[Diy_efi] WB free air calibration

[Rob F]

Bruce,

Thanks for you thoughts.  All your points are well received.  I can't argue
with Bruce^2  =)  I won't worry about it any further.
I also appreciate you explaining the output function of the circuit.  It
would be nice to have this circuit analysis captured in a FAQ.

-Rob


----- Original Message -----
From: <bcroe at juno.com>
To: <Diy_efi at diy-efi.org>
Cc: <bcroe at juno.com>
Sent: Wednesday, May 29, 2002 7:27 PM
Subject: [Diy_efi] WB free air calibration


> Rob,
>
> Glad you got it working.
>
> On Wed, 29 May 2002 09:57:36 -0700 "rob files"
> <rncfiles at hotmail.com> writes:
> > Hi all,
> >
> > After replacing some blown out parts, I got my diy-WB
> > sensor working last night.  Thanks to all the REAL
> > contributors to the project.
> >
> > My free air reading is something like 3.91V.  I was
> > wondering, does this mean all of my readings will be
> > offset by 0.09 volts?
>
> You should first check the voltage from J12 (gnd) and
> J9.  This ideally will be 2.500 volts.  It should be within
> 1% of that value.  Certainly if it measures .090 volt low,
> then all your readings at J8 will contain a ".09V low"
> error component.
>
> This is within the tolerance of the design.  If it bothers
> you, I suggest just adjusting your translation of volts
> to A/F that amount.  If you use an LED bar graph display,
> this error component will be cancelled by the differential
> connection to J8 and J9.
>
> > I was thinking about adding a trim pot to the calibration
> > resistor and adjusting Vout to be the 4.0V free air value.
> > Any reason why this would be a bad idea?
>
> The first thing to understand is that multi digit accuracy of
> absolute readings is neither the design objective or a useful
> feature in actual applications.  Listen to Bruce P.  Next, any
> error at J9 will NOT be correctly compensated by adjustments
> to Rcal.
>
> Third, the sensor was factory calibrated with Rcal to be
> accurate in the operating range.  I suspect free air, while
> a good sanity check, would NOT be considered an important
> operating point.  Adjusting this to a particular point would
> likely increase your error in the useful tuning range.
>
> The function the circuit depends on is
>        Rcal + 2210
> (   ________________   ) 0.3028 Ip + 2.5 V = Vout
>        Rcal + 3580
>
> If you are bent on the most accurate circuit possible, it would
> be necessary to adjust the ratios of R29, R30, R20, & R10,
> and then the absolute values of R21, R22, and R28 to well
> beyond the specified 1% accuracy, and hope the temp
> coefficient is good enough to hold the new accuracy.  Bridge
> circuits can be built to do this matching, but there are much
> better ways to spend your time.
>
> Bruce Roe
>
> > Rob
>
> _______________________________________________
> Diy_efi mailing list
> Diy_efi at diy-efi.org
> http://www.diy-efi.org/mailman/listinfo/diy_efi
>

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi