Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Powering turnouts with servos

14424 views
22 replies
1 rating 2 rating 3 rating 4 rating 5 rating
mlf
  • Member since
    February 2017
  • 18 posts
Powering turnouts with servos
Posted by mlf on Tuesday, May 2, 2017 2:32 PM

Has anyone tried the method linked below to power your turnouts?  If so, how did it turn out? I'm new to model railroading and exploring inexpensive ways to power turnouts.

http://mrr.trains.com/how-to/dcc-electrical/2014/04/operate-turnouts-with-servo-motors

 

  • Member since
    January 2009
  • From: Bakersfield, CA 93308
  • 6,526 posts
Posted by RR_Mel on Tuesday, May 2, 2017 3:01 PM

I’m currently experimenting with SG90 servos (eBay $1.50) to operate my turnouts.  The SG90 has plenty of power and the price is right.  The trick is to keep the motor current low enough so the motor isn’t damaged.
 
The servo can draw as much as 100ma at 5 volts depending on the load.  The motor gets hot at a constant 35ma.  The servo motors drift with pressure against the throw arm without power applied.
 
Mel
 
Modeling the early to mid 1950s SP in HO scale since 1951
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
 
mlf
  • Member since
    February 2017
  • 18 posts
Posted by mlf on Tuesday, May 2, 2017 3:17 PM

Yeah, that was my question.  I guess the motors would constantly be in stall mode, correct?  Don't see how that wouldn't damage them.  Also, would they buzz and be noisy in stall mode?  Keep me updated on how it goes for you.  If it works, I would like to do the same thing.  $1.50 for the servos is great.

  • Member since
    January 2009
  • From: Bakersfield, CA 93308
  • 6,526 posts
Posted by RR_Mel on Tuesday, May 2, 2017 4:12 PM

I didn’t go into detail of the severo operation.  It takes 50 ma for movement and as high as 80ma to operate a micro switch. The motor gets hot at well under stall so the motor can’t be used in stall mode or the motor will be a goner in short order.
 
By using micro switches the servos will work OK but that requires some construction and dinking around with as well as running up the cost.  For proper operation the SG90 servos will require in the vicinity of 80 to 90ma during movement to operate the turnout and micro switches.
 
One of my experiments has been using an Arduino to control the servos but I’ve still run into the high idle current problem.  Constant high current means motor failure.
 
I haven’t given up yet, I still plan on using servos to operate my turnouts.  
 
 
Mel
 
Modeling the early to mid 1950s SP in HO scale since 1951
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
 
  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Tuesday, May 2, 2017 5:25 PM

 Do NOT do what that article had, servos just aren't made to stall constantly. You need a proper controller. However, it's easy to make your own controller. There is a series of articles on MRH by Geoff Bunza on using Arduinos for simple (and very cheap) servo controllers. The article explains all the information you need, you don;t need to knwo how to program, or even wire circuits, just follow the instructions and use the code provided. Mel and I are both workign on our own designs, mine's a little fancy with relays to controlt he frog power and all. I have no plans to sell anything, this is a hobby, not a business, but everything I design will be freely available under open source license - both hardware and software, so anyone who wants to will be able to have PC boards made and buy the parts and put them together, and then load my program and go. There are also commercial options available, on my previous layout I used Tam Valley's controllers for my servos (so this will be my second layout with Servos - I will not go back to Tortoises, and many of those new Tortoise competitors that work the same way draw far more curren than a Tortoise, which takes away one of the side benefits of the Tortoise, being able to wire some LEDs in series with the motor to indicate position, keeping the contacts free for other purposes.). The servos themselves are under $2 each. There are a million and a half ways to attach them to the turnout throwbar. The best option I've seen so far is a nifty mnount made for boats, unlike many of the ones that are 3D printed plastic or cast resin, it's METAL yet only about $2.50 each. There are many DIY methods, one uses some aluminum channel that happens to be the same with as the 9G servos, so I guess that would be an even cheaper option.

 I wasn;t totally sold until i bought by first controller and tried it out. After that I immediately ordered the rest to finish my old layout, and I am totally sold. It's just about the least expensive way to have powered turnouts, and is, depending on how you source the controller, even cheaper than some of those fancier hand throws like the BluePoint.

                        --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    January 2009
  • From: Bakersfield, CA 93308
  • 6,526 posts
Posted by RR_Mel on Tuesday, May 2, 2017 6:04 PM

Randy are you using the Tower Pro SG90 servos?  I bought 20 of them off eBay and their quality control leaves a lot to be desired, but for $1.41 each I guess I can’t expect much.
 
 
 
They all have one thing in common, any pressure against the throw arm draws high current.  With out a load the servo works great, with as little as 2 grams of back pressure against the arm it raises the idle current 15ma and higher.  If I remove the data control to the servo the idle current drops to 9 to 12ma.
 
The motors get very hot (30 to 35ma) with constant pressure with the data control attached.
 
I bought a Turnigy Servo Tester for checking out the servos, all my test have been using it at 5 volts.
 
 
Mel
 
Modeling the early to mid 1950s SP in HO scale since 1951
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
 
  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Tuesday, May 2, 2017 6:42 PM

 Yes, I have a couple dozen of them. On my old layout I tuned each Singlet controller so there was no buzz, and they didn't get hot. I know when I am testing them on the bench I can get them to spike my bench supply to 1 amp if I grab hold of it while it is trying to turn.

 If you look at most of the mouting schemes, they typically use an over-center action, so the servo just has to move far enough to get past the center point, and the geoetry of the wire then serves to hold the points without applying back pressude to the servo - and quite the opposite of Trotoises, you cna use much thinner piano wire, the idea being that it is stiff enough to push the points (and my sample tests are being done with unmodified Pecos - so they still have the point spring in them - I will probbaly remove those for th eactual layout) but then bend once the point hits the stock rail with some bending force less than the servo's capability - so the wire bends instead of the servo stalling. It take very little pressure to hold the points closed.

 I've noticed there is a pretty wide variation in the amount of travel from 0 to 180 on the servo library, some won;t go all the way to 0 and just stall and draw high power, likewise at the 180 end. So I just set my limits well back. I forget which size piano wire is on my samples, but it's strong enough to overcome the Peco spring just a little past half throw, the rest of the distance is just bending the wire.

                                     --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

mlf
  • Member since
    February 2017
  • 18 posts
Posted by mlf on Tuesday, May 2, 2017 7:07 PM

Thanks Randy.  That what I thought when I read that article.  A servo in constant stall mode can not be good.  I will look at the Tam Valley controllers, and also read the Bunza article you mentioned, and then decide which direction to go.

  • Member since
    June 2011
  • 158 posts
Posted by Old Fat Robert on Tuesday, May 2, 2017 7:12 PM

I find it quite interesting as I read the various posts on this forum, particularly in regards to how we allocate and account for our hobby dollars. You gents seem (and that is certainly okay) pleased to have found one economical way to power your turnouts, albeit after some "tweaking" here and there. I prefer the relative security of known devices. On powered turnouts I use the Tortoise machines which I can install in a matter of minutes but at a higher dollar cost. Part of the wonder of the hobby.

 

Old Fat Robert

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, May 3, 2017 6:39 AM

 There's no more tweaking involved with a servo than with a Tortoise. Physical install is practically identical. Music wire actuator up through a hole under the throwbar. The fancy Tam Valley controllers do all the limit setting themselves, and like a Tortoise, once installed they tend to not go out of adjustment. Also handy in double deck situations since a servo is a fraction the size of a Tortoise.

 With more than 100 powered turnouts planned, even saving $10 per turnout becomes a significant cost savings. As it so happens, in putting together the BOM for my circuit design, it comes out to about $10 per turnout all together, but that includes the buttons and lights. That is for the one with provisions for remote control and remote lockout via a CMRI type control system. Ones for yard turnouts and so forth save about $2 per unit, $1 per turnout by leaving off components that aren't needed - the most expensive of which is the stinking connector. A 1 cent each I'll probbaly just leave the 4 resistors I could eliminate just so the unused inputs to the micro are pulled up and not floating.

 I'm cheap elsewhere too - using caulk to lay the track, a $1.19 tube is enough caulk for about 100 linear feet of track AND roadbed. And biggest savings of all, very few of my locos cost over $40, and that's for P2K, Atlas, Bowser level stuff. I am VERY patient with eBay. Even the more expensive ones I have - never paid more than about half price for things like PCM T-1's or an Atlas Gold Trainmaster.

              --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2006
  • 297 posts
Posted by markie97 on Wednesday, May 3, 2017 7:41 AM

I used the SG90 motors with Peco code 83 turnouts and momentary contact push buttons directly wired to the motors. With the Peco over center spring you do not need constant torque on the motor. I believe that I used two 9V power supplies. One button is +9V the other -9V. 

Test the motor beforehand. The throw distance is approximately 180 degrees. Set the actuator arm about halfway through the throw motion. This way the motor will not be actuated to its limits when throwing the points. 

I intertwined piano wire(use a heavier gauge wire) through two holes in the plastic actuator arm. This way the throw arm and actuating wire rotate together to actuate the points. I used two sided tape to stick to bottom of layout. So far they seem to work pretty good. Make sure surface is clean before sticking to bottom of layout.

  • Member since
    January 2009
  • From: Bakersfield, CA 93308
  • 6,526 posts
Posted by RR_Mel on Wednesday, May 3, 2017 8:00 AM

markie97

I used the SG90 motors with Peco code 83 turnouts and momentary contact push buttons directly wired to the motors. With the Peco over center spring you do not need constant torque on the motor. I believe that I used two 9V power supplies. One button is +9V the other -9V. 

Test the motor beforehand. The throw distance is approximately 180 degrees. Set the actuator arm about halfway through the throw motion. This way the motor will not be actuated to its limits when throwing the points. 

I intertwined piano wire(use a heavier gauge wire) through two holes in the plastic actuator arm. This way the throw arm and actuating wire rotate together to actuate the points. I used two sided tape to stick to bottom of layout. So far they seem to work pretty good. Make sure surface is clean before sticking to bottom of layout.

 

I think I understand your servo mounting but a picture would be a great help if you have one.
 
 
Mel
 
Modeling the early to mid 1950s SP in HO scale since 1951
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, May 3, 2017 10:33 AM

 That is definitely a way to do it without a specialized controller. The momentary buttons are the key. Talk about a cheap switch motor. 

 One of the linkage schemes I came across uses paper clips. Cheaper than music wire.

                        --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    January 2009
  • From: Bakersfield, CA 93308
  • 6,526 posts
Posted by RR_Mel on Wednesday, May 3, 2017 11:37 AM

That works great with Pico turnouts, unfortunately I have 18 Atlas Custom Line turnouts.  I converted an Atlas turnout to Pico type a couple of years ago and it also works great.
 
I prefer the Pico snap spring over the Atlas anyway but removing and reinstalling 18 turnouts would be too much for my 80 years old body.
 
Insolently the Pico mod to a Custom Line turnout works very good.
 
 
 
Mel
 
Modeling the early to mid 1950s SP in HO scale since 1951
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
  • Member since
    June 2011
  • 158 posts
Posted by Old Fat Robert on Wednesday, May 3, 2017 12:41 PM

Randy: I did not mean in any way shape or from to imply that you are "cheap". I actually think what you are doing is not only economical but is also valuable to us all as we look for ways to enhance our enjoyment of this (or another) hobby. I am sorry if my post offended you.

Old Fat Robert

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, May 3, 2017 3:03 PM

 I'm not offended - I'm the one who called myself cheap, because I AM cheap. I'm a PA Dutchman. Like my prototype railroad, we are notoriously frugal. Except maybe my car, but then again, by buying it used for half the price it was new, I get to drive a much nicer car than I could otherwise afford.

                  --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2006
  • 297 posts
Posted by markie97 on Wednesday, May 3, 2017 6:38 PM

Here's a short video of the units I described above in operation. Again these are with Peco code 83 turnouts.

By the way, Peco sells an adaptor that will allow allow other turnouts to lock in place. Not sure if they'll work with Atlas but it may be worth looking into.

https://www.youtube.com/watch?v=Tyxa3nqnbhY&feature=youtu.be

  • Member since
    June 2011
  • 158 posts
Posted by Old Fat Robert on Wednesday, May 3, 2017 10:23 PM

I understand perfectly. That is why I drive what I drive. I enjoy every moment in that car.

Old Fat Robert

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Thursday, May 4, 2017 6:34 AM

 Yup, and I don't care if people think I'm snooty because of it. Or other words which typically get assigned to drivers of my brand, which are not fit to post here. Too bad, it's a great car to drive and even just driving the 2-3 miles to work it puts a smile on my face.

                    --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

mlf
  • Member since
    February 2017
  • 18 posts
Posted by mlf on Thursday, May 4, 2017 7:14 AM

I read the article by Geoff Bunza that Randy mentioned above about building a DCC decoder that will control up to 17 servos.  That's what I would like to do.  The article is here:

http://model-railroad-hobbyist.com/node/19446

My question is this though:  I plan to use the NCE Powercab DCC system.  I assume this DCC decoder would accept input from the NCE Powercab and each servo would be connected to the decoder.  I would prefer though to control the servos from a control panel with pushbuttons or toggle switches.  And maybe control them from the Powercab controller also if I wish to.  Can this be done?  If so, would I have to get some sort of add-on to the NCE system like the NCE 524230 Mini-Panel Control for DCC Decoders?

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Thursday, May 4, 2017 7:28 AM

 He has other ones, in the November MRH, that use buttons to control the servos, instead of it being a DCC decoder. You don't get 17 servos per Arduino that way (have to use some of the pins for the buttons) but since the Nano costs less than $5, no big deal. With 2 buttons per servo (norma and reverse) best you can get is 6 servos and 12 buttons on the Nano. If you don't mind toggle buttons (each time you push it, the turnout changes the the opposite position), you can do 9 servos and 9 buttons. The November 2016 MRH has a companion download (also free) that has more details on the circuits he has in his article in the issue itself, which also includes an article on what you need to do to get set up to program the Arduino.

                                            --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

mlf
  • Member since
    February 2017
  • 18 posts
Posted by mlf on Thursday, May 4, 2017 8:05 AM

Thanks.  So i guess if i have 12 turnouts i would have to build 2 of his decoders?  Have you or anyone you know built one of his decoders and used it?  

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Saturday, May 6, 2017 10:23 AM

 I wired up one (as in, one servo, one switch, one LED) part of his and it worked (not the DCC decoder version, the one with the switch or button control). Not that I would have expected any issues, all of the heavy lifting is done by Arduino libraries that have been thoroughly vetted over the years so the code really fall in to the category of "if button 1 is pushed, set servo to position 0. if button 2 pressed, set servo to position 180"  My design uses effectivel the same code, I just added some extras like having it also toggle a relay to control frog power and some LEDs to indicate position. Since mine has so much extra stuff, each one can only control 2 servos. But it's still not a bad deal, the microcontroller chip is less than $2. The most expensive part of my circuit are the connectors to attach the external components like the buttons, LEDs, and the power supply. I'm looking for some alternate sources. 

                   --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!