Speed Pot + DIR switch, accept no substitutes
Dear MRHers,
Both professionally and in modelling, I've used both Pots and Encoders,
and from a "positive deterministic" user-application standpoint, I'll take a Pot every time...
That said:
Quote:
We have some ideas on how to handle the potentiometer when you switch between trains. The train won't suddenly change speeds. We're thinking that you'll need to match the pot position to the current speed of the loco before the pot will become active again.
Echoes of some pro audio consoles I've used (and been involved with the design/manuf of) previously. Commonly such "current pot position VS switched-to SW value mismatch" was "handled" by the pot
(or "encoder with hard end stops" a la Fairlight Constellation console, have seen both used in such applications)
having to be swept _down_below_ the "current software value",
(which, of course, the User has no idea-of in-the-heat-of-the-moment)
and then raised back UP to "catch" the current value and "lock-on"...
(Read: the User is reduced to hitting-the-Minimum-stop and whipping back-up as-fast-as-possible out of sheer blind "regain control NOW" panic...)
...of course, if the Pot A> D stage (or encoder IR stage) wasn't fast enough to keep-up with the "whip" movement, and/or "skipped some values",
(bad/"grainy" pot track, bad A> D stage, low A> D samplerate, bandwidth packet-drops, etc etc)
it could be entirely possible to have-to do the "sweep Down, and back Up" move a few times,
until the incoming stream of Pot values were "caught" and matched as-expected...
(bad in the heat of a Hollywood-level Post Production session,
arguably worse as that $$$$ brass loco heads towards the floor...)
I should also note,
"...one brain, one throttle, one loco/consist at a time..."
words to live buy...
(The UT4/Cab04e switchable "dual cab" throttles always struck me as a
"...we do, because we can, because we're using an encoder..." bonus-feature,
rather than based on any proper-defined "One brain, divided-between two-part-time throttles in one box" use-case...)
Quote:
The potentiometer we chose also has a detent at the half-way position, and we plan to use that for switching mode. Counter clockwise from there will increase the speed in reverse, and clockwise from there will increase the speed forward.
Sorry, bit that's a Hard Pass for me...
0<> Max full/smooth-rotation (NO detent!) with a direction switch/button is the only mode I want or need to see.
Doing a "centre-off" reduces the 0<> Max speed resolution for each direction (IE Deg per SpeedStep),
and introduces "whoops, rotated the knob too-far, now going backwards" errors which simply shouldn't be a designed-in issue. I mean, we left "centre-off" controllers with the H&M Duette of the 1970s...
...having a "centre-off pot" also introduces "pot calibration" issues where the "halves" may not track equally,
with resulting value/speed mismatches in each direction...
(ask anyone who's engineered a L<> R "Pan Pot" control for audio equipment,
it's a complete product torpedo if "100% Left" on the Pan pot =/= actual 100% signal at Left Output Only,
and vice versa...)
...oh, and "Left to go left, Right to go right, Centre off" is as eye/hand/output-device irrelevant as "east/west" is on a DCC system...
(IE where the loco-direction and track-orientation direction are entirely disconnected and independent,
and further seperate from the relative position of the Operator to the current Loco Position).
Quote:
We haven't implemented either of these yet, so need to see how they feel, but we're pretty sure we can make something work for both of these.
I was involved in the dev of the Fairlight (and subsequent BlackMagic Design) "Pyxis" controllers, with electro-mechanical "Jog/Shuttle" knobs. Much like the old Sony BetaCam broadcast VTRs of yore, this knob was a free-rotate encoder under "jog" mode, but a single "punch" would trigger it into centre-off spring-return "shuttle" mode. Sony spent much time/effort/$$$$$$ to get theirs to be the "gold standard" in environments and applications with a lot more on the line than Model RRing, and the FL/BMD experience proved to me why this was the case...
...in short, don't try to 1/2-a** both modes, focus on doing one of the modes really-well, unless you have a Sony sized team and budget to play with...
Happy Modelling,
Aim to Improve,
Prof Klyzlr