NefMoto

Edit: issues for dual bank resolved, see post:
http://nefariousmotorsports.com/forum/index.php/topic,722.msg6469.html#msg6469
______________________________________________________________________
ok, so here we go... I am at the point where I just need to ask for more help on this as I have not solved it with what I feel has been too much time spent in my life on it.

What I am trying to achieve is for my bank 1 sensor 1 to feed data to both banks in regards to fuel trims.  My car is a 2.7t single turbo conversion and i just run one o2 sensor (b1s1).

I have everything else deleted and I have found success in every other aspect of my tuning with the single on me7.1, yet this is a giant road block.  My tune has to stay fat because bank 2 does nothing as far as trims go.  when i log STFT or LTFT bank 2 is always at a consistent 0% and i really wish i could log individual injector on times because I am certain my bank 2 cylinders just dump fuel like when triggered by egts.  It also randomly TCs with no rhyme or reason.  Everything else is good so it must be this.

So now onto where i have been looking.  I started with this line of thought and hunting for ways to shut down the stereo control, from the Bosch documantation:

"FB LRAEB 4110 Functional Description
Notes stereo Lambda policy:
------------------------------------
- The following is FDEF f for a system with stereo lambda control designed.
- The stereo-Gro ¨ ¨ Anha must differ in lengths of 2 (eg B_lra -> B_lra2).
- The f FDEF is also fully suitable for a mono system. In this case, B_lr2
= FALSE. This also B_LRA2 = FALSE.
- For mono-Lambda control (SY_STERVK) = FALSE is set automatically B_lra2
= FALSE.
Description of the function LRAEB:
--------------------------------
- The function returns the switch-LRAEB of adaptations as well as the bit
"mixture adaptation locked" (B_gasp).
- Fu ¨ r Stereo are the switch-fu ¨ r separately, each bank (B_lra, B_lra2)."

and

"528 XXXXXXXX Flag register: DKVS Bank 1 Bits 16 FFh
SY_BDE = 0
Bit 0: = B_syserrk
Bit 1: = v E_frau E_rkat
Bit 2: = Z_frau Bit 3: = Z_rkat Bit 4: = B_lra Bit 5: Bit 6 = B_lrar =
B_rkat Bit 7: = B_frau
529 XXXXXXXX Flag register: DKVS Bank 2 Bits 16 FFh
SY_BDE = 0 & SY_STERVK = 1
Bit 0: = B_syserrk
Bit 1: = v E_frau2 E_rkat2
Bit 2: = Z_frau2
Bit 3: = Z_rkat2
Bit 4: = B_lra2
Bit 5: Bit 6 = B_lrar = B_rkat Bit 7: = B_frau"

So I thought that if I could find those addresses, based of the wording I could set B_lr2 and B_lra2 to false (FF, 00) to achieve a mono lambda system.  I ran this by some mainstream tuners who have said they have run single bank control on stuff like the R32 and they said that I am looking too far into it and that I can achieve this with variables within the map/scalars. 

So that leads me to my never ending revisions.  I have hunted through everything in the public files and tried everything that makes sense to me.  Also combinations.  I cannot bring life back to bank 2  :(

I am positive I am misunderstanding something or overlooking it.  Does anyone here have any idea how to achieve what I am trying to do?  Are there variables that are unlisted in the m-box xdf?  I was under the impression that the current definition has every available map or scalar that exists within the 1024 mb file.

whew... hopefully I explained myself.  if I can ever get this locked down I will post up my files for all to do whatever with.

Thanks in advance for any assistance anyone can offer.

Seems your setup is serious, so i cant understand why you need lambda sensor at all ? If you disconnect last O2 sensor from your car i think you will solve all your problems ;) Without O2 sensors you even can have more better fuel economy :)

with me7.1 dont I have to have some sort of o2 input to the ecu?  I run a wideband for tuning purposes yet nothing is so perfect that the ecu would never want a signal, or is it?

I was not aware no 02s were possible.  i mean doesnt it have to adapt to some extent no matter what?

I could just turn of 02 control, that seems kinda scary though considering I have no egts either.

Heh. Yea. Thousands :/

okay, lol... wishful thinking i guess and i should have known better.

 
Yes for stock car with two or four cats you need O2 sensors, but for extremly tuned car you can remove them without problem

Hmmm how and why you link O2 control and EGTs ?

Unfortunately I have nothing to add to the discussion

But I am watching closely, as my single project should be ready for tuning in the next 3-4 weeks

My worry is that without any O2's, do you become a slave to your environment, I mean if you tune at a given altitude, does me7 adapt fast enough without O2 voltages to travel through mountain ranges

I doubt it will adapt at all without the O2 sensors but you are more than welcome to try it...

No thanks, not looking to sacrifice a built motor for the cause

If you want a single O2 sensor you may have to use 24v N/A SW (that's if they don't have 2 oxygen sensors as well  ;D)

Maybe a patch in the executable code may help you on this but it seems like a lot of work.

Thanks, it really does seem like a bunch of work for minimal payback.  I may have to stick with a Haltech passthrough system in order to retain the safety margin I am looking for

If any of you guys would be willing to explain the reason to go with single over dual O2 sensors I'd appreciate it. I'm at a loss as to why you would want to do that at all.

Thanks.

Because with a single turbo there's only one downpipe?

What about using a loop splitter to split the output of the one sensor into two isolated signals?

I think the two isolated signals would cause the same issues as two primary o2s in the same pipe.

kelesha, i attribute the loss of the last o2s as the last true way for the ecu to even guess at egt and dump fuel in a panic situation.  am I incorrect in this?

Do you have any way of comparing Bank1 vs Bank2 injection?

unless i have missed the variables, i cannot log bank 1 vs bank 2 injectors.  I have ecux and vcds as far as vag specific  logging goes.

Im not familiar with ECUx,

See if you can monitor these 2 ram addresses, at the same time:

0x382366
0x382364

They are both 16-bit values.

I think I can make your file run all 6 injectors as one bank, taking lambda control from B1 lambda sensors.

thanks, I will see what i can make happen.  I may not be able to run the car for a day or two as i have a bunch of work interference.  thanks for the lead and offer to look into it!

I have been meaning to update this but I wanted to drive the car for a length of time and ensure it works under all conditions.  All I did was eliminate my last factory sensor and now I run just the 5 wire VW wideband sensor.  Through the Innovate LC-1 controller I use both analog outputs to simulate both primary sensors, feeding straight into both factory primary plugs.  I have a voltage offset of .001v between the sensors so that the ecu always thinks there are truly two different sensors actively reading the two different banks.

I turned off the interchanged sensor diag functions via:
CDLSHV - from 1 to 0
CDLSVV - from 0 to 1

I also turned off the cat heater diag functions as the heater power and ground go to no where in my wiring now:
CDHSV - from 1 to 0
CDHSVE - from 1 to 0

Thus far everything has seemed to be good, I have active and stable readings on STFT all coming from one cheap and easy to replace sensor that does a much better job at actually reading AFR and that I was using anyway.  Through testing I feel confident this is a good work around for people needing to control both banks through one downpipe/turbo or even just want to put a singe sensor at a y-merge from their two downpipes to get a simpler system.

The banks are split as far as injection goes, thus, you'll need a lambda sensor for each bank.  This is why many people go with a stand-alone or piggy-back with a radical setup... it's easier to make it do what you want.

I don't really see an issue with welding an extra bung onto the test pipe and just running both O2s in the same pipe... it's a compromise, but not having a lambda sensor for each bank is also a compromise.  Running two in the same pipe or patching the signal from one will have the same effect... either way, one bank will almost always be a little different.  You'll just have to tune a bit fat to compensate.  The banks are never perfectly balanced.

Extreme setups ADD wide-band lambda sensors... not remove them.  Some really awesome setups now have a sensor for EACH CYLINDER!!!... as it allows for better, more precise tuning (I imagine Bosch will get there one day, and have thousands of maps for each cylinder!).

P.S... posted before I read page 2.  Happy to hear you solved the issue.

I think the key to doing this as it was intended is in CWKONLS, Codeword: Configuration of lambda sensors. I believe that this value for 1 pre-cat and 1 post-cat sensor should be 3. This should allow post-cat input in lambda regulation. I am not certain, but I think this value can also be set for wideband in lieu of narrowband O2 sensors, I need to investigate this further.

Cool info, I still want to try and getting true single bank working.  I will try out some different values.

So, does CWKONLS set the bits B_lsv, B_lsv2, B_lsh, and B_lsh2?

A value of 1 for B_lsv (B1 S1), 16 for B_lsv2 (B2 S1), 2 for B_lsh (B1 S2), and 32 for B_lsh2 (B2 S2)? The original value of 51 sets the bits for all four sensors?

Have you done any testing with this?

Yes, it sets the bits mentioned. I haven't tested it yet because my car is on a ship.

for lambda fueling or?

I thought somebody here confirmed this was impossible before.

I stand corrected, Jason you are right, I misunderstood something I was reading.

noob response time...  I cant find CWKONLS

help?

code word for configuration lambda sensor
181BB

181BB...
thats what I needed, thanks and found it.  I will try out a few different things tomorrow.

Got a chance to test CWKONLS. no dice, if you set if to a value such as 1 or even 16 to just use one bank, it only trims that one bank, other bank goes dead like running just one sensor plugged in.

When set to something other than 51 it totally shuts down everything for a sensor that's turned off.  this seems like an easier way to delete rear o2s as its just one scalar to edit.

here is a representation, i crank my FPR to exaggerate trims and i can also see if one is correcting and one isnt as the motor runs very unbalanced compared to normal with one bank way different.

This thread is great!   I'm gonna study a little more before I try a primary o2 "splice"  but it looks very cool.

There is a constant that sets a bit for stereo lambda, check out LAMKO or lrs functions, I can't recall where I saw it.

Was there ever any solution to this problem?

I'm working on a software solution now.

There is more discussion about it here:

http://nefariousmotorsports.com/forum/index.php?topic=2112.180

Yeah I have been following that post as well...  I need to make my 1 wideband work for both banks also...
I'll be ready to start tuning soon and I can help.
I helped develop Megasquirt ASM back in the day so I have experience...

Is it not as simple as locating CYL-BANK_CW and setting it to a 0???

No, there is a system constant used at compilation that determines if parts of the code are to be included/excluded.

Ahhhhh I see I see.
Is there any chance of using the Bosch MS 4 Sport to help us?
What if we compared compiling both ways and compare the code?
What if we compare a 1.8 rom to the 2.7?

I'm sure these questions have already been asked before but I'm just coming in here :-)

Hi
Im want to turn off lambda bank2 or bank 1 that no matter, one of front sensor in this software 0261201146 393279 A8 4.2 d3 single turbo
i try to change CWKONLS 51 to 1,2,3,0 and this not working can anyone help ?

The code in the ECU must be changed.
If your ECU is ST based then both the processor and the flash need to be changed.

I can do it, but it is time intensive and I charge for it.

PRJ
I send you pm message and email

yep they're diff vars.

And the 5 year necro award goes to... :D

whoa