Beam Scale Inaccuracy - Check yours

[tl_3237]

During a ‘what if?’ moment I tried something that gave me totally unexpected results.

I use an old beam scale (Herter’s) that has two weights - a small weight on the pan side of the pivot that gives 0.1gn increments and a larger weight on the opposite side of the pivot which gives 5.0gn increments. On the top of the beam is a sawtooth notching that captures each weight as you slide them into position.

One day I was loading a 20.0gn charge. I leveled and checked zero on the scale before starting. As usual, I put the large weight at the ‘20’ notch and the small weight at the ‘0.0’ notch and measured a charge at exactly 20.0gn. Then I set the large weight in the ‘15’ notch and the small weight in the ‘5.0’ notch. To my surprise the same charge that I just finished weighing is now 21.75gn !

I tried a 55gn bullet: ‘55’ and ‘0.0’ showed the bullet at 0.1gn heavy; ‘50’ and ‘5.0’ showed the same bullet as 0.2gn light !

If you have a beam scale that employs a similar weighing technique it would be interesting if you would try this simple experiment and see if you get closure on the results.

I cannot fathom why this is occurring other than a manufacturing defect. Any other theories

[knight0334]

There was one thing that we were taught in chemistry class in high school.. Always set the heaviest counterweight to the highest position within the desired weighing, then use the next highest weight to it its max position within the weighing, and so on until you get your desired weight.


Example: a 3 weight beam
1. the big weight is graduated at 5grs ranging from 5 to 500gr
2. the middle is graduated at 1gr ranging from 1 to 10gr
3. the small is graduated at .1gr ranging from .1 to 1.0gr

To properly measure off 22.7gr you put the weights as follow and will provide you a more accurate weighing:
1. big weight on 20
2. middle weight on 2
3. little weight on .7

Even though putting them as follows is numerically the same(22.7gr), it will generally be off from the actual weight:
1. big on 15
2. middle on 7
3. small on .7

There will be of course variations between manufactures, designs and even individual scales which may reflect differing results.

[tl_3237]

There was one thing that we were taught in chemistry class in high school.. Always set the heaviest counterweight to the highest position within the desired weighing, then use the next highest weight to it its max position within the weighing, and so on until you get your desired weight.

.

Agreed as a matter of retaining consistency but the acuracy issue remains.

My experience would indicate that there is a loss of accuracy as the 'fine' weight approaches its limit. In my example I used a charge that has two possible 'equivalent' weight setting since it allowed me to check the scale against itself. Having experienced a 0.25gn deviation on a 20.0gn charge then I would assume that I would have a similar deviation for a 19.9gn charge which has a unique weight positioning ('15gn' & '4.9'gn). This difference is even more critical for a charge of 4.9gn when the 0.25gn difference is a 5% error.

[Brick]

You should get very close to the same results from either combination. You say it is an old scale. Have you ever cleaned it, or checked the knife edge of the balance to make sure it hasn't rusted or wiped down the pan and pan carrier to get rid of dirt and grease?

Any measuring system will have errors (error in this case does not mean wrong or mistake, it is simply an description of the limits of the measuring system) and the errors might be magnified differently along the range of the measuring system. The examples you cite 1.75 grains over on a 20 gr. sample (8.75% error) and a .3 gr variation on a 55 gr. sample (0.55% error) demonstrate this.

There are check weights available on the market.

[knight0334]

Agreed as a matter of retaining consistency but the acuracy issue remains.

My experience would indicate that there is a loss of accuracy as the 'fine' weight approaches its limit. In my example I used a charge that has two possible 'equivalent' weight setting since it allowed me to check the scale against itself. Having experienced a 0.25gn deviation on a 20.0gn charge then I would assume that I would have a similar deviation for a 19.9gn charge which has a unique weight positioning ('15gn' & '4.9'gn). This difference is even more critical for a charge of 4.9gn when the 0.25gn difference is a 5% error.

How I explained will provide the more accurate and truer measurement. I cant remember the specifics as to why or how that works(been 20-23 years), but it is what it is..

With the bullet that you weighed, they rarely weigh exactly what they are labeled to be. Bullets will vary a few tenth of a grain, sometimes a couple or more grains. I'd trust the "55" notch before I trust the "50+5" as to being true. Same with the the powder at the "20" notch being prefered over the "15+5".

[Enfielder]

Not sure why your scale is acting up. I check my Dillon Eliminator scale with McMaster-Carr brass "Lab weights." They have a tolerance of +/- .002g (P/N 1777T43 for 1g and 1777T44 for 2g), come with an ASTM cert. and are only $4.26 each. 1g = 15.43 grains, so with one of each you can check the range of typical rifle loads 1g = 15.43 grains, 2g = 30.86 grains, or put 'em both in the pan for 46.3 grains. Cheap insurance.

[Flash]

Not sure why your scale is acting up. I check my Dillon Eliminator scale with McMaster-Carr brass "Lab weights." They have a tolerance of +/- .002g (P/N 1777T43 for 1g and 1777T44 for 2g), come with an ASTM cert. and are only $4.26 each. 1g = 15.43 grains, so with one of each you can check the range of typical rifle loads 1g = 15.43 grains, 2g = 30.86 grains, or put 'em both in the pan for 46.3 grains. Cheap insurance.

That, Sir, is an idea which has merit for all reloaders. I'm placing my order tomorrow, just as soon as I get home from the 2nd Amendment gathering in Harrisburg.

Flash

[Farbmeister]

Or buy a scale that is NOT made of metal beams.

Read up on the Lee Safety scale.

Powders have a lot to lot weight variance (as stated from the manufacturer) of 7-10% by volume. So a 20gr charge could weigh as much as 20.2 grains and still be WELL WITHIN specs.

Plus load books are NOTORIOUSLY on the low end of MAX pressure.