Automation test layout

[MarshLane]

Thank you both for your replies. Servos were indeed the other option that I was considering, but I am not entirely sure whether there are any Loconet decoders for servos - am I right in understanding that the Megapoints boards just use the DCC bus?

 

Servos do have the advantage of cost and space, albeit subject to more work in setting them up (which might not be ideal given the number of points that I will be using) and the question over whether there are any Loconet compatible decoders for them.

 

One issue which I have yet to consider is whether, on an N gauge layout, Tortoise motors might be too large to accommodate the density of points required for this layout (i.e., too large to be spaced closely enough for crossovers, fans, and, in the fiddle yards, double slips and scissors crossings).

Hi James,

I think there is a big misconception with some people in the hobby that servos take ages to set up. They don’t. You centre the servo, install it, then hold one button down on the mega points board until it moves as far left as you want, the same for movement to right, job done. In my experience, they rarely if ever need adjustment. Also the speed can be adapted, so like tortoise they are slow moving and don’t have the impact of the solinoid motors.

 

No Loconet decoders needed for Servos. Just plug a connection from your DCC bus into the mega points DCC adapter and all is good. Everything whether on the Loconet connection or on the DCC bus will talk to them fine. Have a chat with Dave at Megapoints if you have any problems as he is really helpful.

 

Also remember that you can run servos in pairs for crossovers the same as you can tortoise, so worth taking into account as it will lower your cost a bit.

 

Rich

[jamespetts]

Rich - to clarify apropos servos and the Megapoints board - is that Megapoints board able to communicate with the base station using the Loconet protocol?

[Nigelcliffe]

Rich - to clarify apropos servos and the Megapoints board - is that Megapoints board able to communicate with the base station using the Loconet protocol?

 

No.

[njee20]

You also need to consider the cost of a multiprocessor board (meaning you only need one DCC module), otherwise you need a DCC module per Megapoints board.

 

Rich is right about combining outputs though. I throw scissor crossings off one Megapoints output - ie 4 servos,and they work fine. I agree they're easy to set up. I use aluminium channel as per Dave's recommendation, but I think next time I'd use 'proper' mounts with attachments for microswitches etc for ease.

[Andymsa]

I had a look at the mega points website, some interesting items, and I will investigate further if I can use any. One question it looks like you can only have 1 mega points controller when connected to DCC, if you require more than 1 points controller you need the multi panel. As the OP uses Traincontroller the two methods may be at conflict when using 2 or more points controller. My understanding could be wrong but this is how I understand it.

[Nigelcliffe]

I had a look at the mega points website, some interesting items, and I will investigate further if I can use any. One question it looks like you can only have 1 mega points controller when connected to DCC, if you require more than 1 points controller you need the multi panel. As the OP uses Traincontroller the two methods may be at conflict when using 2 or more points controller. My understanding could be wrong but this is how I understand it.

 

The Megapoints DCC module works with either one 12-way (or three 4-way) servo controllers,  OR via their multipanel giving 192 turnouts per DCC module. 

You can have multiple DCC modules on a layout, they are just an accessory decoder with a user-settable accessory address (and the subsequent run of 12 or 192 addresses for the servos).   So, no conflict with multiple DCC modules.   

( btw.  Using a Megapoints DCC module onto a single 12-way (or three 4-way) servo board will disable local switch input to that board - its one or the other (DCC OR local switch), not both, unless the Multipanel is used).

 

The one thing Megapoints won't do is feedback into Traincontroller.  Not an issue if treated as a "dumb" accessory device and only Traincontroller, or the DCC system linked to Traincontroller, alter the turnouts, because in those circumstances Traincontroller knows which way turnouts are set.  But, if buttons are fitted to the Multipanel board, then using those buttons will leave Traincontroller with an incorrect picture of the world.   

( For people not planning to use a computer or screen to monitor the position on turnouts (or run the trains), this is pretty much irrelevant - if using a MegaPoints Multipanel, with buttons and LEDs, then operating turnouts from DCC will also change the LEDs on the panel, and the user can still push the buttons on the panel.   But, if using a computer to monitor the state of turnouts/signals, or a screen linked to the DCC system (eg. iPad with turnout positions onto Roco Z21, or the display on an ECoS, and several others), then it is an issue.   )

 

 

I'm electronics advice for a large layout under construction which is using Megapoints stuff, including DCC interface.   I've also used LocoNet extensively on other projects. 

 

 

- Nigel

[Andymsa]

Hi Nigel,

 

Thanks for clarifying, you put my thoughts much more eloquently.

[MarshLane]

Rich - to clarify apropos servos and the Megapoints board - is that Megapoints board able to communicate with the base station using the Loconet protocol?

Hi James,

No but there is no need to.

 

Rich

[MarshLane]

The Megapoints DCC module works with either one 12-way (or three 4-way) servo controllers, OR via their multipanel giving 192 turnouts per DCC module.

You can have multiple DCC modules on a layout, they are just an accessory decoder with a user-settable accessory address (and the subsequent run of 12 or 192 addresses for the servos). So, no conflict with multiple DCC modules.

( btw. Using a Megapoints DCC module onto a single 12-way (or three 4-way) servo board will disable local switch input to that board - its one or the other (DCC OR local switch), not both, unless the Multipanel is used).

 

The one thing Megapoints won't do is feedback into Traincontroller. Not an issue if treated as a "dumb" accessory device and only Traincontroller, or the DCC system linked to Traincontroller, alter the turnouts, because in those circumstances Traincontroller knows which way turnouts are set. But, if buttons are fitted to the Multipanel board, then using those buttons will leave Traincontroller with an incorrect picture of the world.

( For people not planning to use a computer or screen to monitor the position on turnouts (or run the trains), this is pretty much irrelevant - if using a MegaPoints Multipanel, with buttons and LEDs, then operating turnouts from DCC will also change the LEDs on the panel, and the user can still push the buttons on the panel. But, if using a computer to monitor the state of turnouts/signals, or a screen linked to the DCC system (eg. iPad with turnout positions onto Roco Z21, or the display on an ECoS, and several others), then it is an issue. )

 

 

I'm electronics advice for a large layout under construction which is using Megapoints stuff, including DCC interface. I've also used LocoNet extensively on other projects.

 

 

- Nigel

Jon nogel,

Yes that is a valid point. If the OP wants specific feedback off the sensors then he will need to factor in micro switches and feedback decoders, but he same is also true of tortoise motors. There is switching available on the bottom contacts I think for the tortoise but the feedback decoder still needs to be considered.

 

However, as you say, if he’s leaving trainController to operate the points that shouldn’t be a problem. If not, then there are extra things to think about and factor in anyway!!

 

Rich

[WIMorrison]

Even using a tortoise or ip digital motor switch contacts the assumption is being made that because the motor has been activated the point has thrown and this can be a dangerous assumption - especially for hidden points.

 

The only way to be 100% certain the point has thrown is to have a switch attached to the tiebar and for this feedback to be used.

 

I agree that 99.9....% of the time the motor will have switched the point, but that isn’t and will never be 100%

 

It is also true that it doesn’t matter whether you are switching the motor through TC9, iTrain, an iPhone or any command station using DCC or Loconet as none of them know that the point has actually switched - they all assume that as they have issues the command the motor has activated.

Edited September 13, 2018 by WIMorrison

[jamespetts]

I am not entirely sure how important that feedback is - no doubt, it is better to have it than not, all other things being equal, but I am still unsure of: (1) how much effort and expense is involved; and (2) how important that it is on a layout of this sort (in which there are no hidden points, for example, but there are an awful lot of points in the fiddle yard).

 

However, the LocoNet vs. DCC bus for accessories issue is a different one. It does appear at present as if the LocoNet solution is potentially more robust. Is anyone able to point to any significant advantage of using the Megapoints Controllers and the DCC bus over using the CML units or the LocoServo units (as the case may be, depending on the motors used) with LocoNet?

[WIMorrison]

My personal solution would use the control program (TC9) to its full extent and have 2 separate DCC interfaces - one for the tracks and one for your points, you could even create an additional DCC bus for accessories such as signals and any other items.

 

This would reduce the complexity and allow you to use commercial components from Digikeijs throughout which will maximise the compatibility on the control side.

 

The choice for point motors is then more open, personally I have has great success with Cobaly IP Digital and given the number you need the price of a single unit that has the decoder built in would be very competitive compared to a motor and decoder - it would also simplify the wiring.

 

The KISS principle has worked very well for me professionally and in model railways for many years and I would recommend it here also

[Nigelcliffe]

Tiebar (or blade) position feedback is theoretically a good thing.  But in practice, for indoor scales, I don't think its needed given that motors are usually reliable.  I'm pretty sure its not used on the McKinley Railway, and they have a lot of turnouts which are hidden and traversed by multiple automated trains.

 

The only time I've used position feedback was for a level crossing.  That was because the gates took about a minute to complete their movement, so reporting back on completed movement on a control panel (run from a CML DTM30) was useful for the human operators. 

 

Decision between DCC accessory bus or LocoNet is, to my mind, how happy one is with the available devices.   For the main project I'd use one, not a mixture - ie. pick a turnout motor and its control electronics, and stick with that.  It makes future maintenance easier.     For an evaluation project, a mixture would allow the merits of different options to be considered.   I'd be happy to use LocoNet, its been tested over the years, and seems reliable.  Because I've used it for quite a while, I understand what's going on, know my way round the various configuration tools, and find the ability to know the status of devices to be useful if debugging is needed.  But equally, I've built things which use a DCC accessory bus, and those are also reliable. 

[Andymsa]

An advantage of using the cml unit is that it only involves 1 item, using the mega points system involves two items. So more items to check when fault finding.

[Junctionmad]

Even using a tortoise or ip digital motor switch contacts the assumption is being made that because the motor has been activated the point has thrown and this can be a dangerous assumption - especially for hidden points.

 

The only way to be 100% certain the point has thrown is to have a switch attached to the tiebar and for this feedback to be used.

 

I agree that 99.9....% of the time the motor will have switched the point, but that isn’t and will never be 100%

 

It is also true that it doesn’t matter whether you are switching the motor through TC9, iTrain, an iPhone or any command station using DCC or Loconet as none of them know that the point has actually switched - they all assume that as they have issues the command the motor has activated.

 

I think with very few exceptions , determining the point motor has moved is sufficient , as after all its only plastic passengers !! .  I think fitting physical point detection is rather excessive except in a very very few cases and in fact even sensing the tie bar isnt enough as many failures are tie bar failures especially in soldered pointy work

 

 

Dave 

[Fastdax]

( btw.  Using a Megapoints DCC module onto a single 12-way (or three 4-way) servo board will disable local switch input to that board - its one or the other (DCC OR local switch), not both, unless the Multipanel is used).

 

- Nigel

 

It is possible to mix 'n' match switch inputs and DCC inputs to the Megapoints 12-way servo controller.

 

What you need is a DCC accessory decoder board that operates relays. I did it here for Offerston Quay. The relays can then switch some of the Megapoints inputs.

 

On my layout, the Megapoints inputs that are relay-controlled also have a manual switch (normally left open) in parallel with the relay, to override the relay if I want manual control of the point or signal.

Edited September 15, 2018 by Fastdax

[jamespetts]

I note that MERG produce a servo mount with an ingenious positional feedback sensor using microswitches activated by the position of the arm connected to the servo motor. This is not completely failsafe, of course, as it does not protect against certain sorts of mechanical failure, but a model railway generally does not need to be completely failsafe as Dave points out.

 

Do I understand correctly from the LocoServo manual that this unit does have inputs for feedbacks?

[Nigelcliffe]

I note that MERG produce a servo mount with an ingenious positional feedback sensor using microswitches activated by the position of the arm connected to the servo motor. This is not completely failsafe, of course, as it does not protect against certain sorts of mechanical failure, but a model railway generally does not need to be completely failsafe as Dave points out.

I think the intention behind the microswitch mount was for a means to switch turnout crossing (frog) polarity. However, it could be used for feedback if needed. (see below).

 

Do I understand correctly from the LocoServo manual that this unit does have inputs for feedbacks?

 

The LocoServo has eight servo outputs and eight binary input/output lines.  You can do what you want with those eight lines (report turnout feedback, use a track sensor, operate a different output device, whatever is wanted).  So, if turnout movement reporting is important, use those eight lines for it. 

 

 

My experience is that if an instruction goes to a device, the device operates.  And, just about always (ie. so vanishingly small that its irrelevant), a decent turnout motor will always operate.   So, the feedback from a motor operating switch is an unnecessary complication.  

The more likely failure (and still small in a reasonably well designed arrangement) is in the mechanical link to the tiebar, or the fixing of tiebar to turnout blade (a somewhat more common failure place on scratchbuilt track, but again good design can reduce this failure to "vanishingly small chance").

 

 

So, I'd be using those eight binary I/O lines to either operate a relay to throw the crossing polarity (assuming I didn't use a microswitch on the mechanical linkage from motor to turnout), or I'd use them for something else, such as input to system from occupancy devices, or output to drive signals. 

 

 

- Nigel

[jamespetts]

Thank you - that is most helpful.

[NinOz]

The Tortoise motors can be mounted either way round (ie rotated through 180 degrees) if they're too close to another motor. 

As well as can be mounted off-centre to either side.

[Junctionmad]

As well as can be mounted off-centre to either side.

In my opinion of you are going for stall motors , the newer cobalt analog IP units are superior in almost every way to tortoises

 

For servos , I use a MERG derived servo board with integrated relay frog switching and integrated feedback

 

I would try and avoid using microswitch switching with servos as mechanically it can be messy to utilize the movement range of the servo and still switch the switches. The relay units are programmed to switch at the electronic midpoint

Edited September 20, 2018 by Junctionmad

[Andymsa]

In my opinion of you are going for stall motors , the newer cobalt analog IP units are superior in almost every way to tortoises

Can you give more detail as to why they are superior in every way to a tortoise. If memory serves you cannot throw the colbalt by hand to test the throw.

 

Personally I have 190 tortoises installed and rising, never had one fail yet. That doesn't mean they won't but I have found them very reliable

[Junctionmad]

Can you give more detail as to why they are superior in every way to a tortoise. If memory serves you cannot throw the colbalt by hand to test the throw.

Personally I have 190 tortoises installed and rising, never had one fail yet. That doesn't mean they won't but I have found them very reliable

The analog Ip has

 

Bigger voltage range , AC , DC upto 25 v

Gold plated 5A rated switches x 2 ( tortoise rated at 1A , not plated )

Disconnects power at stall, resulting in very low standby current <5 mA ( 15 mA tortoise )

Spring loaded connectors , don’t require soldering , easy of installation

Life time warranty ( 9 years for tortoise )

Slightly cheaper , 12 pack 165 , tortoise 175-185

More widely distributed this side of pond

Arm cant move when power is disconnected , so pressure is not released , this is a good thing as complete power down can be effected at end of travel , ie with push buttons , tortoise will release pressure slowly if powered down

 

Available as full DCC decoded option ( digital IP )

 

And the best feature , smaller

 

QED

Edited September 20, 2018 by Junctionmad

[Andymsa]

The analog Ip has

Bigger voltage range , AC , DC upto 25 v

Gold plated 5A rated switches x 2 ( tortoise rated at 1A , not plated )

Disconnects power at stall, resulting in very low standby current <5 mA ( 15 mA tortoise )

Spring loaded connectors , don’t require soldering , easy of installation

Life time warranty ( 9 years for tortoise )

Slightly cheaper , 12 pack 165 , tortoise 175-185

More widely distributed this side of pond

Arm cant move when power is disconnected , so pressure is not released , this is a good thing as complete power down can be effected at end of travel , ie with push buttons , tortoise will release pressure slowly if powered down

Available as full DCC decoded option ( digital IP )

And the best feature , smaller

QED

All valid points, but your last one personally is an issue for me, as the release of pressure on power down allows precise movement of throw to be adjusted, but that realise is minimal at best and I don't have open blades due to it. The tortoise is not really to be operated with a push button in any case. Spring loaded connectors, ummmm not sure about that I can see wires easily coming out if snagged and how good a connection is it compared to soldering.

 

But at end of day both do the same thing, is one any better than the other I guess it's on what you expect from them.

Edited September 21, 2018 by Andymsa

[Junctionmad]

But at end of day both do the same thing, is one any better than the other I guess it's on what you expect from them.

 

no , not really , some products are clearly an improvement over others , Tortoise in my view have rested on their laurels for far too long , DCC concepts have not . Im no fan of dcc concepts, most of their stuff is too pricey for what it is , but the cobalts ( and Ive disassembled both ) , particulary the analog IP is a better product then then tortoise