I'm also thinking I should bring the charge up to 3.4 or 3.5 to get away from the low end.
Your charge of 3.3 grains is not at the low end. Hodgon's numbers for 147 grain jacketed (Horn XTPs) are a start load of 3.0 grains and a maximum load of 3.4 grains. Your chronograph results of 878 fps (
at 3.3 grains nominal powder charge) are quite a bit higher than Hodgon's 845 fps at
3.4 grains. I'd say that your batch's load of 3.3 grains, as you are measuring it, is plenty. I'd check your scale(s) with test weights, etc. to be sure.
Looking specifically at the squib load, I think that
+1. I think not enough powder charge for efficient powder burn and to fully cycle the slide.
is likely. What on the 650XL might have allowed a case with a low charge to get through the process?
These are things that I've seen on my well-used 650XL:
1) I've been through a couple of buzzers on the low powder check device. I've learned to always visually verify the rise of the check probe on each stroke of the press to back up the audible alarm. It's easy, really; my eyes make sure that the bullet is placed properly, and then shift to the nearby check probe on the second half of the stroke.
2) Variations in case inside dimensions are the normal cause of variations in check probe rise. I can't tell you how many times I've seen a variation in probe rise, stopped loading to check the powder charge, only to confirm that the Dillon powder drop was right on (generally within 0.1 grain). That being said, its never a bad idea to stop the process and check powder charge if you have ANY question.
3) You were tweaking the height of the powder drop in your effort to adjust case bell. Be absolutely sure that your powder drop is travelling all the way up and the slide is all the way into the drop's body after each adjustment that you make to case bell. Also, make sure that the powder drop body is secured to the powder die after you make this adjustment (if you forget, the press will tell you in a few dozen rounds--the powder drop will pop off of the powder die).
4) Be sure that you use the return device that pulls the powder drop back down to its rest position at the end of each stroke. Sometimes, after making bell adjustments, we forget to re-attach that rod/device, and that can result in the odd low charge. If your device has one, adjust the spring tension properly, too.
5) Buy individual press heads if you load for more than one cartridge. Since you're using Dillon dies, once you get everything set just the way you want it for 9mm, then you shouldn't ever need to adjust the dies in stations 1, 2, and 5 again. Adjust the powder check probe when changing powders or power charge, and adjust the bullet seater in station four when you change bullets or COAL but that's it. Having individual press heads for different calibers eliminates redoing all of that setup when you switch calibers.
6) If you have to stop the normal loading sequence for any reason, pay VERY close attention to powder charges in the cases in stations 3 and 4. Look in each of these cases. Take the time to weigh the powder in each of them if needed, before you restart. Do this especially when the indexing mechanism is affected. If you have any question about the charge in any case that you've already dropped a bullet on, pull it.
I've probably loaded 400k pistol rounds on my 650XL with only 1 squib load. That squib occurred when I was first learning about the machine. The machine works great, but there are a couple of finer points to remember about operation and setup.
