Dear MRH Team, Many many
Dear MRH Team,
Many many thanks for the suggestions thus far! from the various thoughts, I haven't yet distilled down a practical "recipe" for a working solution, but there are some logical concepts being presented which are definitely worth chasing.
If I may, however, I'd like to try and "steer" the current stream-on-consciousness brainstorming, by
- looking at the suggestions
- comparing to some research I've already done on this problem,
(Yes, I have been looking around myself... )
- assessing the ideas against the "specification" given from the outset
and then seeing which ideas "show the most promise".
Don A brought an analog DC solution to the table, from Allan Gartner's excellent "Wiring for DCC" website. There are a numbber of such circuits floating around online, one of my faves being from Rick Paisley's "Misc Electronics for Model Railroads". Such circuits can easily achieve the required behaviours, and are a "known quantity".
(extra info: when the 12VDC power feed is dropped to these circuits, the reversing relay resets to an entirely _known_ _predictable_ setting. Therefore, a single "cut power" SPST toggle will allow the track power to pass _thru_ the relay unaffected, and under _direct_manual_ drive control from whatever throttle is feeding the "shuttle route").
However, as noted in the OP, the aim here is to get a DCC-based solution. Without wanting to diverge away from a reasonably sane "Problem> Analysis> Solution" thought path, I've been led to understand that Analog Control has a number of issues which will lead to it's ultimate demise. Ergo, before that "cut-off" point comes, I need to know that the most likely successor Control Option (IE DCC), can perform the tasks required of it.
Rob T introduced the idea of JMRI as an "automated throttle" system. This is entirely logical, and with the ability for JMRI to run on quite tiny/old machines, I _could_ see me building a small rackmount "JMRI box" which plugged directly into any of the show layouts in question, and "just working".
However, I see a number of potential issues, which we'll get to in a moment.
I'll note at this point that for the sake of the "Brainstorm", we're not considering $$$ as a decision criteria. Once we find "the Solution that fits the Problem", _then_ we'll scare ourselves witless working out how much $$$ we're up for/how "cost-effective" it is...
Scarpia recalled that some DCC decoders have the ability to record and playback "action sequences". These initially showed promise when I investigated the early BLI offerings. However, with no form of "positional feedback", I had the experience of a loco
- shuttling on a length of flextrack on a hobbyshop bench
- gradually "creeping" towards one end of the flextrack for every "shuttle cycle"
- and eventually derailing/shorting itself out when it "fell off the end" of the test track
Given that the layouts in question in my case are likely to use decoders from a number of different manufacturers, anything "decoder specific" such as this "sequence play" mode will be limited in it's applicability.
(Cuts accross Point #4)
I also take the point that DCC allows "speed matching" of various locos, but in prototype a geared loco simply _could_not_ match a modern passenger train for speed. Ergo, I'd prefer to have the "shuttle time" adjustable such that it is guaranteed to be longer in duration than the time required for my Longworths Class B 25t Climax #1375 to traverse the "shuttle route" at "1 sleeper/min" speed, and thus be guaranteed-compatible with any shorter "shuttle route", or train travelling at "faster speed".
Russ B made the point that for show work, there are usually a pre-prep'd roster of locos intended specifically for operation on the layout. They therefore can be "custom programmed" to operate on any "custom control system" that the given show layout may employ. This is largely true, but does not allow for the kind of "Panic" situation I described in the 6th post in this thread (which is broadly covered by Point #4 in the OP)
In extreme cases, (although I've never had the misfortune to have to go to this extreme),
if all "pre-prepared show locos" have failed for whatever reason,
(a un-diagnosed "scenery malfunction" during transportation of the layout to the show,
has resulted in all locos getting polyfibre wrapped around their valvegear, for example),
I have witnessed show-layout operators head accross the show hall to the traders,
buy a suitable loco, get it out of the box,
and "save the show" right then and there by deploying it immediately...
To get around the "no reprogramming when replacing a loco" issue, Russ suggested
I think you would need to have a dedicated fleet of "show locomotives" that are all programmed to the same decoder address, in order to change out locomotives and keep the show going. Obviously, the spare locomotives could not be kept on a "live" track, or they would all try to run at the same time.
Again, while logical, I have 2 issues with this.
(both of which are relevant to the assessment of Rod T's suggestions from earlier).
Issue 1- as Russ identified, this would require any currently-not-running locos to be lifted clear of the tracks, or parked on isolated spurs.
Assuming for a moment that the "Shuttle Route" and the "Main Route" are being fed by the same trackbuss/booster, and thus that both the "manual direct drive" throttle _and_ the JMRI "shuttle throttle" are both simultaneously sending commands to _both_ routes, this would seem unworkable. It would be impossible for
- JMRI to drive "loco #3" on the shuttle route
- while a Manual throttle tried to drive "loco #3" on the Main Route
(as both commands are passed via the common booster/trackbuss)
Seperate Boosters, Track busses, and possibly even complete Command systems?
Sure, but you then loose the ability to "drive a loco onto the shuttle route" as per Point #6
The "isolated spur" technique is possibly the simplest for currently-not-moving trains which are "staged in plain sight",
(EG I had a logging layout which had _NO_ "hidden staging", but by parking
- the loaded train at the sawmill
- the MT train at the Log Loading area
- and having a "passing loop" for a 3rd train
I could "stage 2 out of 3 in plain sight", and have the 3rd keep the show running...)
but immediately knocks out things like "onboard sounds occuring while the loco is stationary".
(not mandatory, but an awful shame to loose, bearing in mind that onboard sound is one of DCCs "party piece" advantages)
Issue 2 is more an idealogical one, but stated plainly
- if DCC's primary drawcard is "every loco addressable individually",
- and the "Main Route" (manual) operation seeks to take advantage of this behaviour
then, setting all decoders to "address #3" (for example), seems entirely counter-productive in the greater scheme of things.
Am I trying to "have my cake, and eat it too?" Absolutely!
Not to pull any punches, but if DCC wishes to be "all things to all men", then the ability to
- address each loco individually on demand (which DCC does very well)
and simultaneously
- to "shuttle any loco on a defined track/route without needing to explicitly address said loco,..."
(The fact that it is currently sitting _on_ the shuttle route is explicit enough to define the desired operation),
"...or re-compile/reprogram any "smarts" in the circuit..."
(load JMRI coding, reprogram NCE MiniPanel macros, etc)
is what I need to see.
"Chainsaw" Dave came to the party with an idea that seemed to be able to bypass this "individually addressible VS everyone-responds" issue, using the Consist functions. I had recently had simiilar thoughts, and put them to the NCE-DCC and DCC4EVERYONE YahooGroups. It appears that NCE (my preferred DCC system) has 2 methods of consisting, "Old" (NMRA Compliant) and "Advanced".
"Advanced" consist info is apparently held by _both_ the individual loco decoders _and_ the NCE command system. This means that,
- create an "Advanced Consist" on a given NCE-comtrolled layout
- lift the locos,
- and place them on any _other_ layout (NCE powered or otherwise)
and the "Consist" worth of locos will likely _not_ "just remember" that they are a "consisted set"
The "OLD" consist method however only relies on a given CV to be written into each decoder. Therefore, yes, it _would_ be possible (in theory) to set an entire layout roster worth of locos to "Consist #X", despite the locos never actually locking couplers.
The question was asked "if a loco _thinks_ it's 'part of a consist', will it still obey commands sent to it's _unique_ individual address?"
Response from both yahoogroups says "'Advanced Consist' = NO, 'OLD Consist' = YES"
Hmmm, so this is all looking good. We can have all locos "globally respond" to a given Consist address,
while still being able to be manually overridden/"driven away" on their Individual addresses...
(with the caveat that the “direct commanding” is a function specific to NCE, and is only known to work if the consist was created on a given NCE handset.
IE there is no information RE whether it will work if a Secondary Unit such as NCE-MiniPanel/JMRI was sending the “consist” commands,
while a regular NCE handset was sending the “direct control” commands).
There's only one possible “fly in the ointment”, as mentioned above.
- If _whatever_ device (JMRI, NCE MiniPanel, etc) is sending "Consist commands" thru the same booster/track buss as the "Manual drive" throttles,
and
- that single Booster/Trackbuss is feeding _both_ the Shuttle Route _and_ the "Main (manual drive) Route",
then will the loco that is currently being "manually driven",
recieve and respond to the consist commands?
SO, to wrap, we’ve got lots of great ideas,
but to mis-use a quote from “U2”,
“…I still haven’t found, what I’m looking for…”
Some side-effect info which came to light this weekend:
- using a handful of diodes and a rail-gap can create an “Asymetrical DCC” stopping block,
which is allegedly supported by all NMRA-compliant DCC decoders
Has anyone had experience with this form of DCC train control technique?
- There are a few companies that produce units which sit _between_ the booster and the track,
which do “station stop” and other localized control tricks,
apparently independently of the specific loco/address in question.
(BitSwitch is one such company http://www.dccbitswitch.co.uk/timedstopbitswitch.htm )
Unfortunately, they don’t do the equivalent of the “dumb timer/reverser” mentioned by Don A,
but a combination of such a unit + the above “asymmetrical DCC stopping system” would appear to be a “perfect fit” solution.
- There are also more advanced units available which are DCC system specific. Apart from obliging the use of a given DCC system,
it would seem that they are “single use equivalents” of a JMRI rig,
(IE not greatly different, just replaces the PC you first thought of).
http://www.cmlelectronics.co.uk/index.php?page=shop.product_details&flypage=flypage.tpl&product_id=20&category_id=7&option=com_virtuemart&Itemid=106
Hmmm, lots to consider,
but I don't think we've determined a “working solution” quite yet…
Love to hear any further thoughts from the Collective Wisdom…
Happy Modelling,
Aim to Improve,
Prof Klyzlr