Originally Posted by Bullet bumper
It sounds like your chambers are a bit long if the fired brass is .017 longer than new brass . This is not such an issue with a belted case as the headspace is set by the belt untill the case is a real jam fit and then it might set on the shoulder . Even if there is .017 Approx. extra room at the shoulder the case just blows forward about .011 and back about .006 . So it is not an excessive back stretch as it would be with a non belted case .
My advice is dump the gauge and just size the case to just rechamber paying attention to the fact that sometimes belted cases can get tight just right infront of the belt and a special collet die may be required to size that part down .
As you know, a 7 mag headspaces on the belt, not the shoulder. Here comes the caveat. That all being said, when the manufacture make the brass, they seem to make it with VERY liberal clearance between the shoulder and the chamber. I have seen as much as .030 distance on virgin brass. What this all means is that when the case is fired, it has to stretch that much to contact the shoulder. Sometimes it actually takes two firings to make full contact. Now, the effect, is the shiny ring right above the belt, which is now the weakest part of the brass. From this point on, how you size your brass, and more importantly HOW MUCH you size it, will determine how long it will last. Sometimes, and more often then not, neck sizing is your best bet, but at some point, you will have to full length size it. When that happens, the fl die needs to be set up as a bump die, to just set the shoulder back 3/1000. I have a seven mag that nearly has 9 firings on it. Brass growth has been minimal, and pockets are still tight. Neck tension is very nice, no hardening.
All in all, it is really a combination of somewhat short brass, and liberal chambers. Proper reading, and adjusting the dies will give long life to the brass. Take care.
P/ S, this response was originally intended for the thread opener, I clicked on the wrong reply. ( Ebsarg)