Atlas O has entered onto three-rail scene in a big way. The company inaugurated its O gauge product line by releaseing an exceptionally prototypical track system, a structure kit, a 33,000-gallon tank car, and detailed scale-length center-flow hoppers.

For its first locomotive, the company set its sights on an engine that would look at home on traditional or modern-era layouts, a model that would permit the firm to make variations with minimal changes in the tooling, and an engine with scale-like detailing that also offers top notch prototypical performance. That locomotive is the General Motors SW8/9.

Selection of the singl-stacked SW8 or the double-stacked SW9 is a smart one. This locomotive is a meat and potatoes switcher that would look right at home on a layout that models railroading anytime from the 1940s to 1990s.

The engine has scale dimensions and measures 11 inches long (44 scale feet, matching the prototype). Checking diagrams and prototype photos, the engine captures the major (and plenty of minor) detail items on the 1,200-hp prototype. Atlas O’s commitment to scale detail is apparent within a few seconds of opening the box.

The locomotive has a die-cast metal chassis, truck side frames, and engine hood. A die-cast metal hood on a diesel, you ask? Why not! The tooling is superb andy you can feel the quality in your hands. With smooth tooling and the construction, this engine feels as solid as a rock! Shake the locomotive and you won’t hear any rattling or loose parts. What a concept.

The plastic cab is very nice. It features two crew figures, but no cab light. Instead, it goes one step better – it has illuminated gauges. Yes, little circles of light shine from the control panel. Now that is something you have to see to believe.

The front and rear platforms have a terrific see-through grating, and some versions have a drop step for multiple-unit operations. The engine has delicate add-on handrails that you’ll need to apply, but this is easy and only takes a few minutes. The engine also features illuminated number boards, which enhance the model’s appearance.

The switcher has two directional controls on the belly of the fuel tank (which houses part of the motor). One switch, when on, sequences the engine forwar-neutral-reverse; when in the off position, it sequences in neutral-for-ward-neutral-reverse.

The other switch is a reverse unit lockout. Be advised that these recessed switches are very tiny and you may need the tip of a paper clip or a similar instrument to change their position.

The horizontal can motor is located in the center of the engine and drive shafts connect it to each truck, which gives the locomotive eight-wheel drive. The weight of the die-cast shell, combined with traction tires, make this one sure-footed freight hauler.

Drawbar pull was an impressive 2.1 lbs., which is more than 50 modern, free-rolling pieces of rolling stock, so you can throw this engine at your largest marshaling yard trouble spot. The switcher also has terrific coastability, so you can gently ease to a stop.

Engine performance was commendable. You can actually switch with this switcher. Work in the yard is more like the real thing rather than a session of bumper cars! The engine went through switches as slick as as a whistle, and numerous switching moves at low speeds made for pretty realistic railroading.

In consistent, steady running, we averaged 14.5 scale mph for a low-end speed. You can go through the back-and-forth motions of switching at lower speed, however, and performance was slowest when using a Lionel TrainMaster in the conventional mode.

Our high speed testing averaged 82.3 scale mph, which would be fast enough for you to slap a passenger consist behing and keep your layout’s commuter population happy.

We tested the locomotive on O-27, O-31, and O-42 track. The engine operated on all three diameters, though it looked best on O-42, and coupler swing was limited when running on tighter curves. Atlas O suggests running on O-36 or wider radius curves.

On curves O-42 or narrower the switcher’s couplers occasionally popped open. We traced the problem to the coupler button. On tighter curves the edge of the coupler button rides just below the inside edge of the locomotive’s pilot steps.

If a slight rise or dip in the track is encountered, the coupler moves upward and the button catches on the step, opening the coupler just as if it were above an energized uncoupling section of track. The only remedies are to either cut clearance notches in either the button or the step or re-lay the offending track.

The trucks mount a total of four pickup rollers, so there was always an adequate power supply to the engine, and lights were flicker-free.

CTT had heard some complaint voiced by model owners and hobby shop staffers prior to testing this model, and we thought it proper to address them. First was a concern that the unit would not properly reverse.

We tested an SW8 and two SW9s and had mixed results reversing with the Lionel ZW and TrainMaster, the MTH Z-4000, and the MRC DualPower O27. Though internally the same engine, the SW9s reacted better that the SW8 for some reason. The SW8 liked the MRC and Lionel transformers better that the MTH reverse button, while the SW9s reversed nearly 100 percent of the time with the Z-4000.

We asked Atlas O’s Jim Weaver about this. He suggests that owners just briefly touch their direction control instead of holding down the lever (or pushing the direction button). He said that some operators hold the control too long and the reverse unit cycles back to the original direction.

Another criticism involved that low volume of the sound system. Frankly, the CTT staff does not view this as a major liability since louder is not always better, and we also believe that there is a finite amount of diesel engine roar that most folks care to hear.

At mid-level, we could hear the engine on the other side of our workshop. When switching the volume was adequate, but for constant running, even when set at maximum volume, the sound system was easily overtaken by the noise of a train rattling along tubular track, even on a "sound proofed" layout using carpet padding as roadbed.

A more serious issue is the quality of the sound reproduction. It could have been better. The bell is terrific and the horn is good (but there was a noticeable cutoff at the end of the tone), however the diesel engine sounds were mediocre. The sounds were so static-filled that it was hard to determine if the locomotive sounds were an actual diesel recording or an electronic simulation of a diesel. This was the case with all three of the models was examine.