The Switch Control Topic!!

Oh boy this is exciting!!!

Cost of the Octopus

Hi Les,

If you are using the Octopus then you wouldn't need the fascia controllers, the relays or the lever switch (you use the other pole of the DPDT for the frog wiring. That will save you $42. It works out to about $18.11 per turnout. Alas, with the Octopus you don't get LEDs - the toggle indicates the direction of throw.

commercially produced servo based switch machines

http://www3.telus.net/ChinaRail/Smartswitch.html

Usual disclaimer:  I have no interests in this company other than being a satisfied customer.

Servos Article

Very interesting article.  Currently using twin-coil in hiddeen areas, but this will have to be considered as a replacement.  Thanks for the article

Octopus

 Thanks Duncan, it was not clear as to what parts were required. It is clearer now!

Also the price is right, thanks again for a great article.

Exactally The Type of Article That MRH will become Known for

Duncan this is a really fantastic article & I hope to see more of your electrical mechanical articles for model Railroading. Thank you for this information as I have a few servos but have never used them and never gotten into Model Flying so was unsure as to how to make them work as switch machines or how to use the servos for other applications. I had been wondering why I couldn't get the servos to operate with just a pot and 9 to 12 volts applied.

Joe This issue of MRH magazine has really been a treat. The fact I can load it on my Lap top and read it without any internet connection is great. In fact this article will make an excellent selling point to the advertisers as to the quality and need for MRH as an absolute must for the model railroader

Dan

So much more

As I was reviewing this and watching the animations, I kept thinking of more and more possibilities for the servos.  As a (former) programmer, this kind of thing intrigues me.  I definitely couldn't handle the assembly language, but doing things in BASIC shouldn't be much of a stretch, even though I do little programming anymore.  I certainly hope there are more articles like this and I've bookmarked the Tam Valley site already.  Thanks for wealth of ideas. 

A big thankyou

What a timely article. I have been won over by it and I'm heading off to the Tam Depot to order a couple of Octopus boards. Thankyou again Duncan

Ian Holmes

Microprocessors for model RRs

 Yes I'm interested in a series of articles how to use microprocessors for controlling and animating model railroads.

Joe in Colorado

Servos and microchips

Article was great.  Amazing to see how everything comes together.

As to Microprocessors, I'd love to learn more about what we can do with them and see how creative we can get!

Jason

Hobbyist microcontrollers

Duncan McRee's article is the kind of originality I truly appreciate in MRH!  What a terrific application.

For others who may be inspired to build / program their own controllers, folks might be interested in learning more about microelectronics and the use of simple boards like Arduino (http://www.arduino.cc/).  "Open source electronics" is a hobby onto itself, but it does open the door for anyone to create their own unique project.

I, for one, hope to see more articles like Duncan's in the future!

Many thanks,

Nate

Great article Duncan, I'll

Micro articles

It is good to see my articel has sparked so interest and enthusiasm!  That is just what I had hoped for. I am currently at Hartford with Joe and I will talk to him about doing some articles about microprogrammer programming for model railroad use.

Duncan

another article please

Great article!

You mentioned controlling a turntable with a modified servo, it should also control a transfer table as well.

I will be looking forward to an article on these modified servo projects.

Great article. I am

Great article. I am interested in seeing other articles on using servos and MP programming for MRRs, and also some notes on adapting this to LocoNet.

My current plans are to use BluePoint controllers but that could change. Even if they don't, I got some useful/interesting insights from the article.

servo controlled turnouts

I liked the topic and I am surprised that servo control of turnouts has not taken off commercialy in a big way.

I use the ANE produced servo control product which has a SmartSwitch and SmartFrog that handle 4 turnouts (No association just a user). It compares favourably with the Circuitron products for cost and has more control options and can be used with pushbutton panels or DCC at the flick of a switch. Very useful for local control panels. The four set comes with Smartswitch, four Smartfrogs, a decoder and four small servos with extension leads. There are four toggle switches and all the fixing screws etc for mounting the servo brackets and PC boards. The big advantage is as you say that the size is so much smaller and for shelf units is far better.

Mounting servos as you have shown has a possible disadvantage as the operating wire moves up and down and it requires the length to be carefuly cut to ensure that it does not come out of the throwbar if too short and does not foul the locos etc if too long. Mounting the servo with the drive vertical so it turns side to side overcomes this. A prefer the wire to operate against a fulcrum point as well as it gives better control stability.

Further articles on the electronics would be welcome.

Thanks for publishing your work.

Alan

servo controlled turnouts - Quad controller usage and queries

Hi Duncan,

I'm very thankful for you doing the article.

I have read the manual for the Quad Controller and several things come to mind.

I really don't like components getting hot so whilst a 780x or 790x has thermal shutdown I would much prefer to augment the heatsink and this looks reasonably easy due to where it is mounted - can you confirm?

I would also like to know at what point the frog polarity is changed - I was hoping it would be roughly at mid travel to reduce the problem of overthrow due to flex in the spring wire.

The relays used - are they latching? I guess the question is more one of current draw if all relays are on.

It appears to me that the servos consume power at all positions from your article. I expect that if the tension the point rail is held against the stock rail were to drop, then the servo would increase the tension due to the feedback of the servo. It's more a question of an experiment than anything else.

Also when do the relays trip. I have played with JMRI and tortoises and it is possible to signal unknown state when both tortoise switches are open. I expect that I would need a limit switch actuated at each end of travel to acheive this. Have you considered making the relays dpdt? This would be enough for - if its not switched left then it must be straight logic with JMRI.

I was also surprised that you set the travel limits by jumpers rather than CV's.

The processor is surface mounted. Is it possible to reprogram it on board?

Regards and many thanks,

Greg Palmer

Reply to Greg

It is easy enough to use a larger heatsink.  More important is that you can't block airflow around the heatsink as the heat has to go somewhere.

The frog polarity changes immediately with the relay method.  It would be possible to add a delay in the BASIC code for the Quad but that is not currently done. Polarity changes at the far end of the throw with the micro-switch method.  You might also want to look at the Frog Juicer -an electronic solution to the frog polarity available from Fast Tracks.

They are not latching but they are low current types.  They don't draw any more than an LED when in the hold position.

It is important not to use overly stiff wire to keep current draw down.  That's why I use .025" spring wire.  Servos are fundamentally a DC motor and they have the characteristic of a DC motor that they draw the most current when stalled.  If you are really set to draw teh least power out of the DCC system then you can connect the servo to a seperate power source (4.8-6 V DC).

Programming of stationary decoders is problematic in the sense that they can't be moved to a programming track like a loco.  Also there wasn't enough code space in the micro to put in this without cutting back on something else.  The jumper method works well if you don't need a strong wire to throw the switch.  It is possible to individually set each endpoint individually by setting constants in the code and downloading in to the micro from a PC with the Parallax SX-Blitz programmer.  Craig Bisgeier does this on his layout where he is using solid point turnouts.

Actually in the current model, the micro is now a 28-pin DIP - this allowed me to unify the kit and the assembled product.  However a DIP chip allows for upgrading the micro at a later date.  All of my boards can be easily reprogrammed through a 4-pin header labeled SX-Key surface mounted or DIP.  One of the reasons I like the SX chips is because they are easily reprogrammed in circuit.

Duncan

Quad Servo controller

Thanks Duncan,

I've just been looking at the V2 version which alleviates several concerns - power, and cooling wise. You did pick out all but one question in my previous message. Did you consider DPDT relays to allow one pole for frog and one for detection/primative signalling? DCC is pretty electrically noisey and to have the connector blocks for the detection not adjacent to the frog connectors would seem useful. Either that or the screw connector strip as on the octopus would seem useful in alleviating crosstalk in this area.

Regards,

Greg

Finally! An open mind!

I love this cross pollenation of hobbies.  I often find myself wandering the aisles of general purpose type hobby shops thinking about how I can apply some of the other materials out there with my model railroad.  I think this is a great illustration of what we can accomplish if we don't limit ourselves to the offerings of Walthers and Woodland Scenics.  Kudos to the author, and to Joe for endorsing such a creative approach to model railroading.

Lee

DPDT

Hi,

Sorry for the late reply but I have been watching the Tour de France instead of logging in to the forums...

I did build an add-on board for 4 3A DPDT relays.  It works fine.  You will need an external power supply as these relays draw a lot more power than the little signal relays I used for the original board.  Th add-on board connects into the holes on the Quad to give more real estate for the bigger relays.  The same cold be done with wires and terminal strips but the board is a more elegant solution.  The 2803 driver I use for the relays is quite versatile and provides a neat way to interface between 5V logic signals and higher voltages for relays, motors and so forth.

Another project I just finsihed prototyping is to add three push-buttons so that the end positions of eah servo can be precisely set - it works in the same way as the Dual 3-way (see the video on the Dual 3-way page at tamvalleydepot.com).  The relay slots are still there on this design.

Duncan

RC servos article

A great article.  I would be interested in articles on the electronics for operating the servos.

Ditto!

I'll second that request!