Caution: Because the power is actually closer to AC, many DC motors heat up much more quickly than they ordinarily would on an
analog power source, and some motor types can be seriously damaged with only a brief encounter with DCC track. Many motors will
buzz and hum when presented with this type of power. Don't leave locomotives that are not equipped with a decoder on the track if possible,
to reduce the chance of heat damage to the motor. See note below about coreless motors.
Note: All direct current locomotives will respond to the signals created by Zero Stretching. Which may have unintended results.
As locomotive speed increases, more bandwidth will be demanded of address 00, which can have an impact on response times when
more than 5 DCC equipped locomotives are also in operation. This technique is a bandwidth hog due to the need to constantly send
packets addressed to 00.
Great DCC info on this https://sites.google.com/site/markgurries/home
Fri, 2016-02-12 08:20 — Prof_Klyzlr
As with most technical things, there's
- the Theory (the underlying premise/design-logic, which is worth knowing for diagnostic purposes)
- the Exceptions (the individual situations/specific-combination-of-conditions which may happen to work "onsie or twosie" style, which may give the impression that something is possible,
thus reading to a very slippery slope and widespread pain/anguish...)
- and the Simple Rule (IE that covers the vast majority of practical situations/deployments,
and presents a "fail-safe" approach if/when things don't work the way a previous-experience "Exception" example may have lead the modeller to believe to be "...the way things always work...")
In this case:
- The Theory is that the DCC specification has room within the spec tolerances for a technique known as "zero bit stretching". This is a way of messing with the data packet timing, such that over an given time period, the digital signal sum-result "average" appears to have a "DC Polarity" and "Voltage Potential". (Kind of like a very brute-force square-wave "PWM" analog throttle.).
With this "out of balance" DCC signal on the rails, an analog loco
(motor literally wired direct to rails,
NOT a "Decoder acting in analog mode", that's different!)
will commonly buzz like a mongrel, likely get very hot,
but _should_ move in a reasonably predictable "Forward is forward, speed is variable" way.
- The Exceptions is that very few DCC systems actually implement "zero bit stretching" at all, let alone "properly" (where "Properly" is defined as "with the best interests and smoothest, most controllable, and reliable operation of the Analog loco in mind").
Furthur, even if they do, the high peak voltage at the rail, slamming back and forth in potential polarity, thru the electromagnets which are the motor coils, is a recipe for motor-killing heat.
Furthur, some modellers have identified the "shake" or "judder" of an analog motor rocking back/forth very very fast as a cause of additional mechanical driveline strain.
Even if "zero bit stretching" control is not currently enabled/active/being used,
(IE any/all analog locos sitting on the track are effectively seeing an averaged "zero volts",
because the sum of the DCC signal over time is equally +ve going and -ve going per-sec)
any analog locos sitting on powered rails are still having their motor coils subjected to the aforementioned "rapid polarity-changing voltage", with associated magnetic-field create/collapse Back EMF heating a unavoidable physical result.
- Ergo, the "Simple Rule" to avoid damage to the motors and drivetrains of _Analog_ locos is,
it is very Unwise to place, leave, or attempt to control them for any length of time on a "DCC powered" layout.
(SOME DCC systems may support and allow ZBS control,
SOME analog loco motors and drivetrains may be capable of handling the electrical abuse subjecting them to a DCC signal is in real terms,
but the "Simple Rule" in this case safely covers ALL systems and ALL locos,
with No Risk of damage to anything).
As a wider-application statement, DCC is a Digital Command Control system, It is predicated at both end of the rail-transmission-path on the idea that:
- there is a Circuit (Command-station/booster) transmitting a Digital Signal,
- and a matching circuit (loco receiver) picking up that Digital signal and acting on it,
while also converting the relatively high signal voltage into usable controlable "traction power".
In both cases, the circuit and associated software/firmware design would be sooo much easier for the DCC manufacturers if they did not have-to support:
- Command stations/boosters which had to put out a "pseudo analog control" voltage of controllable average DC voltage and polarity (That many DCC systems do NOT do "ZBS" is a clear and present hint).
- Decoders which had-to-be-able-to recognize when sitting on some form of "Analog Control",
(flatline filtered DC, PWM, etc) and react as-if they were "normal analog locos".
It's worth keeping in mind. Just because one can "bend the basis of DCC",
does not mean it's a good, reccomended, or reccomend-able approach for long-term operation and reliability...
Aim to Improve,
PS in the same "Simple Rule" category:
- Wire Turnout Frogs : even with a simple microswitch, the time spent to wire a frog will repay for ALL locos for the future of the layout, and that goes for both Analog and DCC controlled layouts.
- Always use a resistor when wiring Decoder Function Outputs : If there is _any_ doubt, a 1000ohm series-connected resistor will keep LEDs, low-volt bulbs, and the decoder itself safe from over-volt/over-current conditions. At worst, the LED will only light up dimly, but it will not be _damaged_,
and the modeler is on a "safe ground" confirmed starting platform from which to carefully reduce the resistance to achieve the desired brightness.
You may not need to reset because of ZBS, but I find that you need to reset your Digitrax system from time to time to clean out any "junk" the system has pickup (locos not dispatched, consist no longer in use, ...). I use my system on our show layout, so I reset it after each show.