[Pdelcast]

I guess your question is: How do we make a more reliable controller, and what's the costs and tradeoffs involved?

Well, to start with let's talk about the motor. The motor your pallet jack uses is likely a fairly high resistance motor compared to the motors used for RC. Industrial motors are designed for continuous duty at low power density. The controller can rely on the fact that the DC resistance of the motor will limit currents to a survivable level for the MOSFETs.

Let me go into a little more detail. The current sensors on the board shows are hall effect sensors, which have a fairly low bandwidth (around 50khz or so) -- so it takes the controller somewhere around 20 microseconds for them to respond to an overload condition. The motor that is being driven probably has a resistance of around 40 milliohms, when you include the wiring harness it may be even higher.

The resistance of the motor itself would limit currents to under 800 amps or so. This is low enough that the low bandwidth of the current detection would allow time for the current detection to kick in before the silicon in the MOSFETs got so hot that they were damaged.

Now, on the RC side, we typically want very high intermittent power for short bursts with lightweight motors. The motors we use, consequently, are VERY low resistance and extremely high power density (the Mamba Monster motor is .007 ohms, or 7 milliohms typical.) At 24V, under stall, the Mamba Monster motor would draw around 3500 amps. At that current level, the controller would destroy itself in around 2 microseconds -- so the controller would already be dead by the time the current sense system would be able to respond to the over current condition.

We use a different scheme to determine over current conditions which is faster than hall sensors, but still isn't fast enough to catch every over current condition before damage occurs. The reason is the power density of the controller means that inherently, the controller is much more "on edge" than an industrial controller:

We want controllers that are small and light - - the industrial controller shown is probably, what, about 4 x 6 inches in size? That's 24 square inches of circuit board. It's also attached to a large heat sink (not shown.)

So, that that industrial controller has a power density of about 400 watts per square inch. The Mamba Monster has a power density of about 1800 watts per square inch. Our newest controller, the Phoenix-ICE-HV-160 has a power density of about 3200 watts per square inch.

These power densities mean that the hobby controllers are under considerably higher stress than their industrial cousins.

We also build industrial controllers and military controllers. But they are not controllers with power densities that even come close to the power densities of our hobby controllers.

It's possible to build more reliable controllers for hobby, but there would need to be sacrifices in performance -- higher resistance motors (lower power output, lower power density) significantly larger, heavier controllers with much higher price tags.