Hamish has a posting that mentions how much easier DME arcs are with an RMI display.
Sometimes I feel lucky that I was never taught the official way to a DME arc during my instrument training. We have one nearby, but it just never came up. I flew my first DME arc alone in the plane in actual IMC, and because I was never taught to be stressed about it, it seemed like a simple maneuver. Here’s all I do in the arc itself (assuming that the DME and VOR are already tuned and identified):
- turn perpendicular to the DME source (so that it is off one of my wingtips)
- fly my heading until the DME hits about .2 to .3 miles more than the DME arc distance;
- turn 5 or 10 degrees towards the DME source, and repeat (to allow for winds, turn more if the DME doesn’t start decreasing; less if it decreases too much; don’t let it get less that .2 to .3 miles less than the DME arc distance).
I have one VOR set up as a fence to tell me when to turn inbound, of course. Sometimes I twist the other one to see what radial I’m on, just to relieve boredom, but it’s really not a necessary part of the procedure. I guess that I use the DME as a kind of a digital CDI, and I’ve been happy with the results so far. With my technique, an RMI wouldn’t make much difference, but I can see how it would help if you wanted to track your radial all the way through.
Land and Hold Short 
The DME arc should be (and often is) fairly simple when done properly, but in the US we have to demonstrate one during the checkride using the “approved” method, which with the old OBS and DME means a bunch of error-prone knob twisting. It’s conceptually so simple, but with the requirement that you track your radials all the way around, it’s the sort of thing that goes to pieces as the examiner starts distracting you (even though you’re actually tracking around it fairly well). For me the RMI version just makes it easier to visualise things and keep the track accurate because of the relationship of the wing and the needle… and no, I’ve never done an approach-related DME arc that wasn’t for training purposes. Not yet, anyway, despite there being one just south of here (KWVI). I have to say I rather enjoy doing them, despite the impression my postings probably give…
I read somebody describe your method before I was taught the “proper” way, and having done both of them in the sim (but never in real life), your way is a lot easier. There is only one DME arc on an approach within 50nm of here, and ATC never lets you fly it because it gets in the way of their vectoring, so I’m hoping to never have to do one for real.
Paul — why not try inventing your own DME arc somewhere out of the way. Pick a VOR (any VOR), choose two radials, then fly the arc between them before continuing inbound. That’s really all there is to it.
I found it very annoying a while back when Toronto Centre vectored me around the published DME arc for North Bay (with constant 10 degree course changes) instead of simply assigning it to me. The frequency was busy enough, though, that I didn’t want to get in a long discussion.