whit@milton.u.washington.edu (John Whitmore) (07/18/90)
In article <4145@ash31.UUCP> svoboda@motcid.UUCP (David Svoboda) writes: >bill@videovax.tv.tek.com (William K. McFadden): >> >> It seems that since the carpet has a springy pad underneath, the scale will >> read too low. This is because the scale will only measure the compression of >> its internal spring, and some of that compression gets used up by the carpet. > >Wait a second. Think about this for a moment. What EXACTLY do you mean by >"compression get used up by the carpet"? We are talking about forces here. I took one of these apart (the strain gage kind), and found the load force was transmitted through levers to a calibrated, temperature compensated (four lead) strain gage. Those levers had a fulcrum near each foot, and used the CENTER of the bottom cover plate as the fixed end for the strain gage. So, if you support the scale on a carpet, the bottom plate is bowed (by the carpet pressing in its middle), and the geometry of the levers changes (by the bow angle). The top plate was VERY heavy plastic construction (I'd have thought steel would work better); the bottom plate wasn't, so it would probably deflect considerably on a thick rug. Parenthetically, I found the electronic construction/design to be quite adequate. The mechanical design, though, was only so-so. I disassembled the thing because someone had stepped on it too hard, and broken the strain gage. It looked like that was guaranteed to happen by the attachment mechanism of the strain gage (too high stress at the dowel pin onto the foil of the gage). John Whitmore