Design help wanted - Car Scale
Archiver
Posts: 46,084
Jim, Regarding your query for suggestions on building/buying load
cells for use in a car scale. Perhaps I can be of help.
Yes, you can design and build your own cells but it takes some
education on the use of strain gages and knowledge of
materials/structures. The gages are reasonablely priced [noparse][[/noparse]$5 to $10 /
gage - need perhaps 4 per cell] but other supplies are needed. To
get repeatable/reliable output you must follow the gage producer's
instructions regarding bonding, protection, wiring, etc. And you
cannot expect the same efficiency [noparse][[/noparse]millivolts out / pound weight] as
bought. Bought give around 30 millivolts out at rated [noparse][[/noparse]full scale]
weight whereas homebuilt give around 1 to 10 millivolts.
Therefore noise, amplification, and other problems are greater.
Bought cells are also very linear, but if you calibrate your scale well
and use a generated equation [noparse][[/noparse]Ref: Stamp I manual, Application #7
"Using a Thermister"] or a lookup table [noparse][[/noparse]Tracy Allen's math notes]
you can readily handle nonlinearity with the Stamp. Offset
[noparse][[/noparse]unwanted millivolts out for zero weight - or due to the tare weight
of the empty scale] can also be readily taken care of with a
potentiometer or within the Stamp program.
I can discuss this more throughly if interested.
You aught to be able to buy used cells - I don't have any
information - but you might still want to recalibrate; either by doing
it yourself [noparse][[/noparse]lower accuracy] or by sending them to a testing
laboratory. For a crude calibration you could use bags of sand or
cars which have been weighed on a certified scale. If the used cells
were made by a reputable manufacturer and were not abused [noparse][[/noparse]ie
overloaded] they should remain linear: so only a limited range of
calibration would be needed. By-the-way, you can generally tell if a
used cell has been vastly overloaded by the fact that they have
offset [noparse][[/noparse]do not read zero output for zero weight].
Again, I will be willing to discuss this further if desired.
cheers, and much success, David R. Erskine
cells for use in a car scale. Perhaps I can be of help.
Yes, you can design and build your own cells but it takes some
education on the use of strain gages and knowledge of
materials/structures. The gages are reasonablely priced [noparse][[/noparse]$5 to $10 /
gage - need perhaps 4 per cell] but other supplies are needed. To
get repeatable/reliable output you must follow the gage producer's
instructions regarding bonding, protection, wiring, etc. And you
cannot expect the same efficiency [noparse][[/noparse]millivolts out / pound weight] as
bought. Bought give around 30 millivolts out at rated [noparse][[/noparse]full scale]
weight whereas homebuilt give around 1 to 10 millivolts.
Therefore noise, amplification, and other problems are greater.
Bought cells are also very linear, but if you calibrate your scale well
and use a generated equation [noparse][[/noparse]Ref: Stamp I manual, Application #7
"Using a Thermister"] or a lookup table [noparse][[/noparse]Tracy Allen's math notes]
you can readily handle nonlinearity with the Stamp. Offset
[noparse][[/noparse]unwanted millivolts out for zero weight - or due to the tare weight
of the empty scale] can also be readily taken care of with a
potentiometer or within the Stamp program.
I can discuss this more throughly if interested.
You aught to be able to buy used cells - I don't have any
information - but you might still want to recalibrate; either by doing
it yourself [noparse][[/noparse]lower accuracy] or by sending them to a testing
laboratory. For a crude calibration you could use bags of sand or
cars which have been weighed on a certified scale. If the used cells
were made by a reputable manufacturer and were not abused [noparse][[/noparse]ie
overloaded] they should remain linear: so only a limited range of
calibration would be needed. By-the-way, you can generally tell if a
used cell has been vastly overloaded by the fact that they have
offset [noparse][[/noparse]do not read zero output for zero weight].
Again, I will be willing to discuss this further if desired.
cheers, and much success, David R. Erskine