Title: Method and apparatus for operating a burn-in board to achieve lower equilibrium temperature and to minimize thermal runaway
Abstract: The present invention is a novel method and apparatus for burn-in testing an electronic device. A device under test (DUT) is attached to a burn-in board (BIB), a thermally conductive sheet is placed atop the DUT, the BIB is placed in an environmentally-controlled burn-in oven, and current is applied to the DUT. A test signal can be sent to the DUT and data received from the DUT to determine whether the DUT is working properly. Aluminum, copper, or any material as thermally conductive as aluminum may be used in the sheet. The sheet may have top or bottom surface areas that conduct more heat away from the device and into the ambient environment than that conducted by a flat sheet. The sheet may be applied to a plurality of DUTs simultaneously. Further, a second thermally conductive sheet may be located underneath the BIB and in thermal contact with it.
Patent Number: 6,982,566 Issued on 01/03/2006 to Alam, et al.
| Inventors:
|
Alam; Mohammed (Milpitas, CA);
Hata; William Y. (Saratoga, CA)
|
| Assignee:
|
Altera Corporation (San Jose, CA)
|
| Appl. No.:
|
816773 |
| Filed:
|
April 1, 2004 |
| Current U.S. Class: |
324/760; 324/765 |
| Current Intern'l Class: |
G01R 31/28 (20060101) |
| Field of Search: |
324/754,760,765,762,761
165/802
438/14-18,323
257/724
|
References Cited [Referenced By]
U.S. Patent Documents
Primary Examiner: Nguyen; Vinh P.
Attorney, Agent or Firm: Morgan, Lewis & Bockius LLP
Claims
What is claimed is:
1. A method of burn-in testing an electronic device, comprising:
attaching the device to a burn-in board;
placing a first thermally conductive sheet atop the device such that the thermally
conductive sheet contacts the device;
placing a second thermally conductive sheet beneath the burn-in board and separated
therefrom by an electrically insulating but thermally conductive sheet;
inserting the burn-in board with the device and the thermally conductive sheets
into a chamber, wherein the environment within the chamber is controllable;
applying current to the device; and
controlling the environment within the chamber.
2. The method of claim 1, wherein at least one of the thermally conductive sheets
has at least the thermal conductivity of aluminum.
3. The method of claim 1, wherein at least one of the thermally conductive sheets
is composed of aluminum.
4. The method of claim 1, wherein at least one of the thermally conductive sheets
is composed of copper.
5. The method of claim 1, wherein the first thermally conductive sheet has dimensions
configured to have a top surface area greater than that of flat surface.
6. The method of claim 1, wherein a portion of the first thermally conductive
sheet that is in contact with the device is configured to have a greater contact
area with the device than that of a flat surface.
7. The method of claim 1, further comprising:
sending at least one test signal to the device;
receiving data from the device; and
analyzing the data received from the device.
8. The method of claim 1, wherein the electrically insulating but thermally conductive
sheet is composed of silicon rubber impregnated with aluminum oxide.
9. The method of claim 1, wherein the device is attached to the burn-in board
via a socket that includes an electrically insulating but thermally conductive
slug that contacts both the bottom surface of the device and the top surface of
the burn-in board.
10. The method of claim 9, wherein the slug is composed of silicon rubber impregnated
with aluminum oxide.
11. The method of claim 1, wherein a plurality of electronic devices are burn-in
tested simultaneously.
12. A method of burn-in testing a plurality of electronic devices, comprising:
attaching the plurality of devices to a burn-in board;
placing a first thermally conductive sheet atop the plurality of devices such
that the thermally conductive sheet contacts the devices;
placing a second thermally conductive sheet beneath the burn-in board and separated
therefrom by an electrically insulating but thermally conductive sheet;
inserting the burn-in board with the plurality of devices and the thermally conductive
sheet into a chamber, wherein the environment within the chamber is controllable;
applying current to each of the devices; and
controlling the environment within the chamber.
13. The method of claim 12, wherein the device is attached to the burn-in board
via a socket that includes an electrically insulating but thermally conductive
slug that contacts both the bottom surface of the at least one of the devices and
the top surface of the burn-in board.
14. An apparatus for testing an electronic device, comprising:
a burn-in board to which the device may be attached;
a first thermally conductive sheet that may be positioned atop the device such
that the thermally conductive sheet contacts the device;
a second thermally conductive sheet located beneath the burn-in board and separated
therefrom by an electrically insulating but thermally conductive sheet;
a controlled-environment chamber; and
a current source that applies current to the device.
15. The apparatus of claim 14, wherein at least one of the thermally conductive
sheets has at least the thermal conductivity of aluminum.
16. The apparatus of claim 14, wherein at least one of the thermally conductive
sheets is composed of aluminum.
17. The apparatus of claim 14, wherein at least one of the thermally conductive
sheet sheets is composed of copper.
18. The apparatus of claim 14, wherein the first thermally conductive sheet has
dimensions configured to have a top surface area greater than that of a flat surface.
19. The apparatus of claim 14, wherein a portion of the thermally conductive
sheet that is in contact with the device is configured to have a greater contact
area with the device than that of a flat surface.
20. The apparatus of claim 14, further comprising:
a test signal generator that sends at least one test signal to the device;
a test signal receiver that receives data from the device; and
a test signal analyzer that analyzes the data received from the device.
21. The apparatus of claim 14, wherein the electrically insulating but thermally
conductive sheet is composed of silicon rubber impregnated with aluminum oxide.
22. The apparatus of claim 14, further comprising:
a socket for attaching the device to the burn-in board; and
an electrically insulating but thermally conductive slug inserted through the
socket such that the slug contacts both the bottom surface of the device and the
top surface of the burn-in board.
23. The apparatus of claim 22, wherein the slug is composed of silicon rubber
impregnated with aluminum oxide.
24. The apparatus of claim 14, wherein the burn-in-board is configured so that
a plurality of electronic devices may be mounted on the burn-in-board, and the
first thermally conductive sheet is configured so that the first thermally conductive
sheet contacts the plurality of electronic devices.
25. The apparatus of claim 14, further comprising at least one device for biasing
the first thermally conductive sheet against the electronic device.
26. An apparatus for testing a plurality of electronic devices, comprising:
a burn-in board to which the plurality of devices is attached;
a first thermally conductive sheet that may be positioned atop the plurality
of devices such that the thermally conductive sheet contacts the devices;
a second thermally conductive sheet located beneath the burn-in board and separated
therefrom by an electrically insulating but thermally conductive sheet:
a controlled-environment chamber; and
a current source that applies current to the devices.
27. The apparatus of claim 26, further comprising:
sockets for attaching the devices to the burn-in board; and
electrically insulating but thermally conductive slugs inserted through the sockets
such that a slug contacts both the bottom surface of the device mounted in the
socket and the top surface of the burn-in board.
Description
The present invention relates generally to the testing of electronic devices
in a burn-in process, and particularly to lowering the equilibrium temperature
and minimizing occurrences of thermal runaway of electronic devices being tested
on a burn-in board.
BACKGROUND OF THE INVENTION
Burn-in testing is a process used to screen early failures in electronic
devices (e.g., semiconductor and integrated circuit devices) by operating the devices
at elevated temperatures and elevated voltages over a period of time. This is accomplished
by putting the devices under test (DUTs) in sockets on a printed circuit board
(PCB) designed for such testing. These PCBs are usually referred to as burn-in
boards (BIBs). BIBs are placed in an environmental chamber in a burn-in oven where
they are connected to a current source for testing the operation of the DUTs. Basic
burn-in testing may include providing only a clock signal to the DUTs to watch
for the DUTs' responses thereto. Dynamic burn-in testing is a more sophisticated
form of burn-in testing in that the burn-in test system has the additional capability
to provide input stimuli to the DUTs. In such dynamic burn-in testing, clock and
data signals exercise the device.
During burn-in, an electronic device is subjected thermally to both negative
and positive feedbacks that affect the device's equilibrium temperature. The negative
feedbacks are all parameters that cause an electronic device to cool down. These
include the ambient temperature within an environmental chamber, if that ambient
temperature is below the temperature of the electronic device, and the airflow
in the environmental chamber.
Positive feedbacks include all parameters that cause an electronic device
to heat up. One such parameter is thermal runaway. When a current source is connected
to the BIBs with the DUTs in an environmental chamber during burn-in, the DUTs
draw static current from the current source. "Static current" is the current that
an electronic device draws when the device is turned on, but not in use. The amount
of static current drawn by a device depends on device process, power up voltage,
and temperature of the device. In addition to the static current, DUTs will draw
dynamic current, if the burn-in is dynamic. "Dynamic current" is the current that
an electronic device draws, in addition to static current, when the device is in
use. The amount of dynamic current drawn by a device depends on how much of the
total device is in use and how fast (clock frequency) it is being used. As the
DUTs draw current from the current source, the DUTs produce heat, increasing the
DUTs' temperature, which in turn causes the DUTs to draw more current. More current
produces more heat and raises the temperature of the DUTs, causing the DUTs to
draw still more current. This cycle, known as thermal runaway, can result in the
temperature reaching the melting temperature of the DUTs. Thermal runaway results
not only in the melting of the DUTs, but also the sockets in which the DUTs were
attached during burn-in.
To minimize the risk of thermal runaway, burn-in testing has often been preceded
by the sorting of the DUTs into groups of low, medium, and high static current
devices. Each BIB is then used to burn-in only devices from a single group of devices
at a temperature that minimizes the likelihood of thermal runaway for that device
group. A drawback to this approach is that a device having a low static current
may have reliability specifications that require a chamber environmental temperature
that is higher than that which is used to test the device. This means that some
devices cannot be tested at the temperature that is dictated by their reliability
specifications. Another drawback to this approach is that it requires the additional
test step of sorting the DUTs by static current level, a step involving a considerable
amount of man and machine time that increases the production cost of the device.
SUMMARY OF THE INVENTION
The present invention addresses the aforementioned problems by a novel method
and apparatus for burn-in testing an electronic device. The present invention increases
the negative thermal feedback that works against positive thermal feed back (thermal
runway) in order to lower the equilibrium temperature of electronic devices during burn-in.
In addition to the usual burn-in set up (BIB, sockets, DUTs), a thermally conductive
sheet is placed atop the DUTs. This sheet helps to conduct heat away from the DUTs
and into the environment of the chamber during burn-in. Since the sheet is thermally
conductive and has a larger surface area than the total surface area of the DUTs
alone, each DUT in contact with the sheet achieves a lower equilibrium temperature
during burn-in than without the sheet. In particular, without the sheet, the equilibrium
temperature of each device is typically 15-60° C. higher than the ambient
chamber temperature. By constructing the sheet using an appropriate material, the
equilibrium temperature of each DUT can be lowered to no more than 10-15°
C. above the ambient chamber temperature. This lowering of each device's equilibrium
temperature helps to minimize thermal runaway.
In addition to lowering the equilibrium temperature of each DUT, the thermally
conductive sheet also serves to equalize all the devices' different equilibrium
temperatures. Without the sheet, each DUT typically achieves an equilibrium temperature
different from the other devices on the same BIB in the same burn-in oven. Because
the sheet is in contact with each of the DUTs, it dissipates more heat from a DUT
with a high static current and less heat from a DUT with a low static current.
As a result, all of the DUTs upon which the sheet sits achieve the same or relatively
similar equilibrium temperature. Because all the DUTs achieve the same equilibrium
temperature, DUTs having significantly different static currents can be tested
on a single BIB and at a single chamber temperature, thus making it unnecessary
to sort the devices into different static current device groups. By removing the
requirement for grouping devices based upon their static current, the sheet makes
it possible to test together devices of different static current groupings that
require the same temperature for reliability specifications. This allows for more
efficient and accurate testing of electronic devices.
A second thermally conductive sheet may be positioned beneath the BIB such that
the second thermally conductive sheet is separated from the BIB by an electrically
insulating but thermally conductive sheet. The conduction of heat from the DUT
to this second thermally conductive sheet can be enhanced if the device is attached
to the BIB by a socket and an electrically insulating but thermally conductive
slug is inserted through the socket such that the slug contacts both the bottom
surface of the device and the top surface of the BIB.
BRIEF DESCRIPTION OF THE DRAWINGS
Additional objects and features of the invention will be more readily
apparent from the following detailed description and appended claims when taken
in conjunction with the drawings, in which:
FIG. 1 is a schematic view of a conventional burn-in oven and burn-in system controller.
FIG. 2 is a top view of a typical BIB.
FIG. 3 is a top view of a typical BIB showing channel connections.
FIG. 4 is a schematic diagram of a system controller computer and program modules
and instructions.
FIG. 5 is a top view of a conventional BIB showing spring mechanisms that are
applying pressure to DUTs attached to the BIB's sockets.
FIG. 6 is a flowchart describing one embodiment of the present invention.
FIG. 7 is a perspective view of DUTs being attached to a BIB's sockets.
FIG. 8 is a perspective view of a thermally conductive sheet being positioned
atop DUTs on a BIB.
FIG. 9 is a cross-sectional view of a thermally conductive sheet positioned
atop DUTs that are attached to the sockets of a BIB.
FIG. 10 is a table showing the thermal conductivity of various substances.
FIG. 11 is a perspective view of an embodiment of the present invention in which
the surface area of the thermally conductive sheet is greater than that of a flat sheet.
FIG. 12 is a perspective view of another embodiment of the present invention
in which the area of contact between the thermally conductive sheet and the DUTs
is greater than that of a flat sheet.
FIG. 13 is a cross-sectional view of an embodiment using two thermally conductive
sheets, one positioned atop the DUTs, and the other positioned beneath the BIB.
FIG. 14 is a cross-sectional view of another embodiment using two thermally
conductive sheets.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
As shown in FIG. 1, a conventional burn-in system comprises two basic parts: a
burn-in oven
102 and a system controller
104. Burn-in oven
102
includes environmental chamber
106 for holding one or more BIBs
108
during testing. BIBs
108 each hold one or more DUTs
110 in the BIB's
sockets. Environmental chamber
106 is connected to environmental electronics
112 that control the heating of chamber
106 and any other environmental
test conditions to which BIBs
108 are subjected. Such systems as the ATX
and ATS-12000 systems manufactured by Aehr Test Systems of Menlo Park, Calif.,
are examples of the types of burn-in systems discussed herein.
System controller
104 comprises a computer
114, a chamber controller
116, a driver board controller
120, driver boards
122, and
a current source
126. Burn-in oven
102 is connected to system controller
104 by at least two sets of connection wires: a set
118 from chamber
controller
116 to environmental electronics
112, and a set
124
between driver boards
122 and BIBs
108. Computer
114 includes
software programs, modules, and data for controlling the environmental test parameters
of environmental chamber
106 using chamber controller
116 and environmental
electronics
112.
Each BIB
108 is connected to a driver board
122 via a separate
set of wires
124. Driver boards
122 provide current from current
source
126 and test signals to the BIBs
108 for testing DUTs
110
thereon. Driver boards
122 also receive back through wires
124 any
output or response signals from DUTs
110. Driver boards
122 in turn
are coupled to driver board controller
120 that provides control signals
to driver boards
122. For instance, a pattern generator (not shown) may
be found either on driver boards
122 or in driver board controller
120.
The pattern generator may provide clock signals and test data signals to driver
boards
122 to be transmitted to DUTs
110 for testing. System controller
computer
114 includes software programs, modules, and data to send instructions
to driver board controller
120 and on to driver boards
122 to initiate
and run the testing of DUTs
110. Additionally, computer
114 has memory
for storing test data as well as data returned back from DUTs
110 during
testing. Data returned back from DUTs
110 may then be analyzed and evaluated
in computer
114.
FIG. 2 shows an example of a BIB
108 with edge connector
202 and
device test sockets
204. Device test sockets
204 are coupled to a
PCB that serves as BIB
108. For an example of device test sockets used with
flip-chip type devices, see U.S. Pat. No. 5,419,710, to Pfaff, dated May 30, 1999,
and entitled "Mounting Apparatus for Ball Grid Array Device," which is incorporated
herein by reference. All power, test input, and test output signals are typically
sent from system controller
104 to BIBs
108 through edge connector
202. FIG. 3 shows a conventional BIB
108 including various wires
or channels employed in testing devices. DUTs
110 are coupled to BIB
108
by means of device test sockets such as device test sockets
204. Edge connector
202 provides the connection through which power and signals are applied
to and output signals received from DUTs
110. Clock or input signal channel
308 is coupled to an input pin of each DUT
110. When DUTs
110
are being tested, an output pin for each DUT
110 is connected to a separate
output channel
310. Thus, each DUT
110 drives an output signal on
a separate output channel
310.
FIG. 4 shows an example of a system controller computer
114. Illustratively,
system controller computer
114 comprises CPU
402 connected to user/operator
interface
404 and memory
406. Memory
406 includes operating
system
408, file handling system
410, and test control application
module
412. Test control application module
412 sends instructions
to chamber controller
116, which in turn controls the chamber environment
through environmental electronics
112. Test control application module
412
also sends instructions to driver board controller
120 to provide control
signals and/or current to BIBs
108 via driver boards
122.
FIG. 5 shows a typical BIB
108 that includes sockets
204 in which
DUTs
110 sit. DUTs
110 are held firmly to sockets
204 by spring
mechanisms
502, each of which includes a spring arm
504 that pushes
down on a DUT. During burn-in, current is applied to DUTs
110 and test signals
are sent to and received from DUTs
110 via edge connector
202. Each
DUT
110 on a single BIB
108 usually has a different static current
(i.e., Icc) value. Depending on that static current value and the chamber temperature,
each DUT
110 on the single BIB
108 will have a different equilibrium
temperature during burn-in. In prior art burn-in systems, some DUTs
110
run the risk of experiencing thermal runaway when their equilibrium temperatures
reach the DUTs'
110 melting points. To avoid this, it was necessary to sort
the DUTs into static current device groups, which presented the further problem
of some devices not being tested at the temperature that is dictated by their reliability specifications.
In a preferred embodiment of the present invention illustrated in FIG. 6, DUTs
need not be pre-sorted before being placed in the sockets of a single BIB. Rather,
after the unsorted DUTs are mounted in the socket of a BIB at step
610,
a thermally conductive sheet, serving as a heat pool, is placed at step
620
directly atop the DUTs such that the sheet contacts the DUTs. The entire set-up
(including the BIB with edge connector and sockets, attached DUTs, and the sheet)
is then placed in a burn-in oven at step
630, a current is applied to the
DUTS at step
640, the chamber's environment is controlled at step
650,
and the DUTs are tested. Further, the DUTs can undergo dynamic testing during which
test signals are sent to the DUTs at step
670, return data is received from
the DUTs at step
680, and the data is analyzed at step
690 to determine
whether any DUTs are malfunctioning under the chamber conditions.
FIGS. 7 and 8 show a preferred embodiment of apparatus of the present invention.
In FIG. 7, there is shown BIB
108 with sockets
204 and spring mechanisms
502 for each socket. The bottom row of DUTs
110 are shown being inserted
into their respective sockets
204, and the top row of DUTs
110 are
shown as having been inserted and attached to their respective sockets
204
with spring arms
504 of each respective spring mechanism
502 positioned
to hold attached DUTs
110 firmly in place. Then, referring to FIG. 8, a
thermally conductive sheet
802 is inserted between spring arms
504
of spring mechanisms
502 and DUTs
110. Spring arms
504 thus
apply pressure against the sheet
802 such that the sheet
802 makes
contact with DUTs
110. FIG. 9 is a cross-sectional view of thermally conductive
sheet
802 positioned on top of DUTs
110, which are attached to sockets
204 in BIB
108. Spring arms
504 push down on sheet
802,
causing sheet to contact DUTs
110. After sheet
802 is in place, DUTs
110 attached to sockets
204 of BIB
108 are placed into burn-in
oven
102, chamber
106 of which is controlled by system controller
computer
114 through chamber controller
116 and environmental electronics
112.
Since the efficiency of the thermally conductive sheet depends upon the conductivity
of the material used and the surface area of the sheet itself, other preferred
embodiments include the use of materials that are more thermally conductive than
others, and different shapes of the sheet. FIG. 10 shows a table of conductivity
for various materials. A preferred embodiment is the use of a thermally conductive
sheet where the sheet is made of aluminum or has at least the thermal conductivity
of aluminum. Another preferred embodiment is a sheet made of copper. Since copper
has almost twice the thermal conductivity of aluminum, copper is preferred although
it is generally more expensive than aluminum.
A specific preferred embodiment is a 12 inch by 2 inch by 1/32 inch copper sheet.
Previously, without the sheet, equilibrium temperatures typically ranged 15-60°
C. above the ambient temperature in the environmental chamber. Usually, if the
ambient temperature is 125° C., an additional 45° C. in the equilibrium
temperature will cause the DUTs to melt. With this particular preferred embodiment,
the copper sheet lowered the equilibrium temperature of all DUTs to no more than
10-15° C. above the ambient temperature.
In addition to varying the material for the thermally conductive sheet, one may
also vary the surface area of the sheet to change its efficiency. FIG. 11 illustrates
an embodiment of a sheet
1100 that increases the surface area (above that
offered by a flat-surface sheet) that is available to conduct heat into the ambient
environment of burn-in oven
102. Instead of a flat surface, the sheet
1100
has a crinkled or serrated top surface
1110 that has more surface area than
a flat surface. This greater surface area allows more heat to dissipate from the
crinkled surface than would dissipate from a sheet with a flat top surface.
An additional preferred embodiment is illustrated in FIG. 12 in which the portion
of sheet
1200 that contacts DUTs
110 is constructed to have a greater
contact area with DUTs
110 than a sheet that has a flat bottom surface.
This greater contact area with DUTs
110 causes the sheet to conduct more
heat away from DUTs
110 than a sheet with a flat bottom surface. Instead
of a flat bottom surface, the bottom surface
1210 of the sheet
1200
matches the surface shape of DUTs
110. This allows the bottom surface of
sheet
1200 to contact the top surface and edges of DUTs
110 instead
of limiting the contact to the top surface of DUTs
110.
FIG. 13, which includes all elements of FIG. 9, shows another embodiment using
dual thermally conductive sheets
802,
1320. Thermally conductive
sheet
1320 is positioned beneath BIB
108 and is separated from BIB
108 by an electrically insulating but thermally conductive sheet
1310.
In one embodiment, sheet
1310 is composed of silicon rubber impregnated
with aluminum oxide. Since sheet
1310 is electrically insulating but thermally
conductive, sheet
1310 isolates sheet
1320 from BIB
108 electrically
but not thermally. Sheet
1320 is thus able to conduct heat away from DUTs
110 through sockets
204, BIB
108, and sheet
1320. This
conduction of heat away from DUTs
110 serves as a negative feedback in addition
to the conduction of heat from DUTs
110 by sheet
802. This conduction
of heat is especially useful because the socket is thermally conductive. An example
of this is in Super Ball Grid Arrays (SBGA) and other peripheral ball packages
(packages in which all the device pins or balls are at the periphery of the package),
because all the balls (pins) of the DUT in a peripheral ball package are electrically
connected to the BIB through copper springs or wires that pass through the socket.
FIG. 14 illustrates a further embodiment in which electrically insulating but
thermally conductive slugs
1410 are inserted into sockets
204. Slugs
1410 contact the bottom surfaces of DUTs
110 as well as the top surface
of BIB
108. Since the thermal conductivity of slugs
1410 is generally
higher than that of sockets
204, more heat is conducted away from DUTs
110
using both slugs
1410 and sockets
204 in this embodiment than heat
conducted away from DUTs
110 using only sockets
204 (as in the previous
embodiment shown in FIG. 13). This embodiment is easily implemented in some SBGAs
that have a socket with a hollow core into which slug
1410 can be inserted.
It should be clear to those skilled in the art that the present invention may
apply to any burn-in testing or multiple device testing system, and to any devices
that have bus capabilities whether they are in flip-chip configuration or otherwise.
Multiple input signals, multiple output signals and more than one output enable
signal may be used with the embodiments of the invention disclosed herein. The
devices tested may have solder bumps or other electrical contacts. The methods
of the present invention may include delays between row tests, and rows or columns
or subsets of devices may be tested in any order. While two-dimensional embodiments
(i.e. row and column) have been disclosed, embodiments utilizing methods expanded
from those disclosed above are contemplated here as well. These embodiments may
be referred to as three- or four-dimensional embodiments. A three-dimensional embodiment
may include the use of multiple BIBs, each with its own thermally conductive sheet.
A four-dimensional embodiment may include the use of multiple BIBs, each with its
own thermally conductive sheet, and each BIB may be configured to test DUTs of
a certain physical size range.
While the present invention has been described with reference to a few specific
embodiments, the description is illustrative of the invention and is not to be
construed as limiting the invention. Various modifications may occur to those skilled
in the art without departing from the true spirit and scope of the invention as
defined by the appended claims.
*
