Transcript Document

EEF Contribution to the Business submission
Tony Clohesy, Pera
Submission: work completed on the SMED
prototype (Single Minute Exchange of Die)
My placement
Pera Innovation, Melton Mowbray
I have been at Pera since August, and have done work on a
number of projects such as the one shown, WREED, an airless
dryer for food which uses superheated steam.
I did a lot of work on a temperature control system for a
medical supplies box, MEDICASE.
I have also conducted a number of tests on various types and
constructions of insulation to increase the efficiency of home
delivery – E-FLEX
The project my submission focuses on is SMED, a short-term
prototype building project where I was heavily involved in all
aspects of the project from design and construction to testing
and documentation.
The Task
SMED prototype
Develop a method of improving
the efficiency of food processing
Specifically shortening
changeover time for manually
adjusted equipment
Demonstrate the concept on a box
folding machine
Provide a fully computer
controlled demonstration
prototype for display at relevant
conferences
Overview of the project
•A copy of the box
folding machine
from a large-scale
bakery had been
produced and fitted
with linear
actuators.
•Motor control
circuits were used
to regulate the
positioning of the
various parts.
Linear Actuators
•There are four actuators, two
controlling the distance
between the guiding plates
and two controlling auxiliary
arms behind the guiding
plates.
•They are connected via the
motor control circuit to a 12V
supply, and also have a 5V
current across the
potentiometer to allow
position information to be
read back from it.
Micro-controller
•The micro-controller is
mounted on a demonstration
board which has additional
features such as a
programming port, serial
communications port and
banks of inputs and outputs
including a 5V supply used
by other circuits.
•Outputs include the 5V
supply to the motor
potentiometer and motor
control signals to the relevant
circuit.
Motor Control Circuit
•I designed and built this
circuit around a chip
containing a commonly used
circuit known as the Hbridge.
•The H-bridge accepts logic
signals from the microcontroller and channels a
useable supply to the motor.
Depending on the inputs, the
chip makes the actuators
move in or out.
Buffer Circuit
•Operational amplifiers in this
context are used to ensure the
voltage being read is not distorted
by internal resistances. The
positioning of the actuators is
crucial, and this circuit was added
after preliminary testing to ensure
accurate readings from the
potentiometer feedback.
•I designed and built this circuit
on strip-board and it intercepts the
feedback wire from the
potentiometers to the microcontroller.
Regulator Circuit
•This circuit was built using
linear regulators for the
purpose of protecting the
delicate components of the
control circuits by means of
current limiting capabilities.
•Variable resistors are used
with the linear regulator chips
to control the voltage to the
actuators and hence the speed.
Main challenges (1)
•Basic method worked, but a
few hitches were uncovered in
initial tests:
High current draw burnt out Hbridge chips in the motor control
circuit
Inaccurate measuring of position
Fast movement caused overshoot
of target position.
Slow recognition of target
position by interrupt sequence
Main challenges (2)
•High current draw:
Initial movement of the actuators
required a higher than nominal
current, and when all four moved
at once this current was sufficient
to overheat the H-bridge chips in
the motor control circuit.
•Different sized resistors were
used to reduce the voltage during
testing, but then the power supply
was changed and the regulator
circuit introduced.
Main challenges (3)
•Inaccurate position measurement:
The position was measured by the
potentiometer, but internal resistance
meant that the values entering the
micro-controller were not reliable.
The addition of the buffer circuit
meant that the correct voltages could
be read with no adverse effects from
other resistances.
Main challenges (4)
•Fast movement caused
overshoot of target:
A ‘hard stop’ option is available
where the motor is stopped dead
instead of slowing to a stop. This
draws a lot of power, but is more
accurate. We decided not to use
this, but to run the motors slower
and to increase the frequency of
potentiometer readings by the
micro-controller.
Main challenges (5)
•Slow target recognition:
The sequence in the code which
runs every few milliseconds
performs the actions required by
the serial cable communications,
and checks the position at every
cycle.
This was not checking fast enough
to give high-resolution feedback
during motion, so the code was
changed to increase cycle time.
Extra work
•All circuits were mounted in an aluminium box.
•Mains power socket with lights and switches mounted in
side.
•Connectors from the box to the actuators to allow
disconnection during transit were wired up and mounted.
•All relevant datasheets and technical information was
gathered for the main components used.
•Circuit diagrams and design drawings were produced for all
parts of the prototype. It is our intention to ensure the
prototype is suitable for CE marking.
•Records of purchases and man-hours spent have been kept
up-to-date.
Impact of project (1)
•It is too early to assess the impact in real terms as yet, but
there a number of factors to consider:
Pera retains a valuable client
Prototype displayed for potential investors to see
The changeover time of this particular piece of equipment
has been identified as the main Limiting Factor of the
production line, and has the potential to save a lot of
monetary and labour costs associated with flexibility of
production.
Impact of project (2)
•Increasingly, the food industry is seeing the necessity for
rapid changeover as batch sizes get smaller and flexibility is
crucial for success.
•When manual changeover of equipment such as the box
folder takes up to 10 or 15 minutes to complete, with all the
added safety risks involved, it is not financially viable to
change the settings for batches which alter every hour.
•For a £15,000pw output, every minute is valuable, and if you
are changing batches hourly, every week:
Manual changeover would waste 400 minutes and £2500
SMED changeover would waste only 40 minutes and £250
- a saving of 6 hours of skilled labour and £2250
Conclusion
•My personal development:
Teamwork skills, Communication
Manual skills, Electrical
knowledge
Problem solving, Organisation
•The prototype:
The SMED concept has been realised with a working display
model which can be easily transported and set up using only a
standard mains plug and a laptop. The combination of computer
program, micro-controller program and circuitry presents a
robust user interface ideal for exhibitions and potential new
clients.