Originally Posted by SARDG
Well, that guy sure paints a grim picture. I'm certain that there is some correlation between this and the link(s) I posted - just not sure I'm smart enough to figure out what it is.
The scale addressed in this post is the Dillon and this same kind of conversion error isn't mentioned in the [my] former articles.
That is a very interesting article! It is WRONG, but it is interesting.
In the section "First response from me on Oct. 2, 2010:", there is a table that indicates that certain readings, in grains, aren't possible due to mathematical oddities related to rounding. For example, 4.7 grains should never show up on the scale. If the table is correct...
So, how is it that I can walk out to the garage, load up a trickler and dispense a 4.7 gr charge? In fact, I watched for missing increments on the way up to 4.7 and, except for the occasions where I was overexuberant in twisting the knob, every single reading appears.
There is certainly some technical basis for some of the author's ideas but the problem with 'facts' is that, sometimes, 'experiments' disprove them.
I think that when a reloader moves from a quality beam scale to a digital scale, they have accepted the 0.05 gr error. I know that my Chargemaster and my D'Terminator don't agree to 0.1 gr. If I set the Chargemaster to dispense 42.2 gr, I will, more often than not, get 42.1 gr on the D'Terminator. I figure there are two round off errors going on and I'm willing to accept the variance.
I use check weights in the range of interest. For a 42.2 gr charge, I use a 50.0 gr check weight. Both scales read EXACTLY 50.0 gr. That's great! But the engineer inside me also knows that I have no idea whether or not the check weight is anywhere near 50.0 gr. All I know is that three things tend to agree. They could all be wrong. But they would be wrong together!
Another thing I learned: Your measuring device needs to resolve 10 times better than the thing you want to measure. If you want to measure 0.1 gr, you need to be able to resolve 0.01 gr, accurately. Which means that none of our common reloading measurements are worth a darn - engineering wise.
But, darn, they seem to work anyway... Experiments...