Hello everyone. I'm finally getting around to asking a question I've been curious about for years. Nothing profound, really, but it bears asking. I
use a triple beam balance in my lab for all my mass-determining needs, and prior to each use, I use one of several brass weights to zero it out.
Among the weights are a 50g, 20g, 10g, 5g, 2g, and 1g weights. For years I've selected a weight that was relative to the mass of the sample. For
example, I'd use a 2g if I was weighing only a few grams, or a 50g if I was weighing heavier samples. This was done with the belief that the scale
would be more precise if the weight used to zero the scale was relative to the sample weighed. Is this a reasonable assumption? Does it matter what
size weight is used to zero the scale, as long as one is consistent? Samples appear to weigh differently when different weights are used to
zero the scale, hence my reasoning. Can anyone shed some insight?unionised - 28-1-2007 at 06:10
Checking balances is an art in itself. your approach is fairly reaonable.
Do you trust the test weights more than you trust the balance? If not then get some new weights.
If so then what you should do is zero the balance, put each of the weights on it in turn and read what the balance says. If you are lucky then the
balance agrees with the weight and you can stop worrying . If not then you need to produce a calibration graph (or adjust the balance, but I'm not
sure you can usually do that with tripple beam types).
Another useful check on the health of the balance is to check its linearity- this can be done without any accurate weights (classically, people used
glass stoppers).
Get a bunch of items that won't change weight - coins might be a good example. Label them A to J (or whatever) so you can keep track of them.
Weigh each of them and note the weight.
Weigh all of them together- obviously this weight should be the sum of the individual weights.
Weigh about half of them toegether and compare this to the sum of the individual weights.
Do this with various groups of them and plot the weight indicated against the expected weight (ie the sum of the individual weights). You should get a
straight line. If you don't then at least the differences from the line give you an idea of the inaccuracy of the balance. If that inaccuracy is too
big for your purposes then you know it's time to ask Santa for a better balance.
If they are all perfectly on the line then you just need a "fudge" factor to get the correct reading from your balance, and you only need one accurate
weight to find out what that fudge factor is. Even better provided that the weight is near the top of the balances range it doesn't matter much what
it is. You could buy a bar of chocolate weighing 100g, get someone with access to a good balance to weigh it- say it weighs 105g, take it home and
weigh it on your balance - say it seems to weigh 102g, then you know that all you need to do is convert readings from your balance to the correct
weight is to multiply by 105/102)
The other thing is that if your balance is linear a lot of problems go away. If you want to measure ratios of things you only need the balance to be
linear.Rosco Bodine - 28-1-2007 at 12:28
Class A volumetric flasks have a stated percentile accuracy for volume at a standard temperature ,
and when filled to the meniscus calibration mark with
distilled water at the calibration temperature , a very
accurate reference will be provided , certainly more
accurate than can be resolved by triple beams ,
or quad beams for that matter .unionised - 31-1-2007 at 11:33
Err, in the very real sense that the 0.02ml tolerance on a 10 ml flask is less than the 0.01g resolution of my triple beam balance and that the
tolerances get bigger for larger flasks.
Having said that they are OK for for most of the stuff we expect to do here . The PITA is the tare weight.
Given that the volumetrics are actually calibrated by mass it would be a bit odd if they were more accurate than the balance.prole - 1-2-2007 at 06:55
Thanks to all for your informed replies. I tested my weights and found that my balance was faithfully off by .1 to .3g. I can't bring myself to create a calibration graph yet, as suggested by unionized.
Too depressing. I guess experiments in microscale will just have to wait for my money tree to flower.
This inaccuracy can be safely figured in to most functions in my lab. I've also got mg tare weights that don't even register on my triple beam, so
I've been brainstorming ways of building an old-school 'microbalance' from scratch. It's not a priority yet, so I'm not moving on it. I know I can
purchase a nice and very accurate digital scale, and don't think I'm not tempted - I love to open presents I've ordered for myself. It's one of the
great pleasures in life, and a great luxury. Maybe next year...
-prole
[Edited on 2/1/2007 by prole]
[Edited on 2/1/2007 by prole]unionised - 1-2-2007 at 11:30
Thanks for posting this. I got all the junk together to make one of these and then found an analytical balance on the net at a price I couldn't
refuse. I still plan to make the homemade balance.
As to the original post. I have found the Ohaus triple beams to be accurate within the specs they're made for. I have two; one to 0.1 g, the other
to 0.01g. Both are reliable to these limits ie +/- 0.1 and +/- 0.01. I wouldn't have a balance that called for calibration at each weighing .. like
a pH meter. Not in my shop, no thanks. An Ohaus is less than $200. Electronic scales that are as accurate as the Ohous' are available for less than
$50. They are based on pressure transducers that could change over time. But while less reliable, ultimately, they're cheap.
If you're weighing in the microgram range then another kind of balance is called for. I have a concrete block table set up below street grade for the
occaisional super accurate weighing with the analytical balance. Much effort to calibrate was needed by the way (no pun was meant).prole - 4-2-2007 at 08:19
The balance I have is an Ohaus, and it needs calibrating each time if it has been sitting unused for any length of time. It is perhaps 10 to 12 years
old, so maybe time is the culprit. Yeah, just like my digital pH meter, it needs calibrating all the time.
A bit off topic, but the post about SAS was interesting. Anyone else a member? I used to be, when they sold all kinds of sweet chemistry sets and
such, but they don't sell jack squat anymore (volcano kit, anyone?), probably because of the new paranoid restrictions. And the other science sites
they link to pretty much only sell to teachers. I considered asking my stepdaughter, who is a teacher, to buy some sets for me, but she once saw my
cool steam distillation of orange peels and immediately thought I was 'making cocaine'. So I tried to engage her in a review of her high school
science chemistry classes, but she remembered none of it. I weep for the children in our schools. The future of science in America is bright indeed.S.C. Wack - 4-2-2007 at 17:20
Quote:
Originally posted by chemrox
As to the original post. I have found the Ohaus triple beams to be accurate within the specs they're made for. I have two; one to 0.1 g, the other
to 0.01g. Both are reliable to these limits ie +/- 0.1 and +/- 0.01.
Could have written that myself, except that the 311 has 4 beams. The tolerance is the smallest division over the entire range, and my full range of
calibration weights are accurate to 1/10 mg. Both are over 25 years old and although a school was the original owner, they have been abused some
since. Taking off the metal cover over the agates shows them to be undamaged. I'd expect them to remain as accurate for a few hundred years as long as
this remains the case. I suggest getting a level and checking your surface, and the alignment of the mechanism. It is not unusual to zero any kind of
balance every time you use it, nor is it something that I curse at doing.chemrox - 4-2-2007 at 21:56
I'm not disagreeing with the above. Here's what I have a problem with: I always check zero before weighing but I would be alarmed if the zero
changed and the balance hadn't moved. Even small movements over almost levels surfaces shouldn't change zero on an Ohous triple.
I'm wondering if the parts in the zero adjust have deteriorated in Prole's balance to the point where the zero adjust moves on its own. I'd get an
electronic scale and use the triple to check it from time to time. And wait for a deal on a good used Ohaus.
I can't bring myself to create a calibration graph yet, as suggested by unionized.
Too depressing. I guess experiments in microscale will just have to wait for my money tree to flower.