It is hard to give good advice in this case because it is not clear
what skills you already have. If you are really flat broke then you
will have to build them yourself. This is not as bad as it sounds and
is very rewarding.
1)
The sensor mentioned by the previous Commenter may be the National
Semiconductor LM335:
http://www.national.com/pf/LM/LM335.html
I have used these and find them to be excellent. They are cheap and it
is easy to get an accuracy of 0.1 deg C or less. The output is
extremely linear and could be read directly by an ADC.
For temperature measurement you might also like to look at the
interesting 1-wire range made by Dallas Semiconductor. The interface
(1-wire) is a little odd and may not be appropriate for your PLC
environment. However they do have the advantage that any number of
them can be daisy-chained so that you can easily measure many
temperatures, and only one pair of wires runs around the room.
2) For a very robust, noise free position measurement over this range
(and with this accuracy), consider LVDTs (Linear Variable Displacement
Transducers). They are inductive devices that measure the position of
a ferrite core using a clever phase sensitive circuit. The problem is
that the commercial ones are expensive. If you are good with your
hands it is possible to wind your own, please ask if you would like
help with this.
Because you only want 0.1" accuracy, consider building an optical
sensor out of a transparent strip of plastic. If the plastic is
printed on a laser printer and has dark bands at 0.1" spacing, you can
probably pick these up using very cheap optical sensors. I am thinking
here of the little modules that have an LED and a phototransistor
facing each other across a gap. If several are mounted at a pitch that
is not an even multiple of 0.1" (for example 0.515" apart), the dark
bands will pass them at different times and this can be used to get
even higher resolution than 0.1".
A common 10-turn potentiometer can be set up to give quite good
position measurements. Wind a string around the shaft and then run one
end to your object. The other end has to be on a spring.
I could give you better advice with more background information. For
example is there dust or serious electrical noise? Is vibration a
problem? What are the consequences if the position sensor gives the
wrong reading?
3) Strain gauges are the classic solution to this problem. Strain
gauges themselves are not too expensive, only $USD5 for a card of 4,
enough to do one strain measurement. The expense is in the amplifiers.
I build my own using the INA2126 instrumentation amplifier, these are
about $USD10 each and only a few other cheap components are needed.
The commercial version runs to many hundreds of dollars but may not
provide much more function. A circuit diagram for a bridge amplifier
may be seen in the data sheet for the INA2126:
http://webs.cs.berkeley.edu/tos/design/data_sheets/INA2126_datasheet.pdf
Note that you will need to mount the strain gauges themselves
correctly on the forklift. This will mean checking that the target
load causes a surface strain of at least 10 parts per million. Also it
is important that the strain you measure is only due to the target
load and not to other machine conditions. Also you must protect the
gauges from physical damage, if anything touches them they will break. |
