Can Solenoid Point Motors Affect DCC Bus

[Greynut]

Hi to all,

I'm putting a DCC layout together and it keeps getting itself changed as I find a better way to do something. No, I'm not moaning, I'm quite enjoying the learning curve. Recently, I've been putting on some solenoid point motors and they are working very well.

Some of the motors butt more or less against the DCC bus as it passes by. In the light of quite a lot of discussion about various things that might (or might not) interfere with the signal that the loco decoders are getting, or inducing a spike and perhaps throwing or damaging said decoders - it set me puzzling but I can't find an answer anywhere. Can the electro-magnetic pulse from a solenoid point motor induce a spike or glitch (of any consequence) in the bus and/or DCC signal?

The bus runs are around 30 feet from the eLink to the end and I understand that longer runs might be more prone to induction, although the bus wires are twisted and I'm going to put bus terminators on as well. Main line points are operated via an accessory decoder. The others, sidings etc, by on-off-on switches through a separate transformer/cpu.

I'd be really grateful for any insight or thoughts.

[Rocketcooper2000]

are you implying you are getting some 'spikes' or anomalies '  I would propose there maybe someone more experienced  electronically than me who may help but Ive never come across any of the worries that you describe. my small electrical minds says its impossible [ the electrical launguage is totally different] but like I say Im not an expert! With lots of signals everywhere, radios Tvs etc in most instances, im sure everything still works fine ..  ; ]]

 

[Greynut]

Hi Mark, I'm not asking due to spikes that I am now getting - more to avoid that possibility if I leave the bus routed right next to some solenoids. I'm asking out of general interest too - I freely admit that I'm a bit of a noddy when it comes to things electrical, so all info is useful.

When I started the layout, I had problems with the decoders on the 2 locos that I had, as they kept losing their settings - then one after the other, the decoders blew. It wasn't being caused by derailments shorting it all out but it happened for no apparent reason that I could see. Each bit of track has droppers and it seemed that everything current-wise should have been OK.

But after the decoders were replaced, the settings loss started again. One side of the layout went over (and very close to) a freezer and I noticed that when it cut in, I got the glitches. It happened on several occasions so I didn't think that was co-incidental. OK, it might have been - but the idea of outside interference affecting the DCC bus had crept in. The freezer had to go - it went (the wife will hopefully get used to where it is now) and the settings loss stopped. Again, I could only conclude that the field of the freezer motor had interfered with the DCC signal.

To be safe, not being rich enough to have a supply of decoders on hand, I went into research mode - hence the layout got itself re-wired, using twisted wire bus - and being a tad paranoid about things, I'll be adding the bus terminators as well.

I know what you mean about other day-to-day sources giving off signals and these don't affect the DCC aspect, I'm with you on that - but I just wondered if a short, sharp burst of electro magnetic energy from a solenoid going off right next to the bus might induce a spike or 'jolt' the DCC signal.

[Rog RJ]

If you keep the two DCC bus wires close together (by twisting them for example) then any induced voltage will affect both wire equally, thus cancelling out any induced voltage.

[96RAF]

I don't think I misread your intial post GNut but it seems you are running your points from the track bus.but running a separate (twisted) bus for your points than that for your track is seen as a good idea, even recommended by NMRA, but not set as a standard that people have to stick to.

[Greynut]

Thanks for your replies.

Yup Rog, the DCC bus wires are twisted, so hopefully I'm OK there.

RAF - the 'main line' points work through an accessory decoder powered by the eLink track bus. As I want to twiddle with things as well, the points in the sidings work via a cpu powered by a separate transformer and operated by switches. These work via separate twisted wires (if they run a couple of feet or more) as I thought that if twisting the DCC bus wires cancels out induction, then it might cut down on 'leakage' also. The second tranny has no input into the actual track circuit at all.

[96RAF]

According to NMRA you only need a twist every foot or so (30cm) to eradicate bus wire cross chat.

As an aside:

My points are dual Analogue/DCC in that I have 3 x rotary switches for point selection and 2 x push to make for activation of L or R, wired via a Gauge Master CDU out to each point motor.

The DCC acc decoders are wired to pulse into a cheap car relay coil (one relay for each solenoid side - so yes there are lots of them). The relay closes its contacts for enough time to fire the CDU into the points.

I should have fitted diodes to prevent spiking from the relays but never did and so far not had any failures as a result.

[Greynut]

I have more like 3 twists per foot so it should be OK.

I like your analogue point triggering set up - it seems a lot more compact to what I'm going to end up with when I've finished the point work. I might be feeling one of those 'back to the drawing board' moments fast approaching.

With your DCC, if I've got it right, the same analogue method CDU actually fires the solenoid courtesy of a DCC acc decoder pulse switching it in via the relay. Again, if I've got it right, that gives you the option of DCCing or twiddling the same points because the actual throwing of the point is effectively isolated from the DCC circuitry. That's much food for thought is that ... I'm mulling that one too!!

I see that you have given thought to spiking being generated by the relays. Is that to protect the acc decoder circuit or the potential of a spike to affect things generally? It's encouraging to see that you have not had any problems in that direction though - but I'm still wondering if an electro-magnetic spike can come from a solenoid and which might induce a glitch in the DCC bus which is right next to it.

[Fishmanoz]

Interesting discussion.  Clearly pulsing a solenoid is going to produce an electromagnetic field, again as a pulse.  Unless you could find a reference to someone having done it already, you would need some test equipment to measure the effect in the bus.  There will be some pulse transferred, the question being how much. I can't imagine it being anywhere near as much as the spike produced on recovery from a short circuit but the question is whether it is enough to affect a decoder.  Your freezer experience suggests maybe it can.  It will be producing a field all of the time, but again there will be a spike when it first cuts in.

 

So I don't have an answer for you beyond it is a possibility.  And maximising the separation and fitting bus terminators will both reduce the effect. 

[Greynut]

Cheers Fishmanoz. I won't be in a position to get things up and running to do a test for a couple of weeks but I will give it a go to see. Hopefully, if there is a pulse, it will register on my ancient multimeter. Before I posted, I spent several days trawling the net to see if I could find something on this but couldn't find anything at all - hence my post to see if anyone had come across it.

I'd certainly go for as wide a separation as I can but my layout goes round the walls of the spare room and so is not very wide at all to get a good space between solenoids and bus - especially where there is a run of several points linking across 2 or 3 tracks. Plus I can't stagger the points very much due to limited run length and still include interesting stuff like stations etc. With points switching as trains progress, the one spike becomes several and add to that the potential of two points switching simultaneously at cross overs and magnifying any spike ... it kind of sets you thinking with my set up.

I admit that I might be raising a subject that is actually a non problem but equally, how many problems that get raised about odd things happening to DCC locos might have something like this as a factor. If an encoder loses settings or blows after a derailment short, it's obvious what caused the problem. What about the situations where suddenly the loco doesn't respond or a sudden runaway - could solenoid spikes have a bearing on some of those? I suppose that I'm only stabbing in the dark - no actual proof - but yes, the freezer thing did set me thinking along those lines - rightly or wrongly.

[Chrissaf]

My DCC BUS is about 100 feet of 24/0.2mm wires fed by the controller (Elite) in the middle of a T configuration DCC BUS. I should, in hindsight, have twisted it but I didn't. My 17 solenoid point motors are wired also with 24/0.2mm wires, again un-twisted. I also have both a 5v and 12v DC accessory distribution BUS again un-twisted. All wiring is routed through 20mm holes in the baseboard frame. Some holes, but not all, route both DCC BUS and point cabling. Therefore, there are areas where DCC and point wiring lie side by side following short lengths of the same route. For reasons I won't go into here, my point motors are fired by an external CDU controlled by a bank of relays connected to the R8247 DCC decoder outputs. So basically if there was a fundamental issue of spike interference between point firing and DCC, then I would probably be affected by it. My layout and point control works flawlessly. I did have one doggy R8247 decoder that kept losing DCC addresses after a transient short, but this was resolved by replacing the decoder and fitting DCC BUS Suppressors to the ends of my T shaped BUS.

 

I did an electrical / mechanical apprenticeship followed by college courses in electronics and digital communications systems theory, some time ago I admit, but physics theory is still valid today. In general, it is 'sparking' that is the primary cause of induced interference in digital systems. Sparks generate a broad band of radio frequency radiation, the harmonics of which can be generated in the DCC frequency range (not sure but I think this is in the 8K to 16K range). The firing of the solenoid, yes produces a rapid rising pulse of (DC) current, but provided there is no contact arching (spark quench capacitors fitted across the make contact of the point firing relay) then the likelihood of generating radio frequency interference is minimal.

 

Also, digital systems (I can't comment on DCC specifically) usually employ error correction techniques. Any solenoid generated current spikes that might occur are likely to be of very short duration due to the nature of the way a CDU functions. A fridge / freezer on the other hand is more likely to generate radio frequency interference over relatively longer times, thus exceeding the capabilities of any digital error correction mechanisms built into the DCC specification to recover.

 

I dare say that someone else will come back and say I'm talking b*****ks, that's the nature of forums. But there is no disputing the fact that my layout wiring design (being un-twisted) is theoretically more prone to interference than what you plan to implement, but I have not experienced any issues or adverse interaction between my points and DCC operation.

 

Also, I would take comfort in the fact I don't think issues of point motor / DCC interference has previously been a topic of discussion on these forums. This would seem to indicate that this has not been found an issue with the majority of forum contributors.

[Greynut]

Hi Chrissaf, many thanks for your reply.

The plan for my layout is 4 levels running round the walls of the room, linked by gradual inclines and each level laid out to be different. Each actual level will be a circuit of the room so quite a few trains could be on the move at the same time. Although the levels are 'rings', the bus feeding each is/will be effectively a 'T' configeration with a break in the bus and isolated track joiners. The bus feeding each level will separated at the eLink feed, so it will be a multi-'T' arrangement. I've got two levels on the go so far. I've got nowhere near a 100 foot run per bus, more like 30 feet. Although there will be 4 bus runs, I think I can think of it as 4 x 30 foot bus runs rather than 120 feet overall. I could be wrong though because the eLink is effectively feeding 120 feet - but not in one long length. Powering the eLink side of things, I have a 4 amp tranny - hopefully that will be enough ... we'll see!!

So far, I've got 14 points in, 6 wired working through an ADS8 decoder and 4 working by switches but there will be quite a bit of pointwork to be done. This includes points switching on and off the inclines. The inclines form an outside spiral with each incline getting power from the level it rises from, with droppers feeding up from that bus - then isolated from the level it rises to. And the points on the circuit levels generally. Being a great one for overkill, I'm using 32/0.2 wire for the bus work and soldering droppers to each piece of track. The switch operated points have their own tranny/cdu set up which can be used for other non DCC stuff if required.

I first encountered the decoder blowing problem before I twisted my bus wires and while the freezer was very close to the circuit. I suppose what I should have done is given the locos some more time to chug around before I twisted the wiring to see if simply removing the freezer had solved the problem - looking back that would have been more objective and would have been more step-by-step. Seeing that most of the advice was to twist - I jumped straight in on doing that. It also got me thinking about other possible potential sources of interference.

I've got no real electrical training so I rather jog along and do mini crash courses as and when needed ... and then proceed very carefully. Plus, I find it all quite interesting. Although I got myself puzzling as to whether solenoids very close to the DCC bus would cause me mischief, I am just as interested in the question itself as much as if it might affect me with my particular layout design. So to me, it's just as important to see that you have not experienced anything that you could put down to this happening - and as you are very well versed in this side of things, I certainly take note. Add to that you are also having to run circuits in close proximity and not finding it a problem.

I do agree that as I couldn't find any reference to this anywhere on line, that's probably because it's not a problem. Not knowing bugged me though, so I had to ask. As I said, above, I won't be able to run things for a week or so but I will check to see if I get any readings when solenoids fire ... if nothing else, just for interest's sake. I'll post the findings when I have them.

[choralc]

Excuse my ignorance but I do try to understand the magic of all this technology but when you say twisting the bus do you mean crossing the two separate bus wires two or three times every 'x' amount of distance?

[Chrissaf]

Excuse my ignorance but I do try to understand the magic of all this technology but when you say twisting the bus do you mean crossing the two separate bus wires two or three times every 'x' amount of distance?

 

Precisely !!!

[Chrissaf]

 

My layout is over 3 levels. Your plan to break the track loops with insulated joiners where the potential ring buses are broken into a T is spot on.

 

One other thing to note with the suppressors: The resistors in the commercially bought suppressors run very hot indeed. Too hot to touch (see note 1 below), after lots of research I built my own:

 

I can recommend the following site for DCC theory:

https://sites.google.com/site/markgurries/home

 

and specifically for suppressors (also called a snubber or rc-filter) see:

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

 

The ones I built consist of two 200 Ohm 3Watt resistors in parallel (giving a 100 Ohm resistance of 6 Watts) in series with 0.1uF 100v ceramic capacitor built onto a small piece of strip board held with double sided tape on the sub-frame. Do not encase in heat shrink tubing as the resistors dissipate heat more efficiently in free air (the commercially bought suppressors usually are encased).

 

The two 3Watt resistors in parallel is complete and total overkill, but they now only run warm to the touch and not hot enough to fry and egg on as they dissipate the same amount of heat over a significantly greater surface area. And can be built for a fraction of the price paid for commercial ones.

 

Note1). Using Ohms Law to calculate current flows and power (heat) is not accurate as Ohms Law is only true for pure DC. These suppressors are in a high frequency circuit, complex physics is in action here. The heat actually generated seems far higher to the touch than that that simple Ohms Law calculations would indicate.

V=voltage in Volts, I=current in Amps, R=resistance in Ohms, W=Power in Watts

Ohms Law

V = I x R

I = V / R

R = V / I

W = V x V / R = I x I  x R

 

PS - Don't buy your components from Map***s. Their prices are extortionate. I buy all my components on-line from http://www.rapidonline.com/ free delivery for orders over £30 or about £4 P&P for orders that don't meet the free delivery criteria.

 

 

[Greynut]

Many thanks for the links. I have to say that it was coming across Mark Gurries' website that added to my thoughts on solenoid interaction, as he does think through the potential of DCC bus interference. (Although he doesn't mention solenoids.) And it was his 'Snubber' pages that got me thinking about bus terminators. Once I had the idea of adding 'Snubbers', I did some wandering around the net and read about the heat that they could generate and that it was best not to wrap them. Although having said that, some commercially made ones are shrink wrapped - and selling for up to £8 or £9 a throw, which I thought is a tad steep for what they are.

I bought my components through ... ermmm ... M*pl*ns, as I thought their prices were not too astronomic but I'll check out Rapid next time - thanks for the tip. I've got 0.1uF 50V ceramic capacitors and 120ohm 2W resistors. With the heat that I now think this might generate, from what you have said, I think I'll be warm enough on these cold winter evenings. I'll do the overkill ('tis my way too) and parallel the resistors. Luckily they will get good air circulation where they will be mounted. I did do quite a bit of researching before getting the bits and these components seemed to be a good combination for the job but I'll be interested if you think otherwise.

And I once thought model railwaying was simply sticking your track together and watching your trains chug round ...

[Chrissaf]

Greynut.

 

The 0.1uF capacitors, 50v rating should be fine. Like you, I don't mind over specing even if it costs a little more for a real 'belt and braces' approach. I chose 100v rated capacitors just because they were available at the time of order.

 

I wouldn't double up the 120 Ohm resistors (giving an equivalent 60 Ohm resistance) the value of the resistor controls the resonant frequency of the RC filter, values below 100 Ohms will start to shunt the operating frequency of the DCC BUS. Going slightly above 100 Ohms i.e. your 120 Ohm value is OK though. I have found references to DCC suppressor resistor values between 100 and 150 Ohms being the norm. Mark Gurrie's theory page states power dissipation for 00 gauge to be up to 0.3 Watts. Your 2 Watt resistor shouldn't be at all stressed with that level of power dissipation. It will feel warmer than my two 3 Watt resistors, but still significantly cooler to the touch than the 0.5 Watt resistors used in commercial suppressors. My ‘Mark 1’ home built suppressor did, in fact, use a single 100 Ohm 2 W resistor. I was still unhappy with the amount of heat to the touch, so upgraded to 2 x 3 Watt resistors in my final ‘Mark 2’ design. A re-design that was probably unnecessary but, like you, I wanted real ‘peace of mind’. I did initially buy a pair of commercial suppressors before deciding to replace them with my home built ones. The concentrated heat generated by the commercial ones scorched the heat shrink sleeving with an accompanying burning smell, hence why I promptly replaced them, plus the fact that I burnt my finger touching them.

 

If only buying one or two components, then once you take P&P costs into account then it can sometimes work out cheaper at Map****. I do a lot of electronic projects, so tend to build up an order wish list with Rapid and delay ordering until I hit the free delivery threshold, then the lower component costs at Rapid really mount up. Another very reliable supplier I use is http://www.esr.co.uk/home.htm this supplier sometimes have lower minimum order restrictions than Rapid, but they don't have any free delivery options.

 

If I only want to place a small order and delivery timescales are not an issue, then there are a number of electronic suppliers (China) on that well known auction site with really low prices and free delivery. These China suppliers sometimes do some quite obscure components that you just can't find in the UK. For example do a search (on the auction website - not Google) for a component called an RPR220 this component contains an infra-red emitter and detector in a single plastic package. They make absolutely brilliant loco detectors (mounted between the rails) to trigger things such as signals and level crossings etc. Of course, they need to be used with an electronic control board circuit. This is where my electronic background has been invaluable as I have designed and built my own. The circuits I have built using them don't require ANY modification to the rolling stock such as 'gluing reflective tin foil to the underside' as do some other documented infra-red solutions I have found during my Internet research. And, of course, being infra-red there is no need to muck about with reed switches and magnets, and they work just as well in the dark. There is one company I have found that use these neat little detectors themselves in their own ready made model railway control boards http://www.blocksignalling.co.uk/ worth a look at.

 

Chrissaf

[Greynut]

I hadn't thought through the halving of the resistance, so many thanks for pointing that one out. On the wattage side, I'll see if I'm concerned or not about the heat thrown out - if I am, it won't take much to alter it. Regarding the commercial ones, it's really not on when you can burn yourself and smell scorching with a shop/internet bought item - and that's when it's working properly. OK you can hide it away but I wouldn't exactly feel easy about it. It's handy knowing recommended alternatives to Map**** - much appreciated.

The RPR220 bit is exceedingly interesting btw. I think a new file will be opening for that idea, although I can't see me seriously thinking about signalling and things of that ilk for quite a while. It would be just the thing to work through the second, non DCC transformer - even if and when the Hornby train location detection system comes in, there could be a lot of scope to operate some bits non DCC, if simply to cut down on the DCC loading - as you say, level crossings, for example.

[Fishmanoz]

Just a couple of asides here.  For a start, stop worrying about saying Maplins etc, perfectly acceptable here.

 

Second, if you can find a resonant frequency in a simple series RC filter, I'll defintely eat Graskie's hat.  What we are talking about here is an elbow frequency, above which the impedance of the filter starts to fall, which needs to be above the 15kHz or so DCC frequency but low enough to short out transients arising from shorts etc.

 

If the frequency is such that the filter impedance is just the resistor value, the dissipation  would be around 2.25 watts (the 15 volt DCC voltage squared, divided by the resistance, ignoring peak to peak v RMS volts).  But in fact, the impedance will be significantly higher than that as the C will have a large effect on impedance at the frequency, so the wattage will be less.  Interesting then that your 2 watt resistor got hot, I would speculate it must be to do with the higher frequency components in the  leading and trailing edges of the DCC PWM pulses and the lower impedance they see.

[Chrissaf]

Quite right, incorrect use of terminology on my part. By 'resonant' I just meant the point in the filter frequency response where the filter response starts to change significantly.

With regard to temperature everything is relevant. What is hot to me might be cool to you and vice versa. I just said I was unhappy with the amount of heat given off by the 2W resistor, my personal choice would be for it to run as cool to the touch as posible. I said it was the 0.5W resistors in the commercially bought suppressors that ran too hot to touch.