Note: Descriptions are shown in the official language in which they were submitted.
~66~eD
RESISTANCE WELDER
Background of the Invention
This invention relates to resistance welding In
particular t it relates to improvements in resistance
5 welding that are appropriate for use wi~h robot welders
and automatic and press welders.
Resistance weldin~ is a well-known way to join
together two electrical conductors. It comprises passing
an elec~rical current through the conductors in an amount
10 sufficient to cause localized heating that melts the
conductors, joining them together. This is normally
accomplished by placing a pair of electrodes against the
joined electrical conductors, applying pressure and an
electrical voltage to the electrodes, and timing the I2R
I5 heating to an amount that is sufficient to weld the
materials without creating excessive melting. This can be
accomplished either by a DC or AC voltage in most
applications. However, because of the typical junction F
; resistance between two elec~rical conductors that are to
20 be 30ined, it is normally desirable to reduce the voltage
below the value of typical line voltages by a step-down
transformer to apply voltages of the order of a few volts
or tenths of a volt and currents of the order of thousands
or tens of tho~sands of amperes. The simplest such
25 arrangement comprises a step-down transformer, a switch to
control the application of an electrical voltage to the
step-~own transformer, and a pair of electrodes connected
electrically to the secondary winding of ~he step-down
transformer~. When the electrod;es are placed on opposite
30 sides of the workpieces to be j~ined, closing the switch
applies a voltage across ~he junction ~Qf the electrical
conductors and the resultant electrical heating melts the
spot under t~e electrodes to weld the electrical
conductocs toget~e
3S
,~
`i D3,~
., ~,
Practical considerations of actual welding
operations lead to ~he addition of Yarious refinements to
the process described above. In order to minimize the
cost of ~ransformers used to supply welding currents, it
is desirable to insure that the peak voltage applied to
the transformer from an electrical source places the c~re
of the transformer in saturation, at or beyond the knee of
the B-H curve of the core. Because of this fact, it is
necessary to insure that the state of the core of the
transformer is known whenever a voltage i5 applied to the
transformer. If the last voltage that was applied to the
transformer leaves the transformer magnetized in a
particular direction and the next applied voltage causes
current flow in the same direction, then the magnetizing
current required by the transformer, together with the
load current, may overload the transformer during the
first cycle. This is an undesirable situation that is
readily avoided by making certain that the control circuit
for the welding transformer always applies full cycles of
the input voltage to accomplish welding and that it always
begins the application of voltage on that portion of the
input voltage that ~oes in the same direction. This
a~sures that the first cycle of applied voltage always
encircles the oriyin ~f the hysteresis loop of the
transfo`rmer core, avoiding a current that is far into
: saturation, and also assuring that the peak voltage is
constant during each welding cycle. It then remains only
: to apply the welding voltage for ~ predete~mined number of
cycles of the input voltage to accomplish a weld. The
predetermined number of cycles is determined by experiPnce
but is typically a number that is s~all enough tha~ it
mus~ be con~rolled electronically bec~use the neces,~ary
time peri~d of application is too s,hor~ to be controlled
reliaSly by an operator.
The basic resistance welding system described above
'~L24~i~i9~ -
-- 3
has serious disadvantages when it is applied to resistance welding
in produc~ion lines. A production weld between two pieces of
sheet steel, whether or not they are galvanized, typically
reauires a current of the order of 10,000 to 30,0~0 amperes. A
transformer such as in EP-Al-64 750, EP-Al-l9 747 or BE-A-759 605,
this is wound to supply such currents ln the secondary will
typically weigh of the order of 200 to 600 pounds and will need to
~e cooled by water or other externa] cooling means. Electrical
leads to carry such curren~ts are substantial in size. The usual
means to handle such problems as these in production lines in the
automotive and other industries is to suspend transformers from an
overhead support, to run insulated conductors to a welding head
that includes watercooled electrodes, and to have an operator
place the electrodes at the spot to be welded and apply the
external Eorce to hold the e]ectrodes in place while the weld is
made. The system described above presents a number of problems
when it is converted for use with robot or automatic welders.
Robots are generally limited in the amount of weight that they can
handle, and their operation is normally improved by reducing the
amount o~ that weight. Automatic welders are limited in the
closeness with which they can make adjacent welds, by the size of
their transformers. Robot welders are also hampered qreatly in
operation by being connected to large electrical cables that are
designed~to handle welding currents of thousands of amperes, and
the mobility of robots shortens the useful life of such large
cables.
:~ ` :
~: : :
,
;' ' ' ' ': '
"
:
} ^
:~L2~L~6g~
Accordin~ to one aspect oF the present invention, there
is provided an electrical circuit for a resistance welder that
reduces the weight in the vicinity of the weld.
According to a further aspect of the present invention,
there is provided an electrical circuit for a resistance welder
that is adaptable for use with a robot welding system.
According to a further aspect of the present invention,
there is provided a small, lightweight weldinq transformer which,
because of its size and weight, can be mounted close to the weld
zone, allowing the use of lightweight primary leads and short
secondary leadsO
According to a stil] further aspect o-E the present
Invention, there is provided an~electrical circuit Eor a
resistance welder that is adaptable Eor use with an automatic
welding system.
According to yet another aspect of the present invention,
there ~is provided a method of welding an electrically c~nducting
workpiece by resistance weldinq comprising the steps of rectifylng
an AC voltage~at a line frequency to produce a rectified voltage,
invertlng~the rectified voltage at~an operating frequency that is
higher than the line frequency to produce an AC voltage at the
operating frequency, transforming with a transformer, the AC
voltage at the operating freguency to a lower voltaqe to produce a
step-d~own voltage, characteri2ed by applying the step-down voltage
from~a~center~tap of the transformer through welding contacts to
the~workpiece~.
'
: -
~Z~6~
-- 5 --
Accordinq to a still further aspect of the present
invention, there is provided an apparatus for weldinq an
electrically conducting workpiece by resistance welding comprising
means -Eor rectiEyinq an AC voltage at a line frequency to produce
a rectified voltage, means for invertinq the rectified voltage at
an operatinq frequency that is higher than the line frequency to
produce an AC voltage at the operating frequency, means for
transEorminq the AC voltage at the operating frequency to a lower
voltage to produce a step-down voltage characterized by means for
applying the step-do~7n voltaqe from a center tap of the
transformer means through welding contacts to the workpiece.
According to yet another aspect of the present invention,
there is provided An elec~rical circuit ~or welding an
electrically conductive workpiece by resistance weldin~, to be
connected between an AC voltage source at a power Erequency and a
pair oE welding electrodes, the ciruit comprising a rectifier
connected to the AC voltage source to produce a rectified voltage,
means connected to the rectiEier for invertinq the rectified
voltage of the rectifier at an operating frequency that is higher
:than the -Erequency of the AC voltaqe source to produce an AC
voltage at the operating frequency~ a step-down transformer
connected to the means for invertin~ the rectified voltaqe to
produce~at a pair of secondary term1nals an output voltage at the
operat~ing Er~quency that is lower in amplitude than the voltage
produced~by the~means for inv~ertlhg the rectified voltage, ea~ch of
the s~econdary terminals of the step-down transformer also
- : :
~66g~
-- 6
connected to one of the pair of weld;na electrodes characterized
in that said step-down transEormer has a center tap connected to
the other of the pair of welding electrodes.
According to a still further aspect oE the present
inven-tion, there is provided an electrical circuit for welding an
electrically conductive workpiece by resistance welding, be
connected between an AC voltage source at a power frequency and a
pair of weldina electrodes, the circuit comprising a rectiEier
connected to the AC voltage source to produce a rectified voltage,
means connected to the rectifier for inverting the rectified
voltage of the rectifier at an operating frequency that is higher
than the frequency of the AC voltage source to produce an AC
voltage at the operating frequency, a step-down transformer
connected to the means for inverting the rectiEied voltage to
produce at a pair of secondary terminals an output voltage at the
operating fre~uency that is lower in amplitude than the voltage
produced~by the means for inverting the rectified voltage, each of
the secondary terminals of the step-down transformer also
connected to one of the pair of welding electrodes, characterized
by the step-down transformer having a center tap, and a full-wave
~rectifier being connected to the secondàry terminals of the
~step-down transformer, to the center tap, and to the welding
electrodes to rectify the output voltage of the step-down
transformer and apply a rectified output voltage to the welding
electrodes.
Brief Description of the Drawings
Fiqure 1 1s~a;~block~diagram ~of~a clrcu~it for the practice -
of~the ~present invention. ~ ~ ~
Fiqure 2 1s a~more detalled hlock diagram~of th~e circuit
for~the~practlce of the present invent~ion.
' ~
:
.
; ' , :
Fig. 3 is a circuit diagram that provides more
detail about the operation of inverting uni~s 44 and 46 of
Fig. 2.
Fig. 4 is a detailed circuit diagram of the circuit
for timer 48 of Fig. 2.
~ig. 5 is a detailed circuit diagram of drive
circuit 52 of Fig. 2.
Fig. 6 is a cutaway perspective view of a step-down
transformer that has been built and used for the practice
of the present invention.
Fig. 7 is a set of time plots of vol~acle wave forms
in the circuit of Fig. 4.
Detailed Description of the Invention
Fig. 1 is a block diagram of a circuit for the
practice of the~present invention. In Fig. 1 a source I0
o AC voltage is connected to a rectifier 12. The
connection is shown here as being made~with three leads
which is most likely when source 10 i5 a source of~
three-phase AC voltage, However, electrical energy of any
number of phases co~ld be used. ~ectifier 12 is connected
to produce as an output a rectified voltage which is
appropriate~ly~described as DC voltage wlth an AC
component. ~The output of rectifier 12 is taken~through
cont`ro~l circui~t 14,~ which~is a~controlled~`inverter, to
transormer~ 16. Control circ~uit 14 is used~to convert the
output~o~;~rect;lfier 12 lnto an~AC vo~ltage at a~frequency
tha~t is~higher~than the inpu~t frequency and with an RMS~
~Yalue that is controlla~le.~ ~ ~
Trans~forme~r: 16: is~shown dotted~here because lt may
be useEul~to~change the~voltage that i~ ~upplied to leads
18 and:~ 19~as~an input tQ st~p-down transormer 20
However, it shoul~ be understood that under svlme
circumstances it might be desirable to connect lea~s 18
: ~: : :
~ :
:
`
and 19 directly to control circuit 14. This is a ma~ter
o~ design choice.
St~p-down transformer 20 has a secondary that is
centertapped. The secondary leads of step-down
transformer 20 are connected to rectifiers 22 and 24 to
form a full-wave rectifier. A common connection from
rectifiers 22 and 24 is taken to welding electrode 26.
Center tap 2~ of step-down transformer 20 is connected to
welding electrode 30. When welding electrodes 26 and 30
are placed on opposite sides of a workpiece to be welded
and control circuit 14 is operated to supply current
through welding electrodes 26 and 30, a resistance weld
may be effected between the pieces thus joined. When the
frequency of the output voltage of control circuit 14 is
higher than the frequency of source 10, then step-down
transformer 20 can be made smaller and lighter than it
could otherwise be. This facilitates its use by human
operators and it also makes possible the placing of the
transformer in the arm of a robot welder, thus freeing the
arm for a wider range of motions. In addition, the
smaller transformer makes, it possible to make welds closer
together in automatic welding machines.
Fig. 2 is a more detailed block diagram of the
circuit for ~he practice of the present invention. In
Fi~. 2 t rectifier 12 is connected to source lO to supply
olta~e between bus leads 4Q and 42. Inverting uni~s 44
and 46 are~connected between bus leads 40 and 42 and their
midpoints are~connected to leads 18 and 19 thence to
step-down trans~ormer 20. ~n the block diagram and
circuit of Fig~ 2~ transformer 16 of Fiy. l has been
omitted. As st~ted, this is a matter of design choice.
The b~lance of the circui~ in Fig. 2 includes rectifiers
;~ 22 and 24 that are connected to the secondary willding of
step-down transformer 2Q to produce a full-wave-lrectified
output through ~elding electrodes 26 and 30, through a
~2~
.
workpiece that is not shown, and back to cen~er tap 28.
Control of the circuit of Fig. 2 is initiated in timer 48,
which both sets the frequency of operation of in~erting
units 44 and 46 and also controls relative timing to
control the amplitude of the current flow through welding
electrodes 26 and 30. A c~rrent-sensing element 50 is
connected to supply an input to timer 48 to cut off
operation of inverting units 44 and 46 if current in bus
lead 40 e~ceeds a predetermined value.
Timer 48 prod~ces two outputs that are taken to
drive circuit 52, which is connected to drive power
transistors or other switching elements in inverting units
44 and 46 so that current flows during a first half cycle
through the upper portion of inverting units 44, through
lead 18 and the primary of step-down transformer 20,
through lead 19 to the lower half of invertin~ unit 46,
thence to bus lead 42. The second half cycle sees current
flow thr~ugh the upper half of inverting unit 46, through
lead 19 to the primary of step-down transformer ~l but
flo~ing in the opposite direction. Current continues to
flow through lead 19 to inverting unit 44 where it ~lows
through the bottom half of inverting unit 44 to bus lead
42. Details of this control will ~ecome apparent ~rom
examining more detailed circuit diagrams. Acceptable
; 25 s~itching elements for inverting unit 44 incIude
thyristors, SCRs, gate-t~rnof~ devices, and the like.
Fig. 3 is a circuit diagram that provides more
detail about the operotion of invertin~ units ~4 and 46 of
Fig~ 2. In Fig. 3, AC power from source 10 is takeR
~ 30 through three fuses 60, three Iimiting resistors 62 and
:~ :three co~acts 64 to rectifier 12. Source 10 is typically
a three-phase source at a freq~ency of 60 Hz. ~nd a
convenient power voltage ~uch as 430 volts or the like.
Contacts 64 are here shown as being energized by t~ontactor
: 35 63 under the control of pushbutton 65. ~his supposes that
91:3
~o
pushbutton 65 is operated as a part of the openiny and
closing of welding electrodes 26 and 30 before weld
current is applied and after a weld is completed. In the
alternative, contactor 63 may be controlled by timer 14 o~
Fig. 2.
In Fig. 3, rectifier 12 comprises an appropriate
number of diodes 66 connected to form a full-wave bridge
rectifier. Six diodes 66 are shown here, but it should be
evident that that number may be changed according to the
desired current to be handled, voltage to be applied to
the diodes, and also to a different number of phases than
three. These are mat~ers of desisn choice. The output of
rectifier 12 is a full-wave-rectified voltage that is
positive at bus 40 with respect to bus 42. Inver~ing
lS units 44 and 46 are connected between bus leads 40 and
42. Inverting units 44 and 46 are identical, so only one
inverting unit 44 will be described. In inverting unit
44, a power transistor 68 is connected in series with
another power transistor 70 which in turn is connected to
2G a negative bus lead 42. Power transistor 68 is driven by
a Darlington transistor array 72 which in turn is driven
by drive circuit 52. Power transistor 70 is similarly
d~iven hy a Darlington transistor array 74 which is also
driven by driver 52. Inverting unit 46 comprises power
transistors 76 and 78 that are similarly driven by
Darlington transistor arrays that are not shown here~
Power ~transistor 68 is bypassed by a diode 80 a~d also by
the series combination of resistor B2 ~nd capacitor 84,
which suppresses a rapid rate of rise of volta~e across
30~ po~er tran~istor 6~. This may ~e unnecess~ry with some
choices of power transistor 68. ~ower tran~istors 70, 76
and 78 ~re similarly bypassed. The common poi~ ~6
between~power transistors 6~ and 7C is connec~ed through
lead la to one end of the primary windi~g ~f step-~o~n
trans~ormer 20, and the common point 88 of power
~2~
transi~tors 76 and 78 is connected through lead 19 to ~he
other end of the primary winding of step-down transformer
20. A stabilizing cap~citor 90 is c~nnected through a
resistor 92 between positive bus lead 40 and negative bus
lead 42.
Current-sensing element 50 comprises a current
transformer 94 that senses current flow in positive bus
lead 40 and also in capacitor 90. This combined
connection prevents false trips when capacitor g0 is
charging, when the circuit is first energi2ed. Current
transformer 94 is connected to current sensor 96 which
generates a signal that is proportîonal to the current
measured. This cignal is taken to comparator 9B where it
is compared with a predeter~ined voltage. When current
flow generates a signal that exceeds the predetermined
level, the signal is taken to timer 48 of Fig. 2 to
control operation of timer 4~.
When the circuit of Fig. 3 was built and tested, the
input voltage from source 10 was at a frequency of 60 Hz.,
and the timing circuit of timer 48 of Fig. 2 was operated
so as to generate an input to step-down transformer 20 at
1200 Hz. Under the~e conditions it is appropriate to
ignore the change from cycle to cycle of the voltage
between bus leads 4~ and 42, even though, as the
conventlonal output of a full-wave three-phase bridge
rectifier, it is known to have components of AC voltage at
360 Hz. and multiples of that frequency. Operation of the
c~lrcuit is well approximated by assuming that a DC voltage
is applied between bus leads 40 and 42. An AC voltage is
30 applied to step-down ~ransormer 2D by first causing power
transistors 6B and ~8 to conduct, while power transistors
70 and 76 are not conducting. This gene~ates one
half-cy~le of AC voltage to be applied to step-down
~ transformer 20~ Conditions are then chanqed 50 that power
transistvrs 76 and 70 are caused to conduct, while power
I
transistors 68 and 78 are switched off. This applies a
volt3ge ~o step-down transformer 20 in the opposite
direction, suppl~ing the other half-cycle of AC voltage to
step-down transformer 20. The voltage applied to the
pri~ary of step-down transformer 20 is essentially a
square wave at 1200 ffz. The secondary of step-down
transformer 20 responds to the square wave at 12Q0 Hzo to
produce what is substantially a square wave at 1200 ~Z.
with a slight ripple that is full-wave rectified to be
applied at welding electrodes 26 and 30.
Fig. 4 is a detailed circuit diagram of the circuit
for timer 48 of Fig. 2. In Fig. 4, a single-shot 110
generates a single rectangular pulse that is of the order
of 1.6 milliseconds in duration. Thls pulse is taken to a
weld memory circuit 112. A pulse generator 114 develops
pulses at a predetermined frequency and of a width that is
variable. These pulses are also taken as an input to weld
memory circuit 112. A weld timer circuît 116 generates a
rectangular pulse of variable duration ~hat is connected
to weld memory circuit 112 to enable a weld for a
predetermined time. Weld memory circuit 112 is also
disabled by a signal from current sensing elements 50 of
Fig. 2 and Fig. 3 indicating the presence of an
overcurrent.
An output signal from weld memory circuit 112 is
taken to delayed firing circuit 118, a flipflop that
delays its signal. The output of delayed firing circui~
118, the output of pulse gen~erator 114, and the output of
weld memory circui~ 112`are taken through a NO~ gate 120
~;~ 30 to flip~flop 122. Flipflop 122 generates two outputs that
are rectangular waves of oppos~i~e signs. One of these is
taken to drive transistor 124 and the other is taken to
~rive transistor 126. Output5 of drive transistors 124
and 126 rep~esent the output ~f timer 48 which is taken as
35 two inputs to drive circuit 52 of Fig. 2.
~2~L66~
Considering the circuit of Fig. 4 in more detail,
single-shot 110 includes a switch 128 which changes the
state sf the inputs to a pair of NAND gates 130 and 132.
These are connected to form a flipflop that produces an
output that is taken to single-shot 134. This is an
anti-bounce circuit. Switch 128 is here indicated as a
push bu~ton because that is the form in which it was used
in the circuit that was built. It wo~ld also be possible
to initiate the triggering of single-shot 110 ~ith an
electrical signal from another portion of the circuit or
from a microprocessor used in a system of welding
control. This is a matter of design choice and
convenience,
Pulse generator 114 of Fig. 4 comprises a
re~riggerable and resettable monostable circuit that will
be described for convenience as pulse generator 136.
; Capacitor 138 and a network ~hat includes resistors 140
and 142, potentiometer 144 and diodes 146 and 148 is
; connected through resistor I50 to the positive voltage
~20 supply. The common point of diodes 146 and 148 is
connected tQ pulse genera~or 136 so ~hat one or the other
of diodes 146 or 148 is switched into conduction according
to the sign of the voltage applied at their co~on point.
When diode 148 is switched into conduction, a resistance
25; that is equal to the sum of the resistance of resistors
142 and the right-hand half of potentiometer I44 is
connec~ed in series with capacitor }38 to determine one
pulse time. When~diode 146 is ~witched into conduction,
he resistance that determines the period of the opposite
half of the pulse is that of the sum of~resistQr 140 and
the remaining portion of potentiometer 144. The resuIt is
that a change in the setting of potentiometer 144 chanses
the relati~e len~th of the pulses without changing the
value of their sum. This prod~es at output terminal 152
a square wave at constan~ frequency which here ls 1200 Hz.
/~
and having periods of conduction of each sign that can be
varied by varying the setting of potentiometer 144.
Weld-time timer 116 uses a single-shot 154.
Capacitor 156 is connected in series witl- resistor 158 and
they both are connected to single-shot 154. The sum of
the res~stance of resistor 158 and resistor 160 and
variable resistor 162 combines with the value of capacitor
156 to determine the period of weld time. The resistance
that is selected by adjusting the setting of variable
resistor 162 thus adj~sts the length of time that the
welder stays on. It should be evident that this function
may be controlled on an analog basis by adjusting
resistance and corresponding RC times in a single-shot as
shown here. In the alternative a microprocessor could be
used to control weld time either according to a
predetermined schedule of times or in response to vari~us
otheL items of information s~ch as measurements of weld
yuali~y or the like. These are matters of design choice
that will vary according to the use that is contemplated
for the circuit, In the weld-time timer 116 as shown,
that resuIt is a reotangular pulse at output terminal 164
that is eq~al in duration to the length of the desir2d
weld, typically seconds or fractions of a second.
:: Signals from single-shot 110, pulse generator 114
and weld-timer 116 are all taken to weld memory circuit : :
.
112~ The output of single-shot 110 is applied as one
: : input to NOR gate 166 and the signal from ou~put terminal
1~4 of weld-time timer 116 is applied as the other input
: to NOR gate 166. The output of NOR gate 166 is taken as
an input to the D ter~inal of flipflop 168, which is
clocked by the si~nal from terminal 152. ~lipflop 168 can
be reset by a signal from curr2nt sensing element 50 of
Fig. 2 in case of an overload. The NQT-Q output of
flipflop 168 is ~aken as a reset signal to single shot 154
3S and also as an input to delayed firing circuit 118, where
- ~ -
it is applied to a single shot 170. The output of single
shot 170 is taken to NOR gate 120 where it is applied as
~ne input. Other inp~ts to NOR gate 1~0 are taken from
te~minal 152, the output of pulse generator 114, and the Q
output of flipflop 168. The output of NOR gate 120 is
taken as an input to flipflop 122 where it is applied to
inverter 172 and as one input to each of NAND gates 174
and 176. The output of inverter 172 is taken as a
clocking input to flipflop 178 which provides
opposite-going outputs that are talcen respectively as
inputs to NAND gates 174 and 176. The result i5 to
produce two equal and opposite rec~angular wave forms that
are taken as inputs to drive transistors 124 and 1260
Referring again tO the three inputs to NOR gate 120, the
inp~t from delayed-firing circuit 118 causes the first
pulse in any welding interval to be of a shorter pulse
width than the succeeding pulses. This prevents
saturation of the core of step-down transformer 20 of Fig.
at the start of a weld. The input to NOR gate 1 0 from
weld memory unit 112 determines the total time that
rectangular pulses are allowed to appear at the output of
~OR gate 120. This is the length that is determined for a
: single weld. The input to N~R gate 120 from terminal 152
causes all but the first pulse of any one weld cycle to be
; 25 rectangular pulses at a fixed frequency and o a length
that is determined by the setting of potentiometer 144.
Fig. 5 is a detailed CifCUit diagram:of drive
circuit~:52 of Fig. 2. :In Fig. 5, a timer 184 recei~es as
: ~an input a signal frvm drive transistor 124 of Fig~. 4. An
identical timer 186 receives~an equivalen~ si~nal from
~ drive transistor 126. Since the circui~s in whi~h :timers
:~ 184 and 186 are used are identical, only:~hat as~svcia~ed
: with timer 184 will be described in detail.
~ Ti~e~ ~84 produces a pulse that is taken to power
35 amplifiec 18~ of Fig. 5 wt)ere it i5 used to d~ e
:
6~
,~
negative-going amplifier 190. I~he input pulse from drive
transistor 124 is also taken to a positive-g3i~ng am~lif ier
192. Negative-goi~g amplifier l90 and positive-going
amplifier 1~2 are both connected to p~imary 1~4 of
transformer 196 with one portion of a cycle of currel-t
being supplied b~ each of these amplifiers. Transformer
196 has a secondary winding 198 and a secondary winding
200. Secondary winding 198 is connected to a shaping
circuit 202 and secondary winding 200 is connected to a
shaping circuit 204. St~aping circuit 202 is connected to
the circuit of Fig. 3 to trigger conduction of power
transistor 6B. Shaping circuit 204 is connecteci to the
circuit of Fig. 3 to trigger conduction of power
transistor 78. The corresponding shaping circuits of the
id~ntical portion of Fig. 5 are similarly connected as
indicated, one to trigger the conduction of power
transistor 76 in Fig. 3 and the other, to trigger the
conouction of power transistor 70 o Fig. 3.
The circuit of Fig. 4 that supplied inputs to timers
20 184 and l86 were described as being opposite in sense. It
therefore follows that the voltages to transformer 1~6 and
its symmetrical equivalent in Fig. 5 will be opposite in
phase. Referring to the outputs of shapiny circuits 20~
a~d 20~, in connection with the circuit of Fig. 3~ it can
25 be seen that when pulse shaping circ~its 202 and 204
; produce currents that will cause transistors 68 and 78 to
conduct. Conduction through step-down transformer 20 of
Fig. 3 will be rom left to right. The opposite is true
when the input polarity is reversed so that the inputs to
30 Fig. 3 will turn on power transistors 70 and 76, causing
current flow ~r~m right to ~eft tl)rough step-down
transformer 20 of Fig. 3~ The result of ~his ~peration
will be the application to step-down ~ransformer 20 of
Pig. 3 of a square wave of current at a frequency
deter~ined by pulse generatvr 114 o~ ~ig. 4. The RMS
~2~
,~
value of the cureent in step-down transformer 20 of Fig. 3
will be determined by the setting of potentiometer 144 of
Fig. 3. The number of such pulses, representing the
length of a weld, will be determined by the setting of
variable resistor 162 of Fig. 4.
~ig. 6 is a c~taway perspective ~iew of step-down
transformer 20 that has been built and used for the
practice of the present invention. In Fig. 6,
ferromagnetic core 210 is enclosed by a primary winding
212 and another primary winding 214 that are connected
together. A secondary winding 216 begins from a terminal
218. Secondary ~inding 216 is a sin~le thickness of a
water-cooled electrical conductor, placed to enclose
primary winding 212 and an associated core 21û. Secondary
winding 216 continues to cer~ter terminal 22û which will
compri~e a center tap of the secondary winding 216. In
order to keep the winding sellse of the secondary in ~he
proper direction, secondary ~inding 216 is next taken
around primary winding 214 in a direction to corner 22Z,
then to corner 224, through window 226 to terminal 228,
completing secondary winding 216.
Resistance welding oE sheet metal of gauges in
common use in the automotive industry typically takes
currents of the order of 10,000 or 20,000 amperes. While
25 ~the circuit in Fig. 3 shows two rectifiers 22 and 24, the
realization of that circuit that was built using the
~ :transformer of Fig. 6 shows four, the num~er necessary to
:~ ~ : carry the desired current. In Fig. 6, diode 230 was used
in pa~raIlel with diode 232 to carry the necessary amount
30: of:~current. These two diodes in parallel form the
equivalent of diode 22 o~ Fig. 3. Similarly, in Fig. 6,
di~de 234 is placed in parallel with diode 236 to effect
the e~uivalent of rectifie~ 24 o~ Fig. 3. A c~mon
connestion a~on~ divdes 23~, 232, 234, and 236 is not
shown in Fig. 6 but will be made by clamping an electrical
6~
conductor in the space 238 that now separates them. A
plurality of inlets 240 and outlets 242 carry cooling
water that is passed internally through ducts 244 in
secondary winding 216.
The transformer of Fig. 6 is one that has been built
and ~ested for use in the circuit of Fig. 3. It is shown
here for certain of its features rather than as a
necessary way to build a transformer~ Those features
include a secondary winding that has two turns with a
center tap that is available for a connection. It has
means or placing rectifying semiconductors in a
water-cooled terminal attached to the transformer that
allows them to be clamped readily to the common terminal.
One feature h~wever that represents a particular feature
of the present invention is the ~act that the use of a
frequency above the line frequency to be applied to the
primary of the transformer Or Fig. 6 allows the ùse of
less iron in core 210 that will be necessary at a lower
frequency. The smaller amount of iron, and hence the
smaller amount of copper required, reduces the weight of
the transformer of Fig. 6,and makes it easier to locate
the transformer of Fig. 6 in a robot arm or in an
automatic welder.
Fig. 7 is a set of time plots of voltage wave forms
in the cir~uit of Fig. 4. Each o~ the wave forms is
identifie~ at an appropriate place in the abscissa by the
element number of ~he item of eg~ipment in ~ig. 4 of which
the wave form represents the output. Referring to Fig. 7,
the wave form marked "114" is a rectangular wave ~hat is
generated by the free-running pulse genera~or 114. That
wave ~orm be~ins its rise at a ~ime marked Tl and
repeats with a period of 417 microseconds. The time of
fall of this ~ectangular h~ave form i5 indicated b~y arrows
as being ~ariahle, since that ~ime can be set ~y adjusting
potentiometer 144 of Fig. 4. A second wave form that is
shown in Fig. 7 is that of single-shot 110 marke~ as
'~110,", which is a rectangular wave of 1.6 milliseconds in
duration. That rectangular pulse is shown as starting at
time To in Fig. 7, a time that is determined by
operating switch 128 of Fig. 4 and that can egually as
well be determined by other signals or by programming, as
has been described.
After time To~ time Tl is determined as the
first occurence of a rise in the rectangular wave form of
pulse generator 114. ~his sets the time of the
rectangular p~lse marked "112," which is the output of
weld memory circuit 112. This is a single rectangular
pulse that begins ~t time Tl and continues to the end of
the weld, a time measured typically in tenths of a second
or seconds. Time Tl is also the starting time of the
wave form marked "118." This is a single rectangular
pulse tha~ begins at Tl and ends after 208
microseconds. This is the output of delayed firing
circuit li8 which causes or may cause the first pulse in a
weld cycle to be of shorter duration than the rest of the
pulses.
Consider now the wave form in Fig. 7, marked "120,"
which is the output of NOR gate 120. This is the negation
o~ the logical union of wave forms "112," i'll4," and "118"
of Fig. 7. Time T2 is seen as the fall time of the
rectangular wave representing the output of pulse
generator 114, while time T3 is defined as the time of
fall of the rectangular pulse that is the output of
delayed firing circuit 118. If time T2 occurs before
~0 time T3 as shown herer then the wave form marked "120"
begins at time T3 and thereaf teY is the negation of wave
form "114.`' If time T2 is selected to be later than
T3 then the wave form "120" will be the negation of wave
form "~14O n Wave form ~120~ is ~hen the source of the
wave forms marked ~174" and nl76" which are respectively
~S6
,~
the outputs of NAND gates 174 and 176 of Fig. 4. As can
be seen, wave form "174" comprises alternate p,ulses
selected from "120," and wave form "176" comprises the
remaining pulses of wave form "120." These switcll
inverting units 44 and 46 alternately to produce the
output square wave as desired.
We claim:
: ~
:
