Mathematicians - how to interpret the data?

[Land Owner]

From the Small Digital Scale (https://castboolits.gunloads.com/showthread.php?425406-What-is-your-experience-with-small-digital-scales) thread I tested my three scales, RCBS 5-0-5, RCBS 10-10, and PACT digital, with a test weight set (in grams). Scales were balanced to zero (every ten increments), tabled and leveled side by side, each (except digital) preset to the known and expected test weight (converted to grains), each scale adjusted (if required) to balance, numbers were recorded, deviations were calculated in "plus" or "minus" tenths of grain corrections, and now what do I do with the deviations?

I knew the 5-0-5 was going to be very accurate, but it too required a few corrections to zero. After corrections, it had a few adjustments to balance during the 36 (+/-) test weighing's.

The 10-10 had been problematic after I noted it was purchased damaged, but not disclosed. I have worked to resolve its "stickiness" and have made considerable improvement, but it remains the least reliable of the three. It had a few adjustments to balance and more than once it had to be nudged out of "stuck".

The PACT digital, which I never unplug or turn off its electricity (on purpose), was nearly FLAWLESS. I do not rely on the digital scale for throwing charges in bottle necked hunting rounds, but I do use it to assure the beam scale is set correctly, from time to time checking a thrown charge, and nearly always to weigh a case or a boolit with its near instantaneous readout.

I think, in the Grand Scheme of things, that these tests assure that "reliability" is fleeting, consistency can be "problematic", attention to detail is a constant requirement, and there is more than one way to "skin a cat" (tongue in cheek).

Here is the test data...

https://i.postimg.cc/gk34J1RW/2021-06-23-Test-Weighing-Scales.jpg

[Tar Heel]

You will need to ask a specific question regarding the scales other than "interpret the data" to derive your answer(s) from the data. Remember that your data is "continuous" not "discreet" and analysis in statistical charts will be plotted accordingly. For example an XbarR chart versus and I chart. I would try to determine the Median, Mean, and Range of each scale and then compare the scales with the others to see which one has the lowest range of error. Without reviewing the data, I would have tested the scales using an average weight of those generally thrown by handloaders to get data most useful. Hopefully you did. Accuracy at nominal charge weights vs accuracy at weights never used is an important observation. Accuracy over range is important too. A scale may be balls-on at 22 grains but 30 grains off at 3000 grains.

Also determine how you plan to evaluate and measure "reliability" and "consistency". My personal opinion regarding measurement is that Sigma-6 is a freaking myth. Those that claim it have manipulated their data to eliminate all but that which fits into their created outcome. Sigma-3 is attainable but very hard to maintain. So all you "Sigma-6" believers out there - just keep patting yourselves on the back. I'm the POTUS too.

Anyway, determine your specific question to determine how to lay out the data for digestion and analysis. You may find MS Excel can assist you with this as well.

[pworley1]

The scales say "Accuracy: +/- 1/10 Grain" . From your results it looks like they did that.

[jmorris]

I knew the 5-0-5 was going to be very accurate, but it too required a few corrections to zero. After corrections, it had a few adjustments to balance during the 36 (+/-) test weighing's.


If the scale won’t go back to zero, that is a problem in and of itself. Whatever is causing that difference in zero could easily cause it to be off as weight is added. I would start there. If you are adjusting things to your check weight, that too will obviously effect zero.

It would have been better to leave the scale alone and just record the results.

As far as the 505 goes, when I add a photo electric switch to detect beam movements my eyeball no longer can, they are well within .1 of a grain even when controlling a trickler.


https://www.youtube.com/watch?v=6GYWgAdKzHs

[jmorris]

And are sensitive enough to detect .02 grains of weight added to the pan.


https://www.youtube.com/watch?v=yvF6WJs1zyY

[dverna]

Me thinks too many folks overthink accuracy of charge weights.

[NSB]

Me thinks too many folks overthink accuracy of charge weights.
You are spot on with that comment! In the “real world” those minor variations won’t effect group size at all. Variation is normal in any process and should be acceptable to a point. For those obsessed with this, do a Gage R&R study on your scales and calculate “gage error”. If you can live with the built in error of your measurement system, just go with it and use your check weights once in a while to calibrate it. (I retired as a quality engineer and this stuff is way over rated for real world shooting purposes).

[Hossfly]

I look for .10 gr. For pistol at 25 yards, hunting rounds get dead on as that’s where I get sausage from with my .223. My balance scale matches my digital or the digital gets chunked.

[Scrounge]

From the Small Digital Scale (https://castboolits.gunloads.com/showthread.php?425406-What-is-your-experience-with-small-digital-scales) thread I tested my three scales, RCBS 5-0-5, RCBS 10-10, and PACT digital, with a test weight set (in grams). Scales were balanced to zero (every ten increments), tabled and leveled side by side, each (except digital) preset to the known and expected test weight (converted to grains), each scale adjusted (if required) to balance, numbers were recorded, deviations were calculated in "plus" or "minus" tenths of grain corrections, and now what do I do with the deviations?

I knew the 5-0-5 was going to be very accurate, but it too required a few corrections to zero. After corrections, it had a few adjustments to balance during the 36 (+/-) test weighing's.

The 10-10 had been problematic after I noted it was purchased damaged, but not disclosed. I have worked to resolve its "stickiness" and have made considerable improvement, but it remains the least reliable of the three. It had a few adjustments to balance and more than once it had to be nudged out of "stuck".

The PACT digital, which I never unplug or turn off its electricity (on purpose), was nearly FLAWLESS. I do not rely on the digital scale for throwing charges in bottle necked hunting rounds, but I do use it to assure the beam scale is set correctly, from time to time checking a thrown charge, and nearly always to weigh a case or a boolit with its near instantaneous readout.

I think, in the Grand Scheme of things, that these tests assure that "reliability" is fleeting, consistency can be "problematic", attention to detail is a constant requirement, and there is more than one way to "skin a cat" (tongue in cheek).

Here is the test data...

https://i.postimg.cc/gk34J1RW/2021-06-23-Test-Weighing-Scales.jpg

Are you handling the weights with bare hands? You could be leaving sweat & grease off your fingers each time you touch the weights. Not a lot, but it adds up. Been a long time since I needed accurate weights, but back when I was playing with Chemistry, the scale went in an enclosure that prevented air currents from touching it, and the test weights were handled with gloved hands or a pair of forceps.
Remember that .1 grain is 6.479891 milligrams. That isn't a lot, either. How big a piece of gunpowder is that? What temperature is your work area where you're doing these measurements, and most important of all, what accuracy do you really need? How much variance in the speed of a bullet is really important to you? I understand CDO*, but sometimes you really don't need to go there. Even if you want to.

* CDO is like OCD, but with the letters in the correct alphabetical order.

[NSB]

Go talk to some seriously good bench rest shooters and get their opinion on all of this. You’ll put a smile on their faces.

[Land Owner]

As I thought, scale balancing and testing is "academic" at best and a waste of good time at worst. I am not making volatile "rocket fuel" and I like the results, so far, of all of my handloaded rounds, which have given me countless hours of fun in the making, the shooting, and the gathering of game.

Still, I have these comparisons and through them an empirical sense of which is the "most" accurate, which I can trust, and which I can use as a "2nd opinion" for the others...else, why have three scales in the first place? Maybe that's a rhetorical question, as would be why have more than one gun...but I digress.

[mdi]

Me thinks too many folks overthink accuracy of charge weights.

Agree! Laboratory accuracy isn't needed for reloading so if your scales are within .1-.15 grain you are doing quite well (How do you know how accurate your check weight is?). The most important is repeatable. If your scale says a powder charge is xx.x every time but in fact via lab level scales the weight is xx.yy it won't make much difference. If you are working up a load and find "The Load" as long as your scales say the same every time you're good, no matter what the lab scale says...

If your '-06 load is 45.5 gr of IMR 4064 and it shoots accurately and safely, but a lab quality scale says it is in fact 46.0 gr. so what?

[Tar Heel]

You are spot on with that comment! In the “real world” those minor variations won’t effect group size at all. Variation is normal in any process and should be acceptable to a point. For those obsessed with this, do a Gage R&R study on your scales and calculate “gage error”. If you can live with the built in error of your measurement system, just go with it and use your check weights once in a while to calibrate it. (I retired as a quality engineer and this stuff is way over rated for real world shooting purposes).

Hear! Hear!

[Liberty1776]

An award-winning sharpshooter I met at the range and later at a local gun show was selling resealed boxes of Match Grade rifle bullets.

He had gone through boxes of bullets and analyzed them with some kind of magnetic resonance tool (I forget the exact name) that measured concentricity or internal weight distribution or jacket thickness or some such.

He maintained that ultimate repeatable accuracy comes down to the bullet itself.

So, after he had picked out the "perfect" bullets from several boxes of Match Grade bullets, he repackaged them back in their original Match Grade boxes and sold them at the gun show. They were all still Match Grade. He wasn't cheating or grifting his customers.

It's just that he had found the cream of the cream -- the perfectly balanced bullets that fit his needs -- and sold the also-rans.

I'll never shoot that good, nor will my firearms. Tiny variations in powder loads are a given.

[jmorris]

Me thinks too many folks overthink accuracy of charge weights.

Yep, most pistol shooters likely couldn’t tell a difference in group size at 15 feet with .5 grain charge weight differences. 100 & 200 yard benchrest matches are won every weekend with volume thrown charges, no scale involved.

If you are a guy trying to squeeze a 3” at 100 yard load to something that will shoot even MOA, powder charge to the kernel isn’t your problem or going to get you where you want to go.

[dtknowles]

Yep, a scale that is good to the nearest 0.1 grain is usually more than good enough. That is 0.02% of full scale which is more than great if you are measuring in the middle part of the scales range. 0.1 grain accuracy is not so good (but still probably good enough) if you are weighing charges that are just a couple grains.

Does anyone know of a grains scale that is designed for smaller charges, maybe with a range of 0-50 grains in 0.05 increments?

When I weight sort bullets I use a digital scale that is really only good to 0.3 grains but once warmed up will hold that accuracy and that works for me for now. It is too slow and cumbersome weighing bullets on a beam scale. I think there is more payoff weighing bullets than weighing powder charges but for my best ammo I do both.

Oh, another thing in the really does not matter category. My beam scale has hysteresis, if I am trickling up a charge it will indicate lower than if I take the pan off the scale and set it on again. Said differently if I trickle up the charge to my scale setting and then remove the pan with the charge and put it back it will indicate higher almost every time. I don't know if the stickiness is the damping or the bearings but it is there but it is less than 0.1 grains.

Tim

[dtknowles]

...... 100 & 200 yard benchrest matches are won every weekend with volume thrown charges, no scale involved........


This is not a lie but it is not the whole truth. They are not using just any powder drop and they use powders that meter very well.

Tim

[C.F.Plinker]

Oh, another thing in the really does not matter category. My beam scale has hysteresis, if I am trickling up a charge it will indicate lower than if I take the pan off the scale and set ierently if I trickle up the charge to my scale setting and then remove the pan with the charge and put it back it will indicate higher almost every time. I don't know if the stickiness is the damping or the bearings but it is there but it is less than 0.1 grains.

Tim

Just curious. Does your beam scale have magnetic damping? My 505 does and I have noticed that when the beam is swinging - before the damping is complete - each succeeding peak is lower than the previous one. While this is a sign that the damping is working it could also mean that the final position is slightly above or below the true zero point. This could be due to the magnetic field holding the beam a little bit high or low because the beam movement can't break it loose.

[Ford SD]

Many years ago ... when loading in the basement I noticed when the forced air furnace came on the scale would change when the fan was running

to this day if if the blower fan is on I will not check a charge on the balance beam scale

[Mk42gunner]

... Does anyone know of a grains scale that is designed for smaller charges, maybe with a range of 0-50 grains in 0.05 increments? ...

Tim
The Lee is the closest I know of, and going from memory, it has a 100 grain capacity. I bought one last year when I lost confidence in my old RCBS 10-10.

It did weigh accurately, but the adjustments are a Royal PITA. I ended up boxing it up and sitting it on the shelf. I also bought an older Ohaus (Lyman D-5) that is easier for me to use.

Robert

[jmorris]

This is not a lie but it is not the whole truth. They are not using just any powder drop and they use powders that meter very well.

Tim

That’s not a lie or the whole truth either, the guns, equipment and components they use would likely still shoot better than 99% of out of the box rifles with equal or more precisely loaded ammunition. So you are back to the “many folks over think…” part.

Does make it worth working up loads that use powders that meter well though, if powders that metered crummy but weighed to the kernel won more matches than thrown charges, the winners would be doing that as well as everyone trying to beat them. That’s not the case at the short range matches (100&200 yards).

If one develops a great load, the “node” where multiple charge weights shoot to the same POI will be larger than less great loads that are very picky. This gives a powder “range” where performance is satisfactory vs an exact point. Then even poorly metered loads shoot great.

[Land Owner]

My "competition" is the correct placement of a preferred bullet into game for its swiftest dispatch. There is a LOT of trigger time in hunting hogs. All those chasing "one hole" (and I am [not] immune to that either), may find these data points and this thread a study in futility. Still, the effort to be and stay CONSISTENT in the reloading and shooting of ammunition remains the goal. With time on one's hands (retired), three scales, and a nice set of (tweezer held) calibrated weights (gravity in your location may vary), testing and comparing scales became, "What else was I going to do today?". Now I "know" more (useless stuff) than I knew yesterday.

[jetinteriorguy]

Are you handling the weights with bare hands? You could be leaving sweat & grease off your fingers each time you touch the weights. Not a lot, but it adds up. Been a long time since I needed accurate weights, but back when I was playing with Chemistry, the scale went in an enclosure that prevented air currents from touching it, and the test weights were handled with gloved hands or a pair of forceps.
Remember that .1 grain is 6.479891 milligrams. That isn't a lot, either. How big a piece of gunpowder is that? What temperature is your work area where you're doing these measurements, and most important of all, what accuracy do you really need? How much variance in the speed of a bullet is really important to you? I understand CDO*, but sometimes you really don't need to go there. Even if you want to.

* CDO is like OCD, but with the letters in the correct alphabetical order.
GMO, your comment on OCD is so hilarious.

[Scrounge]

GMO, your comment on OCD is so hilarious.

The line is stolen, but it's pretty real. And I'm definitely the pot. ;) There are places where it comes in handy. Time spent in the military tells me that if you're not CDO when you enlist, you will be when you retire. Only the universe will not cooperate. I'm learning to be a machinist, and at least in that field, you're given tolerances. How close to perfect do you need to be? Read the blueprint, and it will tell you. +/- .005"? +.0000"-.0001"? Whatever. Now I just need a set of drawings for life.

Bill

Bill

[dtknowles]

Just curious. Does your beam scale have magnetic damping? My 505 does and I have noticed that when the beam is swinging - before the damping is complete - each succeeding peak is lower than the previous one. While this is a sign that the damping is working it could also mean that the final position is slightly above or below the true zero point. This could be due to the magnetic field holding the beam a little bit high or low because the beam movement can't break it loose.

I said, "I don't know if the stickiness is the damping or the bearings but it is there but it is less than 0.1 grains."