Finding a Precise Scale for Measuring Human Body Weight Fluctuations

[I_am_learning]

I was lately interested in learning how the human body weight changes through out the day. Like, how much weight (may be in just 10s of grams) is lost during sleeping, after bathing, after workout and like that.
For that I would require a highly precise scale. But most commercially available scales' precision is only upto 100 grams which is not useful for my purpose. When I dug in further, some industrial scales do provide upto 10 grams precision but they (for 100 Kg max) cost in the range of $2500, which is every expensive.
I am looking for some cheaper means of measuring human body weight fluctuations precisely.
There are kitchen scales for just $10 which have precision of 1g but they can measure only upto around 5Kg.

I can see that its much more difficult (and hence expensive) to have the same precise measuring ability over a wide measurement range.

So, If my weight is going to fluctuate by, let's say only about +-2.5 kg, can I some-how modify the cheap kitchen scale to work around the point of my average body weight?

I do have electronics and micro-controller knowledge, but don't have any knowledge about the weight sensor.

And, at last, does anyone know about any research publication/works about this quantified changes of human body weight everyday?

 

[AlephZero]

You should be able to find a load cell that will meet that accuracy requirement cheaper that $2500.. For example this has a range of 100kg and accuracy 20g. That was the first one Google found, so you can probably do better.

http://www.robotshop.com/en/type-s-load-cell-100-kg.html
http://www.robotshop.com/media/files/pdf/datasheet-3138.pdf

The mechanical design of a system so you don't "lose" 20g through a load path that doesn't go through the load cell, is a another problem. That is easy if you are measuring tension in a rope, but not so easy if you want a stable platform to stand on. Maybe you need 4 load cells, one at each corner.

Of course if you get this to work, you may find it responds to the the accelerations caused by heart beats, as well as the "weight" you are trying to measure!

 

[I_am_learning]

You should be able to find a load cell that will meet that accuracy requirement cheaper that $2500.. For example this has a range of 100kg and accuracy 20g. That was the first one Google found, so you can probably do better.

http://www.robotshop.com/en/type-s-load-cell-100-kg.html
http://www.robotshop.com/media/files/pdf/datasheet-3138.pdf

The mechanical design of a system so you don't "lose" 20g through a load path that doesn't go through the load cell, is a another problem. That is easy if you are measuring tension in a rope, but not so easy if you want a stable platform to stand on. Maybe you need 4 load cells, one at each corner.

Of course if you get this to work, you may find it responds to the the accelerations caused by heart beats, as well as the "weight" you are trying to measure!

Thanks. Seems like I need to improve my googling skills; I didn't know about 'Load cell' keyword, and was using 'scale' and wasn't getting desired results.
Now, I am getting lots of "load cell" results, but all of them have around 0.02% non-repeatability
which is 20g form 100kg, but maybe I am not searching right.
Its quite surprising that just for an increase form 100g (most commercial bath-scalse) to 20g precision, things start to get so expensive.
And, since most strain gauge have overload capacity of just 100%, I don't think I can use a 10kg scale and offset it to my avg. weight.

So, achieving like 1g precision seems impossible.

 

[AlephZero]

Its quite surprising that just for an increase form 100g (most commercial bath-scalse) to 20g precision, things start to get so expensive.

Measuring anything to better than 0.1% precision is hard, and not a common requirement.

What use you can make of knowing the mass of an adult human, measured to the accuracy of 1cc of water (i.e. 1g) is another question, of course.

 

[I_am_learning]

What use you can make of knowing the mass of an adult human, measured to the accuracy of 1cc of water (i.e. 1g) is another question, of course.

Lots of things could be done. Like, find out how much air you breathe in and out. How much weight you loose (possibly through perspiration) on hot sun. How much is lost after overnight sleep. etc

Whats the use of knowing all these, is actually what I can't answer. :) Just kill my curiosity, perhaps. :)

 

[jbriggs444]

Lots of things could be done. Like, find out how much air you breathe in and out.

The air you breathe in and out does not affect the down-force you exert on the scale. Air is neutrally buoyant in air, after all.

 

[AlephZero]

The air you breathe in and out does not affect the down-force you exert on the scale. Air is neutrally buoyant in air, after all.

The air you breathe out has a different chemical composition (more CO2, usually more water vapor) and is at a differerent temperature from the air you breathe in. So the buoyancy effects will make a difference.

But trying to measure weight to within 1g, I would be more concerned about random body movements. An acceleration of 10^-5g is enough to cause a 1g apparent change in weight of a 100 kg mass. Compare that with the fact that if somebody is sitting on a cheap and not very rigid plastic chair, you can easily observe their heartbeat rate just by watching the motion of the chair as it deforms with each beat.