Вы находитесь на странице: 1из 11

EECS 501

Submitted by:

Manas Bhatnagar – Team lead


Brian Fitzpatrick – Assembly lead
Wenshuai Hou – Test lead

[POWER SUPPLY WITH LINEAR REGULATOR]


Introduction:

This report presents suitable information and data to design, build, and test a linear regulated
power supply. Power supplies are an important component in the fabrication of electronic systems since
most electronic devices need to be driven by a constant power supply to perform correctly, thus it is vital
to develop a good understanding of their operation. In order to complete this task the AC voltage must be
rectified using the RS402L voltage rectifier. Rectifying the signal will result in a DC signal. The signal is
then to be filtered to reduce the voltage ripple. Finally, the signal will be regulated using the LM723
voltage regulator. This report will take a detailed look at how each component operates and how it can be
utilized to provide the desired output voltage. Finally, proper testing and results will confirm that the
required performance specifications have been met.

Methodology:

I. Project Deliverables: The power supply being discussed in this report will meet the following
specifications:
a. Output voltage of 19 V (within 500 mV) over a range of 100 mA to 1 A of load
current.
b. Over-current protection which limits the load current to1.25 A.
II. Engineering Design:

The first step in designing the linear regulated power supply is to drop the AC voltage provided
by the wall socket from 120 V to 26 V (r.m.s values). This can be done with the aid of a low-voltage
transformer. When powered on, the transformer box has an initial current surge which can result in a
blown fuse. To help prevent this problem a 0.68 Ω power resistor should be inserted in series with the
transformer. This resistance value was deemed suitable by experimentation. Then, to rectify the AC
voltage to having only positive range, the RS402L bridge rectifier should be utilized. Note that the AC
and DC will have their own grounds. The figure below shows how the bridge rectifier transforms the
input voltage and also how it can be designed using diodes.

Figure 1: Functioning of the RS402L rectifier. [1]

1
Once this DC voltage is achieved it is necessary to filter it in order to reduce the voltage ripple on
it. To do this a capacitive filter should be placed between the DC positive and negative terminals. As the
rectified voltage increases, the capacitor get charged and as it begins to drop, the capacitor slowly
discharges through the load, thereby providing a smoother output. For the prototype design, two 1000 µF
capacitors in parallel were experimentally found to reduce the ripple voltage from 1.45 V to 0.475 V. The
figure below compares the output voltage with and without the capacitive filtering.

Figure 2: Capacitive smoothening. [1]

This rectified signal is then fed to the LM723 integrated voltage regulator with the TIP-31
external series pass transistor. The general layout of the power supply circuit from this point onwards can
be given by Figure 3. The sizing of resistors R1, R2 and RSC (refer Figure 3) is discussed later in the
report.

Figure 3: Circuit showing LM723 and TIP-31 arrangement. [1]

The LM723 consists of an error amplifying op-amp which amplifies the difference between its
two input terminals. It also has an internal series pass transistor which functions as a voltage follower by
providing larger current, usually hundred times the current flowing out of the error amplifier, to drive the
load. The figure shown below is the equivalent circuit model of LM723 voltage regulator.

2
Figure 4: Equivalent circuit of the LM723 voltage regulator. [1]

The resistors R1 and R2, shown in Figure 3, form a negative feedback that determines the value
of the output voltage. The data sheet of the LM723 voltage regulator shows that if the output voltage is
higher than the reference voltage, then R1 and R2 can be calculated using

(1)

Given that VREF is 7.15 V, the ratio of R1 and R2 should be 5:3. Since R1 and R2 are voltage
feedback, the current flowing through them is relatively small, thus R1 and R2 should be made large. For
the prototype power supply, they were chosen to be nearly 10 kΩ and 6 kΩ respectively, which satisfies
the 5:3 ratio.

An excessively large output current can be harmful for both the power supply circuit and the load
circuit, thus a current-limiting protection is needed in most cases. This current-limiting protection can be
easily achieved by using LM723. It can be seen in Figure 4 that a „Current Limiter‟ is attached to the base
of the series pass transistor. This current limiter acts as a voltage-controlled switch. As long as the voltage
drop between current sense pin and current limit pin is greater than base-emitter voltage, which is about
0.65 V, part of the regulator output current will flow through limiter instead of the pass transistor,
therefore the transistor output current is limited. From Figure 3 it is evident that RSC monitors the current
flowing out of the external transistor. This makes it function like a negative feedback to control the
current limiter. The following formula can be used to calculate the value of Rsc.

(2)

As per the specifications listed under „Project Deliverables‟, Ilimit and Vsense are 1.25 A and 0.65 V
respectively. Thus the RSC value can be calculated as 0.52 Ω.

Oscillation can easily occur in an uncompensanted amplifier like the LM723, thus a 100 pF
capacitor is used for frequency compensation to eliminate the high frequency component of the output

3
voltage. It exhibits low resistance to high frequency signals, 10 MHz – 100 MHz, so that this noise goes
from the capacitor into the regulator as compensation.

Once R1 and R2 are properly sized, the LM723 should output nearly 19 V, however its maximum
output current as defined by it‟s datasheet is only 150 mA. Therefore, the external transistor TIP-31 is
required to increase the output current. The external pass transistor is a voltage follower, it amplifies the
current flowing into the base by a factor of „β‟, which is typically equal to 100 for transistors. Thus the
maximum output current of circuit increases to (150 mA * 100) = 15 A. In sourcing this current, the
transistor absorbs a large amount of energy due to large voltage drop and current flow. This energy will
manifest itself as heat and thus proper thermal management technique is needed to ensure safe operation.

The data sheet of TIP31C sets the maximum operating junction temperature to 150 °C. When the
collector input voltage is 34 V and the output voltage is 19 V at maximum specified load current (1 A),
the junction to ambient thermal resistance,  ja is given by

T j  Ta 150  25 (3)
 ja max    8.33 C / W
P (34  19)  1

In the above calculation, ambient temperature is assumed to be 25 °C and „P‟ is the power
absorbed by the TIP31C. Knowing that junction-case thermal resistance  jc  3.125 C / W , the

maximum sink thermal resistance can be determined as:


(4)
 sa   ja   jc   cs  8.33  3.125  0  5.21 C / W

This implies that the power supply must be designed with a heat sink in which  sa less than 5.21

°C/W. The TO-220 black anodize heat sink which has a specified thermal resistance of 5 °C/W was found
suitable for use in the prototype [2].

The expected transformation of the voltage signal through various stages of the power supply
design is expressed in the Figure 5, while Figure 6 showss the schematic used to model the prototype
power supply on Pspice

Figure 5: Changes in voltage signal through various stages of design. [1]

4
Figure 6: Pspice schematic.

III. Testing the power supply: The setup used for testing is displayed below.

Transformer

Incoming AC Multimeter
Oscilloscope signal
+
Oscilloscope
probe (10:1)

Linear Regulated
Power Supply
+
Resistor Box
+
- -

Banana cables

Figure 7: Testing setup.

To begin with, the decade resistor box should be connected as load across R1 and R2. This
provides a wide array of load resistances to be used during testing. To properly measure the load current a
multimeter should be connected in series with the resistor box. Next, to measure the output voltage and

5
study its characteristics, the oscilloscope should be connected in parallel across the load resistance. On the
oscilloscope, the „BW limit‟ and „Averaging‟ options should be selected and appropriate horizontal and
vertical axis should be used. This provides a graph of the output voltage which has much less ripple from
noise and is more accurate.

Results

The result of this design process is a linear regulated power supply which provides a constant
supply of nearly 19 V, over a range of 100 mA to 1 A load current. Furthermore, this power supply can
offer over-current protection by limiting the output current to 1.25 A. The performance of the prototype
over the specified current range and its over-current protection can be seen in the data presented below.

Table 1: Output voltage over a range of load current.

Output Voltage Load Current Load Resistance


(V) (A) (Ohm)
19.023 0.00 1.00E+06
19.019 0.10 200
19.023 0.20 91
19.027 0.30 62
19.025 0.40 47
19.025 0.50 38
19.025 0.60 31
19.027 0.70 27
19.025 0.82 23
19.025 0.89 21
19.022 1.00 19
19.016 1.12 17
18.990 1.18 16
16.240 1.23 13
12.630 1.24 10

It can be seen from the above table that the output voltage of the prototype remains within the
required voltage range of 19 V ± 0.5 V over the range of 100 mA to 1 A. Over this range of current, a
maximum percent deviation of 0.142 % is registered from the target output voltage of 19 V. It can also be
observed that the over-current protection specification is met since increasingly larger voltage drops are
registered in the output voltage as load current approaches 1.25 A.

The oscilloscope readings of the AC voltage signal from the transformer, the rectified voltage
signal and output voltage at 100 mA and 1 A of load current are as follows. These readings are in
agreement to those expected in the „Methodology‟ section (refer Figure 5).

6
Figure 7: AC voltage from transformer. Figure 8: Rectified voltage signal.

Figure 9: Output voltage at 100 mA of load curent. Figure 10: Output voltage at 1 A of load current.

A graph summarizing the details of Table 1 can be plotted as follows to better understand the
behavior of the prototype linear regulated power supply. A trendline is added for convenience.

Output Voltage vs. Load Current


22
20
Output Voltage (V)

18
16
14
12
10
8
6
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
Load Current (A)

Figure 11: Output voltage vs. load current

7
It can be observed from Figure 7 through Figure10 that although the output voltage was much
less oscillatory than the incoming AC voltage or the rectified voltage signal, it did have a ripple
associated with it. The peak-to-peak ripple on the output voltage at load current of 100 mA was measured
by the oscilloscope to be 47 mV. The voltage ripple shape can be characterized as having a constant
region followed by a sudden dip and sudden rise; while operating at a dominant frequency of 120.5 Hz.

Finally the prototype power supply was also thermally tested. Using values provided in the data
sheets,  sa  5 C / W ,  jc  3.125 C / W and assuming ambient temperature as 25 °C, the junction

temperature and case temperature can be calculated. Under the approximation that TIP31C case
temperature is equal to heat sink temperature, the junction temperature at 0.1 A of load current is:
(5)
Tc  Ta  (Vc  Ve )   sa  25  (34  19)  0.1  5  32.5 C

therefore, T j  Tc  (Vc  Ve )   jc  32.5  (34  19)  0.1  3.125  37.12 C (6)

Similarly, junction temperature and case temperature at load current is 0.5 A and 1 A was
calculated as follows.

Table 2: Thermal testing results.

Load Collector Theoretical Actual case Percent Theoretical Less than


current voltage case temperature Error junctions 150 °C ?
(A) (V) temperature (°C) temperature
(°C) (°C)
0.1 34.0 32.5 29.9 8.0% 37.1 Yes
0.5 31.5 56.3 52.5 6.7% 85.9 Yes
1.0 28.9 74.5 72.0 3.4% 146.8 Yes

Conclusions

- Internal resistance of the power supply can be calculated by solving a voltage division equation on the
Thevenin equivalent circuit below.

RT
|

|
VOC VL RL
|

|
Figure 12: Thevenin equivalent circuit.

8
(7)

Using equation (3) and Table 1 for values of RL ,VL and VOC the Thevenin equivalent internal
resistance can be calculated as below:

Table 3: Internal resistance calculation.

Internal
Output Load Resistance Resistance
Voltage (V) (Ohm) (Ohm)
19.023 1.00E+06
19.019 200 0.042
19.023 91 0.000
19.027 62 -0.013
19.025 47 -0.005
19.025 38 -0.004
19.025 31 -0.003
19.027 27 -0.006
19.025 23 -0.002
19.025 21 -0.002
19.022 19 0.001

By ignoring the negative values and comparing the positive values it can be concluded that the
internal resistance of the power supply is 0.04 Ω.

- Load Regulation, K, is the change in output voltage due to the output load varying from minimum to
maximum with all other factors held constant. It can be calculated by the using the following equation
and values from Table 1 to be 0.042 %.

(8)

- The efficiency of the power supply can be evaluated by measuring the power delivered to the load as
compared to the power delivered to the input of the power supply (measured using a P4400 electricity
usage monitor). The results of this analysis are summarized in the table below.

9
Table 4: Efficiency calculations.

Load Current Output Voltage Load Resistance Power to Power to Efficiency


(A) (V) (Ohm) Supply (W) Load (W) (%)
0.1 19.019 200 6.5 1.90 29.26
0.5 19.025 38 21.4 9.51 44.45
1.0 19.022 19 38.2 19.02 49.80

It can be inferred from Table 4 that a considerable part of the power is inefficiently lost in the
form of heat dissipated by the TIP31C.

Recommendations

When assembling the power supply, care should be taken to spatially isolate the TIP31C and the
heat sink from other circuit components. This would ensure good conduction to the ambience and also
reduce the risk of heat to other circuit components.

It is also recommended to use a capacitor in parallel with the load to further reduce any high
frequency noise. A suitable capacitor value can be calculated using the formula for capacitive impedance.

Also, if a potentiometer is used for either R1 or R2 or both, the resultant power supply would be
able to output a range of output voltages by changing the ratio of R1 and R2. This would make the power
supply suitable for more applications.

The efficiency of the power supply can be improved by reducing the power dissipation in the
TIP31C. This can be achieved if operating specifications can be met while the TIP31C is functioning in
saturation or cutoff mode, thereby reducing the power dissipated.

References

[1] “Linear voltage regulator”

https://www.cresis.ku.edu/~callen/501/lab1.htm

[2] “Aavid thermalloy Part number 533402B02552G”

http://www.aavidthermalloy.com/cgi-bin/stdisp_print.pl?Pnum=533402b02552g

10

Вам также может понравиться