A jam burnt the motor driver board?

Oh, so you could fry your shield by pulling tight a bit to long? I hoped for there was some sort of stall detection.

It does it for 3 seconds… keep clicking it over and over then, technically, yes, I think you can fry it.

1 Like

I wonder if it would be possible to have a sacrificial part i.e. before the chip is likely to burn out, a sacrificial sleeve between the motor shaft and the sprocket (with larger sprockets) gives way. Maybe easier with a plastic shaft extension.

1 Like

I would put that

on the motor shield and call it fuse. :grimacing:

3 Likes

My other idea was to have motors with keyways so that the grub screw would shear first…a fuse sounds much smarter, especially if replaceable, as I suspect it would be more reliable, and easier to link to the actual burnout current.

Is there a place where we are collecting suggestions for a future board revision?

On Github you create an issue and that will be then tagged as “enhancement”

Edit: The link

Cool/Thanks - done: https://github.com/MaslowCNC/Electronics/issues/50

1 Like

In the mean time, putting an in-line fuse holder into one of the M+/M- wires for the left and right motors wouldn’t be too hard.

not directly, but we do have a patch floating around that detects that the
maslow is at max power and trying to go even faster. Unfortunantly, this
condition is being triggered too frequently to be reasonable to add as a warning
right now (let alone something to stop the machine) becuase we don’t yet have
acceleration planning, so every time we start moving we are in that condition.

There’s also the problem that we don’t control the router, so you can end up
stopping the machine while the router is still spinning, and the spinning router
can then start a fire with the friction of it’s bit against the wood.

or a resettable fuse? (the price is right at less than a dollar per) It would take figuring out what the current threshold would be, and might not actually be possible to provide a fuse that wouldn’t interfere with cutting operations, but I think we should look into it. Does anyone have current data on on normal motor operations versus a wrapping event?

It doesn’t control position, it just applies full power to the motor(s) and
detects via the encoder when they are not moving.

the right fix is to upgrade the chips on the controller board so that they can
handle the worst-case situation, not just the normal-case situation.

If the motors are supposed to draw 2A, use a chip designed to handle 5A, it’s a
couple bucks more, but FAR less likely to burn out.

Would a PPTC resettable fuse allow for the momentary higher currents on acceleration but then still blow if the high currents were sustained e.g. if locked up?

Two things on that point:

  1. if the chains get bound up, the situation is the same (i.e. the router running but not moving), with the added bonus that the motor control shield is potentially overheating to a dangerous level.
  2. The router can be controlled, so perhaps the motor stoppage could be detected and linked to stopping the router (and the two being a package deal, only implemented with automated router control).

So, either way, it seems like a good idea to see if it can be implemented.

The time to trip can be chosen (from among options) and range from a couple hundred milliseconds into the 10s of seconds.

2 Likes

pretty much the same current, just sustained for a long time.

I really think the “can’t keep up” logic is the answer for detecting this, but
even if you detect it and stop the motors, you may still start a fire from the
spindle rubbing against the wood where you stop.

How close is “pretty much”? I am just curious if it has been quantified to the point that a threshold could be identified. Also, not sure if anyone has any data on a motor stalled by being bound up. Given your previous comment about

How often and how long is “full power” applied? A resettable fuse with a long trip time might meet those needs.

This is a good point. Of course, then there is the question of motor damage due to running the motor in a stalled condition. Since all the windings wouldn’t be sharing the current load, there might be some change in the efficiency of the windings affected. The gears should survive if the current stays within the motor ratings, but providing more current through a higher rated chip means that the motor might suffer electrical damage. So, then the question is, do we save a $5 chip and risk a $20 motor?

That risk is not mitigated by not stopping the motors, though.

That is exactly what I was hoping for. During calibration we could use the same loop to check if the chain is feeding. I that possible?

that won’t stop the chain from getting wrapped up, just detect that it wrapped
up to the point that things won’t move.

the “can’t keep up” logic will detect this as well, as the desired position will
keep getting farter away from the actual position, and so the machine will keep
trying to increase the power to the motor to catch up. detecting that we are at
full power for ‘too long’ will catch this case where the motor isn’t actually
turning pretty queickly.

1 Like

Will that logic detect the change in actual vs real position in a a chain wrap situation before the chain binds the motor? It seems like it would not. In which case there is no difference between detecting the encoders not registering movement and the logic detecting a position variation at full power. A sensor that detects the chain looping under (or in the case of a bottom feed, over) the sprocket is the only way I can think of to get ahead of a chain wrap, that said, stopping the motors from trying to continue movement, in whatever form that takes, seems like a good idea.

1 Like

So here’s a concept for a chain guard that mounts directly to the motor mount for 3D printing. Not sure how much thread we have to play with, but I’m sure we can source longer screws if needed. This should keep the chain from wrapping. Also, I added a slot that is directly in line with the center of the chain sprocket so you can put the lip of a tape measure into when you are measuring distance between motors. This is a rough sketch based upon a photo and not using actual measurements. Thoughts?

I drew it in Visio (below… I shaded red with transparency so you could see the mount), saved as svg, imported into Fusion360 and extruded. Easy Peasy.

9 Likes