Pause Button for Shapeoko

  • Document it: check (well sort of, I do mention the Feedhold input in anatomy of a Shapeoko section)
  • Use it: it’s in my TODO list to finally wire it, and I need to go meddle with my controller wiring soon anyway, so I will (and therefore document it better, in the process)
4 Likes

Great! It seems that I was wrong claiming that VFDs limit output current based on parameter settings - thanks @LiamN! But, they should be able to “sound the alarm” and stop after a specified delay.

We’ll get to the bottom of how to talk to these HY VFDs sooner or later, I do not wish to let it beat me…

From RTFM:

PD124:
When the output current exceeds the over-torque detection level and also exceeds half of the set time of the over-torque detection (factory setting 1.0s), the over torque detection will indicate and the corresponding multi-function alarm contact will act. When it exceeds the set time the inverter will turn to self-protection (which sounds ominous…)

This is tied up with PD050 through PD053 where you get to choose which parameters each of the Multi-Function outputs acts on. Value 12 appears to be the over torque alarm.

I guess we could lower the motor current setting to something in range for what the Shapeoko chassis is capable of (for the 2.2kW spindle) and then, if you really jam something up it might trip and hold the feed…

There are also analog outputs PD054, PD055 where we can get a 0-10V analog output representing actual speed, motor current etc. However, I think just probing it via ModBus is likely to be more useful than those analog voltage outputs.

1 Like

If the analog motor current monitor works (does it?) all that’s needed to monitor motor current is a voltmeter like this one (which can simultaneously log 10,000 measurements). Motor current should provide a decent measure of cutting torque and force. (I’ve come full circle on that thinking).

Hmm,

I don’t think that’s the number we’re looking for.

I agree that at our load points on a 2.2kW type spindle we’re in the linear range with minimal non-fixed losses but we still have the reactive (90 degree phase shifted) magnetising current flowing between the VFD and the motor vector summed with the in-phase actual torque real power current with no realistic ability to separate them unless we can measure phase / power factor.

This is why people see about 2.0 Amps all the time on their 2.2kW spindles, that’s the base magnetising current and we don’t load the spindles up hard enough to really move off that stop. There’s nothing wrong with the readout on the VFD, that’s just how the motor is.

That’s what I’m thinking real power into the drive being read by an Arduino and doing some maths in C will help with.

With the possible exception of you, nobody here appears to be loading their spindles much. But these results from someone that does suggest that there is value in monitoring motor current anyway, especially at higher speeds and loads where other losses dominate. Since magnetization current is supposedly independent of load, subtracting the no load motor current at speed from the loaded current to estimate cutting torque should have value, especially at the higher loads when more of the spindle’s available torque is used. I think its a good idea to measure and log both motor currents and VFD input powers to weigh their relative merits.

2 Likes

I really only went 2.2kW to get the ER20 collet and be able to put 1/2 inch bits in the machine, time will tell how hard I manage to load the thing up.

I do need to understand your k factor charts though, they look interesting but I’m not sure what I’m reading.

And yes, I plan to ModBus monitor the output current and put a real power meter on the infeed to the VFD and see what I can learn about what the motor is up to, results beat theories and tend to mess with expectations :frowning:

I used the SFPF spreadsheet to calculate values for the chart from the cutting parameters shown in the posted video.

My Carbide 3D Nomad (old design) has what we call the “Oh-Shit” switch, and I have used it several times. I have a Shapeoko on order and will be adding one, this Off the Shelf Solution? interests me, assuming it does what I think it does.

So where are we supposed to connect a switch? I am guessing that this switch would short the two pins on the carbide board?

@LiamN OOPs, sorry! - I didn’t use the SFPF spreadsheet, I used the one posted (and added some more info) here!

@DanStory Do you just install a Normally open latching type switch to short those two pin on the board?
have you done this, can you please provide a little more info?

Also, then why can’t we use the e-stop pins as well. @WillAdams any toughs?

Folks can wire up any pins in Grbl which they wish to use:

We just feel that an e-stop should be a hardwired stop of everything, esp. the router.

1 Like

Why latching? Just a momentary switch. A true E-stop should cut all power.

I’ll second that,

In my opinion and if I was carrying out any sort of risk assessment for employees;

A software based “Emergency Stop” is not an “Emergency” stop, it’s a “If you’re not busy, the problem isn’t caused by a code bug, you haven’t crashed and when you feel like it, please, only the things you have control authority over” stop.

An “Emergency” stop is an external electromechanical system which is able to override the machinery in question and force a stop, thus the use of NVR switches and contactors to remove power in the case of CNC movement and spindle power. This includes shutting off a misbehaving spindle or VFD. In many cases this trigger circuit is also connected to interlocks such as the doors of electrical cabinets.

There are some exceptions to this external system method, such as conveyor belt systems where it is specifically counterproductive to drop power to the drive because you need the drive engaged to bring the belt to a halt quickly and to keep it stationary whilst just removing VFD power would not even stop the belt in some conditions.

There are many uses for a “Software Stop”, with it’s own button to help you stop the machine, it’s certainly easier than fumbling with the mouse for the Pause button in CC.

Those aren’t the uses for the “Emergency” stop button, on a machine I’m using I have a strong preference for that to electrically cut off the machine and stop everything, whatever the fault cause.

edit:

Perhaps it might be easier to call the software trigger pin on the controller something other than “E-stop” to avoid any confusion?

3 Likes

That’s taken care of: the E-stop header does not exist anymore on v2.4 boards (introduced circa 2017?)

1 Like

Is there still, and will there continue to be, (momentary?) “Pause” or “Feed Hold” inputs? It seems they would be useful to stop the feed more quickly (and safely) in certain circumstance - like alarms from spindles.

Yep the Feedhold input is still there, and talk about the devil, I am literally testing it as we speak on my machine, with a push switch.

3 Likes

Lots of easy and inexpensive ways to do that - search Amazon for “Router Power Switches”, get one, and plug everything into it.

It’d be nice to see the resume pin broken out to the available headers.