VEMS Community Forum
Technical => Wideband & EGT Gauge => Topic started by: Machismo on June 15, 2015, 07:36:25 pm
-
I'm a bit behind the schedule and trying to speed up my first Vems installation, so another 20€ / 100mBTC for anyone who succesfully guides me though any tricks that are needed for correct 4.9 setup. Notice that I don't know anything.. :D
At the moment I have a 1.2.30 firmware, but if needed can update to 1.2.31. Vemstune is the the latest version and ini files have been updated.
I have already done the wiring and measured against car's ground, that 12v is coming to 4.9 sensor pin 4 (gray wire). Other wires:
pin 1 (red) - EC18 / 9
pin 2 (yellow) - EC18 / 7
pin 3 (white) - EC18 / 18
pin 5 - not connected, empty
pin 6 (black) - EC18 / 13
I got the wiring info from here:
http://vems.hu/vt/help/v3/v3_wideband_settings.html
Here are the Vemstune settings, haven't touched those values so they can be totally wrong:
(http://penthouse.homelinux.net/vems/49lsu_widebandsettings_defaultvalues.png)
(http://penthouse.homelinux.net/vems/49lsu_widebandcalibration_defaultvalues.png)
First questions:
Is my wiring correct?
I made no hardware changes to the board, should I do something?
Somebody mentioned that they had some wideband values on a paper that came with the board, I don't have that, should I ask for it?
I connected the sensor and noticed that it does not heat up. Just to test heating I pressed start calibration and it said 'heating'. During that I measured voltage between pin 4 and 3 and there was some small activity there, but only between 0 and 0.4 volts. Pin 4 has constant 12 v so that is not the problem. So it must be pin 3. How to fix that?
-
Try looking here... hardware changes are explaint http://vems.hu/download/sensors/LSU4.9/LSU4.9_pinout.pdf
-
Try looking here... hardware changes are explaint http://vems.hu/download/sensors/LSU4.9/LSU4.9_pinout.pdf
Thanks. I measured the nernst voltage according to the pdf and it was 3.7V, so I installed the 27k resistor, after which the voltage was 5V. I guess this is the only change that needs to be made.
After that I did the sensor ohm measurement between pins 1 and 5 and the result was 102 ohm, so I entered Rcal value of 163.
On the pdf it said something about Ri target value of 165, so I entered that also, but don't know if it's correct.
(http://penthouse.homelinux.net/vems/49lsu_widebandcalibration_defaultvalues2.png)
I also choose set of default values from the bottom menu, no idea if these are any better than the ones I had before by default:
(http://penthouse.homelinux.net/vems/49lsu_widebandsettings_defaultvalues2.png)
At this point I tried if the sensor heats now. It didn't, so I took another look at the heater fet and it was simply installed on the wrong channel! I added a second one, so both yellow locations are now populated. I plan to add a second sensor in the future so that's ok.
(http://penthouse.homelinux.net/vems/wb02_fet_location.jpg)
The heating started to work right away, after a few minutes the sensor tip was too hot to touch.
Lsu 4.9 has no need for freeair calibration, but I guess that not count the initial software calibration?
At first the ohm measurement derived value seemed to be a bit off
(http://penthouse.homelinux.net/vems/wb02calibration.png)
So I changed the Rcal value from 163 to 166. Not a big change, but now it's spot on:
(http://penthouse.homelinux.net/vems/wb02calibration_success.png)
I did some testing and wideband seems to be working without any problems... Correct me if I did any mistakes?
-
Im not and expert but it looks good to me.....
-
, so I installed the 27k resistor, after which the voltage was 5V. I guess this is the only change that needs to be made.
how (where) did you install the resistor ? from where did you get the voltage for the resistor to end up with 5V?
I am planning to test the 4.9 too because we have trouble with 4.2 and firmware 1.2.3X on 3.6 board
-
Im not and expert but it looks good to me.....
Ok. I like the way how it works so incredibly fast, my old innovate LC1 controller was much slower to react. I think it had a 4.2 sensor. I had to calibrate it often to keep it somewhat accurate.
I wonder if adding a second sensor to the other cylinder bank exhaust will make the Live Ve Analyzer autotune work more accurate or faster. There is no mention of that in this guide
http://vems.hu/vt/help/main/tools/live_analyzer.html
how (where) did you install the resistor ? from where did you get the voltage for the resistor to end up with 5V?
I just made a wire that has the 27k resistor in series with it and soldered the other end to EC36 pin 28 (+5v supply) and other end to EC18 pin 13 (wb02 nernst signal).
-
, so I installed the 27k resistor, after which the voltage was 5V. I guess this is the only change that needs to be made.
how (where) did you install the resistor ? from where did you get the voltage for the resistor to end up with 5V?
I am planning to test the 4.9 too because we have trouble with 4.2 and firmware 1.2.3X on 3.6 board
0 problems with that :)
-
Bonus picture of the sensor wiring making
(http://penthouse.homelinux.net/vems/o2_sensor_cablemaking.jpg)
-
thanks... so this should be PIN 12 on M55 connector... I will try this soon and report... LSU4.9 is already ordered..
-
Today I modified my 3.6 to LSU 4.9
Measured the nerst Voltage -> 3,7V
Soldered a resistor 27K from EC36/28 to black wire nerst inside ECU
Measured -> 4,99V
Measured RCal on LSU4.9 Sensor -> 95,5 Ohm should be calibration value 157
Connected Sensor to ECU an started wideband calibration (LSU 4.9 sensor) free air - calibration was way off. Needed to correct the value to 174 to get pointer at the right spot
Engine Start - Value on VEMS Display same as Value on Zeitronix Display - so calibration should be ok
But why is the Rcal calculation so wrong in this case??
best regards
-
Sometimes when when I'm running low battery on my multimeter or there are some problems with measuring wires I get totally false readings. You might want to try to measure for example 100ohm resistor to see if it's working. If it is, remeasure the sensor.
-
Sometimes when when I'm running low battery on my multimeter or there are some problems with measuring wires I get totally false readings. You might want to try to measure for example 100ohm resistor to see if it's working. If it is, remeasure the sensor.
I measured the 27K resistor an it was correct :-/
update... 100Ohm is also correct...
-
Hello Black,
But why is the Rcal calculation so wrong in this case??
The values we provided are ball-park figures based on an average reading on current boards to be used as a starting point for calibration; FYI Bosch calibrates the RCAL resistors per batch, so one sensor might be at the far left (or far right) of the spectrum and still get the same RCAL. There also might be some variation in the on-board resistor values, all in all .. The free-air calibration takes the whole loop response in account and calibrates for this.
You are only 174/157 = 10-11% from the median value of 157 no problem at all after calibration your sensor should work perfectly.
Best regards, Dave
-
perfect.... thanks
-
Now after about 30 hours of use, the 4.9 wideband has started to show erratic readings, it might show too lean or usually way too rich. The engine itself is running fine and other independent wideband at the other exhaust pipe is has showed consistent readings all the time so the Vems controlled 4.9 readings really are erratic... I wonder what has happened. The sensor is supposed to be durable, so I must have done something wrong?
-
It might not be the sensor, because today I connected another, brand new 4.9 and tried to calibrate it. It kept saying 'increase calibration value' even though I maxxed it at 255 already. The config values have not changed, so I guess it's something electrical at the board..
I know it's not the heater since the sensor heated up normally.
-
Hello Machismo,
In cases like this, start at the beginning:
* check connector crimps on both sides
* wiring, shorts, perhaps some wiring melted from running close to exhaust ?
* perhaps one (or more) of the contacts were not inserted all the way and make bad connection ?
All of the above is much more likely than any controller problem.
Best regards, Dave
-
I examined the board closely and found one problem, the nernst voltage 27k resistor wire had worked itself off at the pin 13. It probably happened when I had a passenger couple of days ago and he streched his legs at the wiring. I still have a temporary wiring because of the constant changes and because of the stock ecm backup possibility at this point.
However I connected it back but the problem continues even though nernst voltage is now back to normal. I wonder if the loose wire had the chance to create some kind of shortcut. The wire was pretty stiff, but it might have hit EC18 pin 7 (WB02 pump - ).
-
I didn't find any other faults, so I restested with the new sensor. It was then I noticed that I had made a mistake the first time, the sensor had heated, but only engine running and not when I tried to calibrate it. Im my case the sensor heating is connected to fuel pump relay so it was too easy to make a mistake there...
Anyways, now that I connected the new sensor and the nernst voltage is ok, the wideband is working perfectly again!
Only thing that troubles me is that why the sensor went kaputt... Does too low nernst break it or was it the exhaust backfires/popping that my car has? On the other hand, much older 4.2 sensor on the other pipe is still working normally, so exhaust conditions are not impossible. This is a natural aspirated engine.
The 4.9 sensor that broke was specific to some car, I don't know which. It was quite large opening Bosch LSU4.9 lambda 0 258 017 115 with a short connection wire.
The one I replaced it is Bosch LSU4.9 Universal lambda 0 281 004 148 with much smaller holes and a longer wire. I hope the more protective cover will shield it from backfire?
(http://penthouse.homelinux.net/vems/lsu49sensors.jpg)