Dcc reverse loop module

[Shroomy]

Which is the best dcc loop module to use please

[Nigelcliffe]

The one which is appropriate for the problem.  Which may be "none" as a lot of loops can be switched from the turnout movements and need no additional electronics.

 

You really need to give a bit more information - what works for Z gauge might not be ideal for LGB in the garden, what is simple for a dog-bone reverse may be difficult in a complex junction......

[ianjeffery]

blocksignalling does a dcc reversing module.

 

http://www.blocksignalling.co.uk/index.php/dcc-autoreverser-rev1

 

 

[Phil S]

As Nigel has said - more information is need to give meaningful advice!

IF you can find a 'point-operated/related' solution, and add to that,automatice train detection and switching of the points involved you may have succeeded in your aims without an 'auto-reverser'.

 

older/simpler auto reversers operate by detectiing a 'short-citcuit/ on either RAIL, and at each end of the 'reversing section' which MUST be long enough for ANY TRAIN with METAL WHEEL to be WHOLLY WITHIN the switchable part (or fixed part if that is smaller!**)  They may have an adjustment for sensitivity - to set the threshold between okay and 'short'... this will depend on locos used and the power supply's maximum / overload trip currents.  Relay based versions may be too-slow for some more recent controllers.

 

**WE have a '3 level' layout where the upper and lower level phases are swapped - according to the points (2 Y - sets for each ) linking them to the connecting Helix/mid level -  no need for an auto reverser as originally thought. [ And we have adjacent colour light signals which show red if the phase is wrong - in case the observer cannot see the points,or they failed to change ]

 

A CARE needs to be made to ensure that, if devices like the PSX are used to limit currents in districts, (and these isolate only 1 rail) - that no combination can allow the protection to be bypassed! [The PSX-AR switches both rails ... combining functions of AR and PSX.]

 

NEWER DESIGNS of Sensor-based Auto reverser AVOID ANY SHORT CIRCUIT - however brief, from occuring - by having a small (few centimetres, or less)  'sensor' section at each end of the reversing part: Power is linked onto these by a metal wheel bridging the gap ('as with the old type') from the adjacent [powered] track, BUT this immediately switches the adjacent AR section if its phase is incorrect -  without the possible large current flow of the 'traditional' design, ....  this avoids 10-20A short circuit currents in G Scale down to  whatever the limit is on a Z/N/other type.

Massoth (who make the electronics for LGB) and LDT (kit or ready built) are 2 examples of the newer type.

 

Alternatively or in parallel, optical or other detecting switches CAN be used to trigger the flip-over of phase - and this applies equally to the old types.  As a backup-method, the 'sensor type' will or can also respond the 'traditional way' if the sensing has failed for any reason.

[WIMorrison]

Also consider if you want to use Railcom as most reversing modules do not transmit the Railcom cut-out.

[Shroomy]

It’s a 00 gauge

 

at the moment I am using a Hornby module but it’s seems to stutter causing a short

it had been working fine

using a power cab 

as the loco comes into the reverse loop from points 

larger locos like class 47 seem to have less problem then a 4mt

hope that helps

[Nigelcliffe]

Total reliable and really cheap method:  attach a double pole double-throw switch to the turnout.  There might be such a switch already there within a turnout motor (eg. Cobalt/Tortoise motors have enough switch contacts, as do several other brands).  Use that to reverse both track feeds around the loop depending on which way the turnout is set.    As the turnout needs to be set to run over it, the switch is operated as the turnout is changed.    Also works for DC as well as DCC.  

 

 

The problem being hit at present is the Hornby module is probably tripping at about the same current and same speed as the PowerCab.  So, both detect the short.   That's one of several issues with short-circuit detection based switching, it is relying on the reversing switch being faster to react than the command station. 

 

- Nigel

[Shroomy]

As I say I had a Hornby one fitted it was working well which just seems to be unreliable just want to replace the Hornby auto reverse module with a better one 

Edited February 12, 2019 by Shroomy

[Harlequin]

16 minutes ago, Shroomy said:

As I say I had a Hornby one fitted it was working well which just seems to be unreliable just want to replace the Hornby auto reverse module with a better one 

Hi Shroomy,

 

I suggest you want a solid-state one (that is to say, without the moving parts of a relay) because they will switch faster, causing less stutter of the loco and hopefully switching before your controller detects the short. Should also last longer.

 

Something like a PSX-AR or the BlockSignalling product mentioned above.

 

Edited February 12, 2019 by Harlequin

[HLT 0109]

6 hours ago, Shroomy said:

As I say I had a Hornby one fitted it was working well which just seems to be unreliable just want to replace the Hornby auto reverse module with a better one 

Shroomy,

 

I am using the Hornby one (I have several on my 00 gauge layout and one on my new N gauge layout) and all work perfectly with Power Cab.  At least the new one was but it started to fail intermittently.  I have today discovered a loose solder joint to the rail in the feed from the RLM to the reversing loop.  I am sure this is the problem and expect that reliable service will resume once the connection is properly soldered.  When I say "loose" I mean the joint is broken so the rail is sitting on the feed wire so there is poor electrical conductivity.  I also had the same problem on my 00 gauge layout years ago and found a poor joint on that occasion.  I recommend a thourough inspection of all connections within the reversing loop before paying out to replace your RLM

 

Harold.

[hobbyhorse]

I tried the Digitrax ones but caused lots of problems with shorts, replaced them with the Tam Valley frog juicers that can be configured for auto reverse, and work without any problems.

Simon

[Alan Kettlewell]

I have used Lenz reverse loop units for a long time and currently have two of them on my layout.  They have always worked flawlessly.

 

Cheers ... Alan

 

[melmerby]

I have 2x LDT auto reverse units and they have worked  flawlessly for several years now.

They don't rely on shorting the track but you do need some extra isolated sections of track for sensing.

[Kingzance]

Can this elderly idiot add his request for assistance to this topic please? My layout - my first for many years and my first dabble into the world of DCC - is based on the idea of a terminus - goods facility such as one would find at many steam "end-of-the-line" locations. There is an Up line and a Down line but, in order for trains departing from either the goods or some of the platforms to access the Up line, they need to pass across a facing crossover. Similarly, trains arriving may need to switch from the Down line to the Up in order to access platforms. All my turnouts are Peco SL-E1XX Code 75 Electro-frogs and all are operated by Tortoise motors. My Up line goes into two large loops (allowing hidden storage as the line eventually descends below the level of the station) before returning as the Down line and approaching the same crossover mentioned above, although this time it is a trailing crossover.

I have attached a simplified version as a .jpeg image but the picture may fail to include the connection of the loops at the left end! This arrangement allows me to have what looks like double track throughout and adds the illusion of Up and Down working round the loops. I have laid all the track shown in its correct position and tested the Up line and the loops as far as the turnout/crossing arrangement on the access to the loops and I am pleased with the performance. I haven't ballasted the section to the right of the loops as I am unsure about what I need to do. 
The "problem" section is bounded by my added blue lines. Every rail length has its own feed and I am really confused as to how to deal with the section marked as Down Main between the common part of the upper SL-E195 on the left and the Normal exit of the upper SL-E196 on the right. In reality, this section of track is a little over 1100mm long so it is not the total length of a 4-6-0 steam loco + 6 x 57' coach train! I have, I think, two options:

1 - This is the "manumatic" version where I provide power to the "problem" section through a double-pole change-over switch. I could hold any train before the Up trailing point whilst I switch the track polarity but this prevents a steady approach to the station area for each train. That section of track is laid at a 2.5% rising gradient so I would be loading the motor quite heavily when it restarted.

2 - Could I use a DCC Specialities PSX-AR module (which I have bought) and, if so, would I need to stagger insulated rail joiners to get the device to work?
Sadly, I have received conflicting advice from the manufacturer and from the UK retailer hence - learned friends - my request to you!

Thanks in anticipation!

[WIMorrison]

I would simply isolate the 2 storage loops - insulating rail joiners on both 'red and black' tracks near the points and feed these isolated sections through an auto reverser module. The breaks in the rails MUST be at the same point in each rail and the length of the isolated section should be longer than the longest train you are running (definitely if you have metal wheels on your stock)

 

you can then get rid of your switches and leave the auto-reverser to do all the work - automatically

[Phil S]

'staggering' of the gaps at the auto-reverse entry/exit is not needed or a problem, as such, but may help, or hinder with maintainng the gauge ... as nylon, or no fishplate is not the same as a rigid metal fishplate, for the geometry. [ a few millimetres difference -  but keep it less than the gap between axles otherwise an uneccessary 'dead' section may result' ]

As far as the 'Auto-Reverser' operation is concerned, A 'single break' is all that is needed to TRIGGER the changeover ... BUT BOTH RAILS MUST HAVE BREAKS (just so that no one misinterpretes what I mean by single break....   AS A METAL WHEEL on a single rail makes contact with either side of the gap/insulation, it will EITHER find the phase /['polarity' =bad word for an ac signal!!] wrong or the same .... if WRONG the circuitry rapidly throws a relay/solid state switch to swap the phase of the reversable part (both rails) .. so that the error is removed, and the train can continue .

IF it is the 'short-circuit detector type' - only the 2 gaps at either end are present ... and IF the phase was wrong, then a SHORT CIRCUIT occurs - hopefully only momentarilly - during which the FULL CURRENT POSSIBLE from your controller will try to flow through that wheel !!!  ( this is where PSX type limiters help reduce the problem).

To WORK, your auto reverser must act faster than the safety cutout in the controller .... AND it must have GOOD connections of the dcc bus to the module and adjacent tracks.

The BETTER TYPE now available, uses the idea of a 'sensor track' section which need only be a few millimetres long - but 'instead of being connected directly to full track power, it is really an INPUT to the sensor circuit ... and if the voltages do not match, then it swaps the phase of the reversible section - and does this before the wheel can travel those extra millimetres to reach it .... therefore there is NO SHORT circuit because it has already been corrected.

But both of these require a good feed of the track dcc to the relevent track and module.  A bad connection does not let the maximum current flow - and may not be enough to trigger the change. (Some modules have sensitivity adjustments - a Z or N gauge loco takes less current than most 0 or G scale ... and the maximum possible short circuit current is likely to be differnt too .... you may have a 1A controller (perhaps 2A short circuit current) or  an 8A controller with 20A short circuit current !!  (You can see why LGB/ large scale now promotoes the senssor type to avoid 20A short circuit currents !!!.   a problem can still occur if a train is straddling the break at one end when trck power is (re)applied after a break.

 

Edited February 24, 2019 by Phil S
italics added for emphasis /clarification

[Kingzance]

Iain - thanks.

What you do not see on this image is the amount of station track-work which I have laid and it is all wired so that any section has its right rail wired to red and its left rail wired to black as viewed from a position facing departure from the station. At some point in the plan, it is necessary for the "red" rail to become the "black".
As for the access to the two big loops, I have separate feeds for both loops and insulated rail joiners on all the switched rails of the loops' infeed turnouts and the crossing (just as I have on the out lines from the turnouts on the right side of the bit I questioned). The frogs of the SL-E194 crossing are switched according to a logic table drawn up from the position of the turnout motors. I did this to avoid problems arising through the turnouts/crossing and hoped to move the issue onto a section of straight track. The loops themselves are around 5m long so would allow me to store assembled trains in the hidden areas - I expect that I can hold two trains in each loop so using a reversing module on each would restrict me to having only one train on each, wouldn't it? 

[Kingzance]

Phil - thanks.

 

I think I can see where you are coming from on your response. The layout is OO, the control system is a 2A PowerCab system and I am unlikely to be able to cope with more than two locos operating at any one time. Every section of rail has its own 0.5mm2 dropper feed soldered to its side and soldered to my DCC Bus but the Bus also is split into three Power Districts that are wired in 1mm2, each of which has its own PSX 1 device providing short-circuit and overload protection. The main Bus from the PowerCab supply board is run in 1.5mm2 cable. I think that I have covered volt drop, fault discrimination and good power connections by doing all that.
The PSX-AR module was bought a couple of months ago and is apparently equipped with Software Version N(4.15). Before wiring it up, I wanted to get best advice on exactly where in the layout it should go. Iain's post above suggests one solution but I fear that may reduce my operational flexibility.

[Harlequin]

I think you can have as many trains as you like queued up in the storage loops so long as none of them is standing across the isolating joints or moving across the isolating joints at the same time as another train is doing the same thing at the other end.

 

There's one extra wrinkle related to the above: If you intend to run any parallel movements into the inner loop while another train is leaving the outer (or vice versa) then each loop would need it's own, separate AutoReverser.

 

As far as red-black orientation is concerned, all of the track outside the isolated section that is controlled by the AutoReverser can have consistent orientation - there's no need for any other switching. Inside the isolated section the AutoReverser effectively swaps red and black on demand.

 

Schematic showing Iain's suggestion with my suggestion of separate AutoReversers: 

Edited February 24, 2019 by Harlequin
After thinking about it, re-instated the two AR schematic

[Nigelcliffe]

Kingzance says the two turnouts are controlled by Tortoise motors.  Each of those contains a DPDT switch, which changes when the tortoise is moved.   So, the DPDT switch could be replaced by using the switches inside the Tortoise motors, and the change-over occurs when the turnouts change. 

But, before doing this... .. the contacts on a Tortoise are a bit fragile when subjected to DCC currents in the long term (there's a program of replacing the contacts with external microswitches on all the Tortoise motors in a large 4mm layout I'm involved with), and there is a small dead-zone as the contacts swap position.   The latter could cause locos to stall.

 

So, somewhat more robust is to use one Tortoise internal switch to operate a DPDT relay (available on a nice circuit board with screw terminals for well under £10), and that DPDT relay changes the track polarity. 

 

 

Really simple,  cheap,   and it would work for DC running as well as DCC.   

 

 

- Nigel

 

 

 

[John ks]

Nigel 

With a simple return loop I agree completely with you. drawing below

 

 

Kingzance's loops are double track & the entry and exit points are not the same point.

I suspect that auto-reversers are the easiest way to go.

 

Harlequin, the problem I can see with holding multiple trains in a track controlled by an auto reverser is if the total current draw is more than the auto reverser's trip current

I.E. Several sound locos & lighted coaches with a total current draw of 2.5amps, if the auto-reverser is set to trip at 2Amps then I suspect it would sit there forever changing polarity

 

John

 

[shiny]

The tortoise "dead" zone is supposed to be there. It is to prevent the switch from energising both sides at once and creating a dead short on the frog as the motor travels - or welding your switch and point blades in place. The type of switch is called "Break before make".

You would need two relays for each motor, one one each switch of the tortoise - and not a simply use a single with more poles, otherwise the relay will (de-)energize too soon and you re-introduce the risk of short circuits if the point blades haven't yet caught up with the switch.

It would also be a good idea to wire a snubber across the relay coils, when the current is removed from the relay - back EMF from the collapsing magnetic field may cause sparking and pitting inside the original switch - and further shorten its life expectancy.

 

Stalling is pretty inevitable if relying on switches alone to swap polarity, it is a trade-off between stalling one train or shorting out the entire section - one or the other.

 

I would prefer the auto-reverse solution that Harlequin has described - it is a text-book example of how to do it right. Even more so if switch wear is as much of an issue with those point motors as you have described.

 

The train is powered by the auto-reverser during most of its journey around the loop, so when the points are changed there will be no current passing through the switch at all - and electrical wear is no longer an issue. The only time the switch has to carry a current is when the train is on the short isolated section between the auto-reverser and the point blade.

 

The relays sound like a good idea elsewhere though.

 

 

[shiny]

@Kingzance

 

Are you sure you're not over-thinking this? If you're using DCC, throttle forward and the train goes forwards. Take it off the track and put it back facing the other way - it will still go forwards. The polarity of the track doesn't matter in the slightest because forwards and backwards are relative to the locomotive - not the track.

 

There isn't any need to change polarity of the track anywhere except on the reversing loop itself. As long as your point frogs are wired up properly, you won't have any problems.

 

The plan that Harlequin has shown you won't reduce operational flexibility at all because the flexibility is inherent in the DCC system itself. There isn't anything to be gained from swapping track polarities elsewhere.

 

The only problem I can see comes from having more than one train on each loop track. You must make sure that neither train is parked across the isolated section - otherwise it will introduce a dead short to the whole layout.

The same restriction applies if you have one train entering as one is leaving - the auto reverser is only able to deal with one train at a time. It isn't a problem with the inner track because the crossing junction makes that impossible, but it is easily done on the outer track.

 

I've added my own take if more than one train is using the loop, apologies for making a mess of your artwork!

You would have to split it further if you were adding more trains, but a reversing loop is probably not the best place to store a train because of the polarity switching.

 

 

You mentioned a large storage area under the station, so if you have more space available, think about doing something like this - adjusted to your own layout:

 

 

Excuse the crudity of the diagram - I didn't have time to paint it or build it to scale..

 

You could have multiple storage roads, but as long as the reversing section (highligthed in blue) is longer than your longest train - that is sufficient for an auto reverser. You can store as many trains as you wish in the passing loops to the left, because they're just plain track. Ideally, you should only be driving trains in or out of the reverser - not parking them there.

 

As you can see, there isn't any requirement to make the entire loop reversing, and if you have simplified your diagram to hide more track (and trains) than we thought, then perhaps you should definitely not use the auto reverser for the entire loop. As long as you don't park trains on the reverser, and the reverser is longer then the longest train you intend to run and your power supply has enough oomph - that will do

 

 

Edited February 25, 2019 by shiny
added oomph

[Junctionmad]

just to scotch a few issues 

 

A good quality frog juicer like the Tam valley ons, never lets a short circuit current develop , as it operate much faster then the rise time of a full short circuit current in a command station 

 

Hence it switches at around 1.7A  which is often well below the max current a dcc command station can put out 

 

NO actually short circuit current flows in this case  

 

The other thing is that the total current draw of the locos  in the section being detected by the autoreversers has to be less then the trip current or else the auto reverser will assume its a detection event , equally the command  station current must be capable of rising to the detection current levels , which can be a problem where multiple locos are already running and the headroom of extra current isnt available . ( or the command unit cant generate that current in the first place ) 

Edited February 25, 2019 by Junctionmad

[Harlequin]

If Kingzance never intends to  drive a train (or a loco) completely around either of the loops and back out again without stopping, then Nigel's relay suggestion may be the best answer here.

 

If using AutoReversers and dividing the loops into smaller subsections I don't think there's any need for more than two ARs because you would simply extend the standard red and black connection part way around the loops, leaving smaller AR sections. (The downsides of dividing up the loops are that the storage capacity is then more limited, less flexible and requires careful driving or some system to stop trains exactly within the subsections.)

 

As Shiny points out the topology of Kingzance's loops are a bit limiting, when you think about how they would be used. A better topology would be a balloon loop that connects Up back to Down with a number of storage loops off it. (Almost what Shiny drew.) Then there would be no queuing and trains would be free to enter any storage line while another is leaving any other.

 

Edited February 25, 2019 by Harlequin