The speed steps are not linear in terms of the increased voltage to the motor, with bigger jumps between the later steps. The exact difference in voltage from step-to-step probably depends upon the specific implementation (i.e., AC-motored Hudson vs DC-motored diesel), and even the stall speed. A feature I personally despise that makes your quest even more complicated... the CAB-1 has an "acceleration" feature built into the red knob. If you turn it faster, it may actually command a multi-step increase. If you turn it slowly, it may transmit commands irregularly, or not at all. In my experience, it's not very precise in the sense that x degrees of rotation corresponds to one speed step.
[Like Legacy / CAB-2, the original TMCC had the capability to go directly to a specific speed step. It was never implemented, except in a couple of computer interfaces offered for a short time by third parties. I had one of these, and had difficulty getting it to work; it's a shame Lionel didn't support this feature better during the TMCC era.]
Remember, TMCC is a one-way communications system with no feedback between the hand-held and the base (or the locomotive, for that matter.) If you send a command, there's no way to be sure it was "heard" EXCEPT for that blinking red light. The device tells you a command was received, but not WHAT command was received. The CAB-1 will not keep trying, so if you don't see a reaction, your only option is to send it again by continuing to turn the knob. When you combine the variability of the knob with the possibility that some commands may be "lost" between the CAB-1 and the base, you have a paragon of vagueness.
You can probably tell by now that I never liked the original TMCC. Remote? Yes. Control?? Not so much. With detents in the knob's rotation and an LCD display, Legacy is what the original TMCC should have been -- and still precise and robust enough for most real-world layout environments. Now if they could only address those pesky ground plane issues!
