[DWR]

First, let me say this is an excellent post by TDIwyse. It is a complex subject, not easily covered in a forum such as this. Disregarding wear and poor lubrication - in performance applications, turbos fail in overspeed in which blades or hubs burst OR excessive drive forces that break the shaft. For the HP turbo shaft breakage is most likely. The LP has a robust shaft and is more likely to burst.

Originally Posted by TDIwyse

There's been some compressor map's posted earlier in this thread, but here's (I think) the compressor map for our "big" turbo.

Yep, my compressor maps agree with yours.

Any Pressure Ratio (basically absolute boosted pressure at the turbo outlet divided by the absolute pressure at the inlet of the turbo) over ~ 3300 mbar at low flow rates is off the map. And this is at the turbo, so any pressure drop across the air filter and pressure drop across the intercooler/charge pipe/intake manifold makes this pressure ratio increase... So if we measure ~3300 mbar at the MAP sensor we've definitely gone off the map.

Note the gradient of compressor speed moving up in pressure ratio and then across with volume flow. You can see volume is where we will get in trouble faster. In turbo lingo, this is called the 'choke line'. Rotational speed continues to climb, but additional flow is increasingly difficult to achieve. The work of the turbine goes into heating the air.

If you go off the map, the compressor "drive pressure" or "back pressure" to boost pressure usually increases dramatically, as does IAT's, as the turbine has to be driven harder and harder to make boost because it's going outside of its capability and operating in an increasingly inefficient manner.

Drive pressure is the difference in pressure across the turbine. Exhaust back pressure after the turbine can both increases the pressure in front of the turbine and reduce the drive pressure. From a causual viewpoint, we tend to think about turbochargers backwards, focusing on the compressor. It all starts with the turbine and generating enough drive for the compressor. The turbine in the LP turbo can outpace the compressor, hence we need a wastegate.

I could go on but I don't want to put everyone into a snooze.

[TDIwyse]

Originally Posted by 35d

What is the drive ratiobin testo called stat exh ptrbnus?

I followed DWR's tutorial to find these values and insert them into the custom job.

http://www.e90post.com/forums/showpo...&postcount=140

[TDIwyse]

Originally Posted by DWR

...
Drive pressure is the difference in pressure across the turbine. Exhaust back pressure after the turbine can both increases the pressure in front of the turbine and reduce the drive pressure. From a causual viewpoint, we tend to think about turbochargers backwards, focusing on the compressor. It all starts with the turbine and generating enough drive for the compressor.
...

Thanks for clarifying the nomenclature.

Over the last couple months, after iaknown helped me see that I was actually overboosting the turbo before the wastegate modification, I've been looking for examples of what back pressure to boost pressure ratio's start to become concerning. I've seen everything from 1.5:1 to 2:1 from various places. I'd expect it likely depends on the specific turbo. What's your experience/thoughts on this?

[DWR]

Originally Posted by TDIwyse

I've seen everything from 1.5:1 to 2:1 from various places. I'd expect it likely depends on the specific turbo. What's your experience/thoughts on this?

Oh boy, where to start. This is a case where I think a picture is with a thousand words. This is output from a simulation model (yes it is a mild 335d) that I think illustrates why the question and answer are not straight forward. It varies throughout the system's operating range. My worry in discussing this is many times these pressures are not measured correctly, throwing the whole discussion into a tailspin. The simulation is with an open exhaust, so there is NO exhaust back pressure. It is pure drive pressure across the turbine. And it is just a single LP turbo.

[iaknown]

Originally Posted by 35d

What is the drive ratiobin testo called stat exh ptrbnus?

Iaknown I got a 4th gear pull logged
The map being used at the time was my remap with 3100 boost and a experiment in the n75 maps to try reduce the overboost. Low rpm it seems lower the boost then before but higher it still shoots dangerously high. Not sure if the dips at high rpm is the map or something others experinced while overboost too

That dip appears to be when the wastegate is 1st opening. But you mentioned an experiment in the maps so it appears to be opening too fast. What do you consider dangerously high? I see about 3200mbar, it's on the border but that's a good max. Have you seen higher boost than this at other times?

You have the spike when the turbos are switching but I am not sure that is actually bad for the turbos. Maybe DWR and tdi can chime in on that.

[DWR]

35d

I'm going to ask a couple of questions that may seem odd -
1) What is the IAT?
2) What is the AFR at these points?
3) What is the exhaust temps?

That last graph doesn't look too bad. Certainly mid range rpm is OK because the turbos are compounding. Last spike might be OK if I could see it better and could figure out the airflow. It takes two coordinates to plot it on the compressor map, pressure ratio and air flow.

[DWR]

Originally Posted by iaknown

That dip appears to be when the wastegate is 1st opening. But you mentioned an experiment in the maps so it appears to be opening too fast.

iaknown, you're getting good at this. Must be months of banging your head, huh?

[pheno]

Originally Posted by 35d

Hi so would that be for a test like dis? Or could it be set in the map too. So if you getting overboost low down the Wastagte actuator can open the wastgate say about 10% to bring boost to were it should be?

Yes, you can test various actuators with this but I am quite sure this can not be set in the map at all. I believe this is made just for checking that various actuator work and operate as they should. Modifying map requires different kind of cables and downloading the software from ecu, modifying it and flashing back to ecu.

[pheno]

Is anybody using this ?

http://www.ebay.com/itm/BMW-ENET-Eth...-/130731306086

Ok.. it's not as cheap as cheapest OBD2 cables but.. It looks very appealing since you don't have to play with serial port adapters anymore eliminating potential configuration hassle and loss of speed. Also Ediabas uses ENET "driver" which must be faster and lighter on resourses than any of serial port adapters that rely on USB hardware. Guys, I suggest you buy this and try it out :-)

EDIT: it looks like only F-series support this ENET

[pheno]

Originally Posted by TDIwyse

Well, I've started imbibing ... so it'll have to wait. Can you describe more on what this is supposed to do?

Also, is it possible my underpowered Win 8 tablet is causing some of the slowness? I'm seeing ~0.2msec time steps in the sampling rate. What type of resolution would you expect to see?

The job shows what protocol Ediabas is using to communicate with ecu. If the diagnoseprotokoll_lesen shows 'DS2' for the protocol that is the "slower" protocol using 9600baud rate. 'BMW Fast' is much faster using 115200 baud rate. I don't know if there's some logic in the .prg to find out what's suitable.. for example I believe when running a job Ediabas negotiates best protocol with the ecu.

what tablet are you using by the way ? I was thinking to get Linx7 Win8.1 32bit el cheapo tablet and try how it works. It requires also USB OTG cable because of the USB HW is designed. It could be Win8 is slowing down but I'm sure it couldn't be because of the CPU speed, more likely because of the USB hardware. Actually I could rule that out by making the graph's update speed configurable. At the moment it is updating curve immediately when it receives a job but it might be more reasonable to make a config that tells what update (FPS) speed is wanted. I've tested TestO with 1.4Ghz old laptop and it showed some 10-20% CPU usage.

Anyway if there's a lot of jitter in job's execution speed (bottom right in graph window) it is a sign of a botlleneck somewhere. Maybe Ediabas falls down to using DS2 if it finds 115200 is not working as expected. That's the best theory I can think of... I believe you are using FTDI chip based K+DCAN adapter ?

here's a snippet of 'Debug trace' from iaknown he sent months ago. You can generate the same file from menu Debug->Trace. Select your ecu from the list and press 'run trace'. When done press send (needs internet connection). It runs a set of jobs (mostly STATUS* and *LESEN) and saves them. From the log we can see the sampling speed. Running timestamp in milliseconds is in brackets so here we can see that running one job takes 12750-12719=31ms. That's the speed I am expecting. Ok, this is raw reading speed from the serial port itself - it could be ediabas is adding some milliseconds to the job because it does calculations, etc.



'16.12.2014 21:17:55 <jobstart>'STATUS_LADEDRUCK_IST','
_STD [ 12704] SENDECUTEL
#@1376;P00000A94h;#
_STD [ 12719] I -> D: 14 06 00 00 00 00
#@1378;P00000A94h;#
_STD [ 12719] D -> I: 01 03 00
#@1379;P00000A94h;#
_STD [ 12719] I -> D: 06 0A 00 84 12 F1 2C 10 07 6D
#@1380;P00000A94h;#
_STD [ 12750] D -> I: 01 0B 00 84 F1 12 6C 10 2B 2D 5B
#@1381;P00000A94h;#
_STD [ 12750] ERROR: NO ERROR
#@1382;P00000A94h;#
16.12.2014 21:17:55 'OBJECT','d73n57c0','text'
16.12.2014 21:17:55 'SAETZE','1','word'
16.12.2014 21:17:55 'JOBNAME','STATUS_LADEDRUCK_IST','text'
16.12.2014 21:17:55 'VARIANTE','D73N57C0','text'
16.12.2014 21:17:55 'JOBSTATUS','','text'
16.12.2014 21:17:55 'UBATTCURRENT','-1','integer'
16.12.2014 21:17:55 'UBATTHISTORY','-1','integer'
16.12.2014 21:17:55 'IGNITIONCURRENT','-1','integer'
16.12.2014 21:17:55 'IGNITIONHISTORY','-1','integer'
16.12.2014 21:17:55 'JOB_STATUS','OKAY','text'
16.12.2014 21:17:55 '_TEL_ANTWORT','OKAY','binary'
16.12.2014 21:17:55 '_TEL_AUFTRAG','OKAY','binary'
16.12.2014 21:17:55 'STAT_ID_SG_ADR','18','long'
16.12.2014 21:17:55 'STAT_LADEDRUCK_IST_EINH','hPa','text'
16.12.2014 21:17:55 'STAT_LADEDRUCK_IST_WERT','1.011946E+003','real'
16.12.2014 21:17:55 <jobend>

[35d]

Originally Posted by iaknown

That dip appears to be when the wastegate is 1st opening. But you mentioned an experiment in the maps so it appears to be opening too fast. What do you consider dangerously high? I see about 3200mbar, it's on the border but that's a good max. Have you seen higher boost than this at other times?

You have the spike when the turbos are switching but I am not sure that is actually bad for the turbos. Maybe DWR and tdi can chime in on that.

Iaknown I have seen various actual charge air pressure at top end/ high boost

Stock map - 3300 mbar max
Remap& boost limiter 3200 - 3600 max
Remap& boost limiter 3000 - 3500 max
Constant open Wastagte 3300 max

But I've noticed the last few logs which I've done and posted here they are all like 1-4 gear. When you log higher speeds say 5-6 gear boost on remap is 3500+ and hits limp mode, on stock map 5-6 gear change hits 3300 mbar so no limp mode

I was thinking 3000mbar is safe limit? When remap was done boost was raised to 2.0 from 1.95bar. But after issues tunner raised to 2.1

[35d]

Originally Posted by DWR

35d

I'm going to ask a couple of questions that may seem odd -
1) What is the IAT?
2) What is the AFR at these points?
3) What is the exhaust temps?

That last graph doesn't look too bad. Certainly mid range rpm is OK because the turbos are compounding. Last spike might be OK if I could see it better and could figure out the airflow. It takes two coordinates to plot it on the compressor map, pressure ratio and air flow.

Dwr sorry I did not log those parameters but may have previously in some logs
Here a graph for what I belive is mass air flow again this is a wot run then slow down again wot and so on not one pull

[DWR]

Originally Posted by 35d

Constant open Wastagte 3300 max

Is this with remap?

[TDIwyse]

Originally Posted by pheno

TDIwyse, could you run this job in the picture and post results ? I'm still wondering about your sampling rate which I think that should be much faster.

Here's what it says when I run that job.

Also, this is on a lenovo lynx Win 8 machine with an Intel Clover View 1.8GHz processor. The obd adapter uses a usb interface. The tablet only has 1 micro usb port, so there's also a micro to standard usb adapter in the mix.

[35d]

Originally Posted by DWR

Is this with remap?

Yes with remap.
Without remap and open wastegate I have not checked.

[35d]

This I think is logging airflow

[iaknown]

Originally Posted by 35d

Yes with remap.
Without remap and open wastegate I have not checked.

Your wastegate direct to vacuum sounds fine. You are right on the border but in the safe zone, 3300mbar is the max you want to see in high gear. It is the same max that our tuner prefers and his tune has been proven to be reliable. I forget your altitude (and barometric pressure) but have your tuner raise the setpoint to 2.3bar, tighten up the adjustment on your wastegate a little perhaps, and enjoy your car!

[35d]

Originally Posted by iaknown

Your wastegate direct to vacuum sounds fine. You are right on the border but in the safe zone, 3300mbar is the max you want to see in high gear. It is the same max that our tuner prefers and his tune has been proven to be reliable. I forget your altitude (and barometric pressure) but have your tuner raise the setpoint to 2.3bar, tighten up the adjustment on your wastegate a little perhaps, and enjoy your car!

Tune is set to 3100mbar. The limiter I think is standard at -/+ 0.4 bar.
With constant vacum I reach 3300mbar without 3500mbar so I can't run always with constant vacum?

[DWR]

I agree with iaknown, in part. The actuation rod is obviously not opening the wastegate all the way. I'm not sure it is an adjustment though. It sounds to me like the Electro-pneumatic Pressure Converter (EPDW) is not closing all the way or it is not getting full PWM duty or voltage. When it is bypassed and full vacuum is applied, things work better. My suggestion is to hook up a vacuum gauge, tee'd into the line after the EPDW and compare it to full vacuum... Hmm, I just realized I don't know what the results of this test are in a working vehicle. If I get a chance tomorrow I'll check it out. I would suspect full vacuum at high rpm. If not, then I agree adjust the rod!

[TDIwyse]

Question ... if one is hitting 3300 mbar at max rpm, how are you not completely off the compressor map and in extremely unfavorable operation regions for the turbo?

Using the Not2Fast calculator it would seem at 3300 mbar and 4500 rpm one is flowing ~44 lb/min. Even if you have 0 pressure drop through the intercooler and intake, and 0 pressure drop across your air filter/intake ... you'd be way off the map. Or am I calculating something wrong?

(I put the 3300 mbar/4500 rpm data point on the map below. If you add in the pressure drop issues, it would be even farther off the map... if I'm calculating the flow correctly).

[35d]

Originally Posted by DWR

I agree with iaknown, in part. The actuation rod is obviously not opening the wastegate all the way. I'm not sure it is an adjustment though. It sounds to me like the Electro-pneumatic Pressure Converter (EPDW) is not closing all the way or it is not getting full PWM duty or voltage. When it is bypassed and full vacuum is applied, things work better. My suggestion is to hook up a vacuum gauge, tee'd into the line after the EPDW and compare it to full vacuum... Hmm, I just realized I don't know what the results of this test are in a working vehicle. If I get a chance tomorrow I'll check it out. I would suspect full vacuum at high rpm. If not, then I agree adjust the rod!

I can try that so you are saying out the vacum gauge is between the pressure converter and the actuator and when the pressure converter is activated I should see 500mg it not the pressure converter not working?
I'm not sure if it is not working correct, according to Bmw they say the travel for wastegate rod is 5mm?

Recently the the other pressure converter failed and it just died and I got a error code.

I don't understand why my issue is diffrent to iaknown, tdi wise etc, ok I don't have open exhaust but dpf gutted surely is decreasing back pressure

[pheno]

Originally Posted by TDIwyse

Here's what it says when I run that job.

Also, this is on a lenovo lynx Win 8 machine with an Intel Clover View 1.8GHz processor. The obd adapter uses a usb interface. The tablet only has 1 micro usb port, so there's also a micro to standard usb adapter in the mix.

'BMW Fast' is your protocol so you're all set ! Also your tablet should be absolutely fast enough.