Tomix New Operation Control System (TNOS)

[Vato]

think is much cheaper then install in every train dcc decoder

[kvp]

think is much cheaper then install in every train dcc decoder

The funny thing is, that for automation, you will need all the track sensors and an automation central unit (computer), so you would have to add the same costs to the cost of the decoders. If you want automation with block based traffic, then DCC doesn't really have any benefits over the Tomix cab control with the analog PWM+CL drive. Add Tomix train and station sound units and you get better quality audio much cheaper than with on board DCC sound chips.

 

BTW: Is there any information about any Tomix New Control System controllable block signals? The spare parts seem to have turned up for them on HS, but i can't find any official product announcement. It would be natural to have block/start/entry signals for an automated layout like the one in the video? Is there anyone here who could directly ask Tomytec about this?

[Martijn Meerts]

So, what this really is, is just an analog block system?

[kvp]

So, what this really is, is just an analog block system?

Yes, it's an analog block system with a digital control logic and for each block a pulse width modulated drive with a constant lighting signal. This allows proper speedup and slowdown of each train, while the wheel sensors provide the position feedback. Turnout control is provided and i really hope signal control will be added in the future as the information is already there in the control computer. The analog 'cabs' are digitally switched around to follow the trains too, so if you manually take control of a train, then you will be able to control is regardless which block is it in. (It would be very easy to generate a DCC signal with the block drivers to allow full DCC operation or even mixed traffic with DCC and analog trains chasing each other, so it's just a question of the control software.)

 

On the video, i only see two wires between the daisy chaied i/o control boxes, so i assume the system uses some kind of 1 wire serial bus, very similar to the loconet data/ground bus, while power is provided to each box separately.

[Martijn Meerts]

Yes, it's an analog block system with a digital control logic and for each block a pulse width modulated drive with a constant lighting signal. This allows proper speedup and slowdown of each train, while the wheel sensors provide the position feedback. Turnout control is provided and i really hope signal control will be added in the future as the information is already there in the control computer. The analog 'cabs' are digitally switched around to follow the trains too, so if you manually take control of a train, then you will be able to control is regardless which block is it in. (It would be very easy to generate a DCC signal with the block drivers to allow full DCC operation or even mixed traffic with DCC and analog trains chasing each other, so it's just a question of the control software.)

 

On the video, i only see two wires between the daisy chaied i/o control boxes, so i assume the system uses some kind of 1 wire serial bus, very similar to the loconet data/ground bus, while power is provided to each box separately.

 

Sounds good.. Miniworld Rotterdam uses a similar system (it's called Dinamo, designed by a Dutch guy, not much info available in English I believe), which already does mixed analog and digital and allows computer control etc.

 

Definitely a good way to go if you have a lot of trains. I'm not really sure why I actually never considered using Dinamo instead of DCC when I started building my layout, would've saved a ton of work since I wouldn't have had to install decoders :)

[kvp]

Sounds good.. Miniworld Rotterdam uses a similar system (it's called Dinamo, designed by a Dutch guy, not much info available in English I believe), which already does mixed analog and digital and allows computer control etc.

 

Definitely a good way to go if you have a lot of trains. I'm not really sure why I actually never considered using Dinamo instead of DCC when I started building my layout, would've saved a ton of work since I wouldn't have had to install decoders :)

First, i think Tomix won't add DCC support to this system, so it's up to the hackers to do it.

 

The main reason for DCC is a finer control over the trains. Essentially Tomix CL provides only a minimal F0 support, which is the directional head/tail light and constant internal lighting. With DCC, you can switch head/tail lights on individual units or even individual lights on a single end (often used to switch from 3 white headlights to 1 white on left or right for shunting) and allows turning on and off the lights in each car or even in each room on a sleeper. You can have analog sound, but besides basic startup/shutdown/stop/start/running sounds, that could be programmed based on analog voltage or pwm percentrage, DCC allows manual control over specific sound effects, very much the same way as the lights work. Station stop sounds (door open/close/etc.) have to be added to the platforms instead of into the trains. If you are in N scale and want to run more than one train on a block based layout, these aren't really needed and the Tomix alteratives are not only cheaper but easier to add.

 

Imho the best would be to use both, so one sound/light/motor decoder into each car with at least 8 light outputs, one motor output (optional) and sound with at least 8 channels. For example, for a 113/115 series emu, this would mean proper headlight control, controllable internal and destination board lights and compressor, motor and door sounds coming from each car. Stuffing all this inside is of course not easy in N scale, but for 0 or larger scales it's doable. The cost would be high, but for a 1 or larger scale train, it might worth it. If you have a digital train line (like susi) going through all the couplers, then one main decoder and several small function decoders could cut the costs down a bit. Imho this is where the DCC difference really comes out with japanese trains. Of course, you will still need the wheel sensors and a network to gather the info for proper signal and station stops, but that's not really a problem.

 

ps: Product idea: a drop in car lighting unit with a built in DCC decoder and individually controllable led lights, with an option to connect one or two polarity controlled head/tail light assemblies and/or a motor. The premium variant could have a sound decoder too, with a built in flat speaker.

[Martijn Meerts]

DCC obviously has advantages, but if you just want to run trains in a block system, something like Dinamo or the new Tomix system is plenty. Personally I feel that sound in N-scale is a waste of time and money, and for lighting I'm fine with just head/tail lights and interior lights.

 

However, on the 0-scale trains I have, the sound and various light options definitely add to the whole experience.

 

The question is, when does it get strange.. For example, in N-scale you have the moving bus system which is fun. But when I look at a reasonably large layout and only busses are moving, but no cars, it starts to feel a bit awkward. The same goes for passengers on platforms really, it's nice when they're there, and when a train stops at a platform and the doors stay closed, it's fine that passengers don't move. However, when the doors of a train open, it again feels awkward that passengers are static.

 

Of course, that's all personal opinion, others might think the sounds make things a lot more fun. That's the nice thing about building your own layouts though, you can do what you want with them :D

[railsquid]

The question is, when does it get strange.. For example, in N-scale you have the moving bus system which is fun. But when I look at a reasonably large layout and only busses are moving, but no cars, it starts to feel a bit awkward. The same goes for passengers on platforms really, it's nice when they're there, and when a train stops at a platform and the doors stay closed, it's fine that passengers don't move. However, when the doors of a train open, it again feels awkward that passengers are static.

 

Just imagine how creepy it would be if there were dozens or hundreds of little N scale people moving autonomously ;) Especially if they kept falling off the platform and crying for help in little tinny robot voices.

 

(I wouldn't be suprised if some sort of nanotec-based micro-robotry will make it possible in a decade or two...)

Edited August 9, 2017 by railsquid

[Martijn Meerts]

Just imagine how creepy it would be if there were dozens or hundreds of little N scale people moving autonomously ;) Especially if they kept falling off the platform and crying for help in little tinny robot voices.

 

(I wouldn't be suprised if some sort of nanotec-based micro-robotry will make it possible in a decade or two...)

 

That'd be a great story for a horror movie :D

[kvp]

Just imagine how creepy it would be if there were dozens or hundreds of little N scale people moving autonomously ;) Especially if they kept falling off the platform and crying for help in little tinny robot voices.

 

(I wouldn't be suprised if some sort of nanotec-based micro-robotry will make it possible in a decade or two...)

The obvious solution is platform doors. :)

 

Otherwise i think by gluing small magnets under their feet and using a printed circuit board coil based magnet matrix as the floor, it would be possible to move them around, even load them into the train by having the same matrix installed in the train cars. This would of course also require moving doors and a less than half a feet magnet wide gap between the platform and the doors. Doable with today's thechnology, just it would be very expensive due to the micro manufacturing required. There are however small pcb coil based car and train systems on the market, way smaller than N scale.

 

Personally i would be happy with a moving passenger car system in N.

Edited August 9, 2017 by kvp

[railsquid]

The obvious solution is platform doors. :)

 

Provided your layout is set in a time period after the year 2000 or so ;)

[Khaul]

Just imagine how creepy it would be if there were dozens or hundreds of little N scale people moving autonomously ;) Especially if they kept falling off the platform and crying for help in little tinny robot voices.

 

(I wouldn't be suprised if some sort of nanotec-based micro-robotry will make it possible in a decade or two...)

 

N-scale robots could be programmed to build the layout and drive the trains for you. You could have robotic avatar too :)

[Khaul]

By the way, I don't think the new Tomix control system cannot solve the problem stated here http://www.sumidacrossing.org/Musings/files/110313_Plans_for_an_Arduino-based_Tram_Controller.php. This point to point problem requires two independent throttles with inverse polarities. I don't see it in the default programs.

Edited August 10, 2017 by Khaul

[Kiha66]

I don't see it in the default programs.

I think that's the big question of this control method, if you're limited to the default programs shown in the videos then I'm not sure how helpful it will be unless your layout perfectly matches said plan. If they offer ways to expand the defferent operating options then I feel it will be much more of a threat to DCC conversion.

[Khaul]

The default programs, or modes, seem quite useful for operations in a loop. It's just a matter of how many stations, points and blocks you want to have. It looks like the system can be easily expanded to fit those needs.

[kvp]

By the way, I don't think the new Tomix control system cannot solve the problem stated here http://www.sumidacrossing.org/Musings/files/110313_Plans_for_an_Arduino-based_Tram_Controller.php. This point to point problem requires two independent throttles with inverse polarities. I don't see it in the default programs.

The original demo layout used by Tomix was similar, except it used two crossovers at the end stations and two bypass tracks at the middle. The linked problem is more simple. Btw. inverse throttle operation is shown on the video above.

[Khaul]

The original demo layout used by Tomix was similar, except it used two crossovers at the end stations and two bypass tracks at the middle. The linked problem is more simple. Btw. inverse throttle operation is shown on the video above.

 

Well, of course, inverse throttle is necessary to run a double track loop. Silly me.

 

Edit: I found the relevant video. It is the Suzukawa Ayako video linked by railsquid in page 2 of this thread. Two trains are running in opposite directions in the same loop. Good stuff!

Edited August 10, 2017 by Khaul

[velotrain]

Also, as we have seen, it is quite easy to hack/clone the programs in the 5563 unit . . . .

[brill27mcb]

On 8/9/2017 at 4:41 AM, kvp said:

Yes, it's an analog block system with a digital control logic and for each block a pulse width modulated drive with a constant lighting signal. This allows proper speedup and slowdown of each train, while the wheel sensors provide the position feedback. [....] The analog 'cabs' are digitally switched around to follow the trains too, so if you manually take control of a train, then you will be able to control is regardless which block is it in....

 

I have been re-reading this topic thread, and what kvp said above finally sunk into my head. As I have mentioned before, I run multiple trolleys/trams on essentially a dogbone-shaped loop. The trolleys all follow one another in a parade, and all run from one power supply at one voltage setting. I would like to automate this, and I have been thinking about the TNOS system to do this. Since it would cost about US$500 to have an 8-block setup that allows 4 cars to follow each other smoothly, and I would probably have to figure out how to use included programming software written in Japanese, I have been doing a LOT of thinking.

 

I now see that with TNOS there could be another potential advantage. Presently, at any one time, I only run trolleys that share similar voltage/speed characteristics, so that they follow one another smoothly. Based on what kvp said above, with TNOS I could probably set the control speed for each car (which follows it from block to block) and "speed match" any assortment of trolleys, from fast Tomytec to slow Arnold. Am I right on this?

 

A remaining problem, in addition to cost and language barrier, is that if I vary my layout size (and number of blocks) at different set-ups, I would need to change the TNOS program for each different possible number of blocks in the layout. There is also the issue of messy table-top cabling to central TNOS input/output hubs, instead of a more flexible and neater daisy-chain cabling using a homemade modular distributed control system, where each block and its control circuit would only be connected along the track forward to the next block (to receive detection input) and back to the previous block (to send detection output).

 

Rich K.

[kvp]

If you only need mainline block control, then something simple like the old Berliner TT Bahn automatic block control logic would be enough for you with stop-go relay logic. (parts are still available from Tillig)

 

For speed matching you'll need speed control that follows the train. Each control block would need one traction power output and one or two sensor inputs. And you would like it daisy chained instead of central like in the TNOS. Doable but not simple...

 

Btw. you could daisy chain TNOS io boxes with 4 blocks per box. Or use only one block in each box and use 4 times the boxes.

 

 

[brill27mcb]

Thanks, kvp! I am not familiar with the Berliner TT Bahn automatic block controls - I will have to research them. What do they use for detection?

 

Rich K.

[velotrain]

On ‎8‎/‎9‎/‎2017 at 8:28 PM, Kiha66 said:

I think that's the big question of this control method, if you're limited to the default programs shown in the videos then I'm not sure how helpful it will be unless your layout perfectly matches said plan. If they offer ways to expand the defferent operating options then I feel it will be much more of a threat to DCC conversion.

 

It seems to me that up to a certain point (two trains going the same or opposite ways on one track, with one or two stations - each with two tracks), you can achieve what the new system does with the TCS 5563.  This new system only starts to make a difference when you want to run multiple trains in succession on the same track.

 

I wish we'd stop seeing comparisons to DCC, as multiple people have already stated multiple times that its apples and oranges.  DCC increases your command over engines and can allow multiple engines on the same track at the same time, but still requires one or more operators to control each and all of them.  DCC by itself does not provide automated operation, which - as kvp and others have explained - would require a whole slew of additional components.  At the present time I see the new system as providing a gee whiz factor for solo operators who want to watch multiple trains on a limited number of tracks being controlled via automation, and for my own tastes you'd need a much larger layout than the Tomix demo one for this to be at all appealing.  I don't get the thrill of mindlessly watching a bunch of trains repeatedly starting and stopping and being routed into and out of station tracks.  Sure - it may be fun to watch for the first minute or two, but it gets old fast.  It makes a lot more sense on a larger layout where the stations are more than two meters from each other, with correspondingly longer blocks.

 

Sure, DCC can be used to adjust the speed on multiple engines so they can follow each other around a loop without collisions, but in it's fullest implementation is used to run a reasonably large model railroad in a manner similar to the prototype - which requires one operator for each train.

[kvp]

19 hours ago, brill27mcb said:

Thanks, kvp! I am not familiar with the Berliner TT Bahn automatic block controls - I will have to research them. What do they use for detection?

Threadle sensor tracks. Essentially the same wheel sensing system used by Tomix, but without galvanic isolation, so on contact the DC traction power was momentairly connected to the AC aux power. The relays were bistable self disconnecting types, so throwing one disconnected the selected coil supply to prevent the burning of the relay. This also disconnected the auxiliary AC power fouling the DC traction power after switching is done.

 

Each block had a free and an occupied state held by a bistable relay. One controlled circuit of the relay controlled the block's entry light signal (red or green), while the other controlled the power to the stop section before the signal. Sensors were placed next the block signals. Each sensor set a block further back to free and the current block to occupied. This allowed two blocks of red signals and active stop sections behind each train. (the one just passed and the one behind it to catch the train following the current one)

 

I assume Tomix TCS sensors used with a pullup and a driver transistor could switch the two bistable relays controlled by the sensor, as long as fixed DC power is used for powering the automation instead of AC. And Tomix TCS 2 color signal masts (without the bases) could be used to show the signal aspects. Stopping is a crude on/off, but also works with analog or digital (dcc) brake modules.

 

The only drawback is that train lengths are limited by block lengths and the whole system is single direction only. (double track solves this)

[katoftw]

Hello!  Those point to point programs make me happy.  Especially 9 and 10 with the double crossovers...

 

[JR 500系]

While the TNOS system seems mighty interesting, it can also give someone like me who have close to nothing knowledge about the currents and operating system a big headache... Would someone with minimum or no Japanese language proficiency be able to buy and use the TNOS system as per his requirement?

 

By the way, the TNOS system is out today~