Ask A Scientist General Science Archive

 
name         Tahir N.
status       educator
age          30s

Question -   I am using a top pan 3 decimal point Sartorius electronic
balance.  The balance nominally weighs up to 1000.000 grams.  Before
using the balance to measure say a 100 g of buffer salt, I have carry out
a check weight using a calibrated 500 gram check weight.  The
specification for the check weight is 500.003 to 499.997 grams (i.e., +/-
3 mg).  In other words if my check weight is not within the specification
I am alerted to the fact that something is wrong - whereupon I carry out
a full verification (using other weights eg. 10mg upwards to 1000 grams)
and internal calibration and then re-check.  So far to date the check
weight is always +/- 3 mg.  My question is if I place a beaker with a
weight of 300 g and TARE the balance electronically to give me 0.000 g
then I weigh a much smaller quantity such as 50 mg - do I need to have a
smaller check weight or can I rely on my 500 g check weight.

Is the load cell linear or is there a correction which the balance carries
out when the balance is TARED?
-----------------
I used to just weigh out quantities of distilled water with very precise
pipets.  0.01 cc of distilled water should mass out at 0.01 grams...etc

Peter Faletra
====================================================
The load cell is approximately linear, but there is probably an equation
to correct for any nonlinearity.  Probably the calibration routine simply
sets the coefficients for the linearization equation.

Anyway, tare is just an electronic subtraction after the full weight has
been measured.  If you tare a 300 g container and then weigh a 50 g
sample, you are just weighing the 350 g combination, and then subtracting
300 g from the result.
--
Tim Mooney
=====================================================
Dear Tahir,

I would recommend contacting Sartorius "http://www.sartorius.co.uk/".

Dr. Jim Tokuhisa
=====================================================
The load cell is designed to be as linear as possible, so in principle 
doing zero and 500gm calibration checks should be OK.

TARE is simply arithmetic subtraction of the previous display number from 
all subsequent readings.
A computer chip does the arithmetic after the force-to-numeric conversion 
is finished,
so it has little to do with any failures of the balance's accuracy.
If an operator did not use the TARE mode, and did a subtraction on paper 
after writing down each reading, it would be the same thing.
So what you need to ask yourself is, is the balance linear from 0-500gm?
Is the slope the same over 300-350gm as it is over 0-50gm?
And, is it the same as long ago?

It is imaginable for a mechanical insult or nonlinear electrical 
impairments to create nonlinearity in the balance's movement.
It would also create large zero-offsets or full-scale errors, but if 
someone readjusted the balance to eliminate these,
then your calibration checks would pass but a non-linearity could remain.
But this is less likely than simpler errors.

It would be good to have 1-3 extra calibration weights to use perhaps yearly.
If they read about the same as last year, you have proven to yourself and 
most skeptical observers
that your scale is still quite linear and unchanged.
I might choose 50gm, 550gm, and 250gm.  250 to detect a simple slow 
curvature between 0gm and 500gm,
50 to highlight glitches near 0 gm, and 550 to prove linearity in the 
vicinity of 500gm and a little above.

cordially,
Jim Swenson
=====================================================