Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
SC Analyser2005
Semiconductor Device Tester
Michel Waleczek
230 Ω
1k
100k
100k
(RE + RSW)
PNP C
stands for a connection to +5 V via a
RE
C
VE = RE • β • IB
VE = (5V - VTH) •
RE RB RC RE
230 Ω
1k
100k
4/2005 - elektor electronics Personal Download for Berberena Gonzalez, Juan Carlos | copyright Elektor 35
783600
230 Ω
230 Ω
100k
B
mentioned three parameters by calcu-
D
lation. Of the three principal parame-
D
ters of this static model, which is pri-
VDS1 VDS2
marily used in Spice simulation, it
G G
turns out that only two are actually
VGS1 S
JFET N
VGS2 S
JFET N necessary here, since the third param-
eter (λ) has hardly any effect on the
RSW RSW
ultimate calculation (see inset).
65 Ω
65 Ω
A
In order to determine the values of
ID1 ID2
these two parameters, it’s necessary
RS RG RS
135 Ω
135 Ω
65 Ω
source and drain leads will not change combined with three decoupling directly to +5 V, since the microcon-
the information shown on the display capacitors reduces the voltage to troller has an automatic reset function.
of the SC Analyser 2005, but the exactly 5 V in order to power the The three control signals for the tran-
results of the calculation always corre- PIC16F876, the LCD (with backlight), sistor being tested come from RC4,
spond to indicated lead arrangement. and the three 74HC4502 multiplexers. RC5 and RC6 in Port C. These signals,
The display module communicates which have levels of 0 V or 5 V, are
with the microcontroller in 4-bit mode routed to the three test terminals via
Schematic diagram
via five Port C leads and lead RA5 of three analogue multiplexers in order to
The complete schematic diagram of Port A. The PIC16F876 is clocked at insert resistors with specific values
the circuit is shown in Figure 4. The approximately 1 MHz by the R10/C4 between the microcontroller outputs
circuit draws approximately 6 mA network. The PIC16F876 differs from and each of the test terminals. The
(except for the backlight, which needs the ‘classic’ 16F84 by having an on- resistance values are determined by
around 20 mA), and it is powered by a board A/D converter, which is used in selection signal pairs RB4 and RB5 for
9-V battery. A 78L05 voltage regulator this circuit. The Reset pin can be tied the right-hand signal (J1), RA2 and
Measuring FETs
with the Schichman–Hodges model
The static characteristics of a JFET can be represented by a B = 2·(VK – VGS2 ) [6]
set of formulas that express the drain current as a function of
If the operating point of the FET is in the saturated region, for-
the voltages VGS and VDS of the FET. There are two different
mulas (1) and [3] yield a second-order equation:
formulas, since the model distinguishes between the linear
region 0 ≤ VDS ≤ (VGS – VTH) and the saturated region VTH2 + B·VTH + C = 0
VDS ≥ (VGS – VTH), where VTH is the threshold voltage of the
We now have two solutions for VTH :
FET.
VTH1 = –B + √ [B2 – 4·(C ÷ 2)] [7]
The full model employs three parameters:β, λ, and VTH.
Parameter β is related to the saturation current and the thresh- and
old voltage, while parameter λ represents the channel-length
VTH2 = VTHSAT = –B – √ [B2 – 4·(C ÷ 2)] [8]
modulation factor and is neglected in our calculations (λ = 0).
The mathematical model that we use thus appears as follows: Only one of these solutions is physically possible, and this is
For the linear region, defined as 0 ≤ VDS ≤ (VGS – VTH), the determined by evaluating the two expressions. All that’s left to
following formula applies: do now is to determine V1 and V2 in order to figure out the
actual operating region of the FET in the second configura-
ID = β·VDS·[2(VGS – VTH) – VDS ]·(1 + λ·VDS ) tion:
For the saturation region, defined as VDS ≥ (VGS – VTH), the
V1 = VGS2 – VTHLIN – VDS2
following formula applies:
V2 = VGS2 – VTHSAT – VDS2
ID = β·(VGS – VTH )2·(1 + λ·VDS )
If V1 > 0 and V2 < 0, the FET is operating in the linear
As we have seen, in the first configuration the FET is in the
linear region with an adequately large IDSS , so the first for- region and VTH = VTHLIN .
mula becomes: If V1 < 0 and V2 > 0, the FET is operating in the saturated
ID1 = β·VDS1·[2·(VGS1 – VTH ) – VDS1 ][1] region and VTH = VTHSAT .
In the second configuration, the FET can be in the linear If the FET is operating in the saturated region, the IDSS param-
mode with: eter corresponds to the drain current for VGS = 0. For VDS ,
we take the value corresponding to the boundary of the satu-
ID2 = β·VDS2·[2·(VGS2 – VTH ) – VDS2 ][2]
ration region, which is VDS = VGS – VTH = –VTH .
or it can be in the saturated mode with:
Substituting for VDS in equation [2] or [3] yields:
ID2 = β·(VGS2 – VTH )2 [3]
IDSS = β·VTH2
In this manner, we obtain a set of two equations in which VTH
and β are the unknowns. Depending on the operating region The value of can be easily calculated by rearranging formula
of the FET, we must use either formula [2] or formula [3]. [1] as follows:
If we assume that the FET is operating in the linear region, β = ID1 ÷ { VDS1·[2·(VGS1 – VTH) – VDS1 ]}
formulas [1] and [2] yield: The resistance RDSON corresponds to the slope at the origin
VTHLIN = C – (VGS2 – VDS2 )2/(VDS2 – VK ) [4] of the characteristic curve:
where VDS = f(ID) if VGS = 0
VK = (ID2 ÷ ID1 )·VDS1 Finally, formula [1] yields the following results when VDS
and approaches 0 V:
4/2005 - elektor electronics Personal Download for Berberena Gonzalez, Juan Carlos | copyright Elektor 37
783600
IC1
+9V S1 D1 78L05 +5V +5V
POWER
1N4001
R13 R14 IC3
C1 C2 C3 C5 10
5k6
47k
0 0
9 4x 6
1 3 G4
10µ 100n 10µ 100n
25V 25V R1 1%
12
20 100k 0 MDX
14
5k6 1 13
15
R2 2
1 25 11 +5V
MCLR/VPP RB4 3
26 1
+5V RB5 0
5
IC2 R3 1% 1 3 JP1
9 17 2
OSC1 RC6 1k 2
10 4
16 16 16 C9 C4 OSC2 100Ω 3
2
IC3 IC4 IC5 AN0 R4 1% 74HC4052
8 7 8 7 8 7 100n 100p
C6
IC4
1n 10
0
9 4x 0 6
24 1 3 G4
RB3
27 R5 1%
RB6/PGC 12
LCD1 28 4 100k 0 MDX J1 T.U.T.
RB7/PGD RA2 14
23 5k6 1 13 J2
RB2 15
R6 2
11
15 3 J3
RC4 1
LC DISPLAY 0
5
K1 3 R7 1% 1 3
AN1 2
K A 1k 2
VDD
VSS
R/W
VO
RS
4
D0
D1
D2
D3
D4
D5
D6
D7
PIC16F876 C7
E
100Ω 3
16 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 R8 1% 74HC4052
1n
11
R15 RC0 IC5
12 10
RC1 0
33Ω
13 22 9 4x 0 6
RC2 RB1 1 3 G4
14 21
+5V R16 RC3 RB0 R9 1%
7 12
27Ω RA5 100k 0 MDX
18 16 14
RC7 RC5 5k6 1 13
R10 15
2
6 5 11
P1 RA4/T0CKI AN3 3
C10 1
C8 0
10k 5
R11 1% 1 3
10µ 16V 8 19 2
1n 1k 2
4
CONTRAST 100Ω 3
R12 1% 74HC4052
030451 - 11
RB2 for the centre signal (J2), and RB1 occupied space is used for calculating large for this application.
and RB2 for the left-hand signal (J3). parameter values for FETs. If you The display module can be fitted on
The voltages present on the test ter- would like to program your own micro- the copper side of the circuit board. To
minals are measured by the controller, you can download the nec- make it easy to connect the display,
PIC15F876 via analogue inputs AN0, essary hex code from the Elektor Elec- there is a single-row 16-way pinheader
AN1 and AN3. tronics website (www.elektor-electron- (male) on the circuit board, which
In order to compensate for the effects ics.co.uk). The file is found with the pdf mates with a corresponding 16-way
of the internal resistances of the ana- download for this article (April 2005). female connector on the display circuit
logue switches when making current For those of you without Internet board. Here we chose to use a modern
measurements, these measurements access it is also available on floppy type of display called ‘PLED’ (see the
are made using a second switch in disk from Reader Services, the order ‘OLED and PLED’ inset), but you can
each 74HC4052 instead of directly at code is 0304051-11. Naturally, you can also use any desired LCD module
the leads of the unknown transistor. also order a fully programmed micro- based on (or compatible with) the
The operation of this arrangement is controller from Reader Services (order Hitachi HD44780, although the pin
shown schematically in Figure 3 with no. 030451-41). arrangement may differ from the
resistor R9 connected in the circuit. arrangement for the display used here.
The internal resistance of the micro- Note that the pin arrangement of the
Construction
controller output, which is around display used here is rather unusual,
30 Ω, must also be taken into account. The PCB layout and component layout with pins 15 and 16 for the backlight
Finally, three 1-nF capacitors provide a are shown in Figure 5. Ensure that the being located next to pin 1.
certain amount of filtering for the IC sockets, electrolytic capacitors, 5-V Beside the test leads with alligator
measured signals. regulator and the four ICs are fitted the clips, there is also a special test circuit
right way around. board that is connected to the main cir-
The 1% resistors can be replaced by 5% cuit board and can be used to simplify
Software
metal-film resistors carefully selected testing SMD devices (diodes and tran-
The software is written entirely in using an accurate and reliable multi- sistors). Use three short, flexible leads
assembly language and fills a large meter. The analogue switches must be to connect the SMD holder to the main
chunk of the memory space of the 74HC types, since the internal resist- circuit board, and ensure that the
PIC16F876. Approximately half of the ance of the normal 4000 series is too proper lead sequence is maintained (as
+ - 030451-1
is designed to fit in a standard enclo-
C6 C9 C7 C8 sure with a 9-V battery compartment.
S1
R10
R12
R11
R2
R4
R3
R1
R6
R8
R7
R5
R9
D1
Calibration
C1
C2
3 After the tester is switched on, a wel-
IC3 IC4 IC5
come message appears and displays
C3 IC1 1 2 the software version (SC-Analyser 2005
J1 J2 C5 J3 2 Elektor Rev. 1.0e). If you don’t see this
R15 1-154030
JP1 ROTKELE )C(
14
1 3 message, try adjusting contrast control
16 15 1
IC2 P1 to improve the situation.
R14 LCD1 K1 The first thing you should do is to cali-
C4 brate the internal resistances of the
R13 P1 ROTKELE
analogue switches. If these values are
known as exactly as possible, the
measurements of the various transistor
parameters will be more accurate. If
ELEKTOR this calibration is not performed, a
default value of 65 Ω is used for each
switch. Each time the tester is
(C) ELEKTOR 1 3 switched on, the software checks to
030451-1
2 see whether the calibration has taken
1 2 place, and if it hasn’t, the message ‘Cal
error’ is displayed for approximately
3 one second.
The calibration procedure is easy and
runs automatically. To perform the pro-
cedure, fit a jumper in position JP1
while the tester is switched off and
short all three test leads together, and
then switch the tester on. The mes-
sage ‘Cal Remove jumper’ will be dis-
Figure 5. Track and component layouts for the associated printed circuit board. played on the tester. Now you should
SMD components can be conveniently tested using the small circuit board. remove the jumper, and the calibration
procedure will start. Three resistances
are measured one after the other, and
seen from above: contact 1 to J1, con- identifying the leads. The middle con- their values are displayed successively.
tact 2 (middle) to J2, and contact 3 to tact must always be connected to J2. Next, the message ‘Short RSH XX Ω’’
J3). Although this hardly matters for Naturally, you can also use flexible will appear. After this you can discon-
the measurements, it is important for leads with miniclips. The circuit board nect the three test leads from each
4/2005 - elektor electronics Personal Download for Berberena Gonzalez, Juan Carlos | copyright Elektor 39
783600
other, and the tester will automatically The first line shows the transistor type If the transistor is partially or com-
change to the test mode and display and the value of one of the characteris- pletely shorted, the associated leads
the message ‘No component *–*–*’. tic parameters. The second line shows are marked with an ‘X’ by the
Calibration can be performed when- the lead arrangement of the transistor SC Analyser 2005. A short is consid-
ever desired by repeating the above and, if relevant, the value of a second ered to be present if the resistance
procedure (switch off the power, fit the parameter. measured between two leads is less
jumper, and switch the power on than 50 Ω (Photo 3).
again). Bipolar transistor
The first line of the display shows the FETs
If you experience any problems, check polarity of the transistor (PNP or NPN), The values of the parameters VTH, IDSS
that the five wire bridges are in fact the type of semiconductor material (sil- and RDSON are shown approximately
properly fitted and soldered in place. icon or germanium), and the parame- every two seconds. The drain and
Also check to make sure that the supply ter HFE. The second line shows the source leads are determined arbitrarily
voltage is present at the IC sockets. lead arrangement of the transistor and by the SC Analyser 2005, but they do
the collector current. The current has a correspond to the configuration actu-
value of approximately 1.5 to 4 mA, ally used to determine the parameters.
Operation
depending on the current gain. The If you swap the drain and source leads
After the welcome message showing the SC Analyser 2005 displays the type of of the transistor under test (TUT), you
software version (SC-Analyser 2005 Elek- semiconductor material (‘Ge’ for ger- can obtain the values for the reverse
tor Rev. 1.0e), the display should show manium or ‘Si’ for silicon) according to configuration, but since field effect
the following information (Photo 1). the measured value of VBE (Photo 2). transistors have symmetric structures,
Organic LED (OLED) and polymeric LED (PLED) are recent Glass
Cathode
developments in LED technology. The OLED effect was dis-
covered in the early 1980s by Eastman Kodak, but it has Gluing
only recently been put to practical use in commercial appli- ITO anode
cations such as PDAs and MP3 players. The Kodak LS633 Glass
there shouldn’t be any large differ- shows the voltage and the current
ences in the measured values. through the diode via a resistance of 1
Considering the method used to make approximately 400 Ω. The maximum
the measurements, the accuracy of the current is thus limited to around 12 mA
values provided by the SC Analyser (Photo 8).
2005 is more than adequate. For IDSS it 2
is ±5 %, and for RDSON it is ±5 Ω. How- The second screen shows the same
ever, the measured value of the thresh- information, but this time with a resist-
old voltage can differ from the true ance of approximately 5.9 kΩ, which
value by 0.5 V for certain transistors, causes the maximum current to be lim- 3
although in general an accuracy of ited to approximately 800 µA (Photo 9).
50 mV is achieved. A particular type of
FET from one manufacturer may yield The third screen shows the current in
a value that is very close to the true the reverse direction and the test volt- 4
value, while the same type from a dif- age. The resolution is 100 nA
ferent manufacturer yields a value that (Photo 10).
differs by several hundred millivolts. In
practice, the accuracy is related to the SMDs 5
accuracy of the Schichman–Hodges You’ve probably been wondering how
model and reveals the limitations of to use the SMD portion of the tester.
this model (refer to the inset on this The leads of the component to be
subject). tested must make contact with the cor- 6
responding copper areas on the circuit
Threshold voltage board. SMD diodes must be connected
The threshold voltage is negative for between area 1 and area 3. A small
N-channel FETs and positive for P- plastic rod can be used to properly 7
channel FETs. The maximum value is press an SMD device against the cir-
limited to ±20 V by the software. The cuit board so that its leads make good
resolution is 10 mV for values up to contact.
±9.99 V and 100 mV for larger values 8
(Photo 4).
Conclusion
Saturation current The SC Analyser 2005 is a handy, eas-
The saturation current (with the gate ily constructed instrument with a 9
shorted to the source) is shown here. large number of features, which can
The value of the saturation current be a quite valuable aid to electronics
ranges from 0 to 99.9 mA. The resolu- hobbyists and professionals. A brief
tion is 10 µA for currents up to 10 mA list of its possible applications 10
and 100 µA for currents up to 100 mA includes searching for an equivalent
(Photo 5). type, testing device operation, sorting
devices, measuring unmarked transis-
Drain–source resistance tors, and simply determining the lead
This is the drain–source resistance arrangement without thumbing
when the FET is fully ‘on’ with through a data book. Time to warm up
VGS = 0 V. The measurement has a your soldering iron!
(030451-1)
About the
range of 0–999 Ω (Photo 6).
author
MOSFETs
The displayed voltage corresponds to The author was born in 1965 and
the threshold voltage of the MOSFET studied at the École Nationale
for a drain current of approximately Supérieure d’Électronique et de
2 mA. The measurement range is Internet links Radioélectricité de Grenoble
0–4.5 V, with a resolution of 10 mV. For (ENSERG) in France. He worked as
Author’s website:
proper identification of the device as a an engineer for seven years in the
www.mwinstruments.com
MOSFET, the gate leakage current research department of a large
must not exceed 0.5 µA (Photo 7). FAQ site: French manufacturer of weighing
www.mwinstruments.com/SCA2005/ equipment. Following this he
sca2005.html
Diodes changed to the French branch of
Diodes can be connected between the Author’s e-mail address: Hameg, where he holds the position
left-hand and right-hand terminals of m.waleczek@mwinstruments.com of Director of Development. Between
the tester. The SC Analyser 2005 indi- his travels to the other side of the
Microchip:
cates the positions of the anode and www.microchip.com/download/lit/ globe, he manages to find a bit of
cathode. Three screens are displayed pline/picmicro/families/16f87x/ time to design measurement instru-
in sequence at an interval of approxi- 30292c.pdf
ments for electronic hobbyists.
mately 2 seconds. The first screen
4/2005 - elektor electronics Personal Download for Berberena Gonzalez, Juan Carlos | copyright Elektor 41
3
2 1
2
030451-1
3 1 (C) ELEKTOR
ELEKTOR
non reflected
ELEKTOR
3 1 (C) ELEKTOR
030451-1
2
2 1
reflected