't have a boost map persay, it does everything based on specified load. Under WOT, boost will follow LDRXN. Now for an example:


Absolute pressure=10*(spec load)+300mbar.

LDRXN=195

10*195=1950
1950+300=2250
2250=absolute pressure in millibars.

To determine manifold pressure subtract the barometric pressure. Sealevel=1bar=1000mb.

2250-1000=1250mb. 1250mb=1.25bar

1bar=14.7psi
1.25bar*14.7psi=18.375psi manifold pressure.

Keep in mind, this is a ROUGH approximation. It is dependent on a large number of factors including air density, temperature, elevation, and the alignment of the stars and the moon. Use the above formula to get close, then adjust based on real world observations.

Got it. What's the difference between LDRXN_1_A and LDRXNZK?

LDRXNZK is maximum specified load under continuous knock. Leave it be.

Should I ever use the smooth feature on tables?

I feel as though you've gotten your hooks in me. Being the type of person I am I can't just ignore you as much as I'd like to. Come up with a complete list of questions, I will answer them.

As for the smooth feature, I don't use it. Go into grapical view (tunerpro, squiggly line above the map) and massage things that way until it looks pretty.

I really do appreciate it believe it or not. As for a list, I will try to come up with one instead of just rattling them off, it's just that they seem to come as I go. As for the file earlier, I didn't commit the changes I wanted correctly. So I will post this one that I did. I believe it's more where I want to be. Based on corrado's file, altered specified load, max load, and reversed the EGT delete to use them for enrichment when required. Can you just check this?  THANKS AGAIN

LDRXNZK is NEVER used in S4 M/L box.

B_kfzk is never set because CNOKT is clear and KFSWKFZK is all -96.

When posting logs, post raw CSV's and graphs (that make sense). Most of the time I'm on my phone and can't analyze a log, but can comment on graphed data.

Still thinking something is messed up. I altered the file for lower boost and it still can't keep of with 17 psi requested.

So I'm really thinking something is up with the way the wastegates are being controlled. The one time I disconnected the wastegate lines from N75 I was able to create 21 - 22 psi. With it connected and 95% duty cycle, my max always varies, but never goes near that. So, is the representation of 95% duty cycle the ultimate max physical capabilities of the valve, or is it the max when it comes to representing it in a current tune. Like, the tune's full range that it is setup to utilize. For example, if the boost PID were to be altered, would the ultimate output of the pwm signal at the valve be higher, but still represented as 95%? Do you get what I'm saying?

Sorry if I missed it -- are you at elevation, say 3,500' or above?

Did you try back to back runs with the wg line disconnected and connected?

Did you try unhooking the mani pressure line from the N75?

Did you try venting the atm line from the N75?

Try all of the above on the same day, same conditions, back to back runs.

My elevation is about 800 ft.

I didn't do a lot with the wastegate line disconnected. Just one run really, I was really scared I would break something. Won't the engine go incredibly lean when doing that? Is it safe to do a couple runs like that?

Wouldn't unhooknig the manfild pressure line essentially achieve the same thing?

I did try venting the long leg to atmosphere and it didn't change anything.

I will do separate logs of each scenery for you, if you think it will be safe.

That's definitely not normal.

It almost looks like it's switching over to steady state too early.

What's the rest of the log look like?

I will get some updated logs here very soon.

Start with a bone stock file, not someone else's. Also before you mess with things learn how the engine works and what the hardware limits are.
You have no idea what you are doing, and if you continue this way, you are going to crack your EM's, pieces of which will go through your turbos...

As a start, add ti_w to log and start looking at your fuel injector duty cycle, and you will perhaps understand what you are doing and why you get these kind of responses.