Power Charge Load Dev. Question - Noobie

[Dixie Reckt]

Noob reloader here - based on this limited powder charge test where would you set your powder node at and why? Do I use the 94.5 node or the 97.5 node? STDEV is based on the charge before and the charge after.

In the photo (shot #13) the black stuff around the primer is actually neolube that was sitting in the bottom of my loading block - that's not carbon.

 

[Krogen]

Standard deviation is useless with a sample of 3 shots. On top of that, you're using different loads for the sample population. Standard deviation is a way to measure consistency of a random distribution. You're adding in a systemic variable; the differing powder charges. You can't draw any conclusions from this.

If you're looking for nodes, do a ladder test on paper. How to do that is well cussed and discussed here on the form. Search and ye shall find.

 

[Dixie Reckt]

Standard deviation is useless with a sample of 3 shots. On top of that, you're using different loads for the sample population. Standard deviation is a way to measure consistency of a random distribution. You're adding in a systemic variable; the differing powder charges. You can't draw any conclusions from this.

If you're looking for nodes, do a ladder test on paper. How to do that is well cussed and discussed here on the form. Search and ye shall find.

I'm no expert but to parrot what I heard on Cortina's podcast he and Lou Murdica say the powder charge test is only to test for a velocity node. I don't care about groups on paper at this stage - I'm just trying to pick a powder charge where the velocity seems stable. I don't think you understand the test I did here - I'm intentionally doing different powder charges to look for a velocity node and also see where my pressure limit is.

 

[Krogen]

Oh, I do understand what you did. You're trying to pick a velocity node based on invalid standard deviation calculations.

Try plotting your data instead. Your velocity compared to powder charge should be monotonic. If there are flat spots where velocity changes less with charge increase you may have identified an area where velocity is relatively insensitive to incremental charge variations. It's problematic though, when you only have one shot per powder charge.

 

[ELR LVR]

If there are flat spots where velocity changes less with charge increase you may have identified an area where velocity is relatively insensitive to incremental charge variations.

Look for this, ^^^
then see if these spots are accurate,
Low SD is not indicative of accuracy

 

[RyeGuy]

Those last few shots seem pretty SPICY, I’d try 94.5 or 95.5.
Hunting rig or steel? Do you need the extra FPS for less bbl wear? Or better bbl wear? 94.00-96.00 only jumped 43FPS, I’d say that’s a hell of a window.

 

[ELR LVR]

Those last few shots seem pretty SPICY, I’d try 94.5 or 95.5.
Hunting rig or steel? Do you need the extra FPS for less bbl wear? Or better bbl wear? 94.00-96.00 only jumped 43FPS, I’d say that’s a hell of a window.

Yes, just by looking I was thinking maybe the 95.5 - 96.0 grns window
only 3 fps jump from 95.5-96.0
Maybe split the difference and go 95.7 and have 1/10th grn leeway on either side so a guy doesnt have to be right on the money, and can load faster maybe even without trickling.
---
I try to find loads where I can go 1/10th high or Low and it not make a difference in accuracy
In other words, Forgiving windows
Same with seating depth,
Forgiving windows help allow for changes in outside conditions as well

 

[Dixie Reckt]

Look for this, ^^^
then see if these spots are accurate,
Low SD is not indicative of accuracy

Is that not exactly what I did? I genuinely thought that's what I did.

 

[Dixie Reckt]

Oh, I do understand what you did. You're trying to pick a velocity node based on invalid standard deviation calculations.

Try plotting your data instead. Your velocity compared to powder charge should be monotonic. If there are flat spots where velocity changes less with charge increase you may have identified an area where velocity is relatively insensitive to incremental charge variations. It's problematic though, when you only have one shot per powder charge.

I got ya. Why is this invalid stdev calcs? The STDdev is basically saying how flat the line is if i were to plot it out. I can plot it real quick. But with the limited amount of info here without doing more tests I'm wondering what charge would you pick? Without plotting it you can see the flatter spots are 94.5 and 97.5 right? Ideally I break down those nodes in 1/10 grain increments and plot that right? But for quick and dirty load development...

 

[Clancy]

I’ve been doing this 20 years and didn’t even look at the chart you posted, because it’s useless without group data..

Color some bullet tips with different color sharpies and shoot a ladder test at bare minimum 600 yards.

There will be some groups that overlap. In a cartridge that size I would expect an impact node a couple of grains wide.

Once you have that, if you’re not satisfied with the performance, work on bullet seating depth.

 

[Dixie Reckt]

Look for this, ^^^
then see if these spots are accurate,
Low SD is not indicative of accuracy

The lower an stdev value means its a flatter spot if i were to plot it, I think thats where the confusion is. Added a plot to another reply in this thread.

 

[Dixie Reckt]

Those last few shots seem pretty SPICY, I’d try 94.5 or 95.5.
Hunting rig or steel? Do you need the extra FPS for less bbl wear? Or better bbl wear? 94.00-96.00 only jumped 43FPS, I’d say that’s a hell of a window.

I want most accuracy.

 

[ELR LVR]

The lower an stdev value means its a flatter spot if i were to plot it, I think thats where the confusion is. Added a plot to another reply in this thread.

Yes understood on the SD aspect
All I meant was, the Load could have Low SD yet not be a very accurate load, such as:
the bullet could be exiting at the wrong time during the barrels vibration even when SD numbers are low
providing for worse accuracy, and must be adjusted for somehow
(Low SD's do not always equate to optimum accuracy, even though we want low SD's)
The load may be adjusted for lack of accuracy, say with Seating Depth or with a Tuner or ? Something
---
My point being, if a person is merely loading for low SD's only,
they are not loading for optimum accuracy
at least not yet, but is a good place to start anyhow with collecting load data

 

[Dixie Reckt]

Yes understood on the SD aspect
All I meant was, the Load could have Low SD yet the bullet could still be exiting at the wrong time during the barrels vibration even when SD numbers are low
providing for worse accuracy, and must be adjusted for somehow
(Low SD's do not always equate to optimum accuracy, even though we want low SD's)
The load may be adjusted for lack of accuracy, say with Seating Depth or with a Tuner or ? Something

Basically the reason I'm asking this is I want to move on to bullet seating depth test without doing more powder testing. Freakin tired of poweder testing on this gun, this is like my 4th attempt at load development. Previously I was using mag length limited bullet depth. Now I'm in the lands. The barrel will be shot out prolly in another 500 arounds.

 

[ELR LVR]

I am with ya, and honestly
I pick a safe middle of the road load velocity wise
and conduct seating depth test first
---
This has allowed me to then adjust powder charge, change POWDERS, change primers and conduct further accuracy tests, while using that one seating depth.
---
I would go ahead and move on to seating depth to find your best accuracy and then refine load
---
Sometimes I will then after all that is done
Play with seating depth again but, 99% of the time that one I found initially always works
I may find an additional seating depth later that plays better with the new refined load
but the initial one is often what I go back to when experimenting again

 

[ELR LVR]

Is that not exactly what I did? I genuinely thought that's what I did.

(whispers)
yes I believe so

 

[JFrank]

I got ya. Why is this invalid stdev calcs? The STDdev is basically saying how flat the line is if i were to plot it out. I can plot it real quick. But with the limited amount of info here without doing more tests I'm wondering what charge would you pick? Without plotting it you can see the flatter spots are 94.5 and 97.5 right? Ideally I break down those nodes in 1/10 grain increments and plot that right? But for quick and dirty load development...View attachment 1760905

it’s invalid because it will be different tomorrow.

 

[charlie b]

I tried the 'flatter' velocity curve thing for a while. It happened to be right a couple times in many, many instances.. FWIW, SD doesn't mean much until you have 10 samples or more, ie, 10 per charge. Statistically you need a lot more.

Use the target. I start at 200yd and then verify at 600. As clancy mentioned, you are looking for a spot where the vertical does not change as much with a change in charge/vel.

 

[Krogen]

I tried the 'flatter' velocity curve thing for a while. It happened to be right a couple times in many, many instances.. FWIW, SD doesn't mean much until you have 10 samples or more, ie, 10 per charge. Statistically you need a lot more.

Use the target. I start at 200yd and then verify at 600. As clancy mentioned, you are looking for a spot where the vertical does not change as much with a change in charge/vel.

Problem is, OP is trying to use a standard deviation calculation on 3 shots, each with a different powder charge. As you say, a statistically valid sample is at least 10 shots. Statistics conventions indicate a population of 30. And that is for articles that are ostensibly the same, but with random variations. Standard deviation is meaningless for articles that are >purposely< different. Standard deviation gives an indication of the width of a normal distribution; the bell curve.

To the op, just because a calculation can be made and it appears to correlate with your observations doesn't mean you've arrive at a valid correlation.