Load Cell Parts

You want to build your own load cell? That's a little ambitious, I think. If we forget about hermetic sealing, you need to get a strain gauge and glue it to a metal block or cylinder (stainless steel is preferable). The load capacity is more a function of the metal block's capacity rather than the strain gauge. I have no idea what glue or adhesive to use though.

The cylinder deforms slightly when a weight is placed on it, decreasing its height but also expanding horizontally. The strain gauge follows this movement and changes the values of the resistances (A & B increase resistance while C & D decrease resistance). If you don't exceed the metal's load capacity, it will return to its original shape once the weight is removed.

I think building your own strain gauge would be a little out of your league (it certainly is out of mine). You can buy them though. Omega sells them as well as the adhesives to use.

There's much more to load cells than that, however. We haven't touched temperature compensation yet. We also didn't touch balancing of the wheatstone bridge. One thing nice about Omega Engineering though, is that they also have handbooks and/or tutorials. Go and visit them.

Vulcan is right, it`s not so complicated to build a load cell. The accuracy is mainly depending on how exactly you fix the strain gauges. The surface has to be very clean and the glue as thin as possible to couple the cells strain directly without losses and unlinearities into the gauge. The error follows the cos-function so exact angles are recommended.

The glue chosen best for the project is determined by the cells material (aluminium, stainless steel etc) and has adjusted temperature specs because the output levels of the Wheatstone bridge are in the µV-range. Wiring has best to be done with special materials because thermocouple effects would produce terrible errors otherwise.

Best results can be achieved by powering the bridge with an AC reference voltage in the lower kHz-range and with a small-band lockin amplifier. Sounds expensive but must not be - the more quality your signal has the less must be done with temp-compensating, filtering, shielding, testing .....

When you look for calculation examples then hbm.com (Hottinger Baldwin Messtechnik in Germany) is a good choice, they also sell gauges and glue and have very good application examples with calculations of signal magnitudes as a function of mechanical dimensions and chosen materials.

Good Luck
Uwe

Surface preperation is everything when attaching strain gauges. The surface will need to be mechanically cleaned, degreased, etched with a mild acid solution & cleaned with a mild alkali to give a neutral Ph balance. The glue line then needs to be as thin as possible with no voids. This is achieved by attaching the strain gauge with tape, peeling it back then placing a small fillet of cyanoacrylate adhesive in the corner between the gauge & the mounting surface and using the gauge to squeegee the adhesive along the length of the joint. You then clamp the gauge using a soft silicone rubber pad. Strain gauge measurement only works if the movement of the substrate is exactly transferred to the gauge. Check out the strain gauge suppliers, many will sell a kit of materials needed to attach gauges.

See http://www.nakka-rocketry.net/strainlc.html and http://www.nakka-rocketry.net/strainlc.html#Excel

Don't wast your time or money. You can buy a much better load cell from some of the Chinese suppliers for around $40 US

Do you know of a source in China??

www.ascells.com