chem 454/554

January 21, 2004 (Happy New Year, Chinese)

measurement -- The uncertainty principle recognizes that the measurement device (and measurement process) are part of the system. At the atomic and molecular level, the probe is not likely to leave the system unchanged. On larger samples we may be able to reduce the effect of measurement on the properties of the sample.

most analytical measurements are destructive--

collect sample, grind it, dissolve it and measure the solution.

ideally, sample size is small enough to be negligible damage

surface techniques and spectra of thin films can be nondestructive

sensor can be small, but they still alter the system

a thermocouple has wires that can serve as heat path

a glass thermometer has enough heat capacity to cool a small sample

electrical measurements-- ideal?

A wire ideally has no resistance

voltage is the same at both ends

usually a good assumption (for volts, microamps, 100 Hz or less, 12 inches)

can be a problem at high currents (power lines at 50KV not 110V, low current)

can be problem with small wires, PC Board traces, longer wires

can act as capacitor and or inductor-- a factor as high frequencies (>10 KHz?)

can act as an antenna or magnetic-voltage generator (source of noise)

can have random motion of electrons (thermal noise)

A volt meter (or ADC ) has an input resistance

when you connect it into a circuit it acts as a parallel resistance

this added resistor alters the circuit, currents, voltages

ideally, the input resistance is infinite and therefore no current, no effect on circuit

classical meter (dial) might have input resistance of 10,000 ohm

cheap digital meter might have input resistance of 1-10 megaohm

top line dvm could be 10¹²-10¹⁶ ohm

common worst case is pH electrode

internal resistance is 10 megaohm

signal might be 0.2 volts

with no current, V=0.2V (even with 10⁷ ohms, no current, no voltage drop)

with 90 Megahom meter

R total is 10 + 90 = 100 Megaohm

current is 0.2 / 100 x 106 amps = 2 nanoamps

voltage drop across meter is 2 x 10^-9 x 90 x 10⁶ = 0.18 volts

meter reading is off by 10%

(sample + meter act as voltage divider)

An Ammeter (current) ideally has zero resistance

it is inserted into a circuit to measure the current

typically it is an in-line resistor with a voltmeter that measures the voltage drop

example-- 0.1 ohm resistance

clearly if the rest of the circuit is a few kiloOhm, 0.1 Ohm is negligible change

for a low resistance circuit, the meter can alter the behavior

An Ohmmeter would ideally measure at negligible current

In practice, an ohmmeter applies a voltage and measures the current

The applied voltage generally comes from the battery (1.5-9 volts)

The current is limited to microamps or less

Some small devices may become warm when connected to ohmmeter

This heating can alter the reading or small resistance temperature sensors

Classical Circuits

Voltages (pH) were measured with a potentiometer circuit

a battery and potentiometer (voltage divider) served as a controllable voltage

a very sensitive current meter (galvanometer) was connected

between the sample voltage and the potentiometer

the potentiometer was adjusted until the galvanometer read zero

this means (a) the potentiometer output equals the sample voltage

(b) no current is drawn from the test circuit, hence true V

the voltage is read from the potentiometer dials

six to eight digits is realistic in this type of equipment

calibration is a serious issue, especially at 8 digit precision

instrument is calibrated with a standard cell (Cd amalgam cell)

factors like temperature affect both calibration and measurement equally

calibration voltages-- most electronic circuits are not sufficiently stable over a period of

days, months, years. They need to be “trimmed”-- set to the correct value using a

precise voltage reference.

some devices (most ADC, for example) really compare voltage and reference, so they

compensate automatically for component aging.

Amplifiers

Audio systems are power amplifiers (power is I x V at the output)

Most instruments use voltage (or current ) amplifiers

voltage gain = (voltage out)/ (voltage in)

an amplification factor

Ideally, a voltage amplifier

has a constant gain, unaffected by the value of signal (an error is “nonlinearity”)

has zero output with zero input (the error is called “offset”)

has infinite input resistance (so it doesn’t affect the signal source)

has zero output resistance (so it can deliver current without voltage drop)

behaves ideally over a wide range of frequencies (but there is an upper limit)

that’s really the same as saying it reacts very quickly

Better yet, it is cheap, easy to use, and the gain can be set easily

welcome to the world of operational amplifiers

a standard component in almost all instruments (pH, spectra, temp...)

We might have a signal relative to ground (one wire)

We might want to measure the difference between two wires (differential)

Most amplifiers have output that is measured relative to ground

Power Supplies-- ideally no internal resistance

Battery far from ideal

R-internal could be 0.1-10 ohms

with 10 ma output, 0.001-0.010 volt drop

with 100 ma its 0.01-0.1 V drop

(a flashlight might have 3V battery, but only 2.8 volts at bulb)

Even worse, voltage changes with life

E= Eo - 0.059/n log (Q)

Q changes by factor of 10 over 90% of battery life

E drops by 0.03 volts, perhaps

probably not enough to tell that battery is nearly gone

But most batteries have internal resistance changes

rise in R causes significant loss of power

Batteries are tested under load

simple voltmeter (dead/not dead) isn’t good gauge

LCD/resistance elements on built in battery testers

Series voltage regulator

Vin (6-40 V) [three leads-- in, out ground] Vout (5.00 volts, up to 1 amp)

cost $1

size 1 cm square

switch, run up to 10,000 Hz

senses output, turns on if Vout <5, off at Vout =5V

has its own built in reference

Switching Regulator

more complex, switch drives a transformer

higher power, efficiency (almost all computer power supplies)

Plan for about a week

skip or skim chapters 3- OP amps, 4-computers, 5 Signal-Noise

focus on 6 (light, spectra) and 7 (optical components)

review 8 (atomic spectroscopy)

We will look at spectral instruments for elemental analysis (atoms)

absorption and emission spectra

goals, practical devices, uses

then follow design process

components (ch 1, 6,7)

circuits (op amps, ch 3 revisited)

signal noise concerns and cleaning up (ch 5 revisited)

computers as part of instruments (ch 4 revisited)