Jump to content
 

Newbie Question - DCC/Computer Control


Wordsmith
 Share

Recommended Posts

7 minutes ago, Wordsmith said:

Thanks all for the above comments - I wanted this to be a future retirement hobby that made me think. And it's certainly going to make me do that.

 

I'm beginning to experience the gulf between what you can pick up from books/the Internet and what practical experience can teach you.

 

As to stopping distances, etc., I've found out that you can program acceleration and deceleration rates, etc, into a loco, so that's also on my list of things to play with in the near future.

 

The hobby has come on a long way since I built DC layouts as a teenager....

 

Wordsmith

 

 

When I returned to the hobby about 5 years ago I found out that DCC existed first off, something i had no idea about. But I didnt like the equipment much, so I decided to hold off and just a build a fairly complex DC layout to get to grips with the issues conventionally. So I had power sectors, reversing loops, etc. Now I have decided its time to go into DCC albeit slowly. I feel a lot more prepared being able to apply all my general layout building learning to give me a chance of building something good in DCC. I have been soaking up the advice as to what to do, what to leave well alone, and to identify the areas of the hobby that I find completely uninteresting, and therefore won't bother with. I would say that train operations interests me the most, which i only discovered when I was running 3 trains and a shunting operation by myself. So I am intending to run a layout that will feature automatic running in the future.

 

Give yourself the time to make the decisions, no rush....

Link to post
Share on other sites

A friend recently approached me to help him automate his layout but with one major proviso - he wanted to be able to shunt and make up trains himself and run then from the goods yard to the storage yard.

 

I helped him automate the passenger trains which now run around the layout automatically on a timetable leaving him to manually create goods trains and drive them around. The program ensures that nothing gets entangled and he is a very happy teddy - he says that after 40 years he now has a model railway the way he has been trying to make it work  for all these years :)

  • Like 2
Link to post
Share on other sites

On 09/12/2020 at 11:16, WIMorrison said:

Why do you want to use a DR4018? These motors are meant to have a decode built into them already.

 

Ah thanks - got them working.

 

Remarkably simple when you know what to do. It's the figuring out what to do from a low starting point of knowledge that's more difficult.

 

Only problem I now have is that Traintronics seen to have become a Covid 19 casualty and are no longer trading.

 

Are there any other slow motion point motors with built in decoders? The obvious candidate from a bit of Googling would seem to be a Cobalt digital IP. My gut feel (and I stand to be corrected) is that built in decoders will mean marginally less wiring under the baseboard (and less embuggeration).

 

Cheers,

 

Wordsmith

  • Friendly/supportive 1
Link to post
Share on other sites

The Cobalt Digital IP do have inbuilt decoders and I have 23 of them on the layout however I have now changed to MTB MP1 motors of which I have 26 and I change them using a DR4018. I don't think that the wiring is reduced by Using Cobalts v anything else. I find that the MP1 are quieter than the Cobalts and the range of adjustment is excellent.

  • Like 1
  • Informative/Useful 1
Link to post
Share on other sites

Like Iain, I am a devotee of MTB MP1 motors, driven via DR4018 units. The main wiring is from the DR4018s to the MP1s and is a 3 wire arrangement. I simply use the smallest size 3-core flex wire for this purpose (0.5 mm2), only exposing about 1 inch at each end. This reduces the wire runs to a neat single cable for each motor and it is naturally colour coded. I place the DR4018s centrally within a group of points to keep the runs as short as possible, but I do have some runs up to 2.5metre or so.

 

The wiring for the frogs is much simpler and shorter since I link to the power feed for the droppers local to the point.

 

Yours,  Mike.

  • Like 1
  • Informative/Useful 1
Link to post
Share on other sites

Thanks both - I've put trying out MTB MP1 point motors on my  'to do' list.

 

Another question if I may...

 

I've got the little Peco 'Wiring the Layout' booklet. And they have a section on using link wiring as an alternative to drop feeds. So I rewired the sidings on my little experimental layout in that manner. That resulted in less wiring under the baseboard because the sidings are no longer directly connected to the power bus; essentially they're getting their power from the head shunt drop feeds, and the link wiring is overcoming the self-isolating feature on the points. I can see that power transmission becoming more of an issue as you are more dependent on the rails conducting it (although I assume you could solder the metal rail joiners in place on a permanent layout to help improve the conductivity).

 

Are there any pitfalls to using link wiring?

 

Cheers,

 

Wordsmith

  • Friendly/supportive 1
Link to post
Share on other sites

You should solder the wire to the rails not the rail joiners as they will lose connectivity because the only rely on the push fit that will corrode. 
 

TBH I haven’t a clue what ‘link’ wiring is but I suspect it is simply linking the rail joiners together in a line.

 

you can do the same for droppers by going from one dropper to the next with one going to the bus.

  • Informative/Useful 1
Link to post
Share on other sites

Wordsmith,

 

How you wire up your track is partly dependent on what you want to achieve. So, for example, if you want to have automation, you are going to need some form of occupancy detection. One form of occupancy detection is based on monitoring each section of track for the current taken by a loco. The more sophisticated forms of this detection include getting info from the decoder which enables identification of the decoder address and hence the identity of the loco.

 

If you want to go down this road then each section of track needs its own separate feed that can be monitored independently of any other section.

 

If you're not doing this, then you have more freedom in how you organize the wiring. Clearly, getting a good feed to the track is top priority. Personally, I have not found that there is overmuch wiring involved in having feeds from the DCC track bus to each track section, but I recommend using a quick and easy method of connecting the droppers to the bus.

 

Yours, Mike.

  • Informative/Useful 1
Link to post
Share on other sites

17 hours ago, WIMorrison said:

You should solder the wire to the rails not the rail joiners as they will lose connectivity because the only rely on the push fit that will corrode. 
 

TBH I haven’t a clue what ‘link’ wiring is but I suspect it is simply linking the rail joiners together in a line.

 

you can do the same for droppers by going from one dropper to the next with one going to the bus.

 

Link wiring is essentially directly taking the power from one rail to another, rather than from a power bus. Not the best of diagrams, but you could essentially power a siding by wires from the mainline rails.

 

image.png.df160a04837aa993e753a9952f512a45.png

 

But @KingEdwardII has given me the answer. If you use link wiring, occupancy detection won't work as it wouldn't detect if the train is on the main line or in the siding.

 

Another lesson learned - thanks...

 

Wordsmith

Edited by Wordsmith
  • Friendly/supportive 1
Link to post
Share on other sites

  • 3 weeks later...

So I built myself a 5 ft x 2 ft 6" baseboard over Xmas and am now slowly laying track on it. I can get 14" radius curves, which is probably closer to what would go onto a more permanent layout. (9" radius curves were too tight for my Schools loco). I'm laying the track on 0.8 mm thick cork just to see what difference that makes. If I get all the track laid this month, the plan for February is to get an occupancy detector and try and figure out how that and JMRI work. Much head scratching in prospect.

 

Couple of questions in the meantime if I may.

  1. How often do the major manufacturers redo production runs? for example, if a steam loco goes out of stock, how long before they run off another few hundred/thousand?
  2. How easy is it to renumber steam locos? There's only a limited range of RTR locos for the era I'm interested in, so if I buy several of the same model, they'll need renumbering. Is it practicable to do it yourself. (I note The Model Centre do it for £50, which is a cost I'd like to avoid if possible).

Cheers,

 

Wordsmith

Link to post
Share on other sites

  • 3 months later...
On 05/01/2021 at 18:32, Wordsmith said:

Couple of questions in the meantime if I may.

  1. How often do the major manufacturers redo production runs? for example, if a steam loco goes out of stock, how long before they run off another few hundred/thousand?
  2. How easy is it to renumber steam locos? There's only a limited range of RTR locos for the era I'm interested in, so if I buy several of the same model, they'll need renumbering. Is it practicable to do it yourself. (I note The Model Centre do it for £50, which is a cost I'd like to avoid if possible).

 

I note that these questions were never answered:

 

1.  Models are usually produced in batches and when a specific model is sold out, it tends not to be produced again in the same livery / number / name combination.  The tooling is likely to be reused to create another member of the class at a later date, but the timescale could be a year or it could be several years: it really depends on how popular the model was when it was first issued.  If it sold well, a new batch (but with different name / running number) is perhaps likely in the next catalogue (ie a year later), but if it was a slow seller or the manufacturer thinks that most people who want a loco of that class probably already have one, then it may be several years.  There is also the issue that tooling can get worn or damaged, in which case the answer might be that a particular model is never reissued.

2.  There are plenty of people who renumber models themselves (says he who has never done it).  The challenge may be to find who has produced the transfers for the particular locomotive that you want, albeit if you're not looking for a named locomotive, then you should be okay with a single sheet of numbers for whatever company you are modelling.

Link to post
Share on other sites

I have a couple of questions that aren't really answered in this thread.

 

The first is section lengths - I'm assuming that these should be shorter where something is happening (ie where you want to stop a train, so that the calculation of stopping position is more accurate), but presumably they can be quite long if nothing is to happen?   I'm looking at installing occupancy detectors (or at least insulated joiners to allow occupancy detectors to be installed at a later date) in my fiddle yard loops.  The majority of loops will be around 12 or 13 foot long, so I'm assuming I'd want something like a nine or 10 foot long section, followed by something like a three foot section (that being the area that I want to stop the locomotive).  Would that be correct and if so, is there any way to ascertain what the optimum lengths would be?

 

The second is the question of whether or not turnouts should be connected to an occupancy indicator.  Is it really necessary for every turnout to have it's own section, or can I simply create a section that includes multiple turnouts?  I'm assuming that the principal benefit of using occupancy detection  for turnouts is that it could be used for interlocking.  That is, if the section with the turnout is occupied, then the turnout can't be changed.  Does monitoring the turnouts have any other purpose?

Link to post
Share on other sites

Feedback lengths are pretty much as you suggest. The longer they are the greater the inaccuracy in stopping as programs use a simple time and distance calculation to estimate position. The accuracy of stopping is also affected by how carefully you measure the speed of the loco and assess other parameters.

 

a turnout can never be part of a block ( or sections as you call it) because there can only be one entry and one exit from a block for program logic to work. The program changes turnouts to route a train from one block to the next block, building the logic around this. It is your choice to place feedbacks on turnouts, if you have feedbacks and something is drawing current on the turnout feedback then the turnout will not change until the track is clear, however programs do not need turnouts to be monitored.

  • Like 1
  • Thanks 1
Link to post
Share on other sites

You also may consider different options if you have bi-directional loops in your storage yard.  The attached doodle shows a couple of options - one for bi-directional running and one for single direction.  For the single direction option you can of course just isolate say the last 3 ft or so where the head of train should stop, but for di-directional running then it would make sense, and use fewer sensors, to isolate the whole loop as shown.  Also with 13ft available you may want to consider storing two or more shorter trains in a loop, as per my doodle.

 

These ideas are based on 'occupancy detection', different arrangements would be required for 'passing contact' detection where reed switches mounted in the track are used and no isolation is required.  Happy to elaborate further on that if required.

  

228310280_Occupancyzones.png.ec14906e7c8a986771c649589c7f1c36.png   

Cheers ... Alan

  • Thanks 1
Link to post
Share on other sites

  • RMweb Premium

At a terminal station I have this block arrangement:

 

platforms.jpg.12858baf50b59495e30b4a61839838ca.jpg

 

The 12" & 13" (and traversing table) blocks are for locos to couple & uncouple from trains. My trains can be up to 7 or 8 coaches + loco but usually 4-6 + loco

 

A train enters the platform and runs partly in to the 12 " block, stopping with the rear of the loco over the intersection of 12" & 63" clocks where there is an uncoupler.

Loco uncouples and moves fully into the 12" block leaving the rest of the train in the 63" block which can then be taken back out by a loco coupling on the other end.

All done automatically.

 

(in fact the 13" ones are unnecessary as the loco can couple to a train as it enters the block, without separate detection)

 

  • Like 1
  • Thanks 1
Link to post
Share on other sites

10 hours ago, melmerby said:

At a terminal station I have this block arrangement:

 

platforms.jpg.12858baf50b59495e30b4a61839838ca.jpg

 

The 12" & 13" (and traversing table) blocks are for locos to couple & uncouple from trains. My trains can be up to 7 or 8 coaches + loco but usually 4-6 + loco

 

A train enters the platform and runs partly in to the 12 " block, stopping with the rear of the loco over the intersection of 12" & 63" clocks where there is an uncoupler.

Loco uncouples and moves fully into the 12" block leaving the rest of the train in the 63" block which can then be taken back out by a loco coupling on the other end.

All done automatically.

 

So, is the slow down and stop sequence initiated when current is first detected in your 63" blocks?  That is, as soon as the leading wheels of the locomotive pass from block 73-1 P 1B (13") to 73-5 P 1 (63"), the locomotive starts to slow down and then when the leading wheels of the locomotive pass from 73-5 P1 (63") to 71-1 P 1A (12") the locomotive slows down more quickly (if necessary) to ensure that it stops in the correct location over the uncoupling magnet.  I'm assuming its the detection of the locomotive at the boundary between the last two feedbacks that gives you the accurate stopping over the uncoupler.  The final 12" feedbacks seem quite short, but I'm assuming that's because the trains are by that point travelling very slowly.

 

11 hours ago, Alan Kettlewell said:

You also may consider different options if you have bi-directional loops in your storage yard.  The attached doodle shows a couple of options - one for bi-directional running and one for single direction.  For the single direction option you can of course just isolate say the last 3 ft or so where the head of train should stop, but for di-directional running then it would make sense, and use fewer sensors, to isolate the whole loop as shown. 

 

Most of my loops will be bi-directional because although I want to be able to run trains round and round, I envisage operating most trains as an out and back type operation.  There will be a few shorter loops that will only be able to hold a multiple unit or a short train, so I won't need to split the long loops to accommodate shorter trains.  I'd like to be able to run 20' long intermodal trains, but alas have to accept that I need to compromise and seven 10'+ loops is all I can accommodate.  My initial thinking was therefore as per your first sketch, that I just have one feedback for the entire length of the loop.  However, with that approach, the known point (ie where the locomotive enters the block) is going to be up to 13' before the point that I'd actually want the locomotive to stop.  How accurately can you stop a train at the end of a block that is that length?  The alternative is obviously as per @melmerby's terminus, where each loop is subdivided into three.  That would give more accurate location data in relation to the stopping position, but obviously it becomes more expensive to implement.

 

12 hours ago, Alan Kettlewell said:

These ideas are based on 'occupancy detection', different arrangements would be required for 'passing contact' detection where reed switches mounted in the track are used and no isolation is required.  Happy to elaborate further on that if required.

 

I'd be planning on using occupancy detection - probably the Digikeijs DR5088 (https://www.digikeijs.com/en/dr5088rc-digidetect.html).  At the moment, I'm only looking at laying the track and installing dropper wires, but I want to make sure that I add insulated rail joiners (and track feeds) at the appropriate locations and to route the track feeds to the location at which I'd install a DR5088 before connecting to the DCC bus, to allow me to install occupancy detection at a later date (without too much hassle).  Obviously where I locate a DR5088 would depend on how many blocks I would need to monitor.

Link to post
Share on other sites

I can get accuracy of stopping distance to within 5mm - even with a long brake and stopping distance from the point of detection.  I use Train Controller software and ensure that I do really accurate speed profiling of all locos.  Cleanliness of track, wheels and pickups and mechanical condition of locos are a factor.   

 

Cheers .. Alan 

  • Thanks 1
Link to post
Share on other sites

David,

 

my blocks can have either one, two, three and even four detection sections depending on location. A normal one way line where a train has to stop at a signal has two detection sections, yes you could get away with one but accurate profiling is a must but I have found that not all locos are equal. The first section varies in length but the stop section is normally 6 inches this give accurate stopping every time at the signal. A bi directional line will have this short 6 inch section at either end of the block. Normally you can’t have a point in a block per say but there are work arounds for this and I do have points that are part of a block setup but are not actually in the block image itself. Having detection on points are not strictly needed but personally I do detect them as any trains that possibly lose a vehicle on the points will stop points throwing but of course you need every vehicle to have resistors fitted, having points detected puts the overall cost up of having them detected that needs to been born in mind. As you can see from all the replies we all have our various methods and there is no one way to do things that’s the beauty of these programs there is more than one route to achieve the same objective.

Edited by Andymsa
  • Thanks 1
Link to post
Share on other sites

  • RMweb Premium
1 hour ago, Dungrange said:

 

So, is the slow down and stop sequence initiated when current is first detected in your 63" blocks?  That is, as soon as the leading wheels of the locomotive pass from block 73-1 P 1B (13") to 73-5 P 1 (63"), the locomotive starts to slow down and then when the leading wheels of the locomotive pass from 73-5 P1 (63") to 71-1 P 1A (12") the locomotive slows down more quickly (if necessary) to ensure that it stops in the correct location over the uncoupling magnet.  I'm assuming its the detection of the locomotive at the boundary between the last two feedbacks that gives you the accurate stopping over the uncoupler.  The final 12" feedbacks seem quite short, but I'm assuming that's because the trains are by that point travelling very slowly.

 

I have speeds that reduce in each block as the train approaches the buffer stops, the brake and stop markers for the train are in the 12" block which has a speed limit of 10mph

So there is a gradual reduction 40 - 30 - 20 - 10mph

You can also put temporary speed limits (schedule/train specific) that can take effect anywhere in a block if you so desire

 

The 12" block is just for a loco after uncoupling from the train, although it can also take part of a whole train.

Say you have a 10" loco, you can put the stop marker at 9" leaving the tail of the loco overhanging the previous block which includes the uncoupler at the end. After uncoupling the loco moves forward 2" and is now fully in the block. Train and loco will now show as separate entities in their respective blocks.

 

39 minutes ago, Alan Kettlewell said:

I can get accuracy of stopping distance to within 5mm - even with a long brake and stopping distance from the point of detection.  I use Train Controller software and ensure that I do really accurate speed profiling of all locos.  Cleanliness of track, wheels and pickups and mechanical condition of locos are a factor.   

 

Cheers .. Alan 

So can I, you need it if you want to do automatic coupling/uncoupling:yes:

 

16 minutes ago, Andymsa said:

David,

 

my blocks can have either one, two, three and even four detection sections depending on location. A normal one way line where a train has to stop at a signal has two detection sections, yes you could get away with one but accurate profiling is a must but I have found that not all locos are equal. The first section varies in length but the stop section is normally 6 inches this give accurate stopping every time at the signal. A bi directional line will have this short 6 inch section at either end of the block.

I manage quite well with one detector per block, even with a 103" block.

 

As you say.

Many ways to skin a cat!

 

Link to post
Share on other sites

4 minutes ago, melmerby said:

I have speeds that reduce in each block as the train approaches the buffer stops, the brake and stop markers for the train are in the 12" block which has a speed limit of 10mph

So there is a gradual reduction 40 - 30 - 20 - 10mph

You can also put temporary speed limits (schedule/train specific) that can take effect anywhere in a block if you so desire

are you using adaptive braking ?

Link to post
Share on other sites

31 minutes ago, melmerby said:

No

Unless it's done by default, I certainly haven't set it.

 

If I remember it’s set by default. Look at the loco speed profile under the speed graph there is a tick box for it. The effect is that a block has a max speed of let’s say 50mph then the train will start to slow down in the proceeding block so by the time it reaches the next block it’s doing 50mph.

Link to post
Share on other sites

  • RMweb Gold
1 hour ago, Andymsa said:

If I remember it’s set by default. Look at the loco speed profile under the speed graph there is a tick box for it. The effect is that a block has a max speed of let’s say 50mph then the train will start to slow down in the proceeding block so by the time it reaches the next block it’s doing 50mph.

 

It's not in the speed profile - what you're thinking of there is brake compensation.  Adaptive Braking is set in schedule rules: it's at the bottom under "misc".  It's set on by default when you create a new schedule. 

Edited by RFS
Spelling
Link to post
Share on other sites

10 minutes ago, RFS said:

 

It's not in the speed profile - what you're thinking of there is brake compensation.  Adaptive Breaking is set in schedule rules: it's at the bottom under "misc".  It's set on by default when you create a new schedule. 

Oops my mistake 

Link to post
Share on other sites

I think we should point out to the OP that much of this discussion, certainly the latter parts, is specific to the use of TrainController. Other programs may have similar, called differently - or they may not have the 'function' at all.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...