How Much Precision in a Scale is Required for Precise Weight-Sorting?

[GetReal]

GetReal,
They aren't mocking you, they are just telling you the simple truth of where a group or score benchrest rifle needs to agg.

* I *** am joking about them mocking me. No one here or at any competition has ever mocked me. Gun ppl aren't like that.

 

[Hohn]

I work with fairly precise equipment and tolerances (single digits microns) and general rule is that your measurement system must have about 10x the resolution of the smallest significant difference. So if you care about differences of 0.1gr, you need to resolve down to 0.01gr.

Now resolution is just the ability to distinguish two different but close measurements. Resolution of 0.01 would mean that your measurement system can distinguish 80.02 from 80.01. But if this resolution just comes from an extra display digit and isn’t within the device’s accuracy, it’s not contributing anything unless it’s repeatable enough.

For example, my Hornady comparator set is not very accurate in terms of being close to the stated ID size of the bushings. But it’s very precise in that it has low variation, and this allows good resolution, better than the calipers they affix to. So it works reasonably well as a comparator even if the number I see isn’t the truly accurate number. As long as it’s consistent, it serves the purpose.

 

[Rocketvapor]

@Hohn
I agree with consistency being key. Short term, during one reloading session, and long term for the next session. The reloader is the production staff and his own QC inspector and in many cases his own metrologist. Believing instrument specifications without periodic checks and calibrations might get you safe results, maybe.

I'll use what I've learned about my newest scale, the EJ-54D2. Checks at several points over the instrument's range show it's repeatability and accuracy compared to check standards more accurate than the resolution of the scale. On that day, maybe next week. One day I'll probably be able to detect when the performance degrades using these highly stable check standards. The minimal user calibration procedure recommended by the scale manufacturer isn't going to monitor the true performance in my using environment.

So, to answer the original thread question,

How Much Precision in a Scale is Required for Precise Weight-Sorting​

MORE precision than you REQUIRE, and that's on YOU to determine.

How about another Rocketvapor picture ?
Here's a result of using a cheap, old beat up digital scale.
Resolution of 0.01 gram in gram mode, 0.2 grains in grain mode.
Repeatability seems to be one count over several readings. I get 8.17grams with an occasional 8.16g.
I keep this one around just for sanity checks

Here are the values and avg from a previous post for the case in question #1 of six.
SIZED AND TRIMMED TO 22 NOSGAR 1.765"
1 8.1672 .1674 .1674 .1672 .1672 avg 8.16728
2 8.1694* .1692 .1692 .1690* .1692 .1692 avg 8.16920
3 8.1702 .1704 .1702 .1704 .1704 avg 8.17032
I wonder what case # 2 will weight on this scale?

UPDATE.
Case #2 shows 8.17, case 1 plus 2 shows 16.34g/16.33g
Case 1 + 2 +3 shows 24.51g
Not enough resolution to discern one case from another.
10g shows 10.00g
All 6 cases (49.021g) show 49.03 with an occasional 49.04 on this scale
Probably good enough for initial binning of cases to half a grain?

 

[Rocketvapor]

A year since my last post in this thread.
I don't shoot BR, and not good enough @ F-Class to see some small measurement errors but I do like to eliminate variables.

What effect does primer weight have on performance?
Here are some weights of CCI450 primers.
Granted there were only 2 really heavy ones out of 1000, a few more light ones, but MOST had very consistent weights.


A smaller sort with Ginex primers


First test of the cheaper Ginex SRP UNSORTED.
Did I have a primer flier?

OK, so I'm not a great shooter

 

[RegionRat]

By chance, did you record the individual weight data?

Nothing wrong with your shooting, even counting shot 7 on the CCI 450 side.

 

[Rocketvapor]

Not for those groups.
Screen shot didn't capture velocity.
Ginex were sorted, CCI were not.
Pretty much no wind and enough out to call it a flier.
Got both a right side 9 and a left side (high) 9.
groups get worse (for me) @ 600.
I have the heavy and light primers and one day will shoot for velocity.

 

[Webster]

I realize that the answer to this question may well depend on the components on which we are weight-sorting--more precision for primers, for example, than for cases.

A popular scale with handloaders is the A&D FX-120i. It is advertised as measuring in .001-gram (or .015-grain) increments. Is more precision needed? If so, for what components being weighed?

If you have 10 cases and there is a 2 grain spread low to high and you sorted them into 5 groups. The groups would be about .2 grains spread in the groups.