How many times has this behavior repeated? If only a hand-full of times, I have to question whether your observations are statistically significant. If that is the case, you could just have a weak breaker out of the box. I've seen it happen.

What is the motor rating (on the nameplate0 of the dust collector?

Motor loads have their own special NEC Article.

The chief difference is that conductors are sized based on the motor full load running (FLR) current, and breakers are then multiples of the FLR, often by a couple hundreds of percent depending on the breaker type.

The motor contains its own thermal overload protection.
The breaker is just short circuit protection of the feeder wires to the motor.

It sounds like the starting current (Locked Rotor Amps, LRA) is larger than the breaker can tolerate.

The large rotor being started up from a stop prolongs the start-up surge significantly.

ionized: We have run the cyclone probably a couple dozen times. EVERY time this is the result, without any variation. First attempt - cyclone runs; second attempt - breaker flips. We have tested the Unplug Between Attempts Method probably a dozen times and it starts every time, no variation in the results. I understand statistics, and the likelihood of this exact same failure detail happening 20 times in a row -yet not being significant- is nil. But hey, I can fully appreciate healthy skepticism :) The bad breaker idea sounds very possible, I just dont understand why the load is different between attempts or unplugging.

brickeyee: The machine has a 2hp (14A) induction motor. If the problem were the machine, it seems like it should trip the breaker on any of the other identical 20A 220V dedicated circuits (which it does not). We have 3hp induction motor machines (table saw and shaper) that have never once tripped their breakers. We have not tested these other machines on the dust collector's circuit due to mobility.

Buying another breaker should not break the bank. Or you could interchange the conductors between breakers in the panel to put the dust collector on a different breaker. Per Table 430-52, that dust collector could be on a circuit protected by a 30 amp breaker.

The time delay curve of breaker is pretty far from uniform.

There is significant variation slowed from breaker to breaker.

If it works on one breaker but not another, try exchanging the breakers.

I agree with brickeye that the difference between the receptacles, is probably a difference on the delay curve of the beakers to each receptacle.

I also agree with bus driver, even without knowing the LRA, that it appears it should probably be on a 30 amp breaker.

In addressing the first time/second time issue of start-up, I'd speculate that the start-up capacitor, on the the first attempt is bringing in a large current rush, but on the second attempt it has still retained some charge and now has a lower inrush current.

You probbaly do not have to change the conductor size to use a 30 amp breaker for a motor only load.

You should review article 460.
SHOULD BE ARTICLE 430.

Even the current ratings on the motor nameplate are not used.

The NEC has its own horsepower to full load current table in the article that MUST be used.

This post was edited by brickeyee on Fri, Feb 1, 13 at 10:38

"You should review article 460."

Brick, are you sure you didn't mean article 430-6?

Is this on a Square D Homeline by chance?
I won't go into why I'm asking right now, just curious if it is.

"are you sure you didn't mean article 430-6?"

Article 430 is correct.

Office is packed and memory is failing.

Well we tested the circuit, wiring it to different breaker (known to work) and everything was great - so we bought a replacement. So far so good. Thanks to everyone for your help.
hexus: Yes it was (and still is) a Square D Homeline. Spooky...how did you know (and should I be worried).

"hexus: Yes it was (and still is) a Square D Homeline. Spooky...how did you know (and should I be worried)."

LOL! I KNEW IT!

I've had VERY similar issues with Homeline's. So much in fact that I stopped using them. I used to have several temporaries that I built using Homeline's and I've installed them on a few jobs. I ALWAYS would have issues with motor loads on them (compressors, saws, etc...)
I've had Square D reps swear to me up and down that Homelines are exactly the same internally as QO's minus the little red flag and the indicator window, but I've never had issues with QO's like I have with Homeline's.
I'm sure for most people Homeline's are fine and work perfectly, for me though I couldn't afford all the call backs I was getting with motor loads on them.

Please note I'm not trashing them or bad mouthing them at all, I've just had very similar issues with them on motor loads and couldn't afford all the calls backs I was getting.

hexus: Well good to know. We have another workshop room that has a compressor and vac pump. If we have this issue again, I will go with a different type.

Sounds like they have a tighter than typical inverse-time overload curve.

On anything except motors that would be a mark of better quality.