Has anyone used R/C servos as switch machines as per the article in the Model Railroader January 2011 issue? (Gut the electronics and connect the power directly to the motor.) It seems to me that these would be a cheap switch machine if they will stand the stall current.
I am interested in any information on this subject.
I think I am going to give it a try with some testing before I decide, but at $2 or $3 each, it seems I won’t loose much if it doesn’t work.
I use servos, but I use them the proper way, with a controller board. Still cheaper than a Tortoise. Servos aren;t meant to be stall motors, and I question the long-term durability in using them as such like in that article. Constantly powered and stalled, they draw quite a bit of current.
I installed a pair of old R/C servos that I had from my airplane days on a double ended crossover in a heavily used position on my clubs layout. I took out the electronics board and wired the motor directly. I do not use the servos as stall motors. The switches are Pecos so the switch spring and the gearing on the servos holds the points tight against the rails. The control is with push buttons at the switches location and power is from two 9 volt batteries in series with a center tap. One battery throws the switch reverse and the other throws it back normal. The installation was done over a year ago with no problems. I also put in a couple of micro switches controled by the servos incase we want to add signals at a later date.
Prior to the install I made a test unit with a Peco switch and threw it reverse and normal over 1,600 times over a few weeks. The battery was still indicating 8.6 volts when I took it apart to install on the clubs layout. If you have to buy the servos I don’t know if the cost is worth it instead of commercial switch machines. I’m working on installing several more with the servos I have and some from another club member so there is no cost now.
In low quantities it works out to about $26 per turnout for a servo, control board and mounting hardware. This is based on the Tam Valley Depot singlet system that I used. The control board is a DCC static decoder and panel LED/button system in one. So for that money you get everything needed to bring the servo under DCC control and display the output and manually control it.
The $15 Tortoise will still need the static decoder, and any switch and panel indicators to provide the same functionality. When I priced this out some months back there was considerable savings going the servo route and they are much smaller. Time will tell if they hold up?
If you just want to use toggles then the Octopus board is what you want, not the Singlets. That board controls up to 8 servos for $32. Uses toggles, pushbuttons, whatever you want. So $4/turnout, $2.50 each per servo, and the mounting bit is optional. My friend just gorilla glues his to the bottom of his plywood. I used the Motrak mounts, but he seems to have stopped production, they were quite a bit cheaper than the ones Tam Valley now sells. With the mount kit, $13.50 per turnout.
I was looking at all options, and the Tam Valley octopus III was the cheaper way to go if I needed the electronics to make it work. This works out to about $15 per turnout but includes the switches and LED’s for the control panel.
However, if I can gut the electronics, wire the motor directly, and get the right value resistors in place to limit the stall current and still have the torque needed to operate the turnout, then the cost will go down to about $6 - $7 per unit, including the resistors and switch to operate it with.
The Jan 2011 article says it will work. The real question is; how long will the servo last in stall mode with the power on. Anyway, I just ordered a few SG90 servos from China so I can test it out.
Of course there is always the possibility that one can use momentary switches for control and have no stall power on the servo once the turnout is moved to the desired position. Then the question becomes; Will it stay there? (Peco’s wouldn’t be a problem, but other types may.)
The servos I used were made by World engines for my radio controlled planes. They and the other servos are difficult to turn by hand so I think they will do OK with Atlas or other switches.
OK, I received a couple of SG90 servos and did some testing.
I gutted the electronics and wired to the motors directly.
They need about 4 volts to operate reliably and they draw about 300ma when stalled. If the voltage is left on when the servo is stalled, they start heating up shortly. It takes about a minute to start sensing the heat through the plastic case. (I didn’t let it go to see if it would burn out or how long it would take to burn out.) You can’t add a resistor in series to limit the stall current because the servo wouldn’t have enough current to move in the opposite direction with the resistor in place.
The servos that were used in the article did not draw as much current as the SG90’s that I have. They drew around 10ma. So all servos are different and each should be tested before using them as switch machines.
What I have decide to do is use them wired direct through a DPDT-CO(MOEM) toggle switch. I can use one supply polarity, and the power will be removed after the turnout is switched. You also can not short out the power supply as you could if two separate push buttons were used and you accidentally pushed both at the same time. Resistors will be added as needed to get the voltage down to 4 volts depending on the voltage of the power supply that I will use.
The next test will be when they are installed and connected on a working layout.