Hello, I've checked a few threads on scale accuracy, but can't find a whole lot of specifics on this. I recently got a scale that claims 0.01g accuracy. I figured a low-cost scale claiming that accuracy should, at least, be slightly better than a 0.1g accuracy scale. So, to test, I took a pile of the 2010 Native American $1 coins (via the direct ship program from the US mint). The US mint lists these coins as being 8.1 grams each. Yet, on this shiny new scale, 10 of them weight 80.0x grams (x changes a little based on which ten). A single one almost always weights in at 8.01g. To compare, I took one of our food scales -- accurate to, at best, 0.1g -- and weighed 10 and 20 coins. Still, it comes out to 8.0 grams each, not the expected 8.1g. 20 nickels typically comes out at around 99.98 grams on the "better" scale and 100 even on the food scale. This is far closer to what I'd expected as far as accuracy and tolerance goes. 8.0 grams vs 8.1 grams is a difference of just 1.2%, but I don't seem to have any examples that are different and the readings are consistent regardless of the number of coins or which coins. I would have expected that coins to be a range, but average to the listed weight. 1. The source of the weight specification I'm getting is from the US mint (http://usmint.gov/about_the_mint/?action=coin_specifications) and other sites say the same. However, none of them list acceptable tolerance or variance. Where do I find this information? How about for old coins and world coins? 2. If I'm weighing for a first pass at authenticity testing, what accuracy should I be looking for? I'd assumed more than +/- 0.05g off would be an obvious indicator, since most coins are published to at least thousandths of a gram accuracy or more, and any fake trying to be really tricky would hit the numbers precisely. But, the dollars from the mint are even farther off... If a couple tenths of a gram variance is expected, then I can see why most people use 0.1g scales. But, then, I also don't see the point in publishing numbers of such higher accuracy, such as 2.2268g for a dime, especially if it could actually weigh 2.3 or 2.1g (and, besides, the loss of getting a few fakes would, for most people, be less than the cost of a scale that can accurately read to 0.0001g). 3. How can a calibration weight be trusted any better than the scale itself? Thanks for any helpful info and insight!
1. You might want to contact the mint by phone or e-mail to ask about the tolerance/variations in weight. I have found there to be a decent variance, even in new coins. 2. I think your 0.05 g variance is reasonable for purposes of verifying authenticity. 3. I was going to recommend a calibration weight to you, then I read this. I guess you don't ever really know. Just get it from a good source, and make sure it's in new condition, and you can be pretty certain. If you have access to a chemistry lab, you should be able to find calibration weights pretty readily.
Higher sensitivity scales can get out of wack due to a slight breeze(breath) or even not letting it get to ambient temperature.
If your scale claims 0.01 gm accuracy, it typically should be 0.005 or better at manufacturing. There is no reason you should not be able to at least approach 0.01 gram accuracy, but you will need to watch for any breeze, keep it level, adjust the zero often and calibrate the scale probably every time you move it. The must be an adjustment for the calibration.
The 1.2% difference you are encountering is within US Mint specs for the dollar coins. But if you want to verify the accuracy of your scales take a trip to the local jewlry store. Weigh the same coin on your scales and then have the jeweler weigh them for comparison.
I have been using a Jennings JSVG-40 scale for the past two years with an accuracy of .005. It can only handle up to 40 grams but that's usually fine for most coins. Here is the info... http://www.oldwillknottscales.com/jennings-jscale-jsvg40.aspx For $85 it's as accurate as your likely to find at any jewelers.
Triple-beam balance. Wow....that takes me back to first-year university chemistry. Hiho: It's funny you should mention that. The JSVG-40 is the exact scale I use. And I bought it from the same supplier. Best less-than-$100 I ever spent.
BTW, the best $1 I ever spent was on my 1977 Merck manual from a garage sale. Yes, I keep track of this stuff.
Thanks for all the answers; it seems it was more my misunderstanding than the scales themselves. It surprises me that the tolerance is so wide, but I guess higher accuracy would increases costs too much. And thanks for the tip on the Jennings scales. They look to be pretty nice, including a nice box.
Find a copy of the Coin World Almanac, specifications and weight tolerances for all US coins including the older ones. World coins are a little more difficult. The Krause Standard Catalog is pretty good for the silver and gold coins (no tolerances though) but somewhat hit or miss fo the base metal coins. Tolerance of the small dollars in 8.1 grams +/- .3 grams so almost a 4% variance. I would go with a minimum of .1 gram accuracy. The tolerance of most US coins is less than +/- .2 grams. A scale that only reads in full grams is not accurate enough to detect out of tolerance fakes unless they are WAY off. For example the a Morgan dollar is listed at 26.73 grams with a .097 gram tolerance. A scale accurate to 1 gram would show it as 27 grams. A genuine coin would show 27 grams as long as it was in tolerance (26.6 to 26.8) but a fake has a much wider target. It would show 27 grams as long as it is from 26.5 to 27.49 grams. With a .1 gram scale the counterfeit has to pretty much hit that same weight and tolerance as the genuine coin. For first pass weighing a .01 scale is probably a little over kill. Where accuracy of that extent is needed is if you are wanting to run specific gravity tests. The differences between dry and wet weights are often less than a couple tenths of a gram and specific gravities have to be carried to at least one decimal place and preferably two places. If you want those figures to be accurate you need to have more precision in your weights. You aren't going to get that with a 1 gram or .1 gram scale. The good news is .01 scales are now readily available and pretty cheap. I have seen them at times under $20, about the same as the .1 gram scales. If you want even higher precision then they get more costly. (If you don't mind the slower speed and larger size of the manual scales sometimes you can get great precision cheap. I picked up a Becker analytical balance accurate to .0001 grams for $50 on eBay. An electronic balance that accurate will typically set you back a couple thousand or more.) A true CALIBRATION weight is typically certified to be that weight and have been adjusted to be balanced to several decimal places in comparison to a standard weight from the National Bureau of Standards. (A NBS standard weight has likewise been adjusted to match, to even more decimal places, the Offical Standard weight.) Most "calibration" weights are not true calibration weights but are probably accurate for their marked weight to at least two decimal places
I just weighed a 50 peso on my work scale, a Ohaus Adventure Pro that is calibrated every 6 months and the weight came out to 41.67. I looked up the gross weight of the coin on line and it's 41.666 . The coin was in a bezel and there is some dirt on in the side Independcia Libertad. My "pocket rocket" scale in the car says it weighs 41.3 . Go to a jeweler with a ASE, have it weighed and use it to reference your scale. After that figure your factor +/- .0? and your scale will be fine. That is the difference between a $300 scale and a $30 scale.
Forget weighing coins to calibrate a scale. Get a standard weight. Any kind of standard weight will do. If the scale is accurate then I'd guess your dollars were punched out of an underweight strip but you can't use coins for calibration.
I can think of a few ways to do it with coins and get a very close reading but each has a potential error. Some sellers are known to have scales that overweight and some jewelry stores use different scales for buying and selling. A scale can go out of balance after the fact as well. Of course once you know what anything weighs you canb use that object and this includes coins.
