Assignment:
Uncertainty in a Measurement–2: Applications
(v1)
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
Problems solutions to be submitted
uncertainty-2/v1
W5-3 A temperature-controlled water bath is used in many experiments.
Typically, a heater element/coil and a cooling coil are located at one
side of the bath and the sample at another point. The water is
circulated by a pump. Study such an equipment using catalogues
manufacturers (from their websites).
(a) Study the data sheets and specifications of 3 – 4
manufacturers/makes and interpret the major parameters
relevant to experimentation (e.g., accuracy, stability, etc.). In
particular, note the technique by which the bath temperature
temperatures from 5 to 90 ℃.
is controlled. Select baths that can provide controlled
When an object/sample is immersed in the bath, the experimenter
considers the sample temperature to be identical to the
temperature to which the bath has been set. Consequently,
uncertainty in the sample temperature is identical to that in the bath
temperature and this has to be estimated. Given the location of
Page 2 of 8
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
uncertainty-2/v1
the heating/cooling coil, sample and circulation pump, the water
due to non-uniform temperature is estimated at 0.5 ℃. The water
temperature in the bath is not uniform. The (systematic) uncertainty
temperature is sensed by an RTD probe and the (systematic)
standard is 0.1 ℃. The sensor electronics uses a curve fit equation
standard uncertainty from the calibration of the probe against a
temperature) which introduces an (systematic) uncertainty of 0.2
to represent the calibration data (RTD output vs. sensor
℃.. Another source of error is due to the installation, e.g., electrical
contacts, effects of thermal insulation of the bath, exposure of bath
estimated at 0.1 ℃.
to the ambient, etc.; the (systematic) uncertainty on this count is
(b) Calculate the systematic standard uncertainty in the
temperature of the bath as indicated by the installed
temperature probe.
During operation, a sample is placed in the bath which is set to a
desired temperature and then allowed to attain steady state. The
in ℃:
temperature readings from the probe at 5 s interval are as follows,
64.5 64.7 64.9 65.0 65.2 65.4 65.6 65.6 65.6 65.4 65.3
65.2 65.0 64.8 64.6 64.5 64.3 64.1 64.0 63.9 63.7 63.6
63.5 63.5 63.5 63.6 63.5 63.7 63.6 63.8 63.8 63.9 63.9
63.9 64.0 64.1 64.1 64.2 64.3 64.2 64.3 64.4 64.4 64.5
64.4 64.7 64.6 64.6 64.7 64.7 64.8 64.9 64.9 65.0 65.0
65.0 65.2 65.1 65.1 65.2 65.3 64.4 65.4 65.4 65.5
(b) Systematic standard uncertainty: 0.557 ℃
(c) Random standard uncertainty: 0.081 ℃
(d) Combined uncertainty: 0.563 ℃; Expanded uncertainty (95%): 1.125 ℃
The non-uniformity of temperature (0.5 ℃) is the dominant contributor,
accounting for nearly 90% of the total systematic uncertainty. The random
uncertainty is relatively small due to the large sample size
W5-4 Consider the case where an analog signal in the range – 1.5 to +
1.5 V is input to a 12-bit A/D converter that is calibrated for an input
range of ± 5 V.
(a) Calculate the elemental systematic standard uncertainty due
to digitization.
(b) Repeat the calculation for 8- and 16- bit A/D converters.
(c) Repeat the calculation of part (a) when the input range of the
A/D converter is set at ± 1.5 V. Which is a better option?
(d) What will be the A/D converter output when a steady signal of
– 0.8 V, 0 V, and 0.57 V is input to it? (for part (c)). Does a
Page 3 of 8
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
steady signal confirm steady
uncertainty-2/v1 state in the physical system?
[Hint: if the signal is sinusoidal, and ADC sampling is same as
the signal time period, then!!]
(e) Calculate the ratio of systematic uncertainty (at 95 % CL) to
the value of the input signal in part (d).
The elemental systematic standard uncertainty for the 12-bit ADC at is +/- 5v is 0.000705v,
while at +/- 5V it improves to0.000211V , making the tighter range more accurate. A steady
digital output can be misleading due to aliasing effects, and the relative uncertainty for the given
signals is approximately 0.05% TO 0.07%
Page 4 of 8
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
uncertainty-2/v1
W5-5 A measurement system hardware is shown in Figure 5A-7 below.
The DC power supply gives excitation power to the load cell; the
excitation voltage has to be decided. The load cell output is fed to
an amplifier whose gain can be adjusted. The amplifier output is fed
to the A/D converter; several ADCs are available for a combination
of input voltage and number of bits. You have to decide the DC
power supply voltage, and recommend the optimum combination of
amplifier setting, and A/D converter parameters. If possible, suggest
the make and model of the amplifier-ADC best for this application.
Besides other combinations, comment on the desirability of the
following options:
A: Excitation 10 V, Amplification 100, ADC input 1 – 10 V, 16-
bit ADC B: Excitation 15 V, Amplification 1000, ADC input 1 – 15
V, 20-bit ADC C: Excitation 10 V, Amplification 2, ADC input 1 –
5 V, 8-bit ADC
Select references:
Load cell : [Link]
Page 5 of 8
A/D CONVERTER DIGITAL
8-, 12-, 16-, 20- bit VALUE
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
Figure W5-5. Arrangement for measuring
uncertainty-2/v1 the mass with a load cell.
Answer The optimal configuration for the load cell measurement system is
10V excitation, ~100-200x gain, and a 24-bit ADC (like HX711) for high
resolution without saturating the signal [44, 28]. Option B is the best of the
listed choices due to high bits (20+), though 10V excitation is standard for
stability, and 1000x gain may require careful noise handling.
Page 6 of 8
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
uncertainty-2/v1
W5-6 Lookup the specifications of A/D converters from different
manufacturers and make a comparative assessment of their input
range, number of bits, conversion time, sampling rate (maximum),
and any other important features.
Look up specifications of multiplexers of different makes and
compare their salient features.
Some websites for A/D converters (many more on the web; explore!)
[Link]
Converters-
ADC/_/N-4c43g
[Link]
[Link]
cards/analog- interface-cards/p/OME-PCI-1800
[Link]
module
answer
Key Parameters Of ADCs
Range
Resolution
Definition
Impact on Performance
Sampling Rate
Accuracy
Linearity (DNL and INL)
Differential Non-Linearity (DNL)
Noise Integral Non-Linearity (INL)
Dynamic Range
Relationship with Signal-to-Noise Ratio (SNR) and Resolution
Power Consumption
Input Impedance
Effective Number of Bits (ENOB)
ADC Transfer Function
Page 7 of 8
noc26-me11/Uncertainty Analysis/Assn W-5/Meas
uncertainty-2/v1 ---------- ----
Page 8 of 8