University of Cambridge

Department of Physics: Cavendish Laboratory

E3 Thermal excitation in a
semiconductor

AC signal generator

1.20 k Ω thermistor

V2 V1

NST IA Physics Practicals

Michaelmas Term 2022
 Health and Safety at Work Act, 1974
The following sentences summarize the content of this act as it applies to the day-to-
day running of the practical classes (see also the section on Health and Safety on the
Department of Physics website).
• Each employee of the University and each student working in the University has
a responsibility to take care of his or her own safety and the safety of others.
• By example and training, staff and students alike should be encouraged to adopt
an approach to their work in which good safety practice is regarded as normal.
• An employer is to ensure so far as is reasonably practical the safety of his or
her employees; an employee is to take reasonable care for his or her own health
and safety and that of other persons who may be affected by his or her acts or
omissions at work.
• The Act places the duty on all persons (and it is this duty that applies to students)
that they shall not intentionally or recklessly interfere with or misuse anything
provided in the interests of health, safety or welfare.
• Each experiment is subject to its own risk assessment and safety instructions are
given as required.

Joint work and cheating

The experiments are designed to be carried out in pairs. Discussion among students
and with the demonstrators is strongly encouraged — you may then use any help or
insights gained in these discussions to improve your experiment, your understanding of
the physics and your report. However, your report should be written by you, following
the guidelines on writing reports, and only data collected in your experiment should
be presented as your own. Copying from others’ work and submitting it as your own
report is cheating. Falsifying and/or inventing data is also cheating. If anyone is found
cheating, the matter will be reported to the Head of Department, who will normally
refer the incident to the University Proctors.
3 Thermal excitation in a semiconductor
Aims:
• Skills: to perform an experiment to test a theoretical hypothesis; to gain
experience planning an experiment; to learn how to use basic electrical
apparatus; to gain experience in the use and interpretation of logarith-
mic graphs; to design a spreadsheet to carry out repetitive data analysis;
to estimate the gradient of a graph and its error in Excel.
• Science: to study the non-classical temperature variation of resistance in a
semiconductor; to find the gap energy of silicon.
Preliminary reading: section 3.1 (pages 1–4).
Introductory exercise: section 3.6 (page 12).

———————————————————————–
3.1 Introduction
This experiment studies the variation of resistance with temperature of a small sample
of fairly pure silicon. The silicon is in the form of a thermistor, a small temperature sen-
sitive resistor, whose resistance is measured with an electrical circuit and some digital
meters. The temperature is measured by an “electric” thermometer, a thermocouple.
The thermistor and thermocouple are placed close together in a copper block in the top
of a soldering iron which is used to heat the thermistor.

3.1.1 Theory
Unlike a metal, the electrical conductivity of silicon, a semiconductor, depends very
strongly on temperature. What is perhaps surprising is that its conductivity increases
with temperature — in metals heating usually reduces the conductivity. A simple, non-
classical, model for this behaviour assumes that electrons can only conduct electricity
if there are energy states nearby into which they can be promoted when an electric field
is applied. In a semiconductor there are two sets of electron energy states, the valence
band and the conduction band, separated by a significant energy gap Eg . At absolute
zero all the states in the valence band are full and all the states in the conduction
band empty, with the result that the material is an insulator. However, with increasing
temperature, increasing numbers of electrons are thermally excited from the valence
band to the conduction band; both the electrons in the conduction band and the vacant
states left behind in the valence band contribute to the electrical conductivity, which
itself therefore rises with increasing temperature.
For a semiconductor like silicon, at sufficiently high temperatures (in what is called
the intrinsic region), this model predicts a specific relationship between the electrical
resistance, R, and its absolute temperature, T :
 
T0
R = R0 exp . (3.1)
T

1
T0 is a constant related to the energy gap Eg and Boltzmann’s constant kB (≈ 1.381 ×
10−23 J K−1 ) by
Eg
T0 = . (3.2)
2kB
Equation 3.1 predicts that a graph of ln(R) versus 1/T should be a straight line. The
gradient of the line gives a value for T0 , from which the energy gap Eg can be calculated.
At lower temperatures the resistance no longer obeys the relationship in equation 3.1
so the graph of ln(R) versus 1/T is no longer a straight line — this is called the extrinsic
region.

3.1.2 A method for measuring resistance

A simple way to measure the resistance R of the thermistor is to compare it to a standard
resistance in series, as indicated in figure 3.1. Assuming the same current flows through
each component, Ohm’s law implies that the ratio of measured voltages across the two
resistors gives the ratio of resistances, i.e.
V1
R = (1.20 kΩ) . (3.3)
V2

known unknown
I I
1.20 k Ω R

V2 V1
Figure 3.1: Known and unknown resistors in series. The same current is assumed to flow
through both components.

3.1.3 The thermocouple

Whenever a temperature gradient is maintained in a metal and no electric current is
allowed to flow, a potential difference will appear between two regions at different
temperatures. This “thermoelectric e.m.f.” can be exploited to measure temperature
in a thermocouple as illustrated in figure 3.2. Here, chromel (90%Ni/10%Cr) and
constantan (60%Cu/40%Ni) wires are joined — this is the “hot” junction — and the
other end of each wire is connected with copper leads to a voltmeter at the two “cold”
junctions. Provided that the cold junctions are kept at the same reference temperature
(using a water–ice bath in this experiment), the meter will register a voltage that is a
direct measure of the temperature of the hot junction.
The variation of output voltage with temperature is substantial, but not linear.
To convert from volts to absolute temperature you should use the table for a
chromel/constantan junction on page 11.

3.1.4 The arrangement of the thermistor and thermocouple

Figure 3.3 shows schematically the arrangement of the thermistor and the hot junction
of the thermocouple in the copper block at the top of the soldering iron used to heat

2
 Chromel
Hot
V Constantan

Copper

Cold
Figure 3.2: A schematic cartoon of a practical thermocouple arrangement. For useful
results, both cold junctions must be kept at the same reference temperature. The voltage
recorded by the meter then measures the temperature at the hot junction.

the sample. Copper is a sufficiently good thermal conductor that you can assume the
block is at a uniform temperature, and provided the thermistor and the thermocouple
are well inside the block they will also be at the same temperature. Teflon sleeving
provides electrical insulation. The outer tube is to keep draughts off the apparatus.

thermocouple
themistor
111
000
000
111 000
111
000
111
000
111
000
111 000
111
000
111 1111
0000 soldering
000
111 000
111 0000
1111
0000
1111 iron
000
111
000
111 000
111
000
111
000
111
000
111 000
111
000
111
copper
0000
1111
0001111
111 000
111 1111
0000
000
1110000
0000
1111 000
111 0000
1111 magnetic
000
111 000
111 0000
1111
0000
1111
000
111
000
111 000
111
000
111
alloy

Figure 3.3: A cartoon of the arrangement of the thermocouple and thermistor in the tip
of the soldering iron used to heat the sample.

You can heat the thermistor using the normal element of the soldering iron, controlled
with a variable power supply. The magnetic alloy forms part of a control and safety
device which prevents the soldering iron from heating at all unless it is in place.

3.1.5 Summary of practical procedure

Setting up the circuit to measure the resistance of the thermistor:
• On the small circuit board provided, connect the AC signal generator across the
thermistor and the standard 1.2 kΩ resistor, which are in series (see figure 3.4).
Digital voltmeters connected across the thermistor and the standard resistor will
give V1 and V2 .

3
 AC signal generator

1.20 k Ω thermistor

V2 V1
Figure 3.4: The basis of the circuit used for measuring the resistance of the thermistor.

Setting up the thermocouple:

• Having connected the two cold ends of the thermocouple to two meter leads seal
these cold junctions into two small glass tubes and submerge the tubes in crushed
ice in a large beaker.
• Connect the meter leads to a digital meter which measures DC volts. The output
voltage can be converted to give the absolute temperature of the hot junction as
described earlier.

The main experiment:

• The soldering iron has its own heating element controlled with a variable power
supply. The thermistor and the hot junction of the thermocouple are positioned
in the copper block in the tip of the soldering iron; they are at the temperature of
the copper block.
• Plan and perform a series of measurements of V1 and V2 to give you about fifteen
data points evenly spaced in temperature from 200 ◦ C to 290 ◦ C.
• Transfer the relevant data from your booklet to the spreadsheet and plot ln(R)
against 1/T . This should span what is called the intrinsic region of the semi-
conductor — however you may find that at high values of 1/T the plot becomes
non-linear where the semiconductor has entered its extrinsic region.
• Use the data-analysis facility to carry out a linear regression analysis of the data in
the intrinsic region, recording the value of the slope and its error in your booklet.
From these deduce the value of the gap energy Eg and its error.

———————————————————————–

4
 ———————————————————————–
Detailed instructions for you to follow in the practical
———————————————————————–
In this experiment all measurements, and the data analysis in Excel, should be carried
out jointly with your partner.

3.2 Checking and setting up the apparatus (≈ 45 minutes)

3.2.1 Setting up the circuit to measure the resistance of the thermistor
The thermistors used in the class are made of fairly pure silicon enclosed in a glass
capsule, with copper leads attached.

(green) (red)

potentiometer

1.20 k Ω thermistor

V2 V1
Figure 3.5: The circuit used for measuring the resistance of the thermistor. The AC signal
supplied by the Function Generator is connected across the red and green terminals.

♣ The thermistor is already threaded through the lower of the two parallel holes in
the copper block and positioned in the centre of the block — please do not move
it.
♣ If it has not already been done, connect the 1.20 kΩ resistor to the two “V2 ”
spring contacts on the small circuit board provided: to do this bend the spring
over, insert the copper wire between the coils, and release the spring.
♣ Measure the value of the 1.20 kΩ resistor with a digital meter as accurately as you
can.
♣ Complete the resistance chain, connecting the thermistor leads to the two
“V1 ”springs.
♣ Connect the two digital voltmeters (both set to measure AC) across the thermistor
and the standard resistance as shown in figure 3.5.
♣ The AC signal is supplied by the “Function Generator”. Make sure it is turned
off (the switch is at the rear), and connect it to the circuit across the red and green
terminals. (Ignore the “0 V” and “+15 V” labels on the circuit board.) Set up the
Function Generator as follows:
♣ check the power supply is switched off while you connect up the circuit;
♣ select the output from the “50Ω O/P” socket;

5
 ♣ set the amplitude knob to minimum;
♣ set the DC offset and symmetry knobs STRAIGHT UP;
♣ push the sine function, and the −20dB attenuation buttons IN;
♣ push the 100 Hz button IN;
♣ adjust the frequency control knob to 1.3 (i.e. 130 Hz).

It is useful for you to know why an AC frequency of 130 Hz is used. If the frequency
chosen is too high then capacitative effects can distort the measurements; if it is too low
then the sensitivity of the meters — which are optimised to measure high frequencies
— is reduced. It is also normal to avoid multiples of 50 Hz since these might coincide
with “mains pick-up” and lead to an additional source of error.
As is obvious, the resistance of the thermistor cannot be measured without passing
an electric current through it. This will create heat, and if the current is too large,
the temperature of the thermistor will rise significantly above that of its surroundings,
invalidating your results. Experiments show that this heating has a negligible effect
provided the voltage across the thermistor is less than about 350 mV. You can satisfy
this requirement in the following way.
♣ Set the potentiometer control on the small circuit board fully clockwise, so that
the full voltage is across the resistance chain. (The potentiometer is here for
‘historical’ reasons and requires no further adjustment.) The output rms voltage
of the Function Generator is thus equal to the sum of the voltages across the
thermistor and the standard resistance i.e. V1 +V2 .
♣ Turn on the Function Generator, and as soon as you do so, adjust the amplitude
knob so that V1 +V2 is about 200 mV (the exact value is not important).
Start your report with an appropriate title and the date; describe how you set up the resistance
circuit and the checks you made.

3.2.2 Setting up the thermocouple

The operation of the thermocouple is described in section 3.1.3. To convert from volts
to absolute temperature you should use the table for a chromel/constantan junction on
page 11 — this conversion assumes that the cold junction is at 0 ◦ C.
♣ The thermocouple is already threaded through the upper of the two parallel holes
in the copper block and positioned in the centre of the block — please do not
move it.
♣ Connect the ends of the thermocouple and the meter leads using the springs pro-
vided.
♣ Seal each cold junction into a small glass tube (to provide electrical insulation)
with a covering of water (to provide thermal contact) using a rubber bung to
hold everything in place. Then submerge the cold junctions in crushed ice in a
large beaker.
♣ Connect the meter leads to the third digital meter to measure DC volts at max-
imum sensitivity. At this point you need to check that the thermocouple gives a
sensible room temperature reading. Also, reverse the leads in the meter, to ensure
there is no systematic “zero error” in the thermocouple circuit.

6
 ♣ Make sure your booklet contains a useful record of what you did and measured.
♣ Explain why ice/water is used as the cold junction in the thermocouple system.
Describe how you set up the thermocouple and the checks you made. Explain why ice/water is
used as the cold junction in the thermocouple system.

3.3 The main experiment (≈ 2 41 hours)

SAFETY. The apparatus used in this experiment gets hot enough to give you a painful
burn, so be careful. Good first-aid for a minor burn, if it happens, is immediate immer-
sion in cold water — use the (thermocouple) beaker of ice/water in this event.

Important note: to avoid damaging the thermistors, please do not heat them above
300 ◦ C (which corresponds to an output voltage of 21 mV from the thermocouple).

In this part of the experiment you will need to work together to collect your data. A
single set of data for both of you will suffice.
♣ Check that the thermocouple is positioned in the centre of the block.
♣ The soldering iron is heated by the black DC power supply. Before turning it
on ensure that the VOLTAGE knob is rotated fully anticlockwise and that the
CURRENT knob is rotated fully clockwise.
♣ Switch on the power supply and adjust the VOLTAGE knob until the left-hand
display reads 15 V. Wait for the thermocouple output to reach about 18 mV.
(If the output from the thermocouple does not reach 18 mV at this stage then
ask your demonstrator to look at your set-up before proceeding further.) Then
reduce the heating supply to 12 V (you may need to reduce it a bit further if the
thermocouple output continues to increase), and wait for the thermocouple to
settle down. Roughly how long does this take?

From this point, plan and perform a series of measurements of V1 and V2 to give you
about fifteen data points fairly evenly spaced in temperature from 200 ◦ C to 290 ◦ C (or
maximum, if it is less than 290 ◦ C).
You will need to consider whether you prefer to: (a) adjust the soldering iron heating
voltage by a fraction of a volt, let the temperature stabilise, and then take your readings,
or (b) whether you can let the soldering iron cool down by dropping the voltage by half
a volt or so every few readings. In the latter case you will have to record both V1 and
V2 “on the fly”, as both will be changing continuously as the thermocouple reading
passes each selected value. (Note that V1 +V2 is equal to the output rms voltage of the
Function Generator.)
♣ Make clear in your booklet which strategy you have adopted and why.

Before you start recording any measurements, make sure you have asked yourself what
precision can be obtained with the equipment provided and how you will set about
achieving it. You will also find it helpful to decide roughly what values of the output
voltage of the thermocouple you will require to give evenly spaced data in temperature.

7
 ♣ Prepare the table in your booklet to record your results.
♣ Collect your data. When taking data, always set the voltmeters to the most sen-
sitive scale appropriate to the voltages being measured.

Explain your measurement strategy for V1 and V2 , with respect to the cooling rate and what
advantages it offers. Set up the table in your notebook to record the results; decide what values
of the thermocouple voltage are required to give evenly spaced temperatures, aiming for about
15 points — collect the data.

A systematic error can arise with this method of determining an unknown resistance
if the voltmeters used do not have a resistance considerably larger than those of the
components. The digital meters supplied have input resistances of 10 MΩ: when you
make your measurements of the resistance of the thermistor you will need to convince
yourself that there is no systematic error.
♣ Explain your reasoning in your booklet. (You should already have done this in
the introductory exercise; check with your demonstrator that your answer was
correct — if it was not, then give a full explanation in your booklet at this stage.)

Explain why the resistances of the voltmeters do not significantly affect the determination of the
resistance of the thermistor.

You can now use a spreadsheet to carry out your repetitive data analysis.
♣ Transfer the relevant data from your booklet to the spreadsheet E3data, in the
worksheet E3-data.
♣ Use the spreadsheet to calculate:
♣ V1 +V2 — to check that it is roughly constant,
♣ 1/T , R and ln(R).
♣ A basic plot of ln(R) against 1/T is produced in the worksheet E3-plot. You will
have to adjust the following:
♣ the chart title;
♣ the ranges of the x- and y-axes;
♣ the numbers of decimal places used in the values on the axes.
♣ Check to see if the model for the resistance of silicon (equation 3.1) is supported
by your measurements.
♣ Your data should span the intrinsic region of the semiconductor — you may wish
to ask a demonstrator what this means. At high values of 1/T the extrinsic region
may be entered, and the plot may become non-linear.
♣ Print out E3-data and your graph and stick them in your booklet.
♣ Replot your data using only those values in the intrinsic region as follows:
♣ create a new worksheet;
♣ select the plot from the worksheet E3-plot, right click and copy. Go to the
new worksheet, right click on a cell near the top left and paste in a copy of
the plot;

8
 ♣ right click on one of the data points in the new plot and choose select data.
Edit the chart data range to use the rows that correspond to your intrinsic
region;
♣ update the chart title and check the axis ranges are appropriate.
♣ Fit a trendline and display the equation on your graph. Print out this graph and
stick it in your booklet.
♣ Use the data-analysis facility in Excel to carry out a linear regression analysis of
the data in the intrinsic region; record the value of the gradient and its error in
your booklet. Use this to find the value of the gap energy Eg expressed in electron
volts, together with an estimate of its error (1 eV = 1.602 × 10−19 J).
♣ Comment on what dominates the uncertainty in Eg . Does it depend on the scatter
of your points on the graph? Is it affected by the accuracy with which you can
measure the resistance of the 1.2 kΩ resistor? Why would it be unsatisfactory
to use a fixed resistor with a resistance either much larger or much smaller than
about 1.2 kΩ?
Transfer the relevant data into the spreadsheet and calculate V1 +V2 , 1/T , R and ln(R); comment
on the values of V1 +V2 . Having made appropriate adjustments to the Excel graph using all your
data, print it out and stick it in your booklet. Comment on whether all the points shown fit the
theory — which, if any, are outside the range for which equation E3.1 is valid? If necessary replot
the graph using only those data which fit the theory and fit the trendline. Use the data-analysis
facility to calculate the slope and the error in the slope; hence find the value of the gap energy,
Eg , and the error in this value. Is the exact value of the fixed resistor important? Why would it
be unsatisfactory to have it either much larger or much smaller than 1.2 kΩ?

3.4 Conclusions and tidying up

You have now finished the experiment.
♣ Tidy up your lab book, and write down some overall conclusions. Was the the-
oretical prediction of 3.1 verified — if so, give the temperature range over which
the linear relationship holds. Give your value of Eg and its error. Depending on
the conditions (e.g. the temperature, the purity of the silicon etc.) the value for the
gap energy in silicon is in the range 0.9 − 1.2 eV: is your measurement consistent
with this?
♣ Leave your book in your group box.

Write down some final conclusions.

As usual, before you begin the next practical class, review the summary of “Specific
skills learnt in Experiment 3” and check that you have understood these. If you have
any doubts about these, ask a demonstrator at the start of the next class.

9
 Specific skills learnt in Experiment 3
• Understanding the role of hypothesis testing and parameter investigation
in experimental physics.
• Using basic apparatus for electrical measurements.
• Gaining an appreciation of non-invasive experimental methods.
• Assessing different experimental approaches and deciding which is the
best to follow.
• Gaining experience with logarithmic graphs.
• Designing a spreadsheet to carry out repetitive data analysis.
• Finding the gradient of a graph and its error using Excel.
• Working usefully as part of a team in a way that improves your efficiency.
• Planning an experiment so as to finish up in good time.

10
3.5 Thermocouple table for Ni90Cr10/Cu60Ni40
Variation of thermocouple output voltage with temperature for Ni90Cr10/Cu60Ni40
thermocouple. Data at ten degree intervals from 273 K taken from Kaye & Laby, 14th
edition, p.46, for the reference junctions kept at zero celsius. Data at intermediate
temperatures have been obtained by linear interpolation.

Thermal e.m.f./mV
Temperature/K +0 +1 +2 +3 +4 +5 +6 +7 +8 +9
270 0.00 0.06 0.12 0.18 0.24 0.30 0.35
280 0.41 0.47 0.53 0.59 0.65 0.71 0.77 0.83 0.89 0.95
290 1.01 1.07 1.13 1.19 1.25 1.31 1.37 1.44 1.50 1.56
300 1.62 1.68 1.74 1.80 1.86 1.92 1.99 2.05 2.11 2.17
310 2.23 2.30 2.36 2.42 2.48 2.54 2.61 2.67 2.73 2.80
320 2.86 2.92 2.98 3.05 3.11 3.17 3.24 3.30 3.37 3.43
330 3.49 3.56 3.62 3.68 3.75 3.81 3.88 3.94 4.01 4.07
340 4.14 4.20 4.26 4.33 4.39 4.46 4.53 4.59 4.66 4.72
350 4.79 4.85 4.92 4.98 5.05 5.12 5.18 5.25 5.31 5.38
360 5.45 5.51 5.58 5.65 5.71 5.78 5.85 5.91 5.98 6.05
370 6.12 6.18 6.25 6.32 6.38 6.45 6.52 6.59 6.66 6.72
380 6.79 6.86 6.93 7.00 7.06 7.13 7.20 7.27 7.34 7.41
390 7.48 7.55 7.61 7.68 7.75 7.82 7.89 7.96 8.03 8.10
400 8.17 8.24 8.31 8.38 8.45 8.52 8.59 8.66 8.73 8.80
410 8.87 8.94 9.01 9.08 9.15 9.22 9.29 9.36 9.43 9.50
420 9.57 9.65 9.72 9.79 9.86 9.93 10.00 10.07 10.14 10.22
430 10.29 10.36 10.43 10.50 10.57 10.65 10.72 10.79 10.86 10.93
440 11.01 11.08 11.15 11.22 11.29 11.37 11.44 11.51 11.59 11.66
450 11.73 11.80 11.88 11.95 12.02 12.10 12.17 12.24 12.32 12.39
460 12.46 12.53 12.61 12.68 12.75 12.83 12.90 12.98 13.05 13.12
470 13.20 13.27 13.35 13.42 13.49 13.57 13.64 13.72 13.79 13.86
480 13.94 14.01 14.09 14.16 14.24 14.31 14.39 14.46 14.54 14.61
490 14.68 14.76 14.83 14.91 14.98 15.06 15.13 15.21 15.29 15.36
500 15.44 15.51 15.59 15.66 15.74 15.81 15.89 15.96 16.04 16.11
510 16.19 16.27 16.34 16.42 16.49 16.57 16.65 16.72 16.80 16.87
520 16.95 17.03 17.10 17.18 17.25 17.33 17.41 17.48 17.56 17.64
530 17.71 17.79 17.87 17.94 18.02 18.10 18.17 18.25 18.33 18.40
540 18.48 18.56 18.63 18.71 18.79 18.86 18.94 19.02 19.10 19.17
550 19.25 19.33 19.40 19.48 19.56 19.64 19.71 19.79 19.87 19.95
560 20.02 20.10 20.18 20.26 20.33 20.41 20.49 20.57 20.64 20.72
570 20.80 20.88 20.96 21.03 21.11 21.19 21.27 21.35 21.42 21.50
580 21.58 21.66 21.74 21.81

11
3.6 Thermal excitation in a semiconductor – Introductory exercise
(You are expected to carry out an introductory exercise before each experiment — at
the start of the experiment your demonstrator will check that you have done so.)
A simple way to measure the resistance R of the thermistor is to compare it to a standard
resistance in series. In this exercise you consider the theory behind this and examine
the effect of the resistance of the voltmeters on your measurements.
a) For the circuit in figure 3.6, show that the resistance R of the thermistor is given by
V1
R = Rs . (3.4)
V2
where Rs is a standard resistor.
standard resistor thermistor
I I
Rs R

V2 V1
Figure 3.6: The thermistor R and the standard resistor Rs in series.

b) In practice, to measure V1 and V2 , voltmeters, both with resistance r, are connected

across R and Rs . To investigate the effect they might have on the estimate of R, consider
the following. A voltmeter with a resistance r is connected across the standard resistor
as shown in figure 3.7. Write down an expression for the resistance of the parallel
combination and hence show that the resistance, R, of the thermistor is now given by
Rs r V1
R= . (3.5)
(Rs + r) V2

In this experiment Rs = 1.20 kΩ and r = 10 MΩ. Given that the values of V1 and V2
are only accurate to about 1 part in a 1000, does the resistance of the voltmeter have a
significant effect on the estimate of R?

standard resistor
Rs thermistor
I I
R
r
V1
voltmeter
V2
Figure 3.7: A voltmeter, with resistance r, connected across the standard resistor.

12