My $4 precision weights

[(*steve*)]

I have a digital scale that I purchased when I had a large number of very small objects to count.

It can measure to 200g with 0.1g resolution.

But I've always wondered how accurate it really is.

So, I spent about $4 on a set of "precision weights") from China.

Here are the nominal values, the reading from my $15 scale and the reading from a somewhat more expensive one at work.

  1g   0.99g    0.9971g
  5g   4.99g    4.9989g
 10g   9.99g    9.9957g
 20g  20.00g   20.0005g
 20g  20.01g   20.0043g
 50g  50.01g   49.9906g

106g 106.06g  105.9871g

The last reading is with all the weights combined.

The readings tend to fluctuate +/-1 in the least significant digit, but on the whole I think it's remarkably accurate.

Interestingly, our "office" set of calibration masses cost several hundred dollars. I'll have to pull the calibration certificate for them when we're less hectic at work and see how they compare.

An observation about my masses is that they are all either very close or below the stated weight. Conjecture at work is that these are "rejects" that failed the first test for weight. They are un-trimmed, and clearly you can't trim them up to a higher weight. The two masses slightly over the nominal weight may be within the resolution of the initial test.

It would also be really interesting to weigh one, then touch it and weigh it again. How much does a fingerprint weigh? (answer: about half of the least significant digit of the balance at my work)

 

[shrtrnd]

Sounds like a good $4 deal there (not to mention your $15 scale).
We measure in the 10ths of a mg. We have to watch barometric pressure, wind-
currents in the room, fingerprints/dust.
Our set-up costs a little more than your $19 (total) set up.
Good buy!

 

[(*steve*)]

If you're measuring to tenths of mg, what is the resolution of your balance?

I did note that the temperature (23.3C) was marked on my "calibration certificate" but not the barometric pressure.

I'm not sure if that's because the balance isn't that sensitive or if it's because we don't typically use the full resolution.

I was asked at the time if that was enough precision, so perhaps it wasn't the limit of the device (or perhaps our lab staff have access to a better balance).

 

[shrtrnd]

.0001mg per NIST.
Temperature doesn't affect weight by itself, but it's probably a factor in some
formula involving barometric pressure. I don't know, because we have to keep
our dimensional lab at a constant 68 degree F. So we don't deal with it.
You still got a pretty good deal for the accuracy you got at $4.
We have to keep the item we're measuring in the cold room for at least 24 hours
before we attempt to measure it, because we typically have to check cal on it's
dimensions also, and those ARE temperature sensitive.

 

[(*steve*)]

.0001mg per NIST.

That's pretty amazing -- tenths of micrograms.

 

[shrtrnd]

I worked on atomic apsorbtion spectrometers and gas chromoatographs in a pervious
job. We had an old Ukranian chemist there, who was telling me he was now required
to measure pesticides in parts per billion.
I was standing there, impressed with the technical aspects of equipment capable of
such accuracy while the chemist looked at me in frustration.
Finally he shouted at me, because I was apparently too dense, seeing only the tree
instead of the forest behind it:
"PARTS PER BILLION!", he roared at me.
"WHO CARES? PARTS PER BILLION WON"T EVEN HURT THE BUGS, LET ALONE PEOPLE!"
I think I'm getting used to super-technical stuff. It doesn't bother me much anymore
that my Blu-Ray player can do everything except get my beer for me. I just turn it on
and play my movie, and ignore everything else about it.