*This is the second of a five part series on tolerance in reloading.*

In this part we look at the powder “throw” or drop and try to answer the question, “How accurate does it need to be?”

My load for 45 ACP “wad” ammunition is 4.0±0.1 grains of Clays.

That ±0.1 grain powder tolerance allows a total variation of 0.2 grains, from 3.9 to 4.1 grains. That tolerance permits a maximum of a 5% variation in powder weight from one round to the next: (0.2/4.0) * 100 = 5%.

With the excellent instruments available to us today, I don’t think of that as being very precise. Only 5%? That’s not very good.

After all, when it comes to length, if we’re measure something as 1.200″ and we allow ±0.005″ tolerance from that, that’s 0.833% accuracy or less than 1% error. That’s good!

The 5% weight variation from 3.9 to 4.1 grains is big by comparison.

Can we do better and, more importantly, do we need to do better?

There are *two* factors of interest here. The first is the accuracy of the instrument and the second is the number of significant digits in the answer.

For the throw weight, it’s the second one that’s driving us to accepting the gross 5% tolerance.

Specifically, no matter how accurate the scale, for very small weights at the lower limit of the scale’s ability where it displays a two digit answer, e.g. 4.0, the precision of that answer cannot be better than 1% and, in the case in point, the error could be much larger at nearly 5%.

For example, if I weigh a bullet and the scale says it is 200.2 grains, then I have four significant digits. Since we don’t see the hundredths of a grain weight, we don’t know if it’s a “0” or a “9” so the real weight could be anywhere from 200.20 to 200.29 grains. That’s gives us a range of 0.1 grains or 0.05% when the measured item weights 200 grains. This a *very* precise number and probably at or beyond the limit of what the scale can actually achieve.

With four significant digits in the answer, I can be assured that answer is very precise. And if I measure another bullet and get the same answer, then I am assured those two bullets are probably within 0.05% of each other’s weight.

But that exact same scale, when used for a powder throw of say 4.1 grains, is then showing only two significant digits and a variation by one in that last digit will then be a variation of 2.5%. Because we only have two digits to work with, the possible error has increased dramatically even though the scale’s accuracy has not changed.

With only two significant digits, if I weigh two throws of powder and they both say 4.0 grains, if one was actually 4.04 grains and the other was actually 3.95 grains, then they differ by 2.5% for that 0.1 grain difference. If my tolerance is ±0.1, then the two weights might be even more different, say 3.90 and 4.09 grains. That’s the 5% variation I’ve mentioned.

So, the number of significant digits in the answer is a very important element in building a precision product and, in some cases such as with powder weight, it is much more important than the scale’s accuracy.

Most of the digital powder scales we use in reloading display results down to a tenth of a grain. As we’ve seen, at the small loads we deal with, that becomes the limiting factor in our precision.

Is this bad?

Surprisingly perhaps, no, it’s not.

In fact, it’s perfectly all right.

Many skilled Bullseye shooters punch a terrific number of Xs using ammo made with scales that read to a tenth of a grain and they shoot the same bullets, powder and weight as used in this “real-life” example.

In other words, a 5% tolerance in powder weight apparently makes little or no difference at the target.

*In the next part of this series we will look at COL and crimp, the two measurements we take probably more than any others.*