How big a step to take with Z21 and iTrain, now or later?

[WIMorrison]

May I remind people that the OP will be using iTrain and a Z21 therefore it will be more helpful for the OP if comments are relevant to what he will be using.

 

[Andymsa]

29 minutes ago, jamespetts said:

 

It is slightly more complex than that. TrainController Gold allows for two systems of point position feedback: (1) electronic only; and (2) physical.

 

The electronic only system works by checking to ensure that the correct command has actually been sent out on the relevant data bus, but does not check whether the unit has responded to it. The physical system, which I use on my layout, uses sensors (in my case, micro-switches) to detect whether the point mechanism has actually moved. The former system is of course less effective than the latter, but will at least check that the correct command has got to the relevant data bus. The latter system requires considerably more work to implement, but is more effective and simulates real life modern signalling, in which point position feedback is definitely a thing, more accurately.

 

As to micro-arcing - can I ask how you reach the conclusion that this does not cause oxidation and that running trains does not cause oxidation? It is difficult to find much documented empirical research on this topic: if you are aware of any, I should be interested to read it. My own experiments strongly suggests that running trains does cause oxidation: running a train up and down a clean piece of track for circa 2 minutes will result in black deposits on a previously clean white microfibre cloth that were not there at the beginning of the experiment and are not on adjoining rails that have not had a train run on them. Places where more significant arcing occurs, such as at a spot where a derailment and short circuit has recently happened, causes very large black deposits to build up to the extent that these are visible on the rail head.

 

unless the physical monitoring is connect to the tiebar of the point then any other physical monitoring would be at the mercy of the point blade and tie bar being still connected, that’s how as you say the real railway works.

 

did your research include YouTube as there are a vast amount of videos on the subject, there is a real good one done by a research chemist which if I had the time would look for, plain and simply oxidation is rust, running trains removes this rust hence running trains does not cause oxidation, the black crud you’re seeing on your cloth is build up in the pitting on the rail head. Micro arcing can exacerbate this pitting not just on rails but the loco wheels that pickup the track power. Black crud is caused by using a polar cleaner on your rails ie IPA but using a non polar cleaner reduces this black crud. Some also use a grease type material lightly on the rail head, in Australia it’s known as ionox but this is not very good for traction tyres.

[jamespetts]

24 minutes ago, Andymsa said:

 

unless the physical monitoring is connect to the tiebar of the point then any other physical monitoring would be at the mercy of the point blade and tie bar being still connected, that’s how as you say the real railway works.

 

did your research include YouTube as there are a vast amount of videos on the subject, there is a real good one done by a research chemist which if I had the time would look for, plain and simply oxidation is rust, running trains removes this rust hence running trains does not cause oxidation, the black crud you’re seeing on your cloth is build up in the pitting on the rail head. Micro arcing can exacerbate this pitting not just on rails but the loco wheels that pickup the track power. Black crud is caused by using a polar cleaner on your rails ie IPA but using a non polar cleaner reduces this black crud. Some also use a grease type material lightly on the rail head, in Australia it’s known as ionox but this is not very good for traction tyres.

 

Indeed - real railway grade point position feedback requires monitoring the tiebar, and that is the ideal way of doing it, but that is usually not practical in a small scale. I use feedbacks on the mount itself, which will detect most problems, such as a failed or disconnected servo, mechanical stickage at the mount or an electronic error causing the command not to be sent in the first place. It will not detect a disconnexion between the motor/mount and the tiebar, a failure of the tiebar or disconnexion between point blades and tiebar.

 

I note from iTrain 5's manunal that it does not have point position control built in in the way that TrainController does, so this may be less useful, although this can be customised, I think, from Actions (although I have not investigated this fully - Actions in iTrain are much less powerful than macros in TrainController as iTrain has no user modifiable variables).

 

As to oxidation, it would be interesting to see the video to which you refer. (Rust is a form of oxidation, but not all oxidisation is rust: rust is specifically oxidisation of iron, and model rail track and wheels do not normally contain iron, but rather copper and nickel). What are the keywords with which one can find it? I am not entirely sure that I follow what you mean when you write of the black deposits being pitting - how can a deposit be a pit? Do you mean that this is the material that was once in the pit? It is not shiny and metallic, but black, so presumably the stuff that was in what is now the pit has been oxidised, so these are indeed oxide deposits.

 

I believe that you are correct about using non-ionised cleaning fluids or else risking greater oxide deposits. I do not use alcohol or water for this reason (and yet still get the black deposits as discussed above). Somebody I know has recommended WD40 contact cleaner (not WD40 penetrating oil - the names are very confusing) for this, but I have not tested this myself. I tend to clean the rails dry so far, but might well try WD40 contact cleaner if necessary.

[Andymsa]

52 minutes ago, jamespetts said:

 

Indeed - real railway grade point position feedback requires monitoring the tiebar, and that is the ideal way of doing it, but that is usually not practical in a small scale. I use feedbacks on the mount itself, which will detect most problems, such as a failed or disconnected servo, mechanical stickage at the mount or an electronic error causing the command not to be sent in the first place. It will not detect a disconnexion between the motor/mount and the tiebar, a failure of the tiebar or disconnexion between point blades and tiebar.

 

I note from iTrain 5's manunal that it does not have point position control built in in the way that TrainController does, so this may be less useful, although this can be customised, I think, from Actions (although I have not investigated this fully - Actions in iTrain are much less powerful than macros in TrainController as iTrain has no user modifiable variables).

 

As to oxidation, it would be interesting to see the video to which you refer. (Rust is a form of oxidation, but not all oxidisation is rust: rust is specifically oxidisation of iron, and model rail track and wheels do not normally contain iron, but rather copper and nickel). What are the keywords with which one can find it? I am not entirely sure that I follow what you mean when you write of the black deposits being pitting - how can a deposit be a pit? Do you mean that this is the material that was once in the pit? It is not shiny and metallic, but black, so presumably the stuff that was in what is now the pit has been oxidised, so these are indeed oxide deposits.

 

I believe that you are correct about using non-ionised cleaning fluids or else risking greater oxide deposits. I do not use alcohol or water for this reason (and yet still get the black deposits as discussed above). Somebody I know has recommended WD40 contact cleaner (not WD40 penetrating oil - the names are very confusing) for this, but I have not tested this myself. I tend to clean the rails dry so far, but might well try WD40 contact cleaner if necessary.

 

I did not say the the black deposits was pitting please re read my post. The posts are getting a bit off topic if you wish to start a new topic on this ? as it not fair to the OP. I’ll let you do your own research as I spent many hours doing my research on track cleaning. I’m not here to prove a case as it comes across as such, but track cleaning is a complex subject but think to a real railway and how rails go when nothing runs on them as opposed to a used railway. Yes I use the wd40 contact cleaner as well as other methods.

[njee20]

90% of this thread is now bickering about micro arcing and the functionalities of Train Controller, which the OP isn't using, and indeed I'd counsel against using because of the absurdity of its creator. Andy - James, why not continue this via PM or in a dedicated thread? I don't really get the relevance of the repeated references to turnout monitoring on the real railway either.

Edited December 9, 2021 by njee20

[jamespetts]

20 minutes ago, Andymsa said:

 

I did not say the the black deposits was pitting please re read my post. The posts are getting a bit off topic if you wish to start a new topic on this ? as it not fair to the OP. I’ll let you do your own research as I spent many hours doing my research on track cleaning. I’m not here to prove a case as it comes across as such, but track cleaning is a complex subject but think to a real railway and how rails go when nothing runs on them as opposed to a used railway. Yes I use the wd40 contact cleaner as well as other methods.

 

I have started a new topic as suggested here.

[ITG]

Thanks to those who recognised the subject matter was drifting somewhat. I have to say it was getting difficult to see the wood from the trees!

As I said, this thread is very much in preparation for a future layout, which, unless a significant wave of contrary advice arises, I’m pretty sure I’ll use a Z21 and iTrain. But there are some aspects arising on here, which I wasn’t aware of, or at least didn’t understand. Hopefully, you helpful folk can shed some light…..

 

1. I understand some users of iTrain use Digikeijs DR4088 sensors, but I note there are different versions of this, as notated by different letters at the end of the model code, eg RB, LN. What is the distinction between them?

 

2. Mention has been made of resistive axles on stock. Is this essential? What are the operational implications of not having? All stock? If not, which stock?

 

3. I did wonder (in this thread, before the topic drift) if a list existed of all the physical components to consider.Whilst responses are throwing them up one by one (as far as I can see), a list would help me organise my thinking, and journey of understanding. Has anyone got such a list?

 

As always, thanks for all the help.

Ian

 

 

[WIMorrison]

1. I understand some users of iTrain use Digikeijs DR4088 sensors, but I note there are different versions of this, as notated by different letters at the end of the model code, eg RB, LN. What is the distinction between them?

 

RB = RBus which is the proprietary bus used by Roco and suitable for the Z21

LN =Loconet which is Digitrax proprietary bus and is licenced for use by Roco and Digikeijs

CS = Current Sensing and is an S88 bus product 

Opto = Optically isolated and useful when switches e.g. reed, IR, magnetic, etc are used.

GND - Ground for use with 3 rail systems

LN-CS = Loconet with current sensing an S88 bus

LN-GND = Loconet with GND and S88 bus

LN-Opto Loconet with Opto and S88 bus

 

2. Mention has been made of resistive axles on stock. Is this essential? What are the operational implications of not having? All stock? If not, which stock?

 

Can be useful, however not mandated. If your stock has lighting, then you already have this. Often useful to create resistive axle (5-10k SMD resistor with conductive paint for DR4088xx) on rear wagon to keep feedback and block occupied if train loses rolling stock on layout. It is also beneficial when reversing into sidings, etc. as this indicates to iTrain when the train has entered.

 

3. I did wonder (in this thread, before the topic drift) if a list existed of all the physical components to consider.Whilst responses are throwing them up one by one (as far as I can see), a list would help me organise my thinking, and journey of understanding. Has anyone got such a list?

 

The minimum that is required are;

• command station with a communication bus that feedbacks can react with
• command station that interfaces to computer using recognised protocol e.g. Loconet, Xpressnet, BiDIB, etc
• computer that supports connection from command station
• feedbacks to identify location of trains
• ideally a split bus for track and accessories to enable easy recovery if train overruns turnouts (not mandated but advised, see diagram) 

[peach james]

2:  Resistive Axles- I don't have them on all the stock, about 1/12th of the stock has them.  It depends on how fixed into rakes your trains are, and how much resistance is required.  I use Digitrax (so DB4 and DS 168's), and they will reliably trip over on 12k ohm, so I fit 10k ohm, typically at 1/axle, and 2/vehicle, fitted to all brakes (both passenger and goods), and then any R cars (because they are generally middle of train for passengers.  The goods stock is rather less well fitted- generally about 1/12 of my 16 tonner collection is fitted (I have about 60, 16 ton minerals), and some other stock has been rather more haphazardly fitted.

I'm using RR&Co, and generally have the sensor off times set to 5 seconds, that seems to work for me.  If you have more short sections, then either they need a longer sensor off time, or you need more resistive wheelsets.  I know there is one location (exit from the storage yard) where I have the sensor off time set higher, because of repeated problems of trains getting missed as not completely passed.

I think they are essential for successful operation using current based detection.

[peach james]

To clarify- a lit piece of stock would not need additional resistance :).  Nor would a train that was perpetually banked.   One of my DMU's has resistors, because it is fed through from the powered unit, so to detect the end of the 2nd coach, that has resistors.  (it's the S&C painted up 158, rather than the 108)

 

[jamespetts]

I am not aware of any list in a single external source for all of the hardware requirements or considerations of automation. I and others on this thread have attempted to summarise the relevant hardware requirements as best we can. My own research a few years ago was able to piece together what was needed, although building a small automation test layout also helped.

 

As to the DR4088 devices, Iain Morrison correctly summarises the types. One or two things to know about the S88 bus: first of all, this bus is supported by the DR5000 command station, so you can use this instead of the LocoNet bus if all that you want to use it for are occupancy sensors (implying using a DCC accessory bus for control of accessories such as points and signals). Secondly, it is a less capable bus than LocoNet. Thirdly, it can take one of two cabling types: a bespoke ribbon cable or an RJ45 network cable (at least when used with the DR4088s - I believe that the ribbon cable is the normal standard). Fourthly, its simpler nature may make it more prone to noise/unreliability if you have very long cable runs, although I have not found any problems. Fourthly, it is limited to 16 DR4088CS devices per node. Fifthly, the DR4088LN devices have an S88 bus output, so you can connect a DR4088LN to the LocoNet bus and then connect up to 16 daisy chained DR4088CS devices to each DR4088LN. Finally, the DR4088CS devices are cheaper (even taking into account the need to use slightly more expensive RJ45 cables) than DR4088LN devices. So, it can be helpful, if you need to cluster a few DR4088s together, to use a DR4088LN and connect a DR4088CS right next to it with the S88 bus. You do not need to do any of this and can use all DR4088LNs, but you might find that using a combination of LN and CS types will save a little money.

[WIMorrison]

Ian

 

iTrain can and does work perfectly without any resistive axles, but they can be advantageous however there is also a disadvantage when you have a large number as they are all drawing current from the command station - perhaps only a few mA, but that adds up quite quickly. I personally know of 2 layouts that went overboard with resistive axles, and they needed boosters to compensate - the biggest advantage is reversing into sidings, and if you have stock that uncouples then they are beneficial in showing where the lost end of the train is - but easier (and better) just to sort out the couplings

 

The default 'off time' for DR4088xx cannot be altered, however within iTrain you can set the default on delay (how long feedback should be 'alive' to considerer it as a valid occupancy) and the same with the off delay - how long should it be considered truly off after losing the signal.

 

The iTrain defaults  of 0ms on delay and 250ms off delay are generally acceptable and very rarely need alteration.

[ITG]

5 hours ago, WIMorrison said:

iTrain can and does work perfectly without any resistive axles, but they can be advantageous however there is also a disadvantage when you have a large number as they are all drawing current from the command station - perhaps only a few mA, but that adds up quite quickly. I personally know of 2 layouts that went overboard with resistive axles, and they needed boosters to compensate - the biggest advantage is reversing into sidings, and if you have stock that uncouples then they are beneficial in showing where the lost end of the train is - but easier (and better) just to sort out the couplings

Thanks, that’s exactly the pros and cons response I need. I quite like manual (as opposed to automated) shunting, so knowing the location of the rear of trains (by resistive  axles) is not a key requirement, when reversing in and out of sidings. Uncoupling is part of that same manoeuvre, and at present I’m happy with small tension lock couplings, and servo uncoupler ramps. I’ve spent some time adjusting the latter, and I’m getting quite accurate with placing couplings directly over them. So again, exact location of a wagon, other than by sight, is not key.

At present, most trains are max 4 coaches or goods equivalent, due largely to the overall size of the layout. As I mentioned before, it’s likely to be a new layout in an as yet unspecified and therefore unsized room, before I really get to grips with iTrain. The outcome of the room size options may or may not trigger longer and/or more varied train lengths, which in turn may make last wagon resistive axles more important.

Edited December 10, 2021 by ITG
Spelling correction

[ITG]

Probably a premature question relative to where I am in exploring the possibilities of iTrain, but I am drawn to ask it!

I understand that, for example, iTrain can drive a train departing from, say, platform 1 at station X to arrive at platform 2 at station Y (subject of course to the necessary blocks etc), but if one has a roundy-roundy circuit between the two stations, can you instruct iTrain to complete a set number of circuits before terminating the journey at station Y? And what if station Y is off on a branch and not on the circuit itself?

 

Whoops! Now I’ve thought of another question. How does iTrain deal with a reversing loop, where to complete the above instruction from station X to station Y, the train needs to negotiate a reversing loop to approach from the correct direction? (Assuming reverse loop is wired to accept the required change in polarity, eg by reverse loop module)

 

Ian

[WIMorrison]

Within iTrain there are 3 types of route.

 

The track route where you specify the blocks and turnouts to iTrain then drive the route entirely manually. Not really automation.

 

The train route where you specify the blocks and stations you want the train to visit with the station waiting times and any parts of the route that you want repeated with however many repetitions that you need.

 

Fully automatic  where all you do is tell the train to go and iTrain controls where it goes and when it stops (subject to any restrictions you have placed on the train).

 

the second, train routes, will do exactly what you want. Reverse Loops are just another bit of track and provide no issue or complication for iTrain 

[Michael Hodgson]

21 hours ago, ITG said:

Thanks, that’s exactly the pros and cons response I need. I quite like manual (as opposed to automated) shunting, so knowing the location of the rear of trains (by resistive  axles) is not a key requirement, when reversing in and out of sidings. 

 

 

There is another similar situation on the main line where resistive axles are useful.  If you run steam push-pull trains, when the loco is propelling the train will stop in the wrong place if it lacks resistive axle on the leading carriage - though I am pretty sure you can configure iTrain to handle the train correctly anyway.  My rolling stock has a right old mixture of tension lock coupling types, some of which do work together but are prone to separating, so I have a greater need than most for the rear of divided trains to continue to occupy the section.  So whilst I want resistive axles on a reasonable proportion of my stock, I can live with the fact that it will take time to do the whole fleet.

 

I endorse Iain's posts concerning the bus types the Z21 supports, also the advisability of separating traction and accessory buses.

 

One of the things I've not heard people discussing about automation of a layout is the way it affects the balance of what the operator has to do.  On a conventional layout, you can run one or two trains at once, above which you are trying to concentrate on too many things at once and make mistakes.  I don't want to be a train driver when I grow up, I prefer to just set the route and expect the trains to behave according to signals - that's why I liked the (analogue) "Superblock" controllers designed by a member of MERG - I could act as signalman and leave several trains to do their thing. That system has been described as signalman-centric rather than driver-centric.  With a fully automated layout, this workload balance changes again.  In theory you just switch it on and let it run itself.  However in practice you still hit a limit because there are still things thats have to be done.  Your task now becomes that of fire-fighting all the failures and problems  - ie recoupling divided trains, dealing with derailments, locos failing to start etc.  This is of course why reliability of wiring, trackwork, optimum choice of couplings etc is so important.  My long term goal is to be able to have the layout run itself to a timetable, and my interaction will then be to insert extra ad hoc services as & when I feel like it.

 

 

[WIMorrison]

2 minutes ago, Michael Hodgson said:

... My long term goal is to be able to have the layout run itself to a timetable, and my interaction will then be to insert extra ad hoc services as & when I feel like it.

 

 

 

I helped an 84-year-old implement this exact scenario using iTrain whereby trains are running automatically from his fiddle yard around the layout on a simplified timetable and at his main station he manually shunts and assembles trains between the scheduled trains.

 

He said to me a few months ago that he has been waiting for 40 years to get his layout running the way that he has always wanted, - just like it is now!

[ITG]

As I gradually work through the excellent YouTube videos, another questions occurs. Apologies if the answer arises in later videos, although it is more a question about physical construction than the software.

I note  Bob positions the on-screen blocks and feedback sensors, and inputs the length of the block. But in layout terms, if the block is, say a 150cm long siding, where exactly does the sensor connect to the track? Is it any where on that 150cm? Also, I guess for a siding the end of that block is generally going to be a short distance inside the entry point, but on open running track, what criteria / considerations exists for where blocks start/finish and how long they should be?

Thanks for your help - I am still studying rather than playing, learning rather than actions, but when things do get underway, I’m hoping watching the videos and asking these questions will ease the path forward.

Ian

[WIMorrison]

Ian

 

If the block is 1.5m long, and it has one feedback then you can connect anywhere within the feedback, if the block has 2 feedbacks then each feedback will have its own connection.

 

I am glad you are finding the videos that Bob and I made useful - but one tip, don't try to use YouTube as a support forum, better to post the questions on the iTrain forum or failing that here  

[Andymsa]

42 minutes ago, ITG said:

As I gradually work through the excellent YouTube videos, another questions occurs. Apologies if the answer arises in later videos, although it is more a question about physical construction than the software.

I note  Bob positions the on-screen blocks and feedback sensors, and inputs the length of the block. But in layout terms, if the block is, say a 150cm long siding, where exactly does the sensor connect to the track? Is it any where on that 150cm? Also, I guess for a siding the end of that block is generally going to be a short distance inside the entry point, but on open running track, what criteria / considerations exists for where blocks start/finish and how long they should be?

Thanks for your help - I am still studying rather than playing, learning rather than actions, but when things do get underway, I’m hoping watching the videos and asking these questions will ease the path forward.

Ian

 

generally I try to Position  single feeder wires to the sensors mid point that they detect, double sets of feeders are positioned more towards each end of the sensor section. The start and finish of a block which can contain as many sensor sections as you wish are determined by where physical signals are placed. As to length of a sensor section is a bit harder to explain but they need to be long enough so they can detect a vehicle without the sensor flashing on and off as detected and non detected vehicles pass through the section. But it is possible to set up sensors with delays so flashing doesn’t occur. There are exceptions to the length rule but is more to do with precise stopping of a train at a signal, many just have one sensor and the system will work out where to stop the train, the downside very careful calibration is needed of the locomotive to get precise stopping I use a second sensor section generally about 6 to 15 inches long this enables me to get each train to stop at a signal every time.

[Michael Hodgson]

Don't forget there are different types of sensor.

 

For example a detector that looks for current flow through the resistance of the motor (or an axle) will detect anywhere in the section.

A detector which works by interrupting an infra-red beam will only notice a vehicle standing or moving across the beam.

So position of the detector matters for the latter, but not the former.   iTrain works out where the train is assumed to be by reference to speeds, lengths etc after it has passed a sensor.

[Andymsa]

58 minutes ago, Michael Hodgson said:

Don't forget there are different types of sensor.

 

For example a detector that looks for current flow through the resistance of the motor (or an axle) will detect anywhere in the section.

A detector which works by interrupting an infra-red beam will only notice a vehicle standing or moving across the beam.

So position of the detector matters for the latter, but not the former.   iTrain works out where the train is assumed to be by reference to speeds, lengths etc after it has passed a sensor.

thank you for the reminder, but I was trying to keep things simple for the member who posted the question. 

[ITG]

4 hours ago, WIMorrison said:

If the block is 1.5m long, and it has one feedback then you can connect anywhere within the feedback, if the block has 2 feedbacks then each feedback will have its own connection.

 

4 hours ago, Andymsa said:

The start and finish of a block which can contain as many sensor sections as you wish are determined by where physical signals are placed

so, I think I’m reading that as there can be more than one feedback sensor in a block. Does iTrain then know where the train is relative to each sensor independently? In what I’ve gathered so far, I can understand that a single feedback in a block can detect the presence of a train, and can calculate its position using both train length and block length versus speed, all relative to when it entered the block. But with 2 feedbacks, placed at what seems to be imprecise distances from the start/end of the block, I guess what I’m trying to get clear in my own mind is - how does the second feedback in a block detect any different to the first, if it both are located at imprecise distances?

 

Mmm, I’m not at all sure I’ve explained that well! A classic case of not knowing what I don’t know!

[jpendle]

Assuming you are using current feedback devices, then each feedback is connected to its own isolated section of track, each of these isolated sections could then be configured as a block in iTrain or multiple isolated sections can be included in the same block.

 

e.g two feedbacks, each in a 3" isolated section of track adjacent to each other. This could be configured as two 3" blocks with one feedback each, or a 6" block with two feedbacks, the first feedback starting at 0" in the block and ending at 3" in the block and the second starting at 3" and ending at 6".

 

Regards,

 

John P

[ITG]

1 hour ago, jpendle said:

Assuming you are using current feedback devices, then each feedback is connected to its own isolated section of track, each of these isolated sections could then be configured as a block in iTrain or multiple isolated sections can be included in the same block.

 

e.g two feedbacks, each in a 3" isolated section of track adjacent to each other. This could be configured as two 3" blocks with one feedback each, or a 6" block with two feedbacks, the first feedback starting at 0" in the block and ending at 3" in the block and the second starting at 3" and ending at 6".

 

Regards,

 

John P

Ahah, I think I was assuming that a one block equates to one isolated section. I hadn’t twigged there could be two such isolated sections in a block,

thanks

Ian