Effect of an empty station on a Dillon progressive loader

Having a station on my Dillon XL750 empty had a very noticeable effect on the dimensional accuracy of my loading this afternoon. This is desite the fact that my XL750 is equipped with the Armanov bolted-in toolhead and Armanov free loating die rings.

I was loading my 9mm load and did not notice that my case feed tube had run out of cases (because I was fixated on viewing the powder level in each case), until a case failed to load into Station 1 (sizing) and so the machine stroke felt too light versus normal.

I stopped and added cases, but naturally there were a few rounds made with one or more empty stations until the case feed was again normal.

Because I am testing a load, I am temporarily measuring the "base to a specific point on the ogive" using a Hornady bullet comparator, as a more accurate means of measuring TRUE "COAL". This gives a more consistemt reading than measuring from base to tip of the bullet, as the ogive of a jacketed hollow point bullet is much more closely controlled than the tip is when a bullet is manufactured.

So, in the process of measuring each of the 50 cartridges I made, I noticed a relatively large variance form the norm for a handul of cartridges. Those cartridges were made right after the case feed ran dry, and before all the stations could once again be properly populated.

Take a look at the cartridges marked via red font in this chart:

(You may need to print the chart to be able to read it)



For each group of 5 bullets, except group 8 (Bullets 36 to 40), the average Extreme Variance was just .0016".

But for the group of bullets that involved 1 or more empty stations, it was .0040".

Remember, this is despite the fact that my XL750 is equipped with the Armanov parts and the Redding Competition bullet seater whose insert has been modified to better fit the bullet ogive.

Lesson re-learned: Even with a very carefully set up progressive press, an empty station scenario still creates variances in cartridge dimensions.

Jim G

mcb said:

When I load rifle on my 650 I typically do all the case prep off machine. Size, trim and deburr off press on a single stage press and then do the loading on the 650 without the sizing die installed. No issues have been encountered.

Click to expand...

Hm, something for me to consider.

For those wondering why the desire for precision in a pistol round, it's because the pistol is a SIG P210A Target model that shoots sub one inch groups, at 25 yards, even with my 70 year old eyes looking through progressive bifocals! The pistol makes the effort for further precision worthwhile!

Jim G

EricBu said:

I've noticed this between my 750 and my 550. The 550 shell plate secures the case against the top of the ram, so there isn't much deflection if a station is empty. The case doesn't really have anywhere to go. On the 750, the shell plate also supports the case. So depending on how tight you have the shell plate, you will have some deflection with an empty station. To minimize this, I tighten the shell plate down as tight as possible while still allowing it too index. Keeping all the contact points lightly coated with lithium grease helps.

Click to expand...

Yes, I have my shellplate just a tiny fraction of a turn off the point of not being able to index. But the diameter of the shellplate increases the "leverage" that each station has to tilt the shellplate. But the spacing of the die centers is already too tight for easy wrench access, so I guess Dillon was kind of stuck with the tilting once they adopted the 650 / 750 type of shellplate.

The deflection effect was much worse before I added the Armanov parts and the Redding Competition seating die with a customized seating insert.

Jim G

JJFitch said:

Lots of good info above....

Have you done any accuracy tests validating the .003 variance? If so how dose this variance compare to other variances such as 1/10 grain +/- powder drop or +/- .001 case length, temperature changes, humidity changes, sun angle and so on.

I get the point and I'm also OCD but I kept records of matches at various ranges and recorded temperature, sun angle, wind direction and so on. Each had an affect or POI/POA and knowing this data when shooting the next years matches was helpful at the PPC 50 yard line. My PPC-9 was equal to the 210 in accuracy!

I don't recall measuring every round but at least the first 10 were measured. I have 5 Dillons (550,450) set up for each cartridge I shoot. I have never shot any rounds from a Ransom rest but I was satisfied that my practice ammo was equal to the Winchester ammo we were issued. Our department pistol team was sponsored by Winchester!

All the best, the 210 is worth the extra effort!

Click to expand...

I have been testing at the range, but I still have another issue that masks any variance in COAL now: The muzzle velocity SD = 21 and ES = 93 ! Despite that ES, the P210 continues to produce very fine groups despite my optical shortcomings. I THINK the issue is the VV3N38's slow burn rate. QuickLOAD and GRT both agree that only about 70% of the powder burns before the bullet exists the muzzle, and they both predict that performance might be somewhat erratic. But if groups in the .88" to 1.18" range (including all flyers) represent erratic performance, I'm not sure I want to lose that erratic performance!

I continue to investigate and try things . . .

Jim G

The muzzle velocity SD and ES statistics did actually come out better with this latest batch of cartridges that I described above.

My last earlier muzzle velocity SD and ES were SD = 21 and ES = 93, per the Labradar.

This last batch of bullets was clocked by the Labradar at SD = 18 and ES = 71.

So, I just examined my earlier Labradar files which were generated using differing powder loads (ranging from 7.2 to 8.1 grains), and found that the first batch I fired of the 8.0 grain load in that test, at a COAL (measured the old way using actual base-to-tip COAL (versus base-to-.338 gage via bullet comparator) as 1.142", showed SD =10 and ES = 36, AND the groups fired with that COAL were the most accurate load in that round of testing. The average muzzle velocity of that load was 1328 fps, which is a bit higher than my current average velocity. That makes sense since a shorter COAL raises pressure, and that COAL is shorter than my current COAL.

My CURRENT Base-to-.338 Gage length calculates to a traditional (base to tip) COAL of 1.144". THis is a bit longer cOAL than that earlier load witht he better SD and ES.

Also, the average muzzle velocity of that earlier load was 1328 fps, which is a bit higher than my current average velocity of 1298 fps. That makes sense since a longer COAL lowers pressure, and my latest COAL is longer than my earlier COAL.

Also, my P210 pistol did not seem to like the even longer COALs I have tried.

So, I'm thinking I need to go back to that 1.142" COAL which with my Base-to-.338 Gage measuring technique is 0.9143", or about .0017" shorter than my current Base-to-.338 Gage length. I will try that and see if it reproduces the earlier better SD and ES.

It's hard to believe that a pistol round could be THAT sensitive to COAL, but reloading does occasionally include surprises.

Jim G

jmorris said:

It’s because of the close range of the testing, we are talking about 75 feet and ~ 1300 fps isn’t exactly throwing a loaf of bread.

That said if 93 is the best your getting, I’d try another powder pretty quickly.

Click to expand...

There is no other powder that delivers as much muzzle velocity as the VV 3N38. I want 1300 fps or more, withOUT using a maximum load.

And again, remember, despite the crapp statistics, this load is VERY accurate.

Jim G

JJFitch said:

I must have missed it but what bullet are you testing and what velocity are you seeking and what other powders do you have?

I did extensive 9mm testing around 30 years ago but I was focusing on 135 and 147 grain RNFP! Todays powders should make getting satisfactory data and POI a lot easier! 's

Click to expand...

Bullet = Hornady 115g HAP (excellent quality, and the pure narrowing cone ogive makes bullet seating more accurate)

I want 1300 fps or more muzzle velocity, and I don't want to have to use a "maximum load" to get it. VV 3N38 gives me that velocity halfway up the load table.

I have no other powders yet, and I don't think there is another powder that will give me the above velocity at mid-table number of grains of powder AND burn very clean (I get NO muzzle flash at all), AND run at comparatively low peak pressure (3N38 burns very slow, so the high velocity is a result of longer burn at lower peak pressure).

Jim G

Jim Watson said:

Were you weighing powder charges? First round after a skip on the plate can get more powder as the extra jiggle settles it into the bar.

Click to expand...

Good point to examine. The powder throw has been very consistent (after I cleared up a static electricity issue a few weeks ago when the weather was very cold and furnace was running a lot). But, the velocity variances occur even when there has NOT been an empty station incident. And VV 3N38 flows very nicely because of its tiny particle size.

Jim G

frogfurr said:

.003" play isn't much. There is some play in the tool head groove on my 550. One layer of scotch tape's worth.

Click to expand...

My replacement Armanov toolhead is BOLTED into place on my XL750, and each die is free floated via Armanov free float die lock rings. That's why the normal ES is now so low.

Jim G