I have been very satisfied with these:
Brownells has 236 SINCLAIR INTERNATIONAL products for Sale Up To 91% Off including Reloading, Tools & Cleaning, and Gear.
www.sinclairintl.com
Just remember that when you squeeze a case neck down from the outside with a bushing, the brass spring-back will cause it to open back up around .0005" or so when it is removed from the die. When you open a case neck up from the inside with a mandrel, the spring-back works in the opposite direction, and it will cause the neck to close up around .0005" or so when the mandrel is withdrawn. If using the mandrel as the
final sizing step, I like to re-size the cases first with a bushing that is approximately .001" to .002
smaller than what I would use if the bushing die were going to be the only re-sizing step. That way, the mandrel will actually do some work and open up all the necks equally. If the necks as sized by the bushing die are very close in diameter to what the mandrel will ultimately open them up to, I feel like you get a little more variability in the final product. Due to the brass spring-back closing back up the necks about half a thousandth when using a mandrel, you want a mandrel that is ~.0015" under bullet diameter if it is to be used as the final sizing step.
For example, I would normally use a 0.336" diameter bushing as the
sole sizing step for Lapua .308 Win Palma brass, which gives very close to .002" neck tension (interference fit). In contrast, when I use a bushing die
followed by a mandrel sizing step, I would go with a .335" or .334" diameter bushing first, then finish them off using a 0.3065" diameter mandrel as the final sizing step. It can certainly be argued that using a two-step process such as this works the brass more than a single pass with a bushing die. However, I anneal brass every firing and I have never noticed any issues whatsoever when moving the brass that extra little bit using a slightly smaller bushing followed by the mandrel. In my hands, it's not a limiting source of error. I can imagine it might be more of a problem if you first used a bushing that was .003" or .004" (or more) tighter than the mandrel would open the necks up. By going with a bushing that is only .001" or .002" smaller, the mandrel will open up every neck to a small degree, which I feel improves consistency in its use, but it shouldn't be grossly over-working the brass, especially if you anneal every firing.