DCS with Automated Blocks

To clarify, if it's hands-off, what types of command control functions would you need (e.g., manually changing speed/direction, triggering whistle/bell)?  Another option is to run conventional control with engines in direction-lock so that when track voltage is removed and restored, the engine automatically resumes motion.  And since track voltage controls speed in conventional, you'd get more realistic collision avoidance and station stop approaches/departures.

How many PS2/3 engines are involved?  And if a modest number, would you be willing to modify the engines?  For example, just the power to each engine's motor(s) could be cut-off rather than the overall power to the engine - this would keep the engine alive in command mode so you don't have the problems associated with removing or lowering track voltage to slow/stall a command control engine.

Thanks Stan.  The big function I'm looking for is to stop and than restart in the same direction.   I have a number of engines but may a dozen PS2/3 but as a trainee I know there is always more.  Your idea of just cutting the power to the motors is interesting.  How is that done ?

There are (at least) two basic methods cut or reduce power to the motor while maintaining a sufficient track voltage to keep the engine's DCS electronics alive.  However, I am not aware of off-the-shelf modules that implement either of these so these would require substantial DIY handi-work and persistence.

One method assumes you lower the track voltage in blocks where you want to stop the engine.  Obviously the voltage must still be above that which keeps the DCS electronics alive.  In this case you insert components to reduce the voltage to the motor such as diodes.  So even if the track voltage is, say, 8V the diodes would lower it to, say, 2V or some level where the motor would stall without drawing obscene levels of current.  Then, when full track voltage is restored (to restart motion), with 18V on the track the same 6V diode drop would allow sufficient voltage at the motor to run engines across the desired speed range.  The diode approach might be, say, $3 in parts per engine.  A more refined version of this method would be to sense the track voltage and when it is less than, say, 14V then simply cut power to the motor.  This would take more components than just diodes but still be in the, say, $5 in components per engine.

Another method does not require adjusting the block voltage.  Instead you can have constant DCS level voltage on all blocks but instead generate a "stop" signal sensed by each equipped engine.  It could be electrical such as an audio frequency tone added to the track voltage on the desired "stop" block.  It could be an optical signal such as having the engine detect the red-light in an LED signal control head on the track ahead.  This method would cost more per engine, say, $10 but could provide more interesting capabilities.

In case it's not obvious, messing with electronics is a big part of the hobby for me.  So this is just how I would approach the application.  In any event the key point is, to my knowledge, the insertion of a widget to cut motor voltage from the PS2/3 electronics is a viable method of stopping an engine while keeping the DCS command electronics alive.

Thanks for the details Stan.  I'm favoring voltage reduction With diodes but in addition to the motor stall current issue, there must be a problem if the wheels bridge the block gap and connect 20 vac to let's say 5 vac.

On the surface it seems that auto blocks should not be too difficult to incorporate in DCS. The hard part is position and associating position with a specific engine.  An accessory which could be placed anywhere could sense position and the operator could manually link an engine to the position and/or route.  Auto engine ID could come later.

One issue with cutting the motor power that I can see is the electronics still thinks that the engine should be moving.  So, when it gets power back on the motor, it's going to try to "catch up" and leap out like a jackrabbit.  You can see this if you just momentarily lose power over a switch, after the locomotive coasts over the interruption, you get a lurch as it catches up.  I would guess an extended time where the tack wasn't feeding back would result in full voltage on the motor when power was restored with the resultant high speed departure.

Thanks gunrunnerjohn, jack rabbit starts seem likely when restoring full power to a cut motor.  You must have a similar situation if you return to full power from a stall condition ?  Full voltage could be phased in for a smooth start.

Any recommendations would be appreciated.  I'm beginning to believe the only practical solution to auto stop/starts with DCS other than the trolley function is to run in conventional mode.

Some of the new control systems just coming out may allow for computer control of command locomotives based on feedback from track position sensors.  Personally, I think that's going to be the way to solve the problem in the best manner.

Thanks gunrunnerjohn.  What are these new control systems.   I'm only aware of the Dave Hikel system.

If I understand what you mean by wheel bridge, I think you're OK if you use something like the following to reduce the voltage to the stop block.  Relay closes to short diode string applying full voltage; relay opens to insert diode string applying reduced voltage. The relay is controlled by a timer or sequencing circuit. 

Rollers will momentarily short the gap but once the engine is completely in the stop block, the engine will only "see" the reduced voltage.  Of course you have to mind lighted passenger cars or cabeese...or at least strategically locate them in the consist.  If your lower stop-block voltage comes from an electrically independent source, then I agree there's an issue with two power supplies fighting each other.

GRJ is correct.  Under normal DCS operation, when a stopped engine gets a command to go 40 MPH (or whatever), it ramps up to that speed based on the DCS-settable acceleration constant.  In the starved voltage situation, it will instantly try to go 40 MPH when motor voltage is made available.  The situation is somewhat improved if you put diodes in the engine as that reduces the available motor voltage.  So there would be less jack-rabbiting then what you see in typical DCS operation where the motor has full track voltage when coming out of a stall.

And to your point, yes, at the expense of more complicated electronics you can phase the DC voltage to the motor over several seconds to demote the lurching.

Since you are only running PS2/3, another idea is to ponder if running DC track voltage can help in some way. 

Member Dale H has described using relays in adjacent blocks to ramp up and down conventional engine speed by phasing in different taps of an AC diode-dropping string.  It's pretty difficult to do AC phasing smoothly as this generally means a phase-controlled triac circuit which gets messy for DIY.

Smoothly phasing DC voltage, even at several Amps, is much easier than AC using FETs and related DC control circuitry.  May seem odd running DC on 3-rail but then again this is a discussion forum!

No doubt a computer-controlled, sensor-based, system is the holy grail...especially if starting from scratch.  But with a stable of PS2/3 engines, DCS command control, and a layout/track apparently incompatible with LCS, it does make quite the conundrum.

Stan, thanks for your inputs.  Dropping diodes across the gap are great to provide reduced voltage to the block.  Taps on the diode string could provide phased voltage in subsequent blocks to avoid jack rabbit starts.

FYI I do run conventional and TMCC engines on the layout via DCS, but that just complicates the discussion.  I haven't upgraded to Legacy yet. 

I really want an integrated DCS solution but this does not exist but may in the future.  Most of this can be done in software once you have the location information on the bus.  I not aware of any indication from MTH relative to this future capability.