Force sensors - what to use?

Hi, I am looking into making what will essentially be 3 sets of scales (each one being made of 3 force sensors sandwiched between 2 polycarb discs). These will be used to estimate the amount of beer left in some homebrew kegs so I know when to start brewing again and avoid running out of beer (don't want to move the kegs as this could disturb any sediment that might collect in the bottom of the kegs). I have been looking at some piezo force sensors and force sensitive resistors. The FSRs on Sparkfun are not recommended to be used as scales so that leaves the piezos, and at $20 a pop (from Sparkfun also) that blows my budget by miles. Does anyone know what electric scales use, as I have seen these for about $20, so (assuming they have 1 sensor per foot otherwise they would be hugely inaccurate depending where you stood) this would be $5 per sensor (max).

So, what do you guys recommend using? Could I just grab a few electric scales and use the sensors out of them? or should I go with the Sparkfun sensors (even though they seem a bit expensive), I can't find anything on eBay, etc but I'm probably not searching for the right word! A bit more info on accuracy, the kegs hold about 20 litres each so I'm assuming that's 20 kgs, and I will probably display an output in 5% steps (so 1 kg steps). The empty kegs are about 3.5 kg. So I would want to measure between 3 - 24 kgs in increments of 1 kg +/- 0.5 kgs. Any help or insight is much appreciated. Cheers!

I can't find anything on eBay, etc but I'm probably not searching for the right word!

Load cell is the best word for searching, strain gage next best word. They tend to be industrial and expensive when new, but of course bargains can be found on E-bay at times, so worth a look.

Lefty

Did a quick google search on "how does an electronic scale work" and the results were that the common design is to use a load cell. I might go down this route, either using bathroom/kitchen scale parts if I can't find a cheap sorce of load cells that can handle ~30 kgs. Thanks Lefty ;D

More googling reveals that it looks like there are 4 load cells in one of those glass top bathroom scales (one in each corner). I am assuming that these will be able to take 35+ kgs each as the whole unit can take 150 kgs. I might go grab a cheap $20 set of scales (load cells on ebay were about $17 for a 15kg one) and break it down to see whats inside and do some testing on the sesitivity of the load cells at <5 kgs. I’ll try to keep this updated

I'll try to keep this updated

That would be great. And take pictures during tear down and mock-up. I think many Arduino users would really benefit if given an opportunity of finding low cost source for load cells and learning how these sensors work. They can be quite accurate and sensitive and could be used in so many DIY applications. Also you end up interfacing a these resistance bridge type sensor to an Arduino analog input would be useful to explain.

Lefty

Finally took apart the families bathroom scales (well one corner at least) and this is what I found

seems like a very basic load cell, my scale has four (one in each corner, well I assume that because all corner look the same)
The scale runs off 3v, 2x 3v batteries in parallel (again I am guessing as when 1 battery was remove the unit still worked)
Next step is to buy one and test it :slight_smile:

Hey there. I found this thread by searching for a similar homebrewing project. Funny how that works. :)

Any update on this project? Which scale did you end up buying? How did it work out for you? What was the accuracy of those strain gauges?

Also, I ran across this project on nerdkits.com. The scale they took apart had a single strain gauge attached to a mechanical system that read the distributed weight from a single point. I'm assuming this was so the manufacturer could save a couple bucks on strain gauges.

http://www.nerdkits.com/videos/weighscale/

Thanks.

How about using just one sensor, instead of three?

You can: Make a triangular platform. Two corners supported on posts, the last with a load sensor.

As long as the beer keg's position on the platform remains constant, the load on that sensor will accurately reflect the beer in the keg. (Assuming a keg sitting in the usual orientation, with the usual contours.)

Now... how about not bothering with a fancy, expensive analog load sensor? Put a spring under the "sensing" corner. And fix up simple microswitches to be closed when the spring is squashed "a little", "quite a lot", and "all the way". You can easily fiddle with, fine tune, both the spring and the microswitch positioning. Now you can have the Arduino report "Getting low", "Quite low", and "Mayday! Mayday! Mayday!".

Or, you get a little fancier and use a linear pot to measure how much the spring is compressed, and get an analog value for load on platform.

I did see that article on Sparkfun. That's pretty cool... kinda like the kegbot. I think the OP's concern was the cost of those piezoresistive sensors. They're about $20 each. I was thinking you could get four of the strain gauges from a scale for $20 total. I was also hoping to weigh any number of individual platforms rather than the whole thing.

I hadn't thought of the spring and pot idea. Do you have any links to more information on this?

If a little bit of movement of the keg is ok, then more mechanical setups are possible. If one side of the keg could move up/down ~1", then there are numerous spring/pot seups available to measure that movement.

Hey, I haven't progressed any further with this since the original idea. Yeah the idea about this came from the sparkfun article but not wanting to spend $20 per sensor which could end up cost $60 per keg (about 2x the price of a corny keg in the US!) jpl do I recognise your name from HBT? As I said I haven't gone any further but the scale I took the photo of had 4 sensors and only starts to display a weight when its over ~15 kg, but that could just be the scale programe and nothing to do with the strain gauge. Good luck if you try this out and let us know what you use :)