Do Sidings need Electronic Termination

[RDS]

I seem to remember a while ago on a discussion regarding a Bus system on a DCC layout, that it was recomended that some form of termination was required at the end of the Bus wire.

I don't have a Bus system but just recently, I have had a few decoders doing strange things, i.e. resetting to id3 or failing completely.  When I think about it, these decoders were fitted in Loco's that were 'parked' in my sidings, of which I have 17 on my layout.

Should I have some form of Electronic Termination at the ends of my sidings and may the lack of it have caused the failure problem.

[Chrissaf]

I think this is a question that will generate a wide range of conflicting advice with many people calling other peoples suggestions rubbish. So I'll start the ball rolling, with an upfront apology to start with, if someone else thinks I'm not making sense.

 

If deciding to fit suppresors (what you are calling a termination device) consider these.

.

Note: for any other users reading my reply, remember the OP stated that he did NOT have a DCC Bus installed on his layout. Thus, my comments take that fact into account.

 

1) You say you have 17 sidings, so you might be thinking of installing 17 suppresors. Not a good idea. Each suppressor will draw current from your DCC supply, so 17 of them will add up to a tidy amount of current. Not only that, as each suppressor is a high pass filter, having so many of them in parallel across your layout DCC track might have an adverse affect on your DCC digital signal.

 

2) You say you are a victim of decoders resetting to ID3, this may or may not be as a result of transient shorts on the track e.g forgetting to move a point and running a loco against it. Think back, do you have accidental shorts on a regular basis, if you do, then having at least one suspressor across your DCC track might be beneficial, even if you don't have a separate DCC Bus implemented. If you can't remember the last time you had an accidental short, then short circuits may not be the cause of your lost decoder configurations.

 

3) If you do decide to fit a suppresor just to be on the safe side, then in my opinion it should be located, if possible, close to where the problem area exists perhaps where the bulk of your loco storage sidings are. Failing that, then perhaps across the piece of track furthest away from the controller (this assumes your whole layout is one single power domain). If you have multiple power domains fed by DCC boosters, then a fit a suppresor in each domain.

 

As said before, I daresay others will come back with many conflicting recommendations and suggestions.

 

[Chrissaf]

Further to above have a look also at:

 

https://www.hornby.com/forum/loco-losing-its-programming/?p=1

[Fishmanoz]

Having just sat and thought about the situation from a circuit theory perspective, I can't see anything wrong with what you've said in practical terms Chris.  It can be said that everyone has a DCC bus to which loco decoders and accessory decoders are connected at various points, even if that bus is just the track itself.  The only difference between the track by itself and one with a "real" bus in parallel with it is that the former will have a higher resistance than the latter dependent how well the fishplates, points etc are working electrically.

 

So yes, add a terminator or two, and put them close to any shorting "action".  And no, more is not better and may start to degrade your DCC signal.

[Chrissaf]

Thanks for the supporting comment Fishmanoz, I was just covering my back with the upfront apology in case I missed something obvious or made a mistake.

[Howbi]

With so many sidings being constantly live on DCC would it not be useful to fit each with an isolation switch?

Also, is there any merit, on a large DCC layout, installing the bus wire as a ring main circuit?

[2e0dtoeric]

Howbi - if you really want to, you can isolate all the sidings, but then all loco's and stock parked on them will not work, (ie lights, sounds, etc), and you are defeating the object of DCC!

A 'ring-main' bus is only effective if you have a moderately large (in distance) tailchaser layout. Depending on the width, a single end-to-end bus is adequate, as the dropper wires won't be all that long anyway.

No doubt someone will now post a conflicting opinion!  :-)

[2e0dtoeric]

As a p.s., I have a tail-chaser layout, with an end-to-end bus. I use an Elite controller, with no railmaster or e-link, I have around 30 loco's, some of them sound, and so far (Taps head) I've not had a single decoder chip fail or reprogramme itself!

[RDS]

Hi Chrissaf

Thank you for your comprehensive reply.  My layout was pictured in the thread below and I did have a derailment on the point between Siding 5 and Siding 9 about a month ago and my decoder problems appeared just after that.

https://www.hornby.com/forum/railmaster-layout-plans-share-yours-here/?p=1

[Chrissaf]

RDS.

 

I remember seeing that layout in a previous forum browsing session. Very impressive it is too. A huge amount of interest crammed into such a small space. Assuming it is a relatively recent build I would only make one suggestive comment. With no BUS and only a single DCC controller power connector, you are totally reliant on the mechanical rail joiners for electrical continuity. It may be fine now, but will it still be a year or so down the line (if you'll excuse the pun) when the joiners start to loosen through heat expansion, contraction and vibration.

 

Personally, I would seriously consider adding a DCC BUS with track droppers now, before the layout gets too far developed with scenery and other accessories etc. As the track is already laid, small wire track droppers can be easily soldered to the outside edges of the track and hidden by ballast and connected to thicker BUS wires via a Scothlok or alternative connector. If you add a BUS in a T configuration, then the bottom of the T connects to your controller, and two suppressors can be fitted. One each end of the T cross bar. I found the Scotchloks below are a perfect fit on 24/0.2mm BUS wires, they may suit 32/0.2mm Wires as well.

 

Pack of 100 Scotchlok connectors including 100 Spades

[Chrissaf]

Addendum,

 

If your decoders have indeed been damaged by a short circuit, then there is no guarantee that fitting suppressors will prevent the affected decoders from still resetting to ID3. Fitting suppressors may reduce the frequency of this happening. The use of the suppressors will only help alleviate good working decoders from being damaged by short circuits in the future. You may still need to consider replacing the affected ones along with fitting suppressors for a true long term solution.

[RDS]

 I would seriously consider adding a DCC BUS with track droppers now, before the layout gets too far developed with scenery  ... hidden by ballast .

Hi chrissaf

Thanks again for your further reply.  Personally though, I have no interest in ballast or scenery, except for the odd car that was part of my family history, so I currently have no intention of adding a Bus.

I had not expected that the fitting of the terminations would have any benefit to decoders that appear to have already been damaged but of course I had hoped to prevent premature failures in future.

Having seen my layout, are there any suggestions for the postioning of terminations and what form should they take.

[Chrissaf]

You can buy Suppressors for about £8, but they are very easy to make for a fraction of the cost. The suppressor is typically a 100 ohm resistor in series with a 0.1uF (0.1micro-farad) ceramic capacitor. I would recommend a resistor with at least a 2 watt power rating as they get very warm in operation. The resultant resistor / capacitor series circuit is wired across the DCC BUS (in your case the track) either way round it doesn't matter. Sometimes the resistor in commercial suppressors are found to be 120 ohm rather than 100 ohm but either value will be OK. Personally, I would solder the resistor and capacitor to a small piece of copper clad stripboard rather than just solder them together and wrap with insulation as the insulation wont help with heat dissipation.

 

Note: 0.1uF sometimes referred to as 100n (100 nano-farads) - they're both the same value. Ceramic Disc capacitors and resistors can be fitted either way round as they are not polarity sensitive.

 

As to physical location. Given all the previous post dialogue we've had, I would suggest fitting two and no more. One on the link track between siding 3 & siding 8 to protect the bulk of the siding areas, the other probably on siding 12 to protect the turntable area. Mount the suppressor under the baseboard with short wire tails coming up through the baseboard to solder to the outside edge of the track rails.

 

Just out of curiosity, is your turntable track connected via a DCC Reversing Loop Module (i.e the rotating piece of track is permanently live), or is it dead when the turntable moves off from a connected track. The answer doesn't affect my suggestions, I'm just curious.

 

The resistors and capacitors can be purchased on-line from many electronics suppliers, but I give a couple of links below:

 

3 watt 100 ohm resistors

0.1uF 100v Ceramic Capacitors

 

OR

 

ESR 914-110 2 Watt 100 ohm Resistors

Click on top row far right box titled "2W 5% Carbon" then look for item part number 914-110

 

ESR 883-061 0.1uF 100v Ceramic Capacitors

Click on top row far left box titled "Ceramic Disc" then look for item part number 883-061

[Fishmanoz]

Given my memory tells me there is no reversing loop associated with RDS's TT, why would you want to use an RLM?  Then maybe I've remembered it wrong?

[Chrissaf]

Given my memory tells me there is no reversing loop associated with RDS's TT, why would you want to use an RLM?  Then maybe I've remembered it wrong?

 

Some turntables support (or are modified by users to support) a live track on the turntable platter, say for example to support a sound loco or directional lights. Steam hissing or Diesel tick-over and lights are on continuously as the turntable rotates. If you don't want to lose the sound and lights then the track section on the turntable platter needs to stay live as it rotates. If it then turns through 180 degrees the rail left and right on the turntable platter get reversed relative to the original input track therefore you would need a RLM to prevent a short between the platter track piece and the rest of the track. I was just asking RDS the question was his turntable permanently live or dead out of interest and curiosity, it had no other bearing on the topic

[RDS]

Given my memory tells me there is no reversing loop associated with RDS's TT, why would you want to use an RLM?  Then maybe I've remembered it wrong?

Hi Fishmanoz

You are right, no reverse loop.  I will explain my TT wiring in my answer to Chrissaf.

[Fishmanoz]

Chris, I would think it unusual to have a TT on a DCC layout with a dead bridge, although some of the ways people suggest modifying the Hornby TT for DCC end up like that.  I also understand about slip-ring designs with half the outlets having reversed polarity - see Brian Lambert's coverage of this in the DCC section of his sight.

 

But my memory of RDS's layout is that all his TT outlets are sidings, none go back to the layout.  So there is no chance of his creating a reversing loop, even if he gets his polarity wrong.  If he does, he can just correct the polarity of his siding connection.  Hence my question.

[RDS]

Hi Chrisaff

Many thanks for your reply,  I have a supplier I use for Electronic components and I like the sound of 2 suppressors.  I will go for the locations you have suggested unless anyone makes a different case.

In terms of my TT, it is a Hornby Model R070 DC one that I have modified myself.  The R070 from new, has a clip under each of the rails of the short input track which takes power from the layout.  These clips (cp1) are connected to the rotating track and to some brass clips (cp2) at one end of the rotating track.  These brass clips then connect back to the track (either to the input track or if the TT has been rotated, to a new siding).  

To modify the TT for DCC, I cut both the rails of the short input track, (at the TT side of cp1) and filled the cuts with some non conductive filler.  Therefore when the TT is in the input position, the current flows from the main layout through cp1, to the rotating rail and back up through cp2 to the short input track.  As the TT rotates the rotating rail is kept live and when the TT stops at a new siding, this siding is powered through cp2.  A reversing loop is not required because the TT is only connected to the layout at one position (the input rail).  The output sidings from the turntable are only live when the turntable is lined up to them but this is ideal for me.  At the 180 degree position, the polarity of the track rails are reversed but with DCC, this does not affect the running direction.

[Chrissaf]

Thanks for the corrections to my wrong assumptions. I don't have a TT myself which is why I was asking the question. I had read an article some time in the past about modifiying a TT for DCC, in that article they used a RLM so that is where my assumption that they were all potentially done that way came from. And why I was interested to know if RDS had done it the same way too.

 

They say there are many ways to skin a cat, I daresay there are instances where including a RLM is beneficial such as the one Fishmanoz highlighted where not all TT outlets are sidings.

 

I must learn to ask questions with less detail in them. If I hadn't mentioned the term RLM in my question then I wouldn't now have an 'egg on my face".

 

Question: in RDS explanation all the sidings pick up their power only when lined up to the TT bridge. Hence why a RLM is not needed. If someone particularly wanted all their siding live permanently all the time, thus they wired track droppers on the sidings to a DCC Bus, then would that situation benefit from a RLM?.

[PJR_slo]

Hi

For those interested in the calculations, DCC operates at a frequency of around 8KHz, and therefore the impedance of a 0.1microFarad capacitor is about 200 Ohms.  

This in series with a 100 Ohm resistor gives 300 Ohms which on a 16V DCC system will pass about 50mA current for each terminator, which equates to about 0.85 Watts per terminator, two thirds of which will be dissipated by the resistor.

So as Chris says, a single terminator won't draw much current, but several of them will.

Personally, I would insert link wires from the end of each siding to a single terminator which would also have the advantage of powering each siding from the other, just in case there is a high-resistance fishplate somewhere.

 

Peter

[RDS]

@PJR_slo

Peter, Thank you for the calculations - very useful.  I do favour the idea of joining the sidings together with just one terminator.

Incidentally, could I ask what formula you use to calculate ohms from impedance and frequency.

[poliss]

Everything you ever wanted to know about Snubbers/RC Filters can be found on Mark Gurries website here.

https://sites.google.com/site/markgurries/home/dcc-general-best-practices/wiring-planing/snubbers-rc-filter