Tram line plan

[gavino200]

7 hours ago, inobu said:

 

In any case I stated enough and I don't have anymore to contribute here. 

Inobu

 

 

 

Inobu, your contributions and perspective are always both welcome and invaluable. I'd be very grateful if you'd continue to give your two cents.

[gavino200]

9 hours ago, kvp said:

It's enough to throw the turnouts, they will take care of the polarity. (just power one of the loops only) this makes the wiring more simple.

 

TCS sensor pinout:

-open collector output for direction A

-common ground

-open collector output for direction B

(pullups must be in the control logic, for DCC tie Aand B together)

 

First paragraph: makes sense. Simplicity is good. Have you done this yourself or is it theoretical. Not necessarily a deal breaker. Just curious.

 

Second paragraph: You're talking about DCC just out of interest? Plan A is DC. It would be nice not to need to convert a tram collection, for such a simple line.

 

BTW, what processor are you suggesting? 

 

Do you see any drawbacks with the Azatrax system?

 

How do you have your own personal system set up?

Edited February 13, 2018 by gavino200

[gavino200]

11 hours ago, Kiha66 said:

While looking at sensor options, I found Azatrax already makes a circuit for this track plan!  Seems they beat me to the idea.

http://www.azatrax.com/reverse-loop-system-2.html

 

I'll have to study this. But it looks good. I think Jeff also recommend this option. 

[gavino200]

8 hours ago, kvp said:

I think this might be more relevant

(my setup):

 

 

Sorry, I missed this. Thanks for posting this. That's a great setup you're using. I agree, the Tomix TCS is very nice. I don't want to get such a powerful and versatile system and only use a tiny fraction of it's capability. It's totally overkill for my needs.

[gavino200]

10 hours ago, kvp said:

 

ps: Many modemos could go through 103 mm but some you have to file a bit from the frame and they growl loudly while grinding their gears, imho it's not worth it. On the other hand longer multisection tomytecs bind at their articulation below 140mm and simply derail.

 

That is a pity. A real downside. But the 103R looks great and would be so much better for fitting a street plan. It's a genuine balanced trade-off.

 

Can you give me an example of some of the trams that work on 103 and some that don't, so I can evaluate the trade-off further? Is there any (pre-purchase) way to predict which Modemo trams will have trouble with the 103? I could probably live without the multi-section trams. I assume by 'multi' you mean three or more.

Edited February 13, 2018 by gavino200

[kvp]

12 hours ago, gavino200 said:

Sorry, I missed this. Thanks for posting this. That's a great setup you're using. I agree, the Tomix TCS is very nice. I don't want to get such a powerful and versatile system and only use a tiny fraction of it's capability. It's totally overkill for my needs.

I only got two TCS sensors and no central control, so this was really cheap. The control boards (the two tiny PCB-s at the front) are a low active input Set-Reset flip-flop on the right (with 10 kohm pullups and a led) and an L293D H bridge and the pulsed signal control logic on the left. The 12V to 5V converter IC is behind the Lego speed controller (which uses an LM317 linear regulator). This could be replaced with a single Arduino and a H bridge motor driver shield as it has all extras (pullups, voltage regulator, etc.) built in. The software is just that a low signal from one of the sensors sets the single bit state to 0 and the other to 1. If there is a 0 to 1 change, the motor driver ic pulses the turnouts to one direction, if there is a 1 to 0 change, the same happens with reversed polarity. This is all that you need to have automatic turnouts and changing polarity on the two loops. You can use an analog throttle or a DCC decoder or even full DCC for the tracks, the control logic doesn't care about it. Speed doesn't really matter as the detectors work reliably from 1.5V DC on the tracks and that's more than the minimum to get the trams moving. (btw. getting a good smooth high frequency 12V pwm throttle really helps with slow speed running)

 

11 hours ago, gavino200 said:

Can you give me an example of some of the trams that work on 103 and some that don't, so I can evaluate the trade-off further? Is there any (pre-purchase) way to predict which Modemo trams will have trouble with the 103? I could probably live without the multi-section trams. I assume by 'multi' you mean three or more.

All of my modemos have trouble, so this is a simple answer and most Tomytec trams with more than two bogies won't work on 103 mm. Good examples are the 3 and 4 bogie ones. There is a single 3 section variant that has 2 bogies only, so that actually works. Imho it's not worth it to limit yourself for a 37 mm difference.

 

ps: I have a few modemos, one dauphin tram, 4 kato /pocket line/ and at least one from each tomytec tram motor type (with multiple shells). Also many high floor tomytec interurbans that are actually smaller than some of the trams. And a few bandai shorties, both articulated and normal and two full sized bandai tramway locos. Everything is reliable on 140 mm and about 30% of them can't run on 103 mm, while around 50% runs well. (the rest are grinding gears but somewhat work or don't derail every time)

Edited February 14, 2018 by kvp

[railsquid]

On 11/02/2018 at 1:01 PM, railsquid said:

Mine run round the Tomix 103mm curves fine, the Kato Portram is even rated for 90mm (!), Note I haven't tested the Tomytec 3-part tram mechanism, IIRC it might have a larger minimum radius.

 

 

As I had nothing better to do...

 

 

 

In order of appearance:

• Modemo Tokyo Toden 6000 series
• Modemo Tokyo Toden 8800 series
• Modemo Enoden 100 series
• Kato Duewag Tram
• Kato Portram
• Tomytec LRT-04 tram chassis
• Tomytec TM-10 chassis
• Fleischmann BR212

Controller is an old Tomix one which doesn't have good low-speed control.

I suspect the LRT-04 chassis would be less happy running with an actual body on it.

 

[gavino200]

4 minutes ago, railsquid said:

 

As I had nothing better to do...

 

 

 

In order of appearance:

• Modemo Tokyo Toden 6000 series
• Modemo Tokyo Toden 8800 series
• Modemo Enoden 100 series
• Kato Duewag Tram
• Kato Portram
• Tomytec LRT-04 tram chassis
• Tomytec TM-10 chassis
• Fleischmann BR212

Controller is an old Tomix one which doesn't have good low-speed control.

I suspect the LRT-04 chassis would be less happy running with an actual body on it.

 

 

Awesome! You rock, Squid!!

[kvp]

11 hours ago, railsquid said:

I suspect the LRT-04 chassis would be less happy running with an actual body on it.

Yes mine keeps lifting a bogie on r103. But judging from the slowdown, the bare chassis is binding in the curves too.

[railsquid]

It was definitely sticking a bit on the tram track curves. The section of tramline which which will feature the 103mm curves is operationally only capable of handling the smaller trams anyway so not a problem, was just curious if it would go round at all...

[gavino200]

Time to re-activate this project. Both these track loops are temporary. I'm not even close to having these areas of the layout modeled or even designed. But I'm starting to acquire DC locos and I want to have somewhere to run them. So I'm going to set this up as a simple double loop track connected by a straight section. There's a small chance that I'll change one loop for a simple fork or blind end, if I can't work the loop into a decent looking layout. But I'll try my best to do so. So after I re-read this entire thread I'll have some questions. Hopefully I'll be able to buy components soon.

 

 

[cteno4]

Gavin,

 

your aware you will need to do reversing loop wiring on these end loops, right!

 

jeff

[gavino200]

2 hours ago, cteno4 said:

Gavin,

 

your aware you will need to do reversing loop wiring on these end loops, right!

 

jeff

 

Yes, much of the thinking has already been done earlier in the thread. I need to read back through to refresh my memory and then carry it forward. I believe you made a contribution yourself at one point.

[cteno4]

cool, i couldn't remember if you had gone thru that. they are a little bit of a pain but possible to automate.

 

jeff

[gavino200]

I ordered the Azatrax system. I'll post a little review when I get it set up.

[Kiha66]

Cant wait to see how it comes out, be sure to share a video when you get it working!

[gavino200]

I finally had a look at this today. It's been here for about a month but I've been sort of dreading tackling it. But after wrecking my head over the diagram for a while I finally understand how to wire it. Eventually it should be easy to hide the sensors in some scenery or a building but for the short term I'll have to make some little cardboard frame for them. Also, I'll have to think about the most orderly way to route the wires under the table. 

 

Unfortunately, I forgot to buy Tomix insulated joiners, so I'll have to wait a bit to get it running.

 

 

[kvp]

If you have Tomix turnouts, then there is no need to isolate anything, just feed the power into one of the loops. The turnouts will handle the power routing and polarity switching automatically.

 

You can also leave out the power feed connections through the circuit and just wire the sensors and the bipolar turnout drive outputs (set to pulsed few msec mode).

[gavino200]

2 hours ago, kvp said:

If you have Tomix turnouts, then there is no need to isolate anything, just feed the power into one of the loops. The turnouts will handle the power routing and polarity switching automatically.

 

You can also leave out the power feed connections through the circuit and just wire the sensors and the bipolar turnout drive outputs (set to pulsed few msec mode).

 

I don't understand. In your first paragraph are you saying that all that is needed is to hook up the main power pack to one loop and the switches will do everything. No circuit boards. No sensors? In which case what's the need for paragraph two? What you say makes sense if there were only one loop. But with you'd still need to sense  and switch polarity.

 

Paragraph two is Greek to me.

Edited November 11, 2018 by gavino200

[kvp]

7 hours ago, gavino200 said:

 

I don't understand. In your first paragraph are you saying that all that is needed is to hook up the main power pack to one loop and the switches will do everything. No circuit boards. No sensors? In which case what's the need for paragraph two? What you say makes sense if there were only one loop. But with you'd still need to sense  and switch polarity.

 

Paragraph two is Greek to me.

You need the circuit board and the sensors. But only to switch the turnouts. The turnouts will take care of the polarity if you put the standard feeder from the speed controller into one of the loops.

 

So set up the circuit with the sensors and the turnout control. Set the turnout control output to only briefly pulse the turnouts with power, otherwise it will burn them out. Add the standard feeder to one of the loops. That's all.

 

ps: You only don't need the track isolation and the polarity switching part of the circuit as Tomix turnouts are fully power routing and will switch the polarity when thrown by the sensor circuit.

Edited November 12, 2018 by kvp

[Kiha66]

I disagree, I would very much caution relying on the turnouts to switch polarity.  If one throws slower or fails to throw without using isolation then you could have a short and burn up the control module.  I would recommend using isolation as the diagram shows and not try to cut corners.

[kvp]

2 hours ago, Kiha66 said:

I disagree, I would very much caution relying on the turnouts to switch polarity.  If one throws slower or fails to throw without using isolation then you could have a short and burn up the control module.  I would recommend using isolation as the diagram shows and not try to cut corners.

If you only feed in on one loop, then it's impossible to create a short with the turnouts as there is only one feed point in the system. The Tomix turnouts are fully power routing and if one turnout fails to throw, then the polarity will be wrong and if it happens in the loop opposite of the feed point, then the train just starts to go in reverse. This is the worst case scenario. If the control module is not connected to the traction power at all, so there is no danger of damaging it. Also the throttle has short protction in case of a derailment.

 

ps: The trick is that Tomix turnouts are fully power routing and have built in isolation on both the inside and the outside rails in both directions. Adding another pair with an external polarity switching circuit is both redundant and could actually create a short over the new isolators if the turnout's polarity switching circuit is not in sync with the external polarity switch. The required polarity switching circuit is built into the Tomix Finetrack turnouts. Any other brand or the old 3 wire Tomix turnouts would require the external circuit and the isolation.

[gavino200]

6 hours ago, Kiha66 said:

I disagree, I would very much caution relying on the turnouts to switch polarity.  If one throws slower or fails to throw without using isolation then you could have a short and burn up the control module.  I would recommend using isolation as the diagram shows and not try to cut corners.

 

I agree with this approach (for me). It's just about the edge of my electrical capability to follow the given diagram as is. If it doesn't work I can then rely on tech support. 

[gavino200]

4 hours ago, kvp said:

If you only feed in on one loop, then it's impossible to create a short with the turnouts as there is only one feed point in the system. The Tomix turnouts are fully power routing and if one turnout fails to throw, then the polarity will be wrong and if it happens in the loop opposite of the feed point, then the train just starts to go in reverse. This is the worst case scenario. If the control module is not connected to the traction power at all, so there is no danger of damaging it. Also the throttle has short protction in case of a derailment.

 

ps: The trick is that Tomix turnouts are fully power routing and have built in isolation on both the inside and the outside rails in both directions. Adding another pair with an external polarity switching circuit is both redundant and could actually create a short over the new isolators if the turnout's polarity switching circuit is not in sync with the external polarity switch. The required polarity switching circuit is built into the Tomix Finetrack turnouts. Any other brand or the old 3 wire Tomix turnouts would require the external circuit and the isolation.

 

kv, I very much appreciate your responses. I don't doubt that the circuitry you outline would work for you. I think perhaps you don't fully appreciate the light years of distance between your understanding and mine. I guess thanks for giving me the credit. Alas, it isn't due.

[kvp]

22 minutes ago, gavino200 said:

kv, I very much appreciate your responses. I don't doubt that the circuitry you outline would work for you. I think perhaps you don't fully appreciate the light years of distance between your understanding and mine. I guess thanks for giving me the credit. Alas, it isn't due.

Essentially, what i'm saying is that you can leave out the isolation, the traction power feeds into the circuit and from there into the tracks. Just connect the regular controller to one of the loops. Only set up the turnout control and the sensor parts of the circuit and it will work. Half of it is not needed because of the Tomix turnouts that have everything else inside of them, including all circuits to prevent shorts.

 

If this doesn't work, you can still add the isolators, the feeders and everything else, but they are completly not needed. You only need the 4 sensors, the main control circuit and the turnout control outputs. The track power part could be left off.

 

IDEA: You can try out the Tomix power routing without connecting up the circuit. Just connect your controller to one of the loops. Then connect the two turnouts to a single common Tomix manual control switch. Throw the switch manually before the tram, while it's moving forward at a steady pace. You will see that all polarities are matching and the tram keeps moving forward as long as the turnouts are thrown in time before the tram.

 

If this works, you can see that the optical control circuit is only needed to throw the two turnouts and nothing else need to be connected to it. (the circuit in the video i linked above does that, without isolations or messing with track power, just throwing the two turnouts at the same time back and forth)

Edited November 12, 2018 by kvp