Does this exist? Study of powder weight per piece?

[jmorris]

https://www.thehighroad.org/index.php?threads/a-d-scale-with-auto-trickler.846883/

Here’s why I bought the A&D FX120 and Autotrickler.

I bought mine just to check the trickling device I built, to test with.



If you have any buddies that need one I can hook them up with a very lightly used A&D, never even had the protective keypad cover off of it.

 

[JFrank]

I don’t know what bullet we are talking about but say a Berger 168 gn classic hunter View attachment 1064142
View attachment 1064143

Should make about 4” of difference in elevation according to their calculations. What have you measured in your testing?

1000 yards is much different than 100-200 yard benchrest, where ranges are so short, the disparity in velocity becomes much more difficult to realize. So it’s a good place to answer the question.

I open my big mouth once in a while about load tuning ,barrel timing and once optimized the rounds will print on the same basic plane due to positive compensation, excluding pesky fliers caused by whatever reason beyond this thread we find that once in the node fps changes are not relative to elevation changes much the same as smallest ES doesn’t guarantee smallest groups.
Attached a 1000 yard ladder test showing while not the smallest groups they are in a node 4/10 wide or 24 kernels approx, I have no idea on fps or the es.

 

[Seedy Character]

While I weigh every rifle charge, using a dropped charge, then a trickle to add a grain or 2 or .... to get to my specified charge weight, as most here seem to be doing; what I perceive the OP to be asking about 1 kernel, flake, ball making a difference, would entail NOT weighing a specific charge +/- accuracy of scale, but physically counting a charge volume of xxx kernels. Does +/- 1 kernel matter?


Answer is nobody is going to count 2846 kernels, 5 times, load and shoot.
Repeat using 2847. Chrony each shot compare both groups. Not perfinant, relevant or practical.


A more reasonable question would be: does weight vs volume of powder charge make a difference? Load 20 rounds each, using a Lee powder scoop and on 5-0-5 balance (+/- .2gr) and a precision balance (+/- .02gr).
Shoot 4 groups of 5 with each load and compare


Results? I am not curious enough to do it. I am very unlikely to change my 45 yo routine of loading.
In a hurry for a Cowboy shoots, I loaded 200 rounds by volume. Shooting gongs, not groups, it was sufficient.

Loads near MAX, no way I would trust volume

 

[Varminterror]

For me, it’s the speed, convenience AND accuracy.

The topic comes up occasionally on ASF as to whether one kernel makes a difference at distances where it might matter, and as you can imagine, it’s hotly debated.

You guys check my math on this

If my competition 308 win load is:

43.5gn of Varget yielding 2,725 fps, that is:

62 fps / 1gn
5 kernals / 1gn
1 kernal = 12 fps

Does 12 fps matter at 1,000 yards? Not if your wind reading skills suck

I’m not convinced even linearities like that cross the X/Y intercept. I’ll have to plot out a couple of my loads when I get home (coaching wrestling tournament today), but I know I haven’t observed anywhere near 12fps/kernel. So that first calculation of 2725/43.5 = 62.6 fps/grn doesn’t pan out.

If memory serves, I’m often in the ~10fps/TENTH ballpark - or maybe it’s 10fps per 2 TENTHS, so each kernel would be something on the order of 1-2fps/kernel.

I’ll do the math on a couple of my loads when I get home to my data - one of my Dasher load work-ups a few weeks ago produced almost straight linearity for 2 whole grains, so that data set isn’t even complicated by flat spots of nodes breaking the linearity.

 

[Nature Boy]

@Varminterror

I’m pretty sure it isn’t linear either. I’ll be curious to see what you have observed.

 

[Varminterror]

@Varminterror

I’m pretty sure it isn’t linear either. I’ll be curious to see what you have observed.

I suppose this thread should be motivation to formally study this, but I just coincidentally have a LOT of data and observation from doing my Satterlee tests every few weeks all year.

This is one example of data I have on hand (during a 10 match lull between my wrestlers) - a pic on my phone - when you look at the gain here over individual sections or the entire length, something around 20fps per increment up to 30fps per increment seem to fit relatively well - since each increment is 2 tenths, that’s 10-15fps per tenth. Given my general observation of ~12ish kernels per tenth, this is how I come up with my estimate above of 1-2fps/kernel for that 6 creed load (H4350 under 105H).

But I can get to a BUNCH of data and do some actual regressions after about 25 wrestling matches and a 25mile drive on snow and sleet roads.

 

[Varminterror]

You guys check my math on this

[snip]

5 kernals / 1gn

No criticism intended at all, just correction, as I actually made the same mistake earlier doing the math in my head: I should have caught this earlier since you specifically said “check my math”… you made a decimal error… 5 kernels = ~1 TENTH of a grain, not 1grn. So your 5 kernels/gn should be 50 instead, so then your calculated 12fps/kernel should actually be 1.2fps/kernel.

Comparing this on my own data, so far I have done a simple regression on one recent load development for Dasher - no obvious nodes in that test, almost perfectly linear (how ridiculously annoying ). The regression shows 66.33fps per grain, and using my 0.08grn per 5 kernel approximation, which yields ~62.5 kernels per 1grn, 66.33/62.5 = 1.06fps/kernel. This is Varget under 105H in Dasher… no idea why this one went so sideways, think my neck tension was poop…



So I pulled the raw data, took the velocity change between the average for each 3 shot set, then converted that to fps per grain, then per kernel, range for the increments was 0.29 to 1.39 fps/kernel, with the average 1.04fps/kernel. Pretty close to the regressed slope of 1.06fps/kernel.



I did the same with some of my 6 Creed data, 3 sets of data so far, using the same 62.5 kernels per grain divisor, I calculated 0.92, 0.96, and 0.96 fps/kernel.






I can pull some Varget and H4350 to measure and refine the 62.5 observational estimate, but the result isn’t going to change much. I speculated ~1-2fps/kernel earlier without analyzing data, and given 0.015grn per kernel I’ve observed or the 0.04grn/kernel presented in this thread, the data would reflect exactly that range…

I can keep going, I do these tests for almost every match I shoot. If memory serves, I have something like 35 archived series’ on my Magnetospeed computer, at least half of which would be these tests, and I have at least 9 of them for my 6 creed (6 more plus those 3 above) and a couple more for my Dasher saved already in excel. I’m betting heavily that the same 1-2fps/kernel persists.

Anecdotally, I complained at a match a few weeks ago to another shooter (top level PRS and ELR guy) about that straight line Dasher test pictured above - he responded, “so yeah, you just got a straight like, 7fps per tenth result, eh?”… so he obviously knew the game… since I’m getting 6.5-6.6fps/tenth…

 

[Varminterror]

Thinking more on this topic this evening, I’m reminded of EXACTLY why my load development focuses upon finding flat spots to eliminate velocity sensitivity in my charging process.

If linear regression of change in velocity per change in charge weight reveals ~1fps/kernel, and a typical dispenser is only accurate within ~12kernels - only within ~12fps, and acknowledging that may influence ~3” at 1,000yrds into my groups which might otherwise only be 7”… yeah, I’ll keep looking for flat spots in my Satterlee curves and looking for bunching of POI’s in my Audette Ladders, so that it remains true that I can miss by 30-40 kernels and still impact at the same waterline, rather than loading out of nodes and risk that error.

 

[JFrank]

I bought mine just to check the trickling device I built, to test with.



If you have any buddies that need one I can hook them up with a very lightly used A&D, never even had the protective keypad cover off of it.

I like it

 

[JFrank]

Thinking more on this topic this evening, I’m reminded of EXACTLY why my load development focuses upon finding flat spots to eliminate velocity sensitivity in my charging process.

If linear regression of change in velocity per change in charge weight reveals ~1fps/kernel, and a typical dispenser is only accurate within ~12kernels - only within ~12fps, and acknowledging that may influence ~3” at 1,000yrds into my groups which might otherwise only be 7”… yeah, I’ll keep looking for flat spots in my Satterlee curves and looking for bunching of POI’s in my Audette Ladders, so that it remains true that I can miss by 30-40 kernels and still impact at the same waterline, rather than loading out of nodes and risk that error.

The problem is out of tune/node rifles spit rounds any and all directions. Solve that one and we can both benefit.

 

[Varminterror]

The problem is out of tune/node rifles spit rounds any and all directions. Solve that one and we can both benefit.

Been thinking a LOT on this for the last few years, after riding a barrel as it rotted away and lost over 400fps with the same charge weight over only 1200 rounds… the charge weight node never moved, and the velocity fell like a rock in a creek (ironic wordplay intended). So I’ve reconsidered the idea that these are “velocity nodes,” and am far more convinced they are “mass nodes.” (And equally largely abandoned Chris Long’s OBT theory).

We simplify a LOT when we say “barrel harmonics.” These LR methods really only evaluate vertical dispersion, or lack thereof, but we know we observe - as you mentioned - horizontal spray as well. And not just increased or decreased wind drift dependent upon ToF…

 

[lordpaxman]

The problem is out of tune/node rifles spit rounds any and all directions. Solve that one and we can both benefit.

Thank you! Here I thought I was going to have to count kernels but this is a lot easier to point to.
You guys really don’t count kernels, right? Your equipment is precise enough to “drop” a charge within a kernel or two, and then you autocorrect?

 

[jmorris]

Answer is nobody is going to count 2846 kernels, 5 times, load and shoot.

No, they won’t but we are talking about using measuring equipment capable of resolving the addition of a single kernel of powder.

@Varminterror

I’m pretty sure it isn’t linear either. I’ll be curious to see what you have observed.

We know it isn’t. That’s why we search out areas where the curve is or at least seems to be flat. If you were picking a charge weight and counting on a scale that was accurate to +/- .1, you would rather be shooting for the middle of the red circle vs the blue one. If you were looking for proof that exact charges make a difference you would focus your study inside the blue circle.

Thinking more on this topic this evening, I’m reminded of EXACTLY why my load development focuses upon finding flat spots to eliminate velocity sensitivity in my charging process.

I agree 110% because this leaves us the largest margin for error but that’s different than saying across the board, it doesn’t matter.

 

[Nature Boy]

No criticism intended at all, just correction, as I actually made the same mistake earlier doing the math in my head: I should have caught this earlier since you specifically said “check my math”… you made a decimal error… 5 kernels = ~1 TENTH of a grain, not 1grn. So your 5 kernels/gn should be 50 instead, so then your calculated 12fps/kernel should actually be 1.2fps/kernel.

Yep. That’s why we have peer review as a cornerstone of the scientific process

 

[Varminterror]

If you were looking for proof that exact charges make a difference you would focus your study inside the blue circle.

View attachment 1064310

I agree 110% because this leaves us the largest margin for error but that’s different than saying across the board, it doesn’t matter.

In fairness, for 99.99% of shooters, even if we loaded in that blue circled region, the consequence of +/-1 kernel versus +/-0.1grn is still nill, or very near to it.

*Recognizing first - that blue circled region contains about the steepest slope I have ever seen in this chart, especially for the individual string, gaining from ~3100fps to 3147fps for 0.2grn increase… so let’s play with that data…

Let’s assume a shooter and rifle capable of 1moa at 1000yrds when they have PERFECT ammo, no velocity variability at all.

Loading to the kernel in the blue region would mean +/-4fps, which yields a maximum potential of 1.6” vertical - but when properly estimated by the square root of the sum of the squares, that only actually adds <1/8” to the group vertical.

Loading to +/-0.1grn within the blue region would mean +/-25.5fps, yielding 10.2” maximum potential vertical - which seems like a lot, but again, statistically would only ACTUALLY add 4.1” to our group…

So the difference between loading to within a kernel versus within a TENTH of a grain is ~4” at 1,000 yards… REMEMBER - each click of our scope at 1,000 yards is 2.5” or 3.6”, so the difference of one scale resolution interval between +/-1 kernel loading and 0.1grn loading is only ONE click.

*Interrupting here to point out - very, very few shooters can estimate windspeed within 1mph over 1,000yrds. Wind error of +/-1mph at 1,000yrds means 13.5” potential, with 6.6” increased horizontal statistically expected in the groups (again, on a 1moa raw 1,000yrd rifle). So +/-1mph wind error is literally 50% greater error than +/-0.1grn at 1,000.

Even in the blue section, 1 kernel doesn’t matter. Even in the blue section, 0.1grn doesn’t REALLY matter - it’s just ONE click difference. But we do it anyway.

 

[JFrank]

Thank you! Here I thought I was going to have to count kernels but this is a lot easier to point to.
You guys really don’t count kernels, right? Your equipment is precise enough to “drop” a charge within a kernel or two, and then you autocorrect?

No we don’t although my scale will resolve one kernel. Best path is just load to the middle of the node and you are golden.

 

[Varminterror]

No we don’t although my scale will resolve one kernel. Best path is just load to the middle of the node and you are golden.

I can’t help myself though - when my dispenser drops 41.68, I HAVE to tap the trickler to drop that last kernel to get to 41.70, and when it overspews and hits 41.74, I HAVE to pick those two kernels out of the cup to hit 41.70…

Even though I know (as outlined in my data here in this thread) that over or undershot of one kernel or two won’t influence diddly squat downrange, I still HAVE to do it…

 

[jmorris]

Even though I know (as outlined in my data here in this thread) that over or undershot of one kernel or two won’t influence diddly squat downrange, I still HAVE to do it…

You would be surprised how many competitive shooters do not understand the importance of their mental game.

If one thinks they shoot better if they are sticking their tongue out, they should do just that for every shot.

 

[Engineer1911]

.........................................

Even though I know (as outlined in my data here in this thread) that over or undershot of one kernel or two won’t influence diddly squat downrange, I still HAVE to do it…

But you don't have to proclaim this "correctness" to the reloading world as the only way to load ammunition. The problem is someone might believe you.

 

[Varminterror]

But you don't have to proclaim this "correctness" to the reloading world as the only way to load ammunition. The problem is someone might believe you.

I welcome you to take your baggage to someone else’s airline. I don’t play passive aggressive games.

I’ve pointed out - in many, many threads on this forum and a dozen others, for 20+ years that there are multiple methods to achieve the same goal. I pointed out - in the exact quote you copied here - that despite knowing it’s not critical, I still feel better loading to within 1 kernel. Maybe the irony was lost on you?

The data analysis I offered in this thread is meant to illustrate that multiple methods (very clearly prescribed when I laid out the difference between loading ammo with a Chargemaster vs. an Fx-120i) can yield exceptionally similar results. Maybe you weren’t able to follow the math, but analysis of my existing data confirmed that loading in a node or not, loading to within 1 kernel or 1 tenth of a grain, all remained within 4” at 1,000yrds, only varying by 1 click on the scope’s turret from a true POA.

Data is data. I provided the data. If you read the data and resultingly choose to reload with a Chargemaster, powder drop, or an FX-120i, or pound sand instead of reload at all, it really doesn’t affect my day.

 

[lordpaxman]

If one thinks they shoot better if they are sticking their tongue out, they should do just that for every shot.

I shoot with a guy that does that. He’s really good. I like shiny brass. Whatever floats your boat, more power to you.

 

[thump_rrr]

I have 200 pieces of Lapua small primer brass and I wait till the whole batch is gone through so that all my brass has the same number of firings before annealing.
I had 25 loaded rounds of 6.5CM left over from last year.
I broke them down to double check how all my Chargemaster did compared to my newly acquired A&D FX120-i. I will edit this post tonight with the actual numbers.

While I don’t think that I need to weigh to 1 kernel of powder to get an accurate load, knowing that my powder charges are consistent allows me to focus on other areas where I may be deficient.

My current concern is that my neck tension that isn’t consistent so I am going down that rabbit hole.

 

[Nature Boy]

It takes me 1/2 the time to load to the kernel with the A&D FX 120 and Autoloader than it did with my GP250. As said before, I do it because I can. I can also choose to be precise in my process or sloppy, and if either have no measurable benefit down range, I’ll err on the side of being precise.

There is a benefit however. It provides confidence that I’ll stay in the middle of my accuracy node, regardless of changing conditions. Confidence in your ammo and equipment does show up on target.

Believe that.

 

[Engineer1911]

Confidence in your ammo and equipment is good. A 20 round magazine is better, six 20 round magazines is best.

 

[Nature Boy]

Confidence in your ammo and equipment is good. A 20 round magazine is better, six 20 round magazines is best.

Lots of loaded magazines are best if you’re laying down a base of covering fire, but they’re useless as teets on a boar when you’re shooting a high power match that requires you to singe load. Hell, my Kelbly Panda doesn’t even have a place to stick a magazine.

But I guess I get what you’re saying; volume over accuracy. It explains why the subject of this thread might be so foreign to you.