So you’re a boxer, and you’re weighing in just 80 micrograms too much for your usual weight class. How many eyelashes do you need to pluck out to get back in the ring? Or maybe you’re following the newest diet fad, “microcooking”, and a recipe calls for 750 micrograms of sugar, and you need to know how many grains that is. You need a microgram scale.
OK, we can’t really come up with a good reason to weigh an eyelash, except to say that you did. Anyway, not one but two separate YouTube videos show you how to build a microgram balance out of the mechanism in a panel meter. You know, the kind with the swinging pointer that they used to use before digital?
Panel meters are essentially an electromagnet on a spring in the field of a permanent magnet (a galvanometer). When no current flows through the electromagnet, the spring pulls the needle far left. As you push current through the electromagnet, it is attracted to the fixed permanent magnet, fighting the spring, and tugs the pointer over to the right. More current equals more pull.
The first link, from [Paul Grohe] of Texas Instruments, demonstrated the basic principle. Arrange the meter so that the arm falls into the path of an opto-interrupter, and then build a small circuit that passes more current through the meter’s electromagnet to pull the arm back up when the arm breaks the beam.
When the arm falls, because you’ve put an eyelash on it, it’ll break the beam which drives more current which pulls the arm back up. This will wobble back and forth until it settles with the arm just breaking the beam. Then you can read off the current required to hold it up, and that’ll be proportional to the weight.
The second link, from [Applied Science] goes into a lot more detail. [Ben Krasnow] reconstructs [Grohe]’s circuit and investigates the balance’s linearity (good) and stability (so-so), and demonstrates how to calibrate the scale with squares of paper. He also discovers that his plastic housing is pulling on the mechanism, because of static charge.
[Ben] is easily able to get around 10 microgram precision out of a couple of bucks in parts sitting on his desk. Plus, his hands-on approach to characterizing and debugging the setup really warms our heart.
(And the answers are: 80 micrograms is two eyelashes; ten grains of sugar weigh around 750 micrograms.)
Thanks to [Travis Swaim] for sending us the links.
grains of sugar vary wildly depending on how much they have been processed and/or crushed.
Doesn’t really matter, that’s how much HIS grain of sugar weighs.
That sounds like a reason to desire to know a specific grain’s weight.
What if you cut the interruptor in a triangle shape so it doesn’t block the light all of a sudden ?
Waaay to complicated. Pull 100 eyelashes, place on scale. Divide by 100. Cry.
And if you need the precise weight of a specific eyelash, rather than an average?
Easy.
First weigh yourself. Then pull out the eyelash, weigh youself again and subtract from the first weight ;)
Then your just to damn anal retentive!
No, this is too damn anal retentive:
“you’re”, not “your”;
“too”, not “to” ;)
Matt, Mr. Wibble I think we all just need to smoke a doobie and relax, :)
I agree with you-there is no excuse for ignorance.
Pull 100 eyelashes from, start with left eye. Look in the mirror. Cry.
Scientific american “Amateur scientist” column did a “microgram electrobalence” this way.
[Paul Grohe] is from National Semiconductor not Texas Instruments
One and the same now, no?
That’s what he [Paul] said. :)
Put a small convex lens over the photodiode part, or a drop of transparent glue.
Homemade QCM is the weigh to go.
“The weigh to go”. Ha! You sleigh me!
junkies need very accurate scales
Well, LSD is measured in µg, not mg.
The circuit is curious. Why is he powering the photodiode off the output of the first op-amp? In this case, you don’t need direct electrical negative feedback from the op-amp output to negating input; the feedback loop goes through the mechanical system.
To stabilize the thing, I’d probably try something like this: http://i.imgur.com/0Ud9vSz.png
The first resistor (and fact that the op-amp can’t output more than the rail voltages) limits the current into the capacitor, slowing down the whole system. This essentially turns the whole control loop into an I (integral) loop, instead of the P loop that the video has.
Whenever the arm is too low, current through the panel meter would slowly increase; whenever it’s too high, current would slowly decrease.
I experimented as you suggested but adding capacitors always seems to reduce the frequency of instability at the expense of wilder swings. Regardless of capacitor size. Small caps have negligible effect while larger ones cause the needle to bounce further.
I also modified the circuit to use a non-inverting amplifier with a opto-reciever re-arranged that acts more as a variable displacement transducer and less Binary as in Ben’s case.
“Now we know how many holes it takes to fill the Albert Hall, goo googajoob”