Note: Descriptions are shown in the official language in which they were submitted.
- This invention relates to a method and apparatus
for reproducing operating conditions in variable induced flow
devices, and more particularly to a method and apparatus for
reproducing a predetermined air flow and manifold vacuum
through a carburetor in a carburetor testing system.
Applicants have for many years been engaged in the
production of carburetor test stands and the like for testing
carburetors to determine if they meet ever stricter standards
::
for air pollution. Applicants started in the carburetor testing
. ..
field many years ago when the only test there was for a carbure-
tor was to determine the fuel/air ratio at idle conditions, -
and if such fuel/air ratio was in the neighborhood of plus -;
,~
or minus as much as 6 - 9% of the ideal fuel-air ratio~ the `~
carburstor-was passed ~or installation in the motor vehicle,
with the assumption that it would perform properly in the
finished automobile.
At this early stage of the art, even though each
carburetor on the production line was tested, it was tested
more with a view to making sure that the air and fuel actually
flowed through the carburetor, rather than for accuracy. In
other words, the carburetor was just tested to see if it would
work on an automobile engine before it was installed thereon.
:
,.
'~
~ s need for some type of standards became felt, there
was still no actual testing for accuracy done on the production
line, but only a comparison of the fuel/air ratio of a produc-
tion carburetor with the fuel/air ratio of a carburetor known
to work well on an automobile engine. The carburetor which
was known to work well would be tested under what is known as
laboratory conditions. The tests made in accordance with such
. .
laboratory methods were exceedingly slow, were limited to
laboratory equipm~nt and conditions, were not continuous, and
were not susceptible of being made a production operation.
Such conditions were due to the fact that the ratios
for each carburetor had to be determined by direct measure-
ments of the fuel drawn by the carburetor during a measured
time period at a given flow of air. In order to produce a
definite and controlled flow of air at that point in ~e art,
such flow had to be of a subsonic nature controlled by a
suitable valve, and the measurements of fuel fIow had to be
made directly, such as by measuring the volume of fuel con-
sumed from a graduated glass container.
The carburetor known to work well on the automobile
engine would be tested according to these laboratory methods
to determine its fuel/air ratio. The tested carburetor would
~en be carried to the production line and become the standard
to which the production carburetors would be held. This then
was the first testing for accuracy of carburetors vn the pro-
duction line. -~
As the concern over air pol~ution became greater, and ,-
~he need for greater accuracy in each and every carburetor
: - ' ; ~ ' ' ~ ,' :,
,:' ' . :
.
~ 4~
coming off the production line bec~me felt, the tPsting of
the fuel/air ratio of each and avery carburetor coming off
the production line at an accuracy of that rivaling the
laboratory method of carburetor testing became essential, and ;
obviously new methods of carburetor testing for the production
line were needed.
By the time that testing of a carburetor was more
or less standardized, and testing took place at four points or ~:
more of the operation range of the carbure~or the most common `
ones being idle, off-idle, part throttle, and wide open throttle,
Applicants' assignee was a leading manufacturer in the test
equipment field, and patented the-carburetor testing system
disclosed in the United States Patent No. 3,517,552 to Vernon
G. Converse III, et al. This system was completely satisfactory
for the production rates required of carburetor testing systems
at the time Applicants' assignee produced such a system, and
such a system is still produced for use where production rates
.
are relatively low and accuracy reguirements permit its use.
However, Applicants were soon required by ever increasing
standards of accuracy and ever higher rates of production
to continue to improve on the above mentioned carburetor testing
system. While four or more points of testing were maintained
as standard, it was desired to complete the carburetor test at
least two or three times faster than such a system could per-
form the test.
It should be restated at this point, as disclo~ed in
the above mentioned patent, that carburetor requirements are ;
given in texms of fuel/air ratio permitted at certain engine
manifold vacuums, and that basically a carburetor testing
... . .
- 3 -
'';' ,','. ....
system must produce a given air ~low through the carburetor
and then have a system which will adjust the carburetor throttle
plate to produce a specified manifold vacuum, at which time the -
`; fuel flow through the carburetor is measured. From these values
the fuel/air ratio is computed for the particular point in the
carburetor's operating range.
It can be seen that the great variable feature in the
. carburetor testing system is not a measurement of ~uel flow, as
this can be done with any number of flow measuring devices
presently available in the art, nor is it the measurement of agiven air flow through the carburetor, as this can be done by
any number of devices such as critical venturi meters, variable
area critical venturi meters, laminar flow tubes or subsonic
nozzles, but it is the adjusting of the carburetor throttle
plate to reproduce the predetermined spec;fied manifold vacuum
given by the carburetor manufacturer. While the U.S. Patent
No. 3,517,552 discloses a pneumatic carburetor throttle -
positioner which was and still is satisfactory for carburetor
testing lines with relatively low rates of production a~d
accuracy requirements, it is not satisfactory in some vf today's
current high production carburetor test lines. Therefore, in
an attempt to solve the problem of how to more quickly repro-
duce a required manifold vacuum, we focused our attention on
the part of the operations reproducing system which controlled
the carburetor throttle. `
It should be understood that in performing a test on
an induced flow device such as a carburetor or the like, not
only is it critical to maintaining high production rates that
the test at each point of the operation range be performed
quickly, but also that the carburetor throttle be moved from
4~
one test point to the next test point quickly and accurately.
While the pneumatic throttle posi~ioner may move between test
points quickly, the time it takes to produce the required mani-
- fold vacuum once it gets approximately to the right throttle
opening, under the worst conditions is approximately 30 seconds,
while the present invention, as will be discussed later, cuts
this time better than 50%, which is a significant advance in
the art.
After much work, an electrical system providing two
speeds of movement of the carburetor throttle was devised,
wherein the carburetor throttle was no longer mo~ed by a pneu-
matic cylinder, but was driven by electrical stepping motors
which could be driven at two rates of speed. The carburetor
throttle would be driven at a faster rate of speed when going
between test points, and slow rate of speed when approaching
a test point, so that its overshoot would be at a minimum. ;
However, such a system was still not entirely satisfactory, as
the system, in order to have adequate resolution at low flow
conditions such as idle, even at the higher rate of speed would ~`
take a relatively great length of time to move between test
points, and furthermore would approach such test points ~t a `~
single rate of speed tending to make overshooting such test -`
points a common occurence, and increasing the time the system
would hunt for the proper value. Even though such electrical ;~
~hrottle positioners could be controlled by a computer as dis- `
closed in U.S. Patent No. 3,524,344, and such a disclosed
system was faster than the previously mentioned pneumatic system,
relatively little time saving was gained in the time necessary
to ~et the manifold vacuum.
With the advent of stricter and stricter standards
for accuracy~ coupled with higher and higher production require- ~;
ments, we felt the necessity for, and have now invented a
- 5 -
. . . , : ,
/ - ~
4~
~ .
faster more accurate and more dependable means to r~produce
a predetermined desired vacuum i.n any induced flow device,
such as a carburetor, which requires it.
Therefore, one of the objects of the present invention
is to provide an improved methocl and apparatus for reproducing
conditions in induced flow devices, such as carburetors, whereby
the above difficulties and disadvantages will be overcome
and largely eliminated.
Another object of the invention is to provide an
improved method and apparatus for reproducing predetermined
manifold vacuums in carburetor testing systems or the like,
wherein the carburetor throttle plate can be quickly moved -
from one point of operation range of the carburetor to another.
Anokher object of the invention is to provide that
such movement of the carburetor throttle between test points
is controlled electrically.
A further object of the invention is to provide an
improved method and apparatus for reproducing operating condi-
tions in a carburetor testing system, whereby the carburetor
throttle plat~ can be moved quickly between test points, and
a the carburetor throttle plate approaches a test point, the
movement of the throttle plate becomes proportional to the
difference between the manifold vacuum in the carburetor and
the desired manifold vacuum.
- 6 - i~
A still further object of the present invention is
to produce a throttle setting device of the nature specified ,~
in the preceding paragraph, whereby the throttle se~ting system
is computer controlled.
; A still further object of the present invention is
to provide an improved method and apparatus for reproducing
vacuums, and thus inducing flow, in any devices which depend -
on the presence of vacuum to induce flow therein.
Another object of the present invention is to provide
an improved flow control system which is self contained in
operation, which is suitable for a production environment, can ~-
be placed on a test stand to be operated by a production woxker,
and which does not require for its operation the service of
a skilled laboratory technician.
A further object of the present invention is to pro-
vide an improved method and apparatus for reproducing vacuums `
in vacuum induced flow devices which will operate equally as ;
well with sonic or subsonic flow devices.
A still further object of the present invention is
20 to provide an improved carburetor throttle drive having a j;
stepping motor, wherein the amount the carburetor throttle
moves per degree of movement of the stepping motor becomes
greater the further away you move from the idle position of
,, .
the carburetorO
A still further object of the present invention isto provide a carbur~tor throttle drive as described in the
preceding paragraph, wherein such movement is induced by the
movement of elliptical gears.
... .
- 7 - -
:
. ; ~
.~34~
It is still another object of the present invention
to provide a carburetor throttle drive of the foregoing nature
which is adaptable to computer control and useful in carburetor
testing systems.
It is an additional object of the present invention
to provide a carburetor throttle drive which has faster operation
than previously available, has a non~linear mechanical
advantage, and which has proportional control to reduce the
amount of hunting as you approach a test point.
Further objects and advantages of this invention will
be apparent from the following description and appended claims,
reference being had to the accompanying drawing~ forming a
part of this specification; wherein like reference characters
designate corresponding parts in the several views:
Figure 1 is a perspective view of a carburetor test
stand adapted to test carburetors at several points of their
operation range, and which embodies the method and apparatus
of the present invention.
Figure 2 is a cut-away elevational view of a portion
of Figure 1, showing ~he carburetor which is being tested, and
more particularly showing one embodiment of our improved
carburetor throttle drive.
Figure 2A is a cut-away diagrammatic view of a
basic carburetor testing bench using sonic flow measuring
devices.
Figure 2B is a cut-away diagrammatic view of a basic
carburetor testing bench u~ing sub-sonic flow devices placed
upstream of the carburetor.
- 8 -
.. . . . ..
~ 4~4~
.
Figure 3 is a diagramma~ic view showing four types of
; flow producing devices which may be used in the present inven-
tion, i.e., a variable area critical venturi meter in combina-
tion with an absolute pressure transmitter, critical venturi
meters in combination with an absolute pressure transmitter,
laminar flow tubes in combination with a differential pressure
transmitter, or subsonic nozzles in combination with a differ-
ential pressure transmitter.
Figure 4 is a diagrammatic view of a one system
embodying the present invention.
Figure 5 is a diagrammatic view of aba~c system
embodying the present invention.
Figure 6 is a diagrammatic view of a construction
embodying the present invention, and including an error calcula-
tor and a volta~e to frequency converter.
Figure 7 is a diagrammatic view of the construction
shown in Figure 5, as it may be modified for manual operation.
Figure 8 is a diagrammatic view of the construction
.,.: .
shown in Figure 6, as it may be modified for manual operation.
Figure 9 is an elevational view of our improved
,, .
carburetor throttle drive controller. `
Figure 10 is an end view of the construction shown
in Figure 9.
Figure 11 is a plan view of a modified version ffl
the construction shown in Figure 9.
:, .
/
,
_ g _ ... .
.. . .
., ~ .
Q~
..
Figure 12 is a graph showing a relation between the
degrees of carburetor throttle plate moveme~ versus the degrees
of s~epping motor movement for any of the constructions shown
in Figure 9 ~
Figure 13 is an elevational view of the elliptical
gears used in the construction of Figure 9, at their starting
or closed throttle position.
Figure 14 is an elevational view of the gears shown
in Figure 13 at an off-idle position, and showing that the
gear connected to the throttle plate has undergone a smaller
angular change in position than the gear connected to the stepp-
ing motor.
Figure 15 is yet another view of the gears shown in
Figures 13 and 14, showing that the rate of angular change of
the gear connec~ed to the throttle plate continues to increase
at a rate faster than the change in position of the stepping
motor gear, as can be seen by the graph of Figure 12.
Figure 16 shows the gears of Figures 13 - 15 in their ;~
fulIy rotated position, wherein the carburetor throttle would
be at its wide open or 90 position.
,- . .,
It is to be understood that the invention is not
limited in its application to the details of construction and
arrangement of parts illustrated in the accompanying drawings,
since the invention is capable of other e~bodiments and of
being practiced or carried out in various ways within the scope
of the claims. Also it is to be understood that the phraseology
and terminology employed herein is for the purpo e of des~rip- ;
tion, and not of limitation. ~
-- 10 --
- ~
--\ ~
4~
There is shown in Figure 1 by way of a general example,
one of the uses that Applicants' flow con~rol system may be
-put to, i.e., that of use in one of Applicants' own carburetor
test stands. In this case, the s;tand shown is adapted for
use in a room having a controllecl environment so that the
mass flow rate of air through the carburetor will not be
effected by temperature, and no compensation need be made
therefor. Such stands may have components such as a meter 30
showing the air flow through the carburetor, a meter 31 showing
the fuel/air ration through the carburetor during the test,
manometer tubes 32 for calibrating the ~est stand, and various
other indicating devices and switches depending upon the require
ments set by the carburetor manufacturer.
Of particular interest in this application, as shown
in Figure 2, is the mounting of the carburetor 33 itself, and
the items surrounding it. Shown on the test stand is a means `~
of automatically connecting a fuel source to the carburetor
33 in preparation for a test. The fuel is fed to the carburetor ~;
through a conduit 34 which is connected to the test carburetor x
33 with the aid of a spring pressed coupling 35 which is -
securely held in place during thetest by a solenoid 39 supplied ~-
with electric current through the wire 40. Suitable clamping
hooks 41 hold the carburetor sealingly to the top of the test
chamber. The carburetor throttle drive, to be more fully
explained later, consists of a stepping motor 42 drivingly
connected to a clutch 43 by a pair of spur gears (not shown).
Driven by the clutch 43 is a pair of identical elliptical gears
47 and 48. ,~
Mounted on ~he elliptical gear 48 is a spring pressed
crank 49 which carries a stud 52. Byvirtue of such construction~
.
very rapid connection of the carburetor throttle to the throttle
drive is accomplished. Once ~he carburetor is mounted on top
of the chamber 54, a basic carburetor te~t is ready to begin.
It should be understood that the carburetor test
can be perform2d using both sonic and subsonic air flow measure-
ment, and that the present invention will work equally well to
set the proper manifold vacuum and air flow under either condi- -
tion.
,
In the sonic system, which is the system to which
most of the discussion herein will be directed, due to the fact
that it is the most convenient and most widely used, the
sequence of events in reproducting operating conditions in the
carburetor involved choosing thetest point you wish to repro-
duce, supplying the specified air flow at the test point by
means of a variable area critical venturi meter or critical
venturi meter, and then turning the carburetor throttle until
the desired vacuum is achieved.
This control of the carburetor throttle is achieved
by sensing the absolute pressure downstream of the throttle
plate, by means of an absolute pressure transmitter supplying
an analog signal which is constantly compared with a reference
signal, the throttle being rotated, as described hereafter '~
until the analog signal equals the referenc~ signal.
A related set of signals is compared in the system
wherein subsonic flow is measured, but in this case, ~he
manifold vacuum is preset with the carburetox throttle closed,
and then the carburetor throttle is opened until the desired
airflow is achieved as indicated by a differential pressure
transmitter placed across the flow measuring devices placed
~`~
- 12 -
.
. -, , :. . , ~ ~
upstream of the carburetor, such as laminar flow tubes 65, or
subsonic nozzles 72.
All the operations in t;he system using sonic flow which
are performed with the absolute pressure signal to control the
rotation of the throttle plate to reproduce a desired vacuum
are now performed with the differential pressure signal to
control the throttle plate to achieve a desired air flow after
the manifold vacuum has been preset.
As will be described, the same proportional control
of the throttle plate in response to the difference between a
desired air flow and the actual air flow is available in the
system with subsonic flow, as is available in the sonic system ~` ;
where the air flow is preset and the throttle control is in
proportion to the difference in the actual vacuum and the ~;
desired vacuum.
Shown in Figure 2A i5 a basic sonic flow carburetor
test set up with the carburetor 33 sealingly mounted to the
top of a chamber 54. Inside the chamber 54 is illustrated a
single critical venturi meter 55 connected by means of the
20 conduit 56 to a source of vacuum 57O In the simpliied illus- `
tration ~hown of testing a carburetor at a single test point
in which you are required to set a single flow and a single
vacuum, the vacuum source 57 would be selected to be large
enough to make the venturi meter operate critically, that is
at sonic air speeds. In that condition, for a given upstream
pxessure you would have a definite mass flow rate, such as `~
for example four pounds per minute. ~aving a definite air
flow, the carburetor thxottle plate 60 would be rotated by
the throttle linkage 53 in the manner previously described
until the desired manifold vacuum is achieved. Means of
- 13 -
.. . ~. . . . . ~ , . . -
.. ..
~ ~4~
sealingly mounting the critical venturi meter 55 into the
test chamber 54 are known in the art and one example of such
means is disclosed in U.S. Patent No. 3,517,552.
To test the carburetor at more than one point,
several critical venturi meters would be needed, such as
illustrated by the diagrammatic view shown in Figure 3, wherein
the chamber 54 is shown containing four critical venturi meters ~ -
55, with the flow fr~m carburetor 33 being represented by
the inlet 61. As disclosed in U.S. Pa~ent No. 3,524,344, a :~
single variable area critical venturi meter 62 can replace
the four critical venturi meters 55.
.. Since using either a variable area critical venturi
meter or a critical venturi meter means that sonic flow is
. being used in the system of Applicants' invention, the absolute
pressure upstream of the venturi meter equals the vacuum
present in the carburetor, and thus changesin the vacuum are
indicated by changes in the absolute pressure, which are sensed ~;
by the pressure probe 63, and are transmitted for use in the
remainder of the system by the absolute pressure transmitter 64.
If the use of subsonic flow is desired in the system,
either due to the particular carburetor being tested, the oost
involved, or the fact that it is an item different from a
carburetor that is being tested, the vacuum would be preset
with the carburetor throttle closed by utilizing the signals
received from the differential pressure transmitter 101 connected
to the pressure probes 102. ~-
Since subsonic flow is being used, the signals which
are utilized in the control of the throttle plate are the
ones indicating air flow.
~he flow measuring devices are placed upstream of the
carburetor in chamber 97 as shown in Figure 2B. Air enters
- 14 -
the chamber 97 through the inlet 100, passes through the
laminar flow tubes 65, through the conduit 98 into the carbure-
tor hood 99, and then thr~ugh the carburetor 33.
If subsonic flow is being used, the differential
pressure across the laminar flow tubes determines the flow rate.
To determine the change in flow rate, the change in the differ-
ential pressure is needed. This is supplied by readings taken
by the pressure probes 70 which are connected to the differential
pressure transmitter 71. These in turn are used in Applicants'
system described below.
Another modification of Applicants' system can be
the substitution of subsonic nozzles 72, again selected by the
valves 66, with the differential pressure again being sensed
by the probes 70, being calculated and transmitted by the
differential pressure transmitter 71.
It should be understood that any one of these four
systems or others can be used, and Applicants' flow control
., ~, .. .
system will work equally well wi~h any of them. For the purposes
of ease of description, Applicant will refer to the block shown ~ ~
20 in Figure 3 as the flow measuring system 73, and when the '' ' -
numeral 73 is used it should be understood to mean that any
of the four systems illustrated in Figure 3 could be placed
in the block 73 and used in Applicants' system.
To find the manifold vacuum which will be used for
setting the throttle plate position, in a sonic flow measure-
ment system the manifold vacuum will be the difference between
the controlled room and the pressure measured by the absolute ' ;
pressure transmitter 64. r.
In a subsonic flow measuring system the manifold
vacuum i8 preset and is the difference between the pressure
- 15 -
,~ .
.:
... , .. ,., , , ,, . , :
in the hood 99 and the pressure in the chamber 54.
Referring to Figure 6, in operation, the sonic flow
measuring system 73 supplies, through the absolute pressure
transmitter 64, a voltage signa;L corresponding to manifold
vacuum and represented by the numeral 74, to the direction
comparator 76. Such direction comparators may be such as the
type 19 - 501 made by the Bell & Howell Company, and in any
event are well known in the art and need not be described
herein in detail.
Previous to this event happening, the desired test
points have been determined. Usually ~he desired test points
for the carburetor are the previously mentioned idle, off idle,
- part throttle and full throttle points of operation, and the
carburetor manufacturer has supplied for each of these points
an air flow and a fuel flow at a predetermined manifold vacuum
for each test point. The testing system has to duplicate the
manifold vacuum, the given air flow, and measure the fuel flow
which results through the carburetor, and compare such fuel
flow with the design value to see if the carburetor is acceptable.
Since we are concerned here only with duplicating the
required manifold vacuum and air flow, the rest of a test
system will not be described in detail.
Having the manifold Yacuum and air flow for each test
point, the computer operator will, through a suitable computer
program, feed this information into the mini-computer 75. Such
mini-computers are well known in the art, and neither the
computer, or the computer program need be described in detail
herein, as they are well able to be duplicated by those skilled
in the art. An example of a computer which may be used in the
present invention is the Model PDP-ll, manufacturea by the
- 16 -
L~ 3
Digital Equipment Corporation of Maynard, MassachuSetts. Once
` this information is programmed into the computer, the computer
will do two things. It will first determine the setting of
the variable area cxitical venturi meter 62, if such a meter
i~ being used, which will produce the desired air flow at a
particular test point. If a series of critical venturi meters
55 are being used, the computer will determine and open the '
proper critical venturi meters for the desired air flow. Similar-
ly, if laminar flow tubes 65, or subsonic nozzles 72 are being
used, the proper combination of these will be selected.
Once this is accomplished, the computer will supply
a reference voltage signal 74 corresponding to the desired
manifold vacuum. It should be understood that in a system such
as Applicants, the absolute pressure transmitter 64, or the
differential pressure transmitter 101, indicates the pressure
reading by sending out a signal which is proportional to the '~
pressure being sensed by the pressure probe 63 if absolute
pressure is being measured, or the pressure probes 102 if -
differential pressure is being measured.
For purposes of illustration, using sonic flow measur- -~
ing conditions, assume that a manufacturers specification for
a particular carburetor is that it should flow 0.2 pounds of
fuel per hour at nineteen inches of vacuum at an air flow of
2.0 pounds per hour. In a controlled room, nineteen inches
of vacuum will correspond to 10-1/2" of Mercury (Hg) absoiute,
and for this reading, the absolute pressure transmitter should
supply a reading of two volts D.C. At the starting point of
course, the absolute pres~ure transmitter 64 will be reading
atmospheric pressure, which equals approximately 30" o~ M~rcury;
and will be giving a si~nal equal to 5 volts D.C.
. ~
- 17 -
.
:: 1134~
Before proceeding further, it should be understood
that the systems illustrated in this application are for use
in a room where the environment is controlled ~o a constant ;
temperature and pressure, and, therefore, there is no need
to adjust the pressure readings for the effect of a change in
temperature or pressure. However, if the carburetor testing
is to take place in an environment wherein temperatures fluc-
tuate over a wide range, it is obvious that temperature probes
could be placed upstream of the flow measurement devices wherein
temperature would be taken into account in the calculation of
the mass flow rate. The use of such a temperature probe is
well within the scope of Applicants' invention.
Continuing with the present illustration, opening ;
the critical venturi meter to the point to produce an air mass ;
flow rate of 2.0 pounds per hour immediately causes a large
drop in the pressure sensed by the pressure probes 63. ~n
this case the pressure will drop to 10" Hg absolute, and
the absolute pressure transmitter, which may be one such as
the Series 1331 pressure transducer produced by the Rosemount
Engineering Co., of Minneapolis, Minnesota, will give a voltage
reading of 1.66 volts, instead of the five volts preYiously
shown. This voltage signal 74 will be fed into the direc~ion -`
comparator 76. In the direction comparator 76, the reading
of 1.66 volts, which is in the form of an analog voltage signal,
will be compared with the reference voltage supplied by the
computer 75. It will be recalled that the computer supplied
a reference voltage of two volts D.C. corresponding to the
desired manifold vacuum of lO.S n Hg absolute. Since ~he analog
signal is lower than the reference ~oltage signal, this means
that the carburetor throttle must be opened, so the direction
comparator will supply a signal to the stepping motor translator
- 18 -
`
4~ ~:
(motor control) 79. Such stepping motor translator may be
- such as type STM 1800, manufactl~ed by the Superior Electric
- Company of sristol, Connecticut.
! T~e stepping motor 42 is connected to the carburetor
33 by way of the linkage 53 as previously described. The
- stepping motor will then begin opening the carburetor throttle -
plate, resulting in a decrease in the manifold vacuum simul- ;
taneously with the further opening of the throttle plate. The
absolute pressure transmitter will immediately begin sending
a new voltage signal to the direction comparator. If the motor
control 79 were to have the stepping motor 42 open the carbure-
tor throttle approximately 3 and then compare the voltage
signal with the reference signal, one would find that the
analog voltage signal from the pressure transducer would equal
about 1.90 volts, instead of the previous 1.66 volts or, in
other words, the carburetor throttle will be approaching closer
to the desired position. In actuality, in the simple system
shown in Figure 5, the direction comparator can only give the
motor control a signal to open or close the throttle, and the
motor control drives the stepping motor 42 at a uniform xate
of speed, while the reference and analog voltages are being
constantly compared. When ~he reference and analog voltages are
equal, the direction comparator 76 will send a signal to the
motor control 79 to stop the stepping motor 42, at which point
the carburetor test will take place. The flow control system
just described and illustrated is an excellent systemand is
used in applications where the-volume requirements are relatively
~ow, and provides a speedy and accurate way of reproducing a -
desired air flow and manifold vacuum through a carburetor.
However, we were not happy with its perform~ce on high volume
carburetor testing stands because the single speed of the
- stepping motor still took much longer than desired to travel
-- 19 -- .: .
.'' . , ' f,
between carburetor test points. Therefore, we sought a way
to speed up travel between test points, and thus increase the
production of our test stands.
In order to achieve this increased in speeds, we chose
to have the stepping motor speeds proportional to the error
in the system. By error, we wish it understood that this means
the absolute value of the difference present at any given time
in the values of the re$erence voltage supplied by the computer
75 and the analog voltage signal supplied by the pressure trans- ~-
mitter 64. In other words, the absolute value of the reference
voltage, minus the analog voltage~ equals the error. Referring
to Figure 6, it can be seen that the addi~onal equipment
requixed to make this type of system work takes the ~rm of an
error calculator, which is in actuality an adder-subtractor
which may be of the type 301 manufactured by the Bell and
Howell Corporation and referred to by the numeral 80, and a
voltage to frequency converter 81. Such voltage to frequency
converters are well known in the art and need not be described
herein in detail. A voltage to frequency converter suitable
for our purposes may be the type 19 - 212 voltage to frequency
modules manufactured by the Bell and Howell Corporation.
For purposes of illustration, let us assume that we
still wish a carburetor to flow 0.2 pounds of fuel per hour at
19" of vacuum, which equals 10-1/2 n Hg absolute. In this
situation, the computer 75 will supply a reference signal of
two v~lts D.C. to the direction comparator 76. Again, at the
start of the test, the absolute pressure transducer 64 will
supply a voltage signal 74 of five volts D.C. to the direction
comparator 76~
The computer again will supply the appropriate signal
- 20 -
'.', .
to cause an air flow of two pou~ds per hour to flow through
the carburetor 33. At this point, the absoiute pressure trans-
ducer 64 will be supplying an analog signal of five volts to
,.
the direction comparator 76. Imm~diately upon opening the
variable area critical venturi meter 62 to permit air to flow
through the carburetor, the absolute pressure takes a large
drop to about 10" Hg abs~lute, at which time an analog voltage
signal of 1.66 volts is supplied to the direction comparator ,
76. Since the reference voltage is higher than the analog signal,
we must increase the flow and open the carburetor throttle plate.
The direction comparator 76 therefore will supply a signal to
the motor control 79 causing the stepping motor 42 to open the
carburetor throttle plate. Simultaneously, with the stepping
motor beginning to open the throttle plate, the error between
the analog and the reference signal is computed. In this case
the absolute value of the analog signal, 1.66, minus the refer- ;
ence signal 20, equals .34 volts. This .34 volt error is ed ;' ;
into the voltage to frequency converter 81.
As will be explained later, the voltage to frequency
converter 81 is available in a wide range of values. For pur-
poses of illustration we have chosen to show a voltage to
frequency converter with an output of 1,000 to 1, or in other
words, the .34 volts is changed to pulses of a freque~cy of
340 Hz, or working at a smaller unit of time, the ~oltage to
frequency converter will put out 34 pulses in 1/10 of a second.
These 34 pulses are immediately fed to the stepping motor trans-
lator 79 which then immediately turns the carburetor throttle,
in this case .09 per pulse, or 3 for the 34 pulses. This
3 movement of the carburetor throttle plate is many times
faster than the previous single speed system shown in Figure 5
'~
- 21 -
.
.,
i and, therefore, saves much time in going between test points.
This three degree movement of the throttle plate
causes the flow through the carburetor to increase, which
increasas the absolute pressure to l0.4" Hg absolute r compared
to the lO" Hg absolute previously, this in turn changes the
analog voltage being supplied by the absolute pressure trans-
ducer 64 to l.90 volts.
Since this is a continuous process, this voltage
again goes into the direction comparator where the error is
again computed by the error calculator 80. Since tbe reference
voltage is still greater than the analog voltage, the direction `
comparator again gives a signal to open the carburetor throttle -
further. This signal is supplied to the stepping-motor trans-
lator 79, and in turn to the stepping motor 42. Again,
simultaneously, the error is computed, in this case, the ;.
absolute value of the reference voltage, 2.00, minus the
analog voltage, l.90, equal .lO volts. This is converted to
j a lO0 Hz output by the voltage to frequency converter 81.
Again viewing a small amount of time, in l/lO of a second
10 pulses are supplied to the stepping motor 42 by the stepping
motor translator 79, to open the throttle 0.9 degrees more.
Again this increases the mass flow rate through the
flow controller, 73 which increasesthe absolute pressure .`;
therein to l0.55" ~g absolute, and changes the voltage supplied
by the absolute pressure tran~mitter to 2.0~ volts. The
voltage signal of 2.05 volts is again supplied to the direction
comparator 76, which in this instance finds the analog voltage
greater than the reference voltage, and directs the stepping
m~tor translator 79 to direct the stepping motor 42 to close
the carburetor throttle by way of the throttle linkage 53. ~.
~,:
- 22 -
~ ~4~
Simultaneously with the operation of direction comparator,
the error calculator 80 computes the absolute value of the
reference voltage minus the analog voltage. In this case
the value is .05 volts, which is changed to a 50 Hz output
by the voltage to frequency converter 81. Viewing our .10
second time span, in this amount of time five pulses will be
supplied to the stepping motor translator 79, which in turn
directs the stepping motor 42 to close ~he throttle plate 60
of the carburetor .45 by way of the carburetor linkage 53.
In this case, since we are closing the carburetor throttle plate,
the mass flow rate through the flow controller 73 and the ab-
solute pressure therein are reduced, which reduces the voltage
signal 74 supplied by the absolute pressure transmitter 64 to ``
10.49" Hg absolute.
It can be seen by the example up to this point, that
initially when the carburetor test started and the carburetor
started nn its way to the first test point, the movement of
the throttle plate 60 was very fast, but that as it comes
closer and closer to the desired test point, the movement of
the throttle plate is slower in proportion to the error. This
feature of our invention gives us the best possible advantage ~`
in locating test points, in that during most ~f the travel
between test points the carburetor throttle plate 60 will be
moved ~ery quickly, but as it approaches the test point it
will slow down, so as not to completely overshoot the test
point and bring about a hunting condition, where the throttle
plate continuously overshoots its desired setting in both
directions, but never reaches it, such as could occur with
the system as shown in Figure 5.
In fact, by choosing the ratio o~ the voltage to
23
frequency converter 81 carefully, for instance by choosing
a much smaller conversion ratio, for example 10 to 1 we would
approach the desired manifold vacuum very slowly, and never
overshoot it. The conversion ratio of the voltage to frequency
converter 81 is chosen for each particular carburetor model
to be tested.
To continue with our example, the absolute pressure
of 10.49" Hg absolute which is present after the carburetor
throttle 60 was closed 0.45, causes a voltage signal 74 to
be sent by the absolute pressure transmitter 64 to the direction
comparator 76. This voltage signal has a value of 1.99 volts
D.C. The direction comparator will see that the reference
voltage is now again larger than the analog voltage. The direc-
tion comparator 76 will supply a signal to the stepping motor
translator 79, (motor control) and thus to the stepping motor
42, to open the carburetor throttle plate 60 by way of the
throttle linkage 53. Again, simultaneously, the error calcu-
lator 80 calculates the error as .01 volt D~Co ~ and the voltage
to frequency converter 81 converts this to a 10 Hz output.
Again, in a l/lOth of a second, one pulse is supplied to the
stepping motor translator, which opens the carburetor throttle
linkage in the manner previously described, an additional .09
degrees. This additional opening of the carburetor throttle
plate 60 again increases the mass flow rate through the flow
controller 73, which increases the pressure therein to 10.5"
Hg, which has an analog voltage of 2 volts D.C.
Since the analog voltage now equals the reference
voltage, the direction comparator 76 supplies a signal to the
stepping motor translator 79 to open the carburetor throttle
further. However, since the error is now zero, no pulses are
- 24 -
,.
- - . . . ... .. . . .. . . , ...... , . . . ',:,
3 ~f
supplied to the stepping motor translator 79, and the carburetor
` throttle remains steady, having arrived at the point where
the desired manifold vacuum is supplied. We have thus invented
a computer operated operations reproducing system in which the
speed of change of the carburetor throttle plate during its
travel between test points is initially very fast, much faster
than the older pneumatic or electrical systems, but which is
slowed down as it approaches the test point in proportion to
error, therefore, the test points are obtained quickly, with ~-
a minimum of overshoot, and very little hunting.
Whereas in the old pneumatic systems previously
described r to get a known manifold vacuum to occur at a certain
throttle plate position, say 70, there might be thirty seconds `
or so to hunt for the proper test point, due to the fact that
the old pneumatic throttle positioners were only single speed
devices operating in a very slow manner. Even the old electric
operations reproducin~ system, which had two ~peeds at which
the throttle could be operated, was not-much better than the
pneumatic system, while our new system, with the variable
speed computer controlled throttle positioner, wiIl arrive
at a test point in as much as 50% less time.
Our improved system could even be vperated manually
if this was desired for some reason. Referring to Figure 7,
the only difference between the system illustrated therein
and the system shown in Figure 5 which wss previously described,
is that a manually operated potentiometer 85 will take the place
of the computer 75. In this case the operator would set the refer
ence voltage of two volts on the potentiometer, and would manually-
set the flow controller 73 to the desired air flow. From then
: '
- 25 -
.. . .. ..
,,~,,r)~
on the operation of the system would be identical with that
described for Figure S of the system in which the stepping
motor translator 79 would control the stepping motor 42 at a
single speed.
- A manually operated system in which there is propor-
t~onal control of the stepping motor 42, is shown in Figure 8.
Again, the operator would manually set the reference voltage
of two volts on the potentiometer 85, and manually set the
desired flow through the flow controller 73. -~
If desired, the proportional fl~w control system in
Figure 6 can be simplified as shown in Figure 4. In this
instance a computer 86 would be used to perform the functions
of not only supplying the reference voltage and setting the
desired flow through the flow controller 73, but would also
take over the functions of the error calculator 80 and the
voltage to frequency converter 81.
In any of the systems shownin Figures 4 - 8, the
stepping motor has so far been assumed to be a standard stepping
motor directly connected to the carburetor linkage 53. However,
it can be seen that in any of the systems, travel between
test points could be faster if in addition to the proportional
control provided by Applicants' system, a mechanical advantage
from the stepping motor itself could be had. In other words,
the opening of the carburetor throttle linkage would not be
of a uniform amount for each degree of rotation of the stepping
motor. However, deciding where the mechanical advantage
should be had, and the designing of a throttle drive mechanism
to produce such a mechanical advantage, presented problems of
major proportions. At very low carburetor air flows, such as
',''.' ,
~ 26 ~ ~
at idle and off idle, very fine resolution must be had. In
other words, very close control of the throttle plate must be
had, and no additional advantage over Applicants proportional
control of the stepping motor is needed. However, due to the
nature of the carburetor, the part-throttle and full throttle
test points require much less resolution but much more turning
of the throttle plate to get from the idle and off-idle test
points to the part-throttle and full throttle test points. We
have determined that it is in going to the part-thxottle and
full ~rottle test points where much time could be saved and
where greater speed in the movement of the carburetor throttl2
plate per degree of rotation of the stepping motor would be
. . .
;of definite advantage.
We at first tried the use of spring loaded regular
spur gears with the holes drilled in offset positions. How-
ever, this proved completely unsatisfactory as the gears either
would not stay together, or jammed up, rendering the apparatus `
controlling the carburetor throttle plate completely ineffective. !
Further work on this problem lead to the idea that -~
elliptical cams would give us the desired mechanical advantage
without the problems of the offset spur gears, and we tried
a pair of elliptical cams tied together with a fine cable to ;;
drive the carburetor throttle linkage 53. This indeed, as
shown by Figure 12, gave us the advantage we w~re looking for,
with very close control-of the throttle plate and with the
mechanical advantage we desired. For example, at idle condi-
tion where close resolution is required, if the stepping motor
would change 10, the position of the throttle linkage would
change only 5, or, ~n other words, for each degree of rotation
of the stepping motor, the carbure~or throttle plate would be
- 27 -
.
~'' `,,.
moved one-half degree, while at or near wide open throttle,
for each degree of movement o the stepping motor, the throttle
plate would be moved 2. The advantage of this from the
previous description is obvious, as the movement between test
points would be very rapid, and the test point at wide open
throttle would be very rapidly reached.
The stepping motor has so far been assumed ~o be a
standard stepping motor directly connected to the carburetor ~
linkage 53. Such a stepping motor might be such as the type -
HS-50-P3 Slo-Syn (Trade Mark) Stepping motor manufactured by `~
the Superior Electric Company of Bristol, Connecticut. Such
a stepping motor contains internal planetary gear which has
a mechanical advantage of approximately 100 to 1. Whereas this
stepping motor 42 rotates 1.8 degrees for each pulse applied
by the motor controller 79, the output from the planetary gear
105 rotates the throttle pl~te 0.018 degrees. At maximum
speed, Applicants system could therefore operate 90 degrees i`
from idle to full throttle, in seventeen seconds. It can be
seen that use of a planetary gear 105 which has a mechanical i
.~. . ..
advantage of 20 to 1 will incxease the speed of operation. How~
ever this system, which rotates the throttle plate 0.09 degrees
for each pulse applied by the motor controller 79, does not `
have adequate resolution at idle and off idle flow points. '
But, by using a motor controller 79 which causes `
the stepping motor to increment in half-steps, the stepping
motor movement can be changed ~rom 1.8 to 0.9 for each pulse
of the motor controller. Such a motor controller is described
in the Sigma Stepping Motor Handbook published in 1972 by Sigma
Instruments, Inc. of Braintree, Massachusetts, starting at
page 25, and is commercially available as Model No. 30003,
- 28 -
? " ' , . ' . ~ ~ -i " ' . . . ' ' ` , ,, ~ . . .
~ i
manufactured by Scans Associates, Inc. of Livonia, Michigan~
However, the use of the motor controller does not change the
angular speed of the stepping motor, but only allows closer
control of it. -
Thus, we can now use a stepping motor 42 having an
internal planetary gearset 105 with a 20: 1 ratio.
The advantage of this motor controller and non-linear
mechanical advantage from the previous description is obvious,
as the resolution at idle flow points will be 0.022 degrees per
pulse, the resolution at full throttle will be 0~0~8 degrees per ;
pulse, and the test point at full throttle would be rapidly
reached. ~`~
While our system without the elliptical arrangement -
could go from idle to full throttle in 17 seconds, our new
system with the non-linear mechanical advantage of the elliptical
gear arrangement could perform the same operation in 3.5 seconds.
In other words, almost five times as fast. Saving this much
time in high production carburetor test work indeed is a sig- ~ `
nificant advance in the art.
In this system, the throttle stepping motor 42 is
mounted on a frame mem~er 87, which may in turn be mountea on
a carburetor test stand 29. Drivingly connected to the sha~t
88 of the stepping motor 42 is a clutch assembly 89. The
clutch assembly is in turn connected to a first elliptical gear
90. This first elliptical gear 90 is drivingly engaged with
a second elliptical gear 91 rotatably mounted on a shaft 92
carried by the frame member 87. Fixedly mounted to the gear
91 is an adaptor plate 93, on which is mounted a pin 94, such
pin 94 would correspond to the stud 52 shown in Figure 2, and
would engage the carburetor ~hrottle linkage 53 for movement
_ ~9 _ :
,
.
of the carburetorthrottle plate 60.
A modified version of this arrangement shown in
Figure 11 and Figure 2 would have the stepping motor 42 and
the clutch assembly 89 offset from each other inst~ad of in
line as shown in Figure 9. In this case, the stepping motor
and clutch would be drivingly connected by a pair of spur gears
95 and 96. The clutch 89 would have the elliptical gear 90
mounted on the opposite end of the shaft from the gear 96 and
drivingly connected to the second elliptical gear 91, on which
the stud 52 would be mounted. If required, as shown in Figure `~
2, the stud 52 could be mounted on a separate crank 49.
A more graphic illustration of the mechanical advantage
offered by the elliptical gears can be seen by referring to
Figures 13 - 16, which are based on the graph shown in Figure 12.
.
Figure 13 shows the relationship of the first elliptical gear i~
- 90, and the second elliptical gear 91, at their starting or
carburetor throttle idle position. ,~
Figure 14 shows the two elliptical gears 90 and 91 at
a position where the carburetor throttle plate has been opened
but is still at a near idle condition. When the gear 90 has
turned 30, as shown at point A in Figure 12, the gear 91 has
- moved only 15, while in Figure 15, as the gear 90 continues to
move away from the carburetor throttle closed position, due
to the elliptical shape of the gears, a mechanical advantage ~-
starts coming about in the movement of the gear 91. When
the gear 90 has turned 45, as shown at point B, the second
elliptical gear has now come to a position 30 from its closed
position. In other words, for gear 91 to move the first
fifteen degrees, 30 of stepping motor rotation was required,
- 30 -
.~ :
or 2 degrees of stepping motor xotation equals one degree ~
of throttle plate. The next 10 degrees of throttle plate -
opening took only 15 degrees of stepping motor rotation.
By referring to Fig~re 12, by the time the stepping
motor is moving from the 70 degree to 80 degree position, each
degree of stepping motor rotation results in two degrees of
throttle plate opening. Thus the very fine resolution needed at
idle is achieved at the same time, very fast movement of the
; throttle plate, where required, is provided.
There is thus provided an improved method and apparatus `
whereby the objects of the present invention and numerous
additional advantages are attained.
~;' '.
- 31 - ~
