I have been working on some precision loads for my two precision rifles. One is a 6.5 Creedmoor the other 224 Valkyrie both bolt action. I am issuing a Dillon beam scale and IMR 4350 for the Creedmoor and CFE223 for the Valkyrie. I have been working velocity node testing hoping to find a powder load that gives me single digit standard deviations. This has been less than successful. The best SD I have seen so far has been 14.7 and that was not repeatable. My velocities are all over the place. I get a couple rounds close then rounds that are several hundred fps different resulting in SD’s of 200 fps +\-. I don’t think this is my load weights as variations that large in fps would require over a grain difference in powder charges which is not happening. So that leads me to wonder if it is the powders or is my magnetospeed giving me bad numbers? I have had several times in the past when doing incrementally larger charges in sequence where the magnetospeed would give me a slower round in the middle of a string when it should have gone up in velocity. This has happened several times. I am just wondering if maybe I need to invest in a better scale, or chrono? Or maybe give a different powder a go when it becomes available again. Any thoughts?