But, I found that it helped me to make sure the secondary mirror edge and the primary reflection in the secondary were concentric with the focuser tube.
That's because Cheshires can also be used as collimation caps.
But these are less accurate for placing the secondary with respect to the focuser axis and for tilting the secondary correctly than a specific tool for it -- usually you see quite a bit of the opposite end of the tube around the secondary, and so it's not that easy to see whether the edge of the secondary and the edge of the reflection of the primary are exactly centred.
You can, however, already see gross errors in the tilt of the secondary if you rack the focuser in and out. If the Cheshire ring reflection and the image of the primary centre spot stay concentric over the entire travel of the focuser, then the focuser axis is indeed pointing close to the middle of the primary. If they seem to move with respect to one another, then it isn't, and then you need to do the Cheshire reading with the focal plane halfway between the pupil of the tool and the physical Cheshire ring location in the tool for the tool to still work well.
You have a "passive" tool for setting the tilt of the secondary, a sight tube with cross-hairs, and an "active" one, the laser collimator. With a sight tube you set the tilt of the secondary so that the primary's centre spot appears behind the cross hairs. The tricky part is that focusing the eye on both the cross hairs and the centre spot is hard (the cross hairs are close and the centre spot roughly at one focal length).
Laser collimators don't suffer from this and are more precise when they're good (i.e. sit well aligned with the focuser with the laser parallel to the focuser axis), but cheap ones usually aren't -- in my experience the first good one is the FarPoint laser, even though it often needs a really good 2"-1.25" adapter if you have a 1.25" laser collimator and a 2" focuser. Mind you, there are expensive ones that I don't consider reliable (the HoTech and the the Baader come to mind).
Once you've set secondary tilt with a sight tube or a laser collimator and set primary tilt using the Cheshire, you can also evaluate the placement of the secondary by using the Cheshire with the focuser racked out enough so that the image of the primary fills the secondary completely; if the secondary is liberally sized you might need extensions to place the Cheshire's pupil far enough. From that vantage point the image of the primary and the edge of the secondary should be concentric. If they are not completely concentric (i.e. when you rack out the focuser the primary is clipped first at one spot on the edge of the secondary) then the secondary's not ideally placed, but small errors are not that relevant (they just shift the fully illuminated field at low power a bit).
Once the secondary is tilted correctly (by using a sight tube with crosshairs