Beware, there be dragons...

Dear Joe,

OK, I see where you're going with this, but let's drop a bit of reality for a sec...

- At the basic "layout wiring" level, sure, a properly wired layout can handle being powered by either Analog or DCC

- Quite a number of modellers set up their primary "layout track feed" on a connector of some description,
and just "plug in whichever system (analog or digital) they want to run (today)"
(IE Physically "just swapping between systems" is easy to do...)

However...

- RE Locos:
          - A pure analog-wired loco may take damage if left to buzz/cogitate/overheat on DCC

          - A DCC-decoder-equipped loco should handle being controlled on Analog, but not all do
            (and the "controllability experience" typically suffers, before we even consider Consist/MU-ing locos
            together)

          - and DIS-abling "Run on Analog" mode within the loco decoder is S.O.P. for most DCC users,
            as leaving it ENabled can leave some Decoder+Control-system combos open to "unexplained runaway"
            behaviours.

- RE Layout Wiring
          - Many modellers use active-switching units, such as Tam Valley Depot "Frog-juicers",
            to power turnout frogs, handle reversing loops, turntables, and other "difficult" wiring-polarity situations.
            Such systems will NOT work when powered on Analog, and indeed may take damage.
            (These situations can be handled by using Microswitches and similar Passive switch alternatives,
            although such options seemingly flummox a great many modellers?!?!?)

          - Ditto for any "DCC Accessory Decoders" which may be wired to the Track Buss,
            such as Turnout decoders, DCC Circuit breakers, etc

          - Oh, and if you're running a hardwired Throttle Buss, designing and deploying one which is "cross-
            compatible" with both your < insert given DCC system HERE> and (presumably tethered walkaround)
            Analog throttles is no small feat...
            (Sure, DCC might allow WiFi throttles, but we have to consider what exactly will be your "throttle" system
            under Analog, and wire appropriately...)

Guess what I'm saying is, if you design and build the entire layout from the outset with wiring configured for lowest common denominator (Analog) power,

AND

you can corral/isolate/swap out any "not the right format loco for the control-system in-use" locos with extreme predjudice and diligence,

then it might be worth going "dual-mode"...

...Otherwise, the Better Advice is to just pick One system or Other,
Commit, and wire all downstream components (inc locos) in consequence...
(Hint: DCC is not going anywhere, and is the option that keeps your options OPEN going-forward)

BTW, there are quite a number of threads on exactly this "DCC or Analog? What about 'Dual-mode'?" topic,
the Search Box at top right of this page should provide enough prior-discussion to cover most any "but what about...??" you might be considering...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

PS as you are considering a Digitraxx system, it's worth mentioning that Digi has a "Address Zero" mode which kinda allows the Digi system to appear to control a "pure-analog-wired" loco in-parallel with your DCC locos. It does this by abusing the DCC signal timing specs in a method called "zero bit stretching". In short, it deliberately creates an out-of-balance (Asymmetrical) DCC signal which appears to a "wired straight accoss the rails" motor as "really spiky, bumpy, nasty DC". The trade-off is that this "pseudo Analog DC" voltage is actually quite harmful to many types of motor, commonly causing overheating, mechanical-stress in the drivelines, and in extreme cases, pre-mature mech-death... 

One or the other

Over the years, half my locomotives are DC and the other half DCC.  I have a small simple layout.  I plug the DC controller in and run my DC locos (one at a time, naturally).  Or I plug my DCC controller in and run my DCC locos.

I run both DC or DCC

I built my layout using DC and blocks, toggle switches.

Now I run DCC, all blocks are ON, throttle controls which engine to run. Other engines sit and make noise(the ones with sound boards). Most turnouts are manual, some are electric and on a standalone power supply.

I still run DC locos on occasion. A toggle switch disconnects the DCC and a second toggle connects my analog power pack.  All DCC locos get parked in dead blocks.

isolate all other locos

couldn't you just run a DC loco around a continuous loop of track without any extra electronics?

couldn't all other DC and DCC locos be put on (switched) unpowered track while doing so?

i built a PWM throttle that plugs into an NCE PowerCab plug for testing DC locos.

Are there DCC breakers or Frog Juicers involved?

Dear Gary, Dennis, Greg,

In each of your cases, are there any DCC circuit breakers, Frog Juicers, slave DCC boosters, or Accessory (turnout) decoders wired to the track buss?
(IE downstream of the "unplug the DCC system, plug in the analog throttle" connection-point?)

If Yes, I'd be interested to hear how these "common DCC wiring" devices handle having their "diet of DCC" taken away, and being force-fed "raw analog" power?

Happy Modelling,
Aiming to Understand,
Prof Klyzlr

The first train is multilingual

Thanks everyone.

The loco I bought works on analog and DCC per the manual (also has capacitor so no frog juicer should be needed), and there will be a dead storage spur when I get the next vehicle (track maintenance =D  ).

The track will run around 3 sides of a room, point-to-point with turnouts at either end, effectively like the "Trolley/mining circuit" Circuitron has a layout for (#367.) The automatic polarity reversing is what makes analog automation so appealing.

The automation would be optical sensors controlling two turnout switches and an automatic polarity changer. The polarity reverser sits electrically between the track and the analog power source, so if I disconnect DC at the track it’s out of the circuit, and if I use block voltage drops I can bypass them. I just have to think about the switch actuator signals…

Mine

is just a simple double loop with two sidings, no reverse loops, so everything is wired to one lead wire with a quick connect so I can swap between DC and DCC just by disconnecting the currently connected system and connecting the other.  About 5 seconds of work to switch.

run both on the same layout ? (DC and DCC)

Yes, there are all kinds of ways to do it but in practice I think most folks end up running just DCC and put new sound decoders in their favorite DC engines. If one had a large collection of older brass and wanted to be able to run it they might find it worthwhile to add DC but in most cases newer engines are preferred so DCC is the easy default way to go. Keep in mind there's  only so many hours in the day and modeling time can be hard to find so simple control system means more time for other things like building a layout or rolling stock or running operations....DaveB

The prof is as always on top

The prof is as always on top of it. My club is building one of our layouts to handle DC and dcc. There is a lot of planning going into how to wire it according to accommodating the circuit breakers in particular.

For a small layout with simple dcc needs, it is just a matter of a toggle switch. For a bigger layout it is much more complicated, but certainly doable. 

You can do it

I'm in the process of joining a club that has both DC and DCC available, which they had to do in order to keep the peace when they installed DCC, and some members didn't want to be forced to convert large loco collections. In DC mode control was/is done by putting plugs in sockets around the layout fascia (no automation!) and they just wired up DCC to replace one of the cabs. For a start, this made the DCC system try to function on a layout with common-rail wiring, which you're not supposed to do, and apparently there were initial problems, but they sorted those out. Now they say the DC throttles are getting less and less use, and as faults develop, they're just disconnecting them one by one, and as new track is installed, a lot of it is DCC only. You could say it's a slow gentle transition. It seems as if the two systems have managed to coexist for a good few years, though.

Multi-lingual?

Dear Joe,

In order of appearance:

As per most RTR DCC-equipped offerings...
(backwards compatibility is a thing, analog still lives! )

Hmm, unless it's a current-spec ESU V5 Direct with integrated KA, there's something about this which does not ring right... Which specific Make/Model of loco model are we talking?

...which makes sense, I totally get it. Indeed, waaaaaaaay back when (circa 2011, LOL),
I was thoroughly convinced that DCC couldn't achieve the kind of simple, compatible, bulletproof "train shuttle" operation of a "Dumb Timer" unit + single diode-stopping section...

https://model-railroad-hobbyist.com/node/5122

Since then, we've had "standalone" options come (and unfortunately go) from Tam Valley Depot, Iowa Scaled, and a number of other manufs.

(Search box, top right of this page, search term "Train shuttle"...

...or, for the most-recent and thorough discussion on the subject, check this thread
https://model-railroad-hobbyist.com/node/40240)

We also have various options from the DCC system manufs which integrate into their systems,
IE inject the "commands" into the command-system, rather than try to intercept/modify the track-based commands before they get to the loco. IMHO, NCE's "MiniPanel" is one of the better-integrated examples. Combined with ABC-capable decoders and appropriate diode-arrays, the resulting solution is almost the equivalent of the "analog dumb timer", but not quite. (Still requires explicit programming to address the loco currently on the shuttle route).

Logical, but not necessarily the only (or best) game in town, esp in terms of "re/configuration on the fly".
In many cases, eliminating excess complexity and circuitry such as the need for "active end-stop detection" can be beneficial. There is a temptation to "deterministically detect, control, and sequence every element", but tightly following this only seems to create "holes that one can drive a train thru"...
(EG locos can only ever face a given direction,
Loco X must end up at Track position n or the sequence fails, etc etc).

This is why I'm a big fan of "catch-all" systems like Asymmetrical Braking,
easily deployed with simple cheap passive components, 
that work first-time, every-time, all-the-time, whether the train is being driven automatically or manually...

It also blurs the line between:
- "Full manual drive" (literally nothing done automatically)
- "Automation" (which for the US modeller, typically implies "human does Nothing, system runs itself")

This hard binary between the 2 states does Not Have To be This Way, and tasteful deployment of "operational assist" automation gear is the headspace which allows that to happen...

It's actually simpler than you are thinking. Given that such Analog shuttle systems use an electro-mechanical relay as a DPDT reversing switch, when UNpowered, they drop to a detreministic consistent IN --> OUT wiring state, aka they work as a "straight-thru piece of wire". Ergo, a simple toggle switch on the "DC-power IN" to the shuttle circuit will act as an effective, simple, single "shuttle bypass/disable" switch. 

As an example, check the homebruise analog tethered throttle in the YT. It is a simple single-transistor Throttle circuit, with a "dumb timer" circuit (a la Rob Paisley http://www.circuitous.ca/AutoRevCheap.html)

The RED toggle switch located on the TOP of the case is the Shuttle Enable/Disable switch.
- Pull towards the User = Manual drive (shuttle disabled)
- Push away from the User = Shuttle enabled (Orange LED = shuttle relay state)

As a slightly-more complex example, still using Analog Traction control and basic "dumb timer" reverser + diode-stopping, check out the "Bindle Mine"...

If by "switch", you mean "turnout", again, this is where either:
- a MiniPanel solution (DCC)
- an Arduino solution (DCC or Analog, albeit needs more User R&D to deploy)
- 555-based "JK-flop" Tortoise-driver circuits
( http://www.circuitous.ca/556Stall08.html
/> http://www.circuitous.ca/3556StallAltern.html
/> http://www.circuitous.ca/ReverseLoopController.html )

kick into the discussion...

I hope this helps...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

Capacitor, Diode, switches, and DCC

Hi Prof,

Thank you. I would love to talk about this! I will do it in chunks as I've got to head home in a bit. 

First, the capacitor: from the manual (MTH R17 Subway Train):

"The super capacitors provide backup power to the electronics, allowing the locomotive to better negotiate track sections suffering from poor electrical conductivity. This ensures that sounds continue to play, the locomotive continues to move forward and lighting remains lit. During power shutdowns, the super capacitors provide power to allow the locomotive shutdown sounds to continue playing through their conclusion."

Maybe I'm reading that wrong. I have no idea what a ESU V5 Direct with integrated KA is.

Next, the diodes:

I agree that optical sensors are not fool-proof, and I'll incorporate a dead zone at the end of each leg in case of failure. But my reasoning for using them as two fold: a) the circuit is advertised that way from Circuitron and the sensors come in their kit components, andb) I know that photo sensors won't affect DCC operation since they're not in the circuit. Maybe diodes are OK, I just haven't thought much about it and haven't personally used them enough to understand them yet. I was going to look into current sensors, but I think the photo sensors will work short-term and give me some quick and easy satisfaction and motivation to keep progressing. I have zero trains or experience at the moment, so getting something up and running fairly quickly would be a good win. So using Circuitron's equipment and layout seems like a decent choice. But that’s not to say I’m not open to the diode idea…

Re switches:

The way I understand it, the turnouts will have solenoid actuators, and the actuators need a signal from something - that's what I was calling the switch - which lies between the sensor and the solenoid. Anyway I was just thinking that if the 'switch' stays physically connected, then any DCC control signal to the solenoid can make its way back to the switch's internal circuit. It's probably insignificant and maybe even open-ended inside the switch, but I felt that it would just be better to play it safe and avoid a digital signal running inside an analog component. I guess I was thinking the same for the automatic reverser: if I only have one track connection, unplugging the analog at the track takes the reverser out of the circuit, and insures I never mistakenly have analog and digital running into the tracks at the same time.

Then, DCC Programming:

I was going down a rabbit hole of searching out DCC components I would need/want to make the automation work. Granted I don't know a ton about what's out there, but it seemed like it was getting pretty hardware intensive. Then I realized, with the Zephyr, I could probably build a Raspberry Pi, load JMRI on it, program it, and run it with the Zephyr and some position sensing hardware. My assumption is that JMRI can make route/track adjustments based on sensor input.

Prof...

Exactly what Dave said. 

"Mine is just a simple double loop with two sidings, no reverse loops, so everything is wired to one lead wire with a quick connect so I can swap between DC and DCC just by disconnecting the currently connected system and connecting the other.  About 5 seconds of work to switch."

I have six turnouts and a bit more track.  But no fancy electrics.  The turnouts are switches like those from John Saxon. And that's it.

I was thinking more about the

I was thinking more about the track connection. With Prof's suggestion, there could be two 'hot' and one common connection at the track, with the hots running through a DPST switch. Then nothing needs to be disconnected, and both systems electrically isolated. 

My thought (not saying better or worse) was to use a multi-pin connector, disconnecting/connecting whichever system I felt like using. The advantage to that is that I was thinking I would include wiring for the switches, and whatever else needs to be changed over, into the multi-pin, even if there are just jumpers or empty spaces. Of course I could also use an xPxT switch, assuming there aren't too many other things that need to be changed over. Flipping a switch does seem a bit more 'polished' than a plug, but that's subjective and there are advantages to each I guess.

Once I actually start building layouts things will become much more clear, maybe I'll have a preference.

Video using a DPDT Switch

Here is a YouTube video using a switch to go from DC to DCC.