Wideband Calibration Unstable

[AntRoyal]

Hi,

I'm finally getting around to installing VEMS into my Fiat 131r. I have the ignition system running, but want to get the WBo2 sensor working before tackling the injection system.

If someone could help, I'd appreciate it.

I've read all the documentation I could find and was following basically http://www.vems.hu/wiki/index.php?page=GenBoard%2FManual%2FWideBandHardwareTest

Using the 100 ohm resistors ...

I've set pump_pw_zero fine (-0.0002v).

But on calibration with the heater turned on (I'm using the new WBO2 calibration in VEMSTune to turn on the heater) .. the nernst to Pump- voltage fluctuates from 0.398v to 0.413v.

If I alter the wbo2_nermst_dc_target I have to increase the value from 140 to 156 to get it near 0.450v, and it is not stable. (VEMSTune warns that the max should be only 140).

If I attempt to calibrate with the resistors in place, I can't get it to stay in the green, it reads anywhere from 20.7 to 21.1 several times a second.

Calibrating with the sensor, the gauge jumps all over the place.

Can anyone point me in the right direction as to how to troubleshoot this ?

I have a V3.3 GenBoard, firmware 1.1.97 and VEMSTune 0.40.13 (2011.11.15)

Configurations posted below





Note also that VEMSTune says that the battery calibration is out of range.

Thanks
Looking forward to getting this one sorted !!

[GintsK]

Are your pump (-) are 0.4V ?

If your batt calibration are so drastically out means serious problem with board or corrupted firmware.

Are your grounds, especially signal ground, are connected as per manual?

Are you sure about your DVM readings finally?

Gints

[AntRoyal]

Hi Gints

Thanks for the reply.
Yes the Pump voltage was fine at 4V, and stable.
I'll double check with another DVM, but I figured that that the same problem with the calibration instability is reflected in the O2 readings.

As regards firmware, I tried this about three months ago and had the same problem.. left it for a while .. then tried again yesterday, updated to the latest firmware and VEMSTune, still same problem. Didn't have any problems doing the firmware upgrade.

Grounds are connected as per manual, I'll double check today.

Maybe I should try to troubleshoot the battery calibration problem first, I'm not sure if that would affect the O2 problem.

Where can I find out info on how the battery calibration works. I imagine that it takes a signal either from the 12v or 13.8v line through a voltage divider and uses the Battery Calibration value to calculate actual voltage. (I'll double check that DVM today)

Thanks
Antony

[AntRoyal]

Double checked earth and seems OK.
Double checked the battery calibration with the scope and battery calibration spot on (with the high calibration value), agrees with DVM.
The VEMS seems to create a ripple on the supply power line of over 500mV. Is this normal ?

Thanks
Antony

[GintsK]

Yes may be start with battery cal issue. It is ADC0 on chip. With voltage divider from R13 and R6.
http://www.vems.hu/wiki/index.php?page=GenBoard%2FVerThree%2FSchematic

Try to measure current drawn by board supply. Or simpler - check for very hot parts on board. May be some damaged IC makes voltage drop or something like that.

I just measured AC component on 12V supply -  0,035V.

Gints

[AntRoyal]

Well doesn't sound too good, much lower ripple.

Current is 107mA.. will pull it out of the case and check battery voltage at ADC0 input and check for hot chips, I prefer ones with salt and vinegar on them !

Antony

[AntRoyal]

OK .. making some progress.

No hot chips but measured VCC, AVCC and AREF

VCC fine at 4.03V .. (ripple at 360mV)

But AVCC and AREF are both 4.37V

A quick search on AREF and VEMS found a reference to a missing 33 ohm resistor on AVCC (definitely not one installed on my board), if this goes to 5V instead, I'm sure will alter the battery calibration, and also the O2 calibrations (I'm guessing without checking schematic). The reference in the Wiki also talks about not using a resistor but tieing directly to 5V with a wire.

Any advice ?

Antony

[GintsK]

Seems you have either bad DVM or have some global error.

VCC and rest two get supply from different voltage regulators. Both are +5.0V.

IIUC 33 Ohm resistor are dropped long ago. There is 0 Ohm trace instead.

Ripple also terrible!

[AntRoyal]

Thanks ..

Readings were taken from a scope.
There is no trace (0 ohm or 33 ohm) to the +5V supply (I think this is what I need to add to bring AVCC to 5V) from my reading the AVCC is being supplied internal from the AVR, only coupling capacitors installed.
I'll add the trace and recheck, then look to deal with the ripple problem.

Thanks
Antony

[GintsK]

I think leave AREF as is. Especially if your board is newer than year2006. seems they left internal supply there.
AREF description http://www.vems.hu/wiki/index.php?page=BuildProcedures%2FSectionThree But note - modifications are related just for very old boards.

May be some similarity with your problem:
http://www.vems.hu/wiki/index.php?page=MembersPage%2FDavidBlades%2FRegulatorProblem

[AntRoyal]

Update on this topic ..

It turned out that the Battery Calibration was the clue. I add small wire from AVCC to +5v. (I left AREF alone).

Immediately I could reset the calibration and this was within limits.

The wideband o2 also responded and I could set this within the prescribed limits. I re-calibrated with resistors and then the sensor and while the readings are rock solid, I was able to get reasonable results.

Sensor is now in the exhaust and reading OK (as far as I can determine)

I still have some ripple on the VCC but it doesn't seem to be affecting anything as yet.

Thanks Gints.

Antony

[GintsK]

Originally AVCC comes from different regulator, right? May be it is overloaded or damaged?

Gints