KingEdwardII

Why not treat the two lines coming out of the terminus as two single lines rather than as a double track? I am prompted to suggest this since this is in fact the organization of Aberystwyth station until the 1960s. One of the single lines ran to Carmarthen while the other went to Shrewsbury. Aberystwyth had an MPD with a triangle for reversing engines between those single lines, rather than a turntable. The triangle was however separate from the main running lines and was closely associated with the MPD. And in case you think that Aberystwyth might have been on the small side - in reality there were 5 passenger platforms and a substantial goods yard. Platforms were quite long allowing for trains with substantial numbers of carriages. Operations got "interesting" on summer Saturdays...

Indeed. The prototypes seem to vary a lot with respect to the distance between the dolls. In cases where the arms are at different heights, they can often be very close together. It seems fairly typical to allow a bit more space in the cases where the arms are at the same height. I assume Ratio have chosen the common case where the heights are different, but it is a pity that they don't make any provision for some variation in the doll spacing with this kit. I think that the round post version (Ratio 468) provides more space between the dolls. Yours, Mike

How does your system compare with a system using a motor shield combined with an Arduino board or Raspberry Pi? Yours, Mike.

How are you building the DCC Controller? Which technology are you using?

I don't use solenoid motors - I use MTB MP1 motors in the main - slow action. I switch pairs of points like crossovers using a single channel on my DCC accessory decoder (Digikeijs DR4018 in my case). Works a treat, 100% reliable. The current draw of these motors is fairly low - max 150mA - so that a pair never tries to draw more than is available from the accessory decoder. Yours, Mike.

Wordsmith - I'm glad it's working out for you. Regarding the soldering, for "small" jobs like soldering dropper wires, I never hold the solder wire. I leave a roll of it on my workbench with the end pointing upwards and then rub the tip of the soldering iron across it when I'm ready to do the work. The tip collects more than enough solder for a small job and that leaves me with a hand free to assist getting the wire and track positioned. I'm not against using croc clips to hold things, but I'd only hold the solder wire for jobs requiring large quantities of solder - more in the plumbing line ;-) As for flux - I have belt & braces approach and apply flux to the work and also use flux solder - I don't get any failures of my solder joints. Tracklaying is steady work and you need to take some pains with it to get smooth running. I've moved over to a system of using track layout software (xtrkcad is my current program), getting the design sorted out on there and then printing out the design on paper at 1:1 scale, which I then stick down to the baseboards and use as a guide for laying the track. The design software helps ensuring that things fit together properly and that you avoid curves being tighter than necessary - xtrkcad lets you specify a minimum radius and then gives a warning if that radius gets breached anywhere. The software also lets you play with variations like changing between different radius points.

Iain, I think you have a great idea for us collectively to put together a set of info on the various busses. I think that given that the topic here is "automation", this info must include some details relating to some of the pieces of software used to achieve the automation. e.g. If you have iTrain + XpressNet, what set of capabilities are going to be possible? I don't think we should swamp this with too much detail on the software, but knowing that a particular combination can provide info on the location & identity of locos on the track (say) is something important to know when you're planning a system. Yours, Mike.

Mine has arrived - very neat piece of kit. Some pictures of my Pi 400 package as delivered from The Pi Hut. The two boxes contain the mouse and the power supply. All the connections are at the back of the keyboard, excepting the WiFI, of course: I show the legend for the connectors, as printed on the box. The MicroSD card comes already installed in the slot - this contains the operating system. I need to add an HDMI connected Monitor and a USB attached SSD drive for data etc. Yours, Mike.

PS Now for more of the red stuff. My favoured tipple is a Barossa Shiraz, say a St Hallett.

I'm glad you've fixed the problems, Roger. No doubt you have also learned plenty as well, which is all part of the fun. Yours, Mike.

Roger, If your picture shows the multimeter measurement places correctly, then I think that the left hand point shows that the frog has the wrong polarity when the turnout is set to the diverging (i.e. curved) route. In this case, the extreme left measurement should be 0V, i.e. the frog should be the same polarity as the bottom stock rail - but it is showing up as 18V. So as your loco passes over this turnout, once the driving wheels straddle the gap between the lower switch rail and the frog, 18V passes across the driving wheels on that side since they are electrically directly connected. Result: a short. You need to examine the mechanism for providing the feed to the frog - it isn't working correctly. I think you said that you're using Cobalt iP Digital point motors - you should cross-check your wiring with the iP Digital manual - it looks like terminal 3 is the feed for the frog, while terminals 1 & 2 are the DCC input. It may be that you need to swap the connections to terminals 1 & 2 to get the correct polarity. I'm not a Cobalt user, although the MTB MP1s that I use do have a similar kind of arrangement and it is possible to get the polarity the wrong way around. I find that my multimeter is my best mate when wiring up the layout - and I check all the polarities and voltages before I let a loco loose on the track. Good luck, Mike.

I've ordered a Pi 400 from Pi Hut - I've been thinking about buying a Pi 4 for some time, but I was fretting about the required cooling arrangements. The Pi 400 solves this aspect and the integrated keyboard concept is very neat. I plan to use mine not only to run model railway software, but also as a server for our house, running 24/7 with a large capacity SSD attached. I also fancy the idea of attaching a large touch screen (> 22 inch) and using that rather than the mouse for interaction. I'll report back on my progress. Mike.

Roger, Looking at your most recent pictures above, and the fact that you get separate shorts for District 1 and District 2, I think that the problem is most likely caused by the frog of each point having the opposite polarity to the curved switchblade. I say this since it is suspicious that the loco driving wheels appear to straddle the gap that exists between the curved switchblade and the rail leading up to the frog, in both cases. It should be straightforward to test this hypothesis with a multimeter - check if there is a voltage between those two positions on the track - it should be zero. If there is a voltage here, then the power feed to the frog is wrong and needs switching over. Mike. PS. Which red is in your glass??

Likewise - I drill 1mm diameter holes in sleepers for the pins. I have a standard cordless Bosch drill with keyless chuck and it has no problems gripping the 1mm drill bits. I have both ply and MDF in my baseboards and I've never needed to drill pilot holes to drive home the pins - I use a small & lightweight jewellers' hammer to avoid damaging the pins & hold the track pins with tweezers while hammering them in to keep them vertical. Yours, Mike.

Leon, Assuming your new layout is going to be static in the one room, I can recommend building some very simple legs out of planed timber - I used a mix of 34 x 44mm and 44 x 44mm for mine. Straightforward to attach to the underside of the baseboards and you can add horizontal stiffeners where necessary... Yours, Mike.

For controlling my MP1 motors, I use the Digikeijs DR4018, as mention above by Iain Morrison - that is indeed a cost effective solution if you're using DCC. The DR4018 has 8 channels for turnout motor control and a single channel can drive multiple turnout motors in situations where 2 (or more) turnouts need switching at the same time, such as crossovers, since the MP1s only take 150mA. I've not used the MP1s in an analog setting, so I can't comment on that, although the feed to the MP1 motors is an analog 12V, as you mention. I think that the following video shows an analog arrangement using DPDT switches, if I have understood it correctly: Meanwhile, it is also possible to control mimic LEDs (for feedback) for the MP1 using the arrangement demonstrated in the following video: Yours, Mike.

James, I can't agree with your comments about turnout motors, at least as it applies to my experience using MTB MP1 motors. First, the MP1 footprint is quite small, 28 x 40 x 13mm, with the longest dimension along the direction of the track, so that the motor sits underneath the turnout it is operating. I have had no problems installing MP1s in "crowded" situations including 3-way points, double junctions, double slips, etc. The MP1s have a further virtue in allowing for a good deal of adjustment, particularly perpendicular to the track direction. Second, MP1s don't draw any current when they are idle - they cut out when they reach their end position. Indeed, if they don't cut out, this indicates a problem such as an obstruction preventing them reaching their end position, which I find a useful characteristic since the small buzz they emit brings my attention to the issue. And yet, despite having the cut-out, the MP1 design still means that the turnout blades are fimly held in place at the end position. A further advantage of MP1s is that they have a built-in micro-switch that can be used for controlling power to the frog of the turnout. Regarding wiring, I use 17A / 2.5mm2 stranded wiring for my main bus and 10A / 1.0mm2 stranded wiring for my accessory bus and I don't find them "impractically thick" - they are very easy to work with. I use 0.75mm2 and 0.5mm2 stranded wires for connections from the bus to accessories and track and they too are easy and practical - for example, connecting straightforwardly to the screw clamps on both MP1s and the Digikeijs DR4018 units. I think that protection against short circuits is advisable whatever the wiring used - but larger wiring itself is actually a protection against fire once the wire can handle more current than the power supply can deliver - it is undersized wires that can overheat and burn.

Wordsmith, First, regarding the wire types. I use stranded wire everywhere - even for dropper wires. I find that solid core wires are much more difficult to handle, especially when soldering, so I never use them. I generally use wires from mains flex (flex, not mains cables!) for larger size wires, since they are readily available and relatively cheap and are colour coded. Flex wires are available from 2.5mm2 down to 0.5mm2, with nominal current carrying capacity over distance (i.e. in terms of metres) as follows: 28/0.3 == 14 AWG == 2.5mm2 17A 19/0.3 == 16 AWG == 1.5mm2 13A 32/0.2 == 17 AWG == 1.0mm2 10A 24/0.2 == 18 AWG == 0.75mm2 6A 16/0.2 == 20 AWG == 0.5mm2 3A I have used the 3 alternative formulations for describing the wires here - this is a very confusing topic. I prefer the cross sectional area formulation since it does not depend on the wire construction, unlike the stranding formulation. I use 2.5mm2 wire for the power bus - more to keep the voltage drop low across my layout, since I doubt I'll ever be using 17A! I use 1.0mm2 wire for my accessory bus. Connections from the bus to units like the Digikeijs DR4018 for point motor control are typically done with 0.75mm2 wire while connections to my slow-action point motors (MTB MP1s) are in 3 strand 0.5mm2 flex. Dropper wires, typically no more than 200mm long, are done with black stranded wire 0.25mm2. Regarding your power supplies, I was surprised to see you mention a 10A supply. You must be using some very power hungry accessories to need that, although such supplies are readily available from the world of laptops and similar equipment. I run my layout using a 3A supply for the controller and a 2A supply for the accessories. Yours, Mike.

Bob, Congratulations on a superb and very professional set of videos. Mike.

Regarding "Tap Splice Connectors" (as RS Components dub them), otherwise called Scotchloks after one of the leading brands, I have very good experience with them. There are different sizes of connector - and the right size must be used for the wires concerned. My DCC power bus uses 2.5mm2 flex cores (== 14AWG) and I use the Blue RS Components connectors to tap into the power bus with 0.75mm2 (== 18AWG) wires. These tap wires I typically then connect to a pair of terminal connectors which then have two or more pairs of dropper wires attached - the dropper wires are smaller still, to make the connections on the track unobtrusive. The terminal connectors allow me to detach the droppers if required. Meanwhile the DCC accessory bus uses 1.0mm2 flex cores and I use the Red RS Components connectors to tap into this, linking to 0.75mm2 or 0.5mm2 wires to whatever accessory device is involved, such as the Digikeijs DR4018. Using the tap splice connectors is quick - stick the relevant wires into the connector and a quick squeeze with a pair of pliers and it's done. I've never had one fail on me. The bright colours used are also helpful in finding the connections when messing around in the dim light under the board. The tap splice connectors are really cheap - a recent 50 pack of the red connectors was £1.65. Yours, Mike.

@Robin, The power supply for the accessory bus on my layout is currently a 12V DC supply redeployed from an old piece of computer equipment. I use this to feed the Power terminals of the DR4018 units, which actually drive the point motors. The DCC control signals for the DR4018s come from the power bus at present. However, this is only a temporary arrangement - I am moving towards computer control for the accessories and this will involve a separate low-cost DCC controller driving the accessory bus, like a SPROG, with a Raspberry Pi issuing commands via JMRI driven by a touch screen display. The current wiring on my DR4018s is shown in the picture. The power bus is at the top, the accessory bus at the bottom and the connections to the MP1 point motors are the various white 3-core flex cables. Mike.

Regarding the MTB MP1 point motors. These have two completely separate sets of 3 connectors each. One set drives the motor itself to switch the point, with a left/right pair and a common, which in your case come from the DR4018 decoder. The other set relate to the 1 amp SPDT switch - which can be used to provide power to the point frog. In this case, one connector is for the wire to the frog, while the other pair are for feeds from the right and left rails, typically from the power bus in a DCC setup. The SPDT switch flips between right and left in conjunction with the motor moving the point. The picture attached shows the wiring related to one of my MP1s operating a PECO electrofrog point. The brown & blue wires are from the power bus and feed droppers to the rails at the toe end of the point on the right hand side, plus feeds to the SPDT switch of the MP1. The frog connection is made with the green wire connected via the terminal block on the left. The motor feeds from the DR4018 to the MP1 are the three wires passing horizontally from the left towards the top of the picture, with the green wire being the common. I power the DR4018 decoders using a separate accessory bus, independent of the power bus. I use terminal blocks a lot for the final connections to anything on the track to allow for removal of any piece of track with minimal disturbance to the wiring. Terminal blocks are dirt cheap and very reliable in my experience.

For Peco Electrofrog points in a crossover, put insulating joiners on both rails where the two points connect. Switch the polarity of each of the frogs independently, as shown on the web page linked in Sol's post. I use MTB MP1 point motors, switched via Digikeijs DR4018 DCC decoders, and for crossovers I use a single channel on the DR4018 to drive both motors, so that both points switch together. Each motor has a built-in switch used to set the polarity of the frog of the point it controls. With respect to frog switching, the setup is the same for all electrofrog points irrespective of whether they are part of a crossover.

I use mains 2.5mm FLEX for the main DCC power bus, rather than 2.5mm T&E cable, since the flex is multi-stranded and is much more flexible than the solid core present in the cable. Here is an example from Screwfix, which is a good price: https://www.screwfix.com/p/nexans-3183y-white-3-core-2-5mm-flexible-cable-10m-coil/7723t The flex has another advantage in that the earth wire is also sheathed, unlike in T&E cable, and can thus be used for other purposes when wiring up your layout. It is also available in a variety of lengths to suit your needs & budget. The only downside of flex is that you have to strip the outer sheathing before use - but this only takes a couple of minutes for a 10m length. I also use 1mm flex for the DCC accessory bus - this is available in a 5-core version with a wider variety of sheathing colours that can help distinguish between different uses under the board. For connections, I use Splice Connectors, which I source from RS Components, to avoid the need to have soldering under the layout. These connectors are available in 2 sizes, one good for the 2.5mm power bus and the smaller one good for the 1.0mm accessory bus. Each connection is a 5 second job with a pair of pliers and is extremely reliable. Yours, Mike.

I recently had to make this change for the Zimo MX638D 21-pin decoder I installed into my new Heljan 47xx 2-8-0. The default installation had the loco going in reverse when "forward" was selected on the controller. In my case, bit 0 of CV29 was initially "0" so I flipped it to a "1" by adding 1 to the original value of CV29. It is important to leave the value of the other bits of CV29 unchanged, since they affect other functions of the decoder. Now my 47xx goes the right way when I select "forward" on the controller! Mike