ok so im stuck.

based on having put in a brand new sensor, and upgrading the firmware on my lc1, i think i have narrowed my problem down to an evoscan issue.

i have the lc1 wired directly to the ecu since i have a mac with no serial port. I hooked up logworks using a friends laptop and the car idles at ~15.5 when fully warm.

on evoscan (im using 2.7) fully warm idle reads 7.35 steady for a bit then will swing to the 9's sometimes.

during cruise, the values look somewhat appropriate and the readings are very responsive.

when doing a 3rd gear pull there is no "tip in", it just jumps to a value in the 10's and holds and (i know my tune is very rich, but the readings dont seem to correlate with cells being hit in the map, and yes i know they should not be matching since the maps are just a request number.)

basically i cant trust this to do any tuning.

below are a map trace and clip from virtuyal dyno. what the heck could be causing this?

 

Check your formula for how it reads.

 

double and triple checked it, the formula is exactly what it should be out of the how to thread.

 

What formula are you using? 0.05859*x+7.35 is the formula I have been using and works with evoscan 2.6

Also you should make sure all grounds are good and no loose connections. Did you recalibrate the sensor after logging these values? Sometimes the calibration gets set to something such as "custom" and will throw off the readings. Make sure its set to gasoline.

 

yes thats the formula im using.

i havent checked grounds, but i have no reason to suspect them...yet, because in logworks i appear to be getting appropriate/proper readings.

the part i cant explain is why logworks is reading 15.5 at idle and evoscan reads either a flat 7.35 or somewhere in the 9's.

 

Here is a log i did with multiple driving conditions as well as idle and a 3rd gear pull.

is it possible im paranoid, and for example the run up the 240 load column reading 10.3 on the wideband with requests in the 8's is just because it cant go any richer? IDC's are in the mid to upper 90's.

but i then still worry about the 80-120 load cell trace since they are VERY far off from the request.

 

I never ever EVER got my lc-1 to work with evoscan. no matter wtf I did. Now, I log with an openport 2.0 with the lc-1's 5v connection feeding into pin 8

good luck

 

well obviously id like to find a solution to the current setup but...

an openport 2.0 works on an evo 9?

 

yes, it has an option to accept a 0-5v input on pin #8 of the plug. I have my lc-1's 5v output (the brown wire i think) going into the female plug.

I think there is a way to log it into evoscan but I havent put much effort into searching. I just do a bunch of runs and then hook my laptop in to look at all the saved logs on the sd card.

 

after a suggestion from tephra i finally got out and played with the car on evoscan 2.8.005 tonight. results were similar if not identical.



map trace is still very similar to the one in the 1st post.

some thoughts though

-i noticed the readings are pegging at 7.35 when IDC is 0 and im coasting. shouldnt i be jumping to 22.5 as in full lean not full rich under that condition?

- im hitting 98-99% IDC in evoscan so again it comes into question that when im hitting cells 8.6-9~ish that the car may just not be able to get richer than the 10.2 its reading...i dont think i have the ***** to change the fuell values and see where it reads, id rather not blow up the car over this friggin lc1 garbage.

unfortunately im probably just gonna have to wait till next year to tune, i have no time to deal with trying to RMA the damn lc1 until the car is away for the winter, and i dont want to waste 500 bucks to pay for a tune with things i have been seeing lately.

this sucks

 

heres a short log with pull if anyone would be kinda enough to take a look at it.

 

just to further rule out a computer/logging speed issue, i ran a few logs today with log to screen off.

made no difference and the wideband readings in the log look even worse.

looks like im buying a new wideband, since apparently no one has seen this happen before?

 

have you gone into LM programmer and verified that the correct voltages are ouput at 0,1,2,3,4, and 5V by setting each as the only output for all wideband readings? I would start with programming the wideband differently, and change your formula. Rule out that first

also, theres a new lc-1 firmware out. might want to try that as well..

 

no, i never did any voltage tests, because that would require a windows machine, which i dont have access to very often. due to running a mac with virtual windows machine, i cannot connect to the damn thing via serial cable. thats why the inline UEGO has become so appealing...

how new is the firmware you speak of? i updated it to the latest that was available about a month ago.

the formula in evoscan is correct like i said earlier. the one test i did when i had a windows machine in my hand showed logworks reading ~15.5 at idle when connected via serial cable, and evoscan showing anywhere from 7-9ish. that tells me that the sensor is indeed functioning semi correctly, but there is a disconnect somewhere between it and evoscan, OR the output wired to the ecu is all effed.

either way, i really need something more user friendly and not requiring so much setup. i dont have workspace i used to, so rewiring the whole thing and going through all the programming is really not a possibility at this point unfortunately, as much as i know that is the next step.