Created with Sketch. Created with Sketch.

Review of Cobalt Turnout Motor

Posted by Richard S. Johannes, NMRA HUB Division on Dec 31st 2014

Review of Cobalt Turnout Motor

Buy Cobalt Turnout Motors

When I started model railroading some 5 decades ago, the hot ticket was the Atlas Snap-Switch. These turnouts used inexpensive solenoids that echoed their name. They snapped audibly and the turnout followed suit. Solenoids improved and I eventually graduated to NJ international solenoids with two pair of contacts attached. This approach to turnout motion and wiring was the standard for a long time until the rise of the slow motion turnout motor.

I was introduced to these by the late Lorell Joiner and was wowed from the moment I first saw one. Circuitron’s Tortoise has become the de facto standard for this type of turnout motor and Tortoises can be found everywhere. The Tortoise, like its earlier solenoid predecessors, also has a pair of single pole – double throw (SPDT) internal switches that can be put to any number of uses. Personally, I use one pair to power the frog particularly on hand laid all-rail turnouts and the other to determine the position of the points. We’re all used to the whine from these green boxes found under so many layouts.

Well there’s a new kid on the block, the Cobalt Switch Motor. These devices still share many of the features we’ve come to expect in a slow motion switch machine. The Cobalt like the Tortoise has an external fulcrum pivot and two internal SPDT switches. It even looks a like a Tortoise although it’s smaller and cobalt blue instead of green. When Gary Paulino of Traintek, LLC asked me to evaluate and write review of the Colbalt machine I decided to do so by comparing several features, size, quietness, speed, ease of installation, the spring wire, power consumption and cost. Of note, both devices have options for single DCC Accessory Controllers attached as single unit. I’m not going to cover this issue as I’m taken by the fact that it would appear that single address accessory decoders must not take advantage of the economies of scale as the cost per turnout is invariably lower with multi-address accessory decoders than with single address decoders.


The Cobalt machine is definitely more compact in all directions. As can be seen in Table 1 below, all dimensions of the Cobalt are smaller than the Tortoise. The dimensions in Table 1 are for the size of “box” itself.

CobaltTortoise% Difference
Depth1 -1/4"

Table 1. Switch Machine Dimensions in Inches

Evaluating this is somewhat subjective despite the objective differences. I’ve definitely had places were an extra ¼ inch would have made a mountain of difference, allowing me a more standard installation rather than using some kind of extension or horizontal mounting. The footprint of the mounting surface (plus linkage clearance) is larger than external dimensions of the box for both machines. The Cobalt machine’s footprint is 2 inches by 1½ inches whereas the Tortoise’s is 2 inches by 2 inches a 25% difference. If the horizontal mounting flange were removed from the Cobalt machine this difference would increase to 34%. These machines will definitely fit into some places were a Tortoise will not.


First of all any comparison that involves sound level should make mention of the SwitchMaster. These machines are in a class by themselves when it comes to low noise. You really have to watch the points to know if the machine is active, they’re that quiet. Since the Cobalt machine markets itself as a quieter machine, I just connected one at its nominal voltages and ran it. I immediately thought it was quieter than a Tortoise but I must admit the degree of quietness didn’t strike me. The Cobalt’s tone is lower in pitch than the high pitched sound of the Tortoise. Not trusting my ears, I recorded the sound levels at a variety of voltage settings for both machines. Remember that a 3 dB difference is approximately twice the sound power. Most humans begin to have some problems with differentiating sounds in the 1 dB range. I found the Cobalt machine to have about a 2.2 dB difference which would be consistent with being quieter but not hugely so which is just what my ear suggested.

Slow Action

I tested both machines using a number of voltages. The Cobalt runs between 9-12V DC and Tortoise runs on anything less than 12. I tested both at 6V, 8V, 10V, 12 and 14V. I used two measures, one was the time of full excursion and the other was in cm/s. The reason for both is that the Cobalt, owing to its smaller frame, has a smaller total excursion, 0.50 inches versus 0.8125 inches for the Tortoise machine. By either measure the Cobalt moves slower. Figure 1 below shows the time to full excursion for both machines using a variety of driving voltages. The difference between the two is between 0.193 seconds and 0.803 seconds with higher voltages showing a larger difference.

Cobalt Switch Machine Voltage Time Relationship

Figure 1 Effect of Driving Voltage on Total Excursion Time

Ease of installation

The Cobalt has very short installation instructions printed on the back of the blister pack. If you’ve ever used a Tortoise they’re sufficient. The Manufacturer, DCC Concepts, refers you to the Web for more detailed instructions. There you’ll find a 21 page color PDF with a lot of useful information including the details of the internal switches which are laid out identically to the Tortoise. Circuitron provides a 2 page black and white instruction sheet packages with each machine. The approach to installing them is identical with four screws provided with both machines for vertical mounting. The fulcrum slides up and down in a virtually identically manner on either machine. However, a nice feature of the Cobalt fulcrum is that it has a series of 5 holes drilled in it rather than the single center hole found on the Tortoise fulcrum. This is no doubt useful for some tight and off-center installations. Perhaps the nicest feature of the Cobalt machine is its use of a built-in wire connector for the eight possible connections (2 power and two sets of 3 pins for the SPDT switches). This has always been an issue with the Tortoise. While there are third party slide-on terminal blocks for the Tortoise, none really fit quite right and there’s always a risk of shorting. Many just solder wire or Molex connectors to Tortoises and this does add to the difficulty of installation when compared to the Cobalt machine. The Cobalt Motor features a spring action device that allows you to put a stripped wire into the hole, just push back the release to allow entry, let go and poof your wire is attached. I do wish the slots were numbered as they are on a Tortoise.

Spring Wire

Both machines come with a three inch piece of spring wire. The Tortoise’s wire is thinner. It fit into my wire case in the #70 (0.028 inches). The Cobalt’s wire fit into the #68 slot (0.031 inches). I mounted each of them horizontally into a vise and added the same weights at 2 inches from end of vise. The Tortoise’s spring wire deflected ¾ in but the Cobalt’s wire only deflected ½ in. With smaller frame of the Cobalt requires a stiffer spring wire but I’ve felt the standard wire that comes with a Tortoise not stiff enough especially with hand laid turnouts. I long ago bought some phosphor bronze wire from Tichy Train Group and routinely replace the wire that comes with a Tortoise.

Power Consumption

I tested the stall currents for both machines and both matched their published ratings. The Cobalt drew 30 ma and the Tortoise 14 ma. While 30 ma may not seem like much current, you’ll be at ½ amp with only 17 Cobalts whereas it will take 36 Tortoises to draw half an amp. The Cobalt’s stall power draw is twice that of a Tortoise.

Not Tested

I did not attempt to test the longevity of the internal switches in either device. I know some people have had issues with the sliding switches in the tortoise failing. I’ve had this happen as well but only 2 in 10 years of use with about 20 machines. The Cobalt has gold plated contacts and they may be better but I’m not in a position to say. The documentation states that the Cobalt internal switches are non-shorting, but I didn’t construct a definitive test for this. The Tortoise documentation does not describe whether they are shorting or non-shorting but in use with frog power, they appear to behave as though they are non-shorting. Lastly, I did not test the force that will lead to stall across the range of voltages. However, I did test the ability of each motor to lift a fixed weight using the same distances between the motor connection and fulcrum and distance from the fulcrum to the weight. I used the same spring wire for both tests. My goal was to find a stall weight. Both machines were able lift 6.4 ounces. I stopped there because I felt any more weight was about to permanently bend the spring wire. While I don’t think I could say one or other generated more force, I think it is fair to say either can provide sufficient force to move and hold even a sticky pair of switch points.


Here’s the summary. (See Table 2). While the Cobalt wins 5 to 2 on check marks, the score is actually closer than that as I’m not sure each checkmark merits equal weighting. The final word for me is that if if you can handle the higher cost but you want a somewhat quieter easily installed slow motion switch machine to fit into a tight space that won’t hold a Tortoise, you definitely now have a new option.

Cobalt Tortoise
Smaller Size X

Quieter X

Slow Motion Action X

Low Power Consumption
Ease of installation X

Heavier Spring Wire X

Lower Cost

Table 2. Summary of Features

Buy Cobalt Turnout Motors