wideband calibration

[Tony C]

is there any chance of this ever being made possible within MT?
as i thought may make things easier when replacing old wideband sensors with a new one,
or am i talking out of my arse 

[[email protected]]

MegaTune doesnt have the way of turning the heater on, so no its not possible.

Rob

[fphil]

The VT calibration tool used set the free air reference of the wideband nersnt controler  did not work on my install of the first of my dual w02 sensor (lol) last week end.
It stayed on the "heating" mode for 5 minutes before I switched off.
I use a USB/RS232 coupling to talk with Vems.
I noticed that Hyperterminal does not work through that link.

Could it be that the debug status sent by VT to start the heating of the sensor does not go through that type of USB link?

Meanwhile, when running, the lambda value saturates at 1.28 until I zero the integral action of the nersnt PID.??!

any hint?

[fphil]

++++++++
I read the post "WideBand problem" from DevInAz after writing these comments  .
"I had a similar problem, but since i had been seriously messing around with it, I ended up replacing the board too"
My Vems is also a brand new one. Does someone know the technical matter with the Vems board which makes them replacing each time the calibration function does not work?
+++++++++

I finally bought a "Round WB02". This way my double WB02 need will be solved.
I was puzzled for a while when I see no way to calibrate the controller / the sensor as the calibration options of the tool menu of VT are disabled.

Anyway when plugged, it seems to work.
After while (sorry I do not have ethernet connection on the spot where the car is parked) I read at wiki page AfreshTiny/FirmwareUpgrade that "Round WBO2" is self-calibrating. Nothing was said in the Round box apart the set point of the Heater controller and of the Nersnt controller, which is at this point useless. (A short notice could save 1 hour from the customer time)

Then I turned back to the WB02 controller of the Vems 3.6 using 1.1.98 firmware. I launched the auto calibration of the tools menu, keeping the sensor in free air, still heating for minutes .... no more. I also tried the terminal mode (the 02 value is given page mlp07 now on), started the debug mode, the sensor went heating … then …no outcome.

Finally I got the car running, on dual WB02. At idle, on left bank Round shows 0.90, on right bank Vems shows 0.88. Swapping Round and Vems gives same bank result. Hurrah !!

- My temporary conclusion: Vems firmware 1.1.98 has got a WB02 self-calibration function as well.
Could you confirm this point ?. I am not sure of that since the line for the calibration resistance of the sensor is not wired on Vems -as it is on Round)

Of course if it is not, how one makes the VT/tools/calibration to work?

I any case it could be fair to check from time o time that Round or Vems WB02 is given 20.9 in open air … How to do that?

Then I the played a little with the Nernst PID parameters about the actual factory set up (on my own risk because I have been with such things for some tenth of years). The only behaviour which can be readily observed is the lambda  response curve to a throttle step. I noticed a large overshoot, lambda reaching the end of scale of 1.8. I see no good reason for that
Thank you for your comments.
Philippe

[GintsK]

"self calibration" is VT feature, not firmware. If calibration value is way off, like 120 instead of 220 or opposite, sensor do not start to read - ECU keeps it in warmup phase because of unexpected output. So I suggest to try input manual values and see what happens.

[[email protected]]

The Round is self calibrating, in that you plug it in and it works, it needs no work from VEMSTune or anything else.

[fphil]

Thanks for the idea Gintsk. The value shown by VT/tools/calibration is 184, I would try to download other values to cheat the firmware (1.1.98).

btw this is mlp07 page I got while waiting for the 02 value in debug mode

mll
Ri:0097 Nernst:03F7
Heat:B6 Pump:7A
AFR=19.03 ??.?
AFR=14.71  P      ON
 
The 2 AFR were stable, Heat and Nernst varying, which is ok I think.

Yes Rob, I understoood  afterwards the round is self-calibrating, althougi could be fine to be able to check the free air value.

Philippe

[GintsK]

btw this is mlp07 page I got while waiting for the 02 value in debug mode

mll
Ri:0097 Nernst:03F7
Heat:B6 Pump:7A
AFR=19.03 ??.?
AFR=14.71  P      ON
 
The 2 AFR were stable, Heat and Nernst varying, which is ok I think.
 

IIRC one of those "AFR" should be "O2". The first one.

[fphil]

You think so ? Hmm ... AFR should be infinity for open air .
I was expecting "02 = .." instead of the "?...?"

[Sprocket]

++++++++
I read the post "WideBand problem" from DevInAz after writing these comments  .
"I had a similar problem, but since i had been seriously messing around with it, I ended up replacing the board too"
My Vems is also a brand new one. Does someone know the technical matter with the Vems board which makes them replacing each time the calibration function does not work?
+++++++++

The real reason I replaced it, was that fact I broke it, not something that failed of it's own accord 

Then I turned back to the WB02 controller of the Vems 3.6 using 1.1.98 firmware. I launched the auto calibration of the tools menu, keeping the sensor in free air, still heating for minutes .... no more. I also tried the terminal mode (the 02 value is given page mlp07 now on), started the debug mode, the sensor went heating … then …no outcome.

From memory you need to ground the metal casing. That is what I was told a while back, not sure why, but I have done it ever since.

I had a similar situation. Free air calibration, the sensor stayed in the heating phase, and wasn't getting hot. No idea why as the sensor was a good one. swapped the sensor and the problem went away. I think I need to check it again before I bin it.

In Vemstune 2011.11.15 there are two free air calibration menus, the 'new' one tells you exactly what is going on, and calibrates the sensor automatically, while the other require some manual intervention. not sure why there are two, exceptemaybe for trial purposes

- My temporary conclusion: Vems firmware 1.1.98 has got a WB02 self-calibration function as well.
Could you confirm this point ?. I am not sure of that since the line for the calibration resistance of the sensor is not wired on Vems -as it is on Round)

Vems Round uses the calibration resistor in the sesnor connector. The sesnor is calibrated at the factory when the sensor is made, and the resistor is burned with the calibration value that the controller measures, thereby, when you replace the sensor, it will work the very same as the one it replaces with no need for any other calibration requirements. V3.x uses software calibration instead which i am led to believe is as good as the resistor, but with the benefit of being configurable. replacing the sesnor will require the new sensor to be calibrated, due to the small variations in manufacturing which is the real reason for the resistor in the plug in the first place. I asked the question why V3.x does not use the calibration resistor, and the answere is that it would use up another analog input.

[GintsK]

Sensor heating up is dumb function. If sensor stays cold - in most cases it is simple problem with hardware or wiring. The only thing where simple measuring of sensor resistance gives clear answer about sensor - heater is around 3 Ohms. Also for ECU/wiring side bulb gives clear representation of heat-up process.

Heat up phase is just time controlled function with no feedback. Later when sensor gives signal back to ECU sensor temperature is controlled by measuring impedance of some already hot part inside sensor.

Sensor metal part is isolated. No matter is it grounded or not during free air calibration.

[fphil]

the sensor stayed in the heating phase, and wasn't getting hot

I agree Gintsk. In my case , Sprocket, the sensor was getting hot.
I guess there could be some software bug inside the calibration function (VT or Vems ?) since afterwhile I tested ok the Vems controller against the Round (for one point only, car idling)

[fphil]

"self calibration" is VT feature, not firmware. If calibration value is way off, like 120 instead of 220 or opposite, sensor do not start to read - ECU keeps it in warmup phase because of unexpected output. So I suggest to try input manual values and see what happens.

Alright, this is the trick to have the wb02 calibration to work.
BTW I calibrate the temperature inputs. There the VT tool is really useful and works well, even if one needs to exit from time to time to clean up the memory garbage which slows the thing down.