At last, we can provide you with an ‘at first hand’ info about the new elements of the Power Functions system:
The lights




Their full length is shown above. The lights are powered via a new-type-only plug, and a surprisingly huge brick that separated their wire, which has the exact dimensions of a classic 2×2 brick. Pity it has no electric contacts at all, otherwise it would be a way to connect the lights to the old 9V system. An advantage of the lights’ design are long, quite flexible wires that connect the forementioned brick to the actual lights. The lights can be inserted into any 1-stud wide hole, and they take up to about two thirds of a stud in depth, which allows to put a transparent 1×1 brick on the other side of the hole. The lights give strong, white light, that strongly brightens the transparent brick of any color, thus trans-orange brick may appear yellowish, etc. The lights’ design seems to protect them from any damage caused by the cables being pulled. It’s hard to tell how much power the lights need, but when powered from regular, not fully charged batteries together with a motor, they become apparently dim while the motor is running.
The switch



Not too big, it has a lever that can have an axle put through. The lever itself is very much like the one used in the PF handset, except it can swing roughly twice as much. Nice thing is a pole reverser on every switch, identical with the ones used in the handset. But there is a surprise too - unlike the PF IR receivers, the switches are sensitive to the direction in which the battery box switch is put in. It works like a pole reverser on them, which seems inconvenient (you need to keep in mind how to turn the battery box on in order to avoid all the switches working in the opposite way), but may have some use in future.
The linear actuator


Their range is not their main advantage, as it is merely 5 studs compared to the 4 studs of a traditional pneumatic cylinder, and 8 studs of two connected pneumatic cylinders. But the actuators make up for it with their excellent strength and accuracy - it takes 26 full revolutions to extend or retract them fully. Compared to the traditional Lego worm gears, which need 6 revolutions to extend by 2 studs, it means that we just got the most accurate Lego element ever: with accuracy of 5.2 revolutions per stud against 3 revolutions per stud of the regular Lego worm gear.
The speed of full extension / retraction of an unloaded linear actuator for various motors (in seconds):
- PF Medium: 5.6
- PF XL: 10.6
- NXT: 13.3
- RC (lower output): 1.7
- micromotor: 97.5
- 47154: 5
- 71427: 6.2
- 2838: 0.8
The actuators need their tips to be locked in order to extend or retracted - when let loose, the tip is simply revolving. It may be the consequence of some kind of an internal clutch that protects actuators from the damage when they can’t be extended / retracted further and they are still driven. We have been testing them with PF XL motors, and observed that when an actuator reaches its extreme position with the tip locked, something starts to snap rapidly - obviously something inside the actuator, as no snapping PF XL motors were ever recorded. It seems to be an intended behaviour, as the actuator is apparently not damaged, and continues to work normally as soon as the motor changed direction.
A pleasant surprise is to see two kinds of braces that come with actuators - one allows an articulated mount, the other a rigid one. The first type will be obviously more popular, but it’s a good thing to have them both.



Both braces are 3 studs wide, and need actuator to be driven with at least 3 studs long axle (the latter, rigid brace allows longer axles). With the articulated brace, the actuator can be driven via the new, shorter universal joint, or with two possible setups of the bevel gears (shown above). The older, 4-studs long universal joints can’t be used here.