Fantastic detective work here. I think we're about 98% confident we've found the problem (bad connector between node 10 and ramp motor). Add me to the list that just cut the connector out and soldered.

The long-term solution is a redesign of the node 10 board to add the (missing) protection circuits to the output of the motor drive chip. Short-term solution is make damn-sure the connection between the node 10 and motors is solid with no interruptions.

IMG_0589 (resized).JPG

Gomez comments are pretty much Stern's confirmation of what Peak-Pin deduced. Huge thanks. Bonus: easily fixed if you so desire. No need for that plug at all. Cut it out and hard-solder the wires. If later you need to remove board or motor, you can do so via the plug on the node 10 board.

I would point out another fact that reinforces that this fix solves the problem. Same board is used in Godzilla, with a motor that also moves a large and heavy part (the Drake building). AFAIK, there's been no reports of blown node 10s on GZ. This is supporting evidence that the issue was intermittant electrical connection to the motor (via the wrongly-sized and wrongly-installed crimp-on plugs w/Stern's solder-on-wires hack).

Exactly this. Well stated. We can finally put this issue behind us and enjoy the game with one less worry.

Some basic physics may be relevant here. Induced current (induction) is a function of current rise (amps/second) and distance (meters). A quick, sharp pulse on a wire close to another conductor can induce a current. This falls off as the cube of the distance, so it's very sensitive to things like crushed insulation. Wires in close proximity (via crushed insulation under a tight zip tie) are considerably more prone to stray currents.

The ramp motor is driving a considerably higher load than the clock and hence, surely has higher current loads. Its possible that stray currents from this kludged assembly produce issues with the motor drive chip if the wires are too-close. This is just speculation, but would explain why you might see problems with tight zip ties even after you've confirmed that resistance is nominal.

Just a thought.

In short, no. Not without seeing the internals of the motor drive chip. From an electronics standpoint, these are RLC (resistance, inductance, capacitance) circuits. There's well established means to tune these circuits for specific effects, like fast rise times or low induced currents. You'd have to look at these as a system (motor drive, cabling, and motor) in order to properly design this circuit. In theory, the chip designer has already done this. Clearly, Stern did not follow best practice and left various protection elements out of their design. They'll have to engineer an ultimate fix. Until then, we can apply common sense and ensure 1) that the motor connection is solid, and 2) try to limit noise/induced currents. Suggestions above serve both of these goals.

Tons of discussion about issues (or lack thereof) with the Molex connector and pins/crimps. I just removed the damn connector and hard-soldered the wires. Why would anyone go to any trouble at all with the connector? It's not like you can't remove the motor once you bypass the connector, it still connects directly on the node 10 board. The only reason I can think of is somehow keeping the game "original" or "stock". If that's your reason, great. Me? I'd rather just eliminate a known problem once-and-for-all.

For all the hassle that this connector presents (including people ordering tooling and spare pins), you still have a weak point in the system. We know definitively that intermettent connection can blow node 10. TMK, there's not been a single report of a blown node 10 on a game with the connector removed. There HAS been reports of blown boards on games with replacement connectors. Why not do everything to remove any possible weakness in this connection by eliminating that plug?

I have several spare stepper motors (from upgrading Prusa 3D printers), and these motors have the same specs as the LIN motors Stern uses for the ramp mech. Importantly, these motors have robust 20 gauge wiring (compared to the flimsy 24 gauge on the Stern motor). I was thinking of swapping. Would be good to know if alternate motors work. The key specs are number of poles and the degree of rotation per pulse. The Prusa stock motors are the same in this case. Then again, if its not broke, don’t fix it…

I doubt it’s low line voltage. The Meanwell power supplies are conservatively rated at 100-240 VAC. They’ll actually tolerate as low as 88 VAC and perform to spec.

https://www.meanwell.com/qa.aspx

If it's the growing concensus that the problem is external to the node 10 board and motors, then a possible solution might be an isolated power supply? Shouldn't be that hard to work up. Leave the communications in place, share a common ground, and wire up the node 10 to its own PS. For some of the folks doing testing, what are the power requirements for the board?