Weird Dillon 550 45acp powder drops

Background - I've been having a light load / FTE issue for a while - but only with the 45ACP. 38 is fine. Both use WST.  

Last week - out of sheer frustration - I measured every load of 100, rejecting out of spec ones. I'm aiming for 3.7 gr. Measuring on an electronic scale with +/- .02 gr accuracy. Didn't write them all down but I was typically seeing around 10 drops at 3.68-3.72 (in other words accurate to within the bounds of the equipment) then I'd get an outlier like 3.88 or, memorably, 3.54. Tip that back in - the next drop is right in range again. So around 1 in 10 overall was out of range. I could understand it if a light load was followed by a heavy one, or vice versa, but that's not happening.

I've been swapping/installing stuff for a while to try and fix the problem. I currently have a powder baffle fitted, Arredondo powder bar with Uniqetek micrometer and one of Alex's powder drops. Everything has been polished to the best of my ability (limited). None of these modifications have affected the problem.  The setup is pretty much identical to the .38 - which isn't exhibiting the issue.

The outcome was on Sunday's match I had no ammo issues. We'll not talk about trigger control.

Anyone encountered anything like this before?

Thoughts and suggestions welcomed.

I suggest checking Fail Safety Mechanism.  No questions issue is related to the range of powder bar motion.  Specifically if bracket is worn out and flexes powder bar doesn't go all way, and hence you'll have low charge.

AP

Could be. I have adjusted the screw setting on the bottom of the actuator rod and the little square insert is hitting both ends of the 'channel'.

I might just get a new power measure body and move everything across. That's about the only thing that hasn't been swapped out.

straybrit wrote:Background - I've been having a light load / FTE issue for a while - but only with the 45ACP. 38 is fine. Both use WST.  

Last week - out of sheer frustration - I measured every load of 100, rejecting out of spec ones. I'm aiming for 3.7 gr. Measuring on an electronic scale with +/- .02 gr accuracy. Didn't write them all down but I was typically seeing around 10 drops at 3.68-3.72 (in other words accurate to within the bounds of the equipment) then I'd get an outlier like 3.88 or, memorably, 3.54. Tip that back in - the next drop is right in range again. So around 1 in 10 overall was out of range. I could understand it if a light load was followed by a heavy one, or vice versa, but that's not happening.

I've been swapping/installing stuff for a while to try and fix the problem. I currently have a powder baffle fitted, Arredondo powder bar with Uniqetek micrometer and one of Alex's powder drops. Everything has been polished to the best of my ability (limited). None of these modifications have affected the problem.  The setup is pretty much identical to the .38 - which isn't exhibiting the issue.

The outcome was on Sunday's match I had no ammo issues. We'll not talk about trigger control.

Anyone encountered anything like this before?

Thoughts and suggestions welcomed.

Some thoughts; please take what you think may help and discard the rest:

1) Is the loading deck full when you are running the press? I've found that powder variance can swing wildly if all stations are not occupied.
2) Is there any undo sizing effort needed on some occasional piece of brass?
3) If so, is the brass a particular headstamp?
4) How sturdy is the bench?
5) Are you making the same movement with the handle every time?
6) Have you checked the failsafe rod spring adjustment?
7) Any difficulty priming? What brand of primers? Unis Ginex are tight, tight, tight.
Is the press catching anywhere on the way up- a fresh case perhaps not seated all the way, or bullet tipped just enough to induce a slight stoppage going into the seater?

My experience has been that when these, or conjunctions of these occur, I may find the reasons for powder drop variability. Using WST, which meters like water, will be easier to jostle and increase or decrease powder charge due to machine movement being more or less severe than "normal."

Again, take that for what it is worth. Hope it helps.

-Chase

If I remember correctly, WST is a flake powder. My suggestions are based on this being the case. Flake type powders are typically harder to load than extruded or flattened ball. The problem is it sticks to everything. So tap the side of the powder hopper to help it fall. Someone suggested a small aquarium air pump to vibrate the powder to the bottom.

If your loading to within 0.02gr that's as good as your going to get. I use 3.7gr of N310 for 45. I'll get a weight differential of 0.10 on the total (37.0gr) of ten throws. That 0.10gr could be anywhere as you have found out.

I really don't think the problem lies in the machine. Unless your willing to weigh and trickle each charge, 3.7gr in every case every time is just not doable.
I would suggest instead of chasing the weight variable, try using 3.8gr to eliminate the failure to eject problem. (Where I'd start).

You could also take a look at the spring weights in the pistol, but that's an entirely different topic.
Hope this helps in some small way.

It's extremely dry during the winter in many locations. Try using a drier sheet to wipe down the hopper to eliminate static electricity. In some extreme instances it may be necessary to ground the powder mechanism.

Ask me how I know this?!!

I think I may not have explained myself clearly - let me try again.

We can see that the press is capable of producing identical (within the limits of available measurement) loads on a repeated basis. Yet - every so often I get an outlier. It's all the same brass (when this started I got anal enough to sort by headstamp) and as far as I can tell I'm using the same physical action. The bench is pretty sturdy 4x4 legs, 2x4 frame with no more than 12" between joists and topped with 3/4 ply and lots of screws attaching it. I have wiped everything in the powder path with a drier cloth.

Let's go back to the numbers - the low load of 3.54 is a massive outlier. If I recall my Mech Eng lessons from many decades ago that's >6 SDs away from the mean. A healthy system just doesn't do that. Now - I can accept that there may have been a static hangup stopping powder coming through and on the next cycle the bar filled correctly. However - let's look at the other end of the curve. 3.88. Again that's a > 6SD deviation - and - where did that extra mass come from? The loads on either side were nominal.  Is there some bizarre inconsistency in powder density? I do make sure to vibrate the powder prior to putting it in the hopper because, per Winchester tech support, the graphite coating on the flakes can congeal and it needs to be broken up.

Apart from anything else this shows the weakness of the 'measure 10 loads and average them out' approach. The press is quite capable of producing 10 loads of x weight and you're good - it's also as I've seen quite capable of producing 9 loads of x and 1 load of x - 5%. Over the ten loads it still looks adequate when totaled - but you've got one very light load in there.

I'm still no nearer to solving this - though I did adopt the 'add another .1g to make sure the lighter loads function' approach. Which, although irritating, is a functional short line workaround. I guess I'll just have to start individually measuring each long line load come summer. In the meantime I see that UniqueTek have a hopper grounding system so I'll try that - although the idea of connecting a powder filled container to the electricity supply does seem somewhat counter-intuitive (yes - I know it's just the ground connection - still feels wrong). Maybe I'll try the pump/vibrator after that.

Edit to add - while on the Uniqetek website I grabbed and re-read their accuracy tips file. There's a whole bunch of ideas in there. Methinks I have a busy few evenings ahead of me.

Texasref wrote:If I remember correctly, WST is a flake powder. My suggestions are based on this being the case. Flake type powders are typically harder to load than extruded or flattened ball. The problem is it sticks to everything. So tap the side of the powder hopper to help it fall. Someone suggested a small aquarium air pump to vibrate the powder to the bottom.

If your loading to within 0.02gr that's as good as your going to get. I use 3.7gr of N310 for 45. I'll get a weight differential of 0.10 on the total (37.0gr) of ten throws. That 0.10gr could be anywhere as you have found out.

I really don't think the problem lies in the machine. Unless your willing to weigh and trickle each charge, 3.7gr in every case every time is just not doable.
I would suggest instead of chasing the weight variable, try using 3.8gr to eliminate the failure to eject problem. (Where I'd start).

You could also take a look at the spring weights in the pistol, but that's an entirely different topic.
Hope this helps in some small way.

WST is Ball Powder

Sometimes that Arredondo powder bar will stick, flex or bind and also the adjustment bar can come loose or wobble.  Make sure you have fee movement throughout the range of motion.  Disconnect the actuator and cycle by hand pressure only to see if there are any tight spots, etc.  Too tight or too loose will cause issues.  And it there are any loose grains lying around, they can get trapped under the spaces on the bottom of the Arredondo bar and drag / stick.  I have swapped out the upper spacer bar sometimes to get better / more consistent operations / stop powder leaks, etc. 

Next, check that little plastic square that rides in the slot to actuate the powder bar. It can fall out, slip or get in a bind as well. And also a couple of the cap screws that hold all that fail-safe and actuator stuff together can back out a little.

Suggestion is to swap back to the standard Dillon powder bar and use 2 wavey washers to keep the adjustment from creeping and see what you get. 

CR

OK - I realize this is turning into something of an epic - but hopefully it helps someone else.

Changes made :
Polished the inside of the powder funnel (600 grit, 1500 grit then aluminium polishing compound) - not quite a mirror finish 
Swapped the Arredondo for a Dillon small bar - with UniqueTek micrometer
Replaced the 'wavy washer' making sure it's the right way around
Cleaned everything with 90% isoprop
Wiped down the inside of the powder hopper and delivery mechanism with drier cloth

Not yet grounded - waiting on that

Moved everything by hand to make sure there's no tight spots.

So - to test - 109 powder drops, 94 were in the 3.68-3.72 bracket - in other words within the measurement tolerances
min observed 3.66 [8]
max observed 3.78 [3]

Interestingly the max and most of the min came in the last 20 drops - so I rather suspect there's some static build up happening. Given that it's currently 25 below outdoors here and any moisture that was in the atmosphere is now lying on the floor that isn't surprising. 

I'll get the grounding kit in the next couple of days and see what that does - if anything. Even without that it's reduced the spread to about 3SD and the variance to < 1x10e-3 which I think for a system with this number of uncontrolled variables is pretty good. I'm sure Wobbly will be along any time now to correct my long forgotten eng stats :-)

Well, the data are very narrow.  90%+ are within .04 grains and, unless you have a very expensive laboratory scale, the resolution is +/- .01.  So the data can never be anything but relatively narrow.  Hard to even define a decent analysis of variance.


Since you have a method of measuring powder to +/- .05 grains, that is VERY LIKELY good enough.

Electronic or beam scale?

Electronic. Quoted as having a +/- 0.02g accuracy. Hence my accepting anything in 3.68 - 3.72 range as within the requirement.

My experience with electronic is that you have to be on the lookout for drift all the time.

It would be great, given the non-random (about every tenth drop) nature of your outliers, if you figured this out and got back to us after you validated.  A total range of .3 grains is pretty much what we all already know to be the limits of the Dillon measure with some powders. Since this equates to nearly 100 ft/sec velocity delta and that equates (according to Hornady’s free ballistic app) to a bullet drop (POI) shift at fifty yards of two inches, I’m interrupting my load process to use my Redding #3 to drop all loads.  This old workhorse gives me a total charge variance of .1 grains. I’d like to go back to the Dillon measure if you find the cause.

I wouldn't have said it was non-random - it was about one in 10 - not every tenth drop. 

So far the 'solution' appears to be a number of things. Polishing the inside of the powder funnel helped, putting the powder baffle in at 90 degrees so that there is a constant, low, compression also helps. Using the UniqueTek micrometer seems like a good thing as well - the Dillon adjustment screw is too coarse (mandated by the low quality metal) and needs a threaded hole in the side of the bar to put in a grub screw to hold it in place. The movement of the powder bar to get a 0.05g change in drop is really really tiny. There's no way the stock adjuster is going to hold that on its own.  I did swap between an Arredondo and a Dillon powder bar - both with micrometers, and that didn't have any measurable effect on its own.

The other big thing, for me at least, seems to be static build up. I'm waiting for a grounding kit to arrive but if I empty out the powder, wipe everything down with a drier cloth, refill and start I get about 80-100 rounds before I see flakes sticking to the side of the tube. At that point I see the powder drop variance increases. Not as much as before I made the other changes - as I said earlier it's about half of what I was seeing originally and seems to be well within .06 gr - but it's there. Prior to that point all the drops are within .02g - which is the accuracy limit of the electronic scale I'm using.

I should point out that I've done all the actual measurement tests with WST.  Going by 'feel' of the round when discharged though - the same seems to hold for N310.