VFD spindle control safety

[Tangoterrier]

I have bought a CNC machine which has a Huanyang vfd and a spindle pre-wired. At the moment, the vfd is powered up when the control box is switched on so there is no independent control over the power to the vfd. The spindle can be controlled by software so when a tool change is needed the spindle stops.

I would like to be 100% sure that when I am changing the cutter the spindle cannot start and in my inexpert opinion this can only be done by disconnecting the vfd from the power or the spindle from the vfd with a switch. Switching off the vfd (which I cannot do at present but is easily achievable) feels wrong but the manual is very clear that the vfd must never be powered up with the spindle disconnected. If I went down the spindle disconnection route I cannot imagine that I would never accidentally power the vfd without switching the spindle back on.

So, as far as I can work out, the only safe way to do a tool change is to turn off the vfd. Is that what people do?

[Doddy]

So, as far as I can work out, the only safe way to do a tool change is to turn off the vfd.

You are correct - as far as providing a completely safe solution you would require to disconnect the supply. The manual does indicate an electromagnetic contactor (relay) but notes that this should not be used as a general on/off control.


Is that what people do?

Nope, life's too short.

[Neale]

No, I never turn off the VFD during tool change. Would probably double tool-change time, waiting for VFD to power down/run up. I'm with Doddy - life's too short...

[Kitwn]

A little less certain that powering off the VFD, you could incorporate a switch in the control cable from the computer to the VFD preventing it being remotely activated by accident. This doesn't protect you against a control fault in the VFD itself. The danger then is that you forget to switch it back on and dig a non-rotating tool into your workpiece. How you protect against that is up to you.

Kit

[Neale]

I realise that I have to trust some amount of hardware and software. However, I am more concerned about triggering the "start" function by accident, which would move the spindle to the next position and then run up the spindle. So, I would get some warning as the machine would start to move from the tool change position, allowing me to let go the spanners. In addition, my F360-generated gcode seems to need the start button to be hit twice - first moves spindle to tool change position, second moves it to start position before turning it on.

I control the spindle by supplying a signal to the VFD control input, so a loose connection would stop it running rather than start it. There are other possible failure modes, of course, but these seem to be less likely and the associated risk is less.

In addition, I am the only user of a hobby workshop so my approach to risk is different to a commercial workshop. I'm now curious to know how commercial machines address this kind of issue

[Doddy]

There will be interlocks on a commercial machine. I was contemplating my little SX2.7 but realised that I'd previously removed the swarf guard from around the spindle - of course its a requirement to move this to change tools and this provided complete isolation from the incoming supply - I expect any commercially available machine will have similar interlocks. God help you within a commercial environment if you defeat such an interlock and the HSE get involved.

[Clive S]

Also you can factor in that there is little to no torque on start up of the spindle with these type of spindles .

[magicniner]

If the VFD is not of exceptionally poor design you will be able to hard wire an E-Stop switch directly to the VFD data inputs and configure the function in parameters if required, when activated this will over ride any other signal trying to power on the spindle until the E-Stop switch is reset.

[Tangoterrier]

I have done the non-rotating tool trick a few times with my previous machine!!

[Kitwn]

I have done the non-rotating tool trick a few times with my previous machine!!

Haven't we all! My normal process now is to add 8 second delay to the G-code after starting the spindle to make sure it comes up to speed. As the same relay starts the dust extractor I cannot hear the spindle start but keep a wary eye on the VFD display to make sure it doesn't stay at 00000. 8 seconds is enough to realise I've done something stupid and stop the cut before it starts.

Kit

[m_c]

Also you can factor in that there is little to no torque on start up of the spindle with these type of spindles .

This.
Unless you have a spindle with lots of torque (I.e. milling machine/lathe spindle that's been geared down from a multiple KW motor), then holding the spindle with your hand will be enough to stop it from spinning up.

Plus most VFD manufacturers advise against regularly power cycling VFDs.


If you were worried about damage, I'd put the risk of crush injuries on any reasonably sized/powered CNC, far higher than a spindle starting up when you've got your hands near it.

[Neale]

If you were worried about damage, I'd put the risk of crush injuries on any reasonably sized/powered CNC, far higher than a spindle starting up when you've got your hands near it.

Probably right. Thinking about it, I've stalled my 2.2KW spindle by trying to push it too fast/deep through steel - with a 4mm cutter. That was at around 5K RPM - no torque worth talking about. A spanner on a spindle would easily hold it.

[routercnc]

I’m in the minority here and I start and stop my spindle manually through large push buttons on the control box main panel. These are linked back into the VFD inside the control cabinet through the built in terminals under the front cover.
Same goes for the speed control which is a potentiometer on the control box main panel linked back to the VFD.

Some jumpers in the VFD have to be set to use external controls.

This means the PC is not in charge of the spindle at all except for an e-stop where it will stop and inhibit the spindle running, again using terminals on the VFD.

I’ve never had the spindle come on when changing tools, and have never forgotten to start the spindle before hitting cycle start. In the same way that I have never tried to drill a hole on a drill press without starting the motor! It’s a preference thing and I just like to use it that way, plus it feels a bit safer.

[Doddy]

I've had my G&T for the night so am feeling somewhat ambivalent. The OP questions safe operation - and he's correct to do so. I hold a view which I think is shared by some that, gauging an acceptable level of personal risk against convenience that I/we are prepared to cut some corners. But, please understand that doesn't make it safe. I wouldn't pretend for a second, or promote that nipping up a cutter in a spindle with your hands/body/spanner anywhere near is "safe" in the event of either hardware or software failure.

Right!, where's that blasted cat?

[Kitwn]

At present I'm following Doddy's principle of taking what I see as an acceptable risk when tool changing but I'm aware my machine has a built in risk which is that the relay that starts the spindle will pulse briefly during power up/down of the controller. I use a cheap Chinese breakout board and LinuxCNC. The relay on the board operates a relay on the wall that activates the spindle and the dust extractor.

Not a problem when it's all switched on but if I wait long enough there'll be a power cut just as I have a firm grip on a tool to pull it out of the (still quite tight) collet and there'll be enough stored charge in the VFD capacitors to modify my finger prints.

It's all a balance of probabilities but reading this thread does remind me I should add some kind of interlock. Perhaps a nice big switch in series with the VFD control relay which includes a back contact for a warning light. I have all the necessary bits in the shed so it's not exactly beyond what is 'reasonably practicable' as the legislation says.

I would never criticise routercnc's KISS approach but know myself well enough to predict the odd oversight, especially towards the end of a tiring day and am therefore in favour of the more automatic control for my own machine.

[ericks]

Been using my new diy machine for a while but must admit i have not completed the safety design...too busy with other stuff:)

I should know better!!

[AndyUK]

You are correct - as far as providing a completely safe solution you would require to disconnect the supply. The manual does indicate an electromagnetic contactor (relay) but notes that this should not be used as a general on/off control.


I ordered a relay the other day intending to use it for exactly that function, then I read this in the manual. I'm sure I'll find another use for that relay, so the £5 I've spent on it is immaterial, but whats the reasoning for this exactly? Is it the noise that the VFD is spewing out, or is it more of a surge current thing? :confusion:

[Doddy]

I ordered a relay the other day intending to use it for exactly that function, then I read this in the manual. I'm sure I'll find another use for that relay, so the £5 I've spent on it is immaterial, but whats the reasoning for this exactly? Is it the noise that the VFD is spewing out, or is it more of a surge current thing? :confusion:

I wouldn't like to speculate, however, wording from the manual:-

"No contactor should be installed between the power supply and the inverter to be used for starting or stopping of the inverter. Otherwise it will affect the service life of the inverter."

My guess (and it is a guess) is that these have been engineered by an accountant and power-up surges, and inductive EMF stress the power components.

[A_Camera]

I have bought a CNC machine which has a Huanyang vfd and a spindle pre-wired. At the moment, the vfd is powered up when the control box is switched on so there is no independent control over the power to the vfd. The spindle can be controlled by software so when a tool change is needed the spindle stops.

I would like to be 100% sure that when I am changing the cutter the spindle cannot start and in my inexpert opinion this can only be done by disconnecting the vfd from the power or the spindle from the vfd with a switch. Switching off the vfd (which I cannot do at present but is easily achievable) feels wrong but the manual is very clear that the vfd must never be powered up with the spindle disconnected. If I went down the spindle disconnection route I cannot imagine that I would never accidentally power the vfd without switching the spindle back on.

So, as far as I can work out, the only safe way to do a tool change is to turn off the vfd. Is that what people do?

The way I do it is simple. I have a 3 phase safety switch installed between the motor and the VFD. Simple and no electronics involved.

26163

Every time I have to change tool I simply flip it to off and when ready I flip it to on. This procedure makes sure that regardless if my VFD, or control software, or whatever else would start the spindle due to an error, the spindle won't spin and my hands and fingers are safe. This may be regarded as overkill by some DIY machinists, but in my opinion the only safe way to work with these machines. It also gives me a sort of ultimate emergency stop possibility, but it should never be used as an on/off switch. However, I have to admit that that I did start the spindle a few times with it because I forgot to flip it to on after a tool change before starting milling and because the movement already started when I realized my mistake. Everything is still fine, but I know this is risky and should push the e-Stop and restart from there, instead of doing this.

Anyway, this is what I use for safety and it is 100% safe as long as I remember to flip it to off before starting the tool change. I do not turn off the VFD because it takes too long to start it ever time.

[Kitwn]

Connecting/dissconnecting a motor from a VFD or stepper controller when it is powered is generally considered a risky business due to the voltage spikes that can occur. Power transistors have been described as 'the fastest fuse on three legs' for good reason. If your switch has a back contact that can be used to prevent the VFD from trying to start when the motor is not connected that might be a good move.

[ericks]

Using a field isolator on drive driven motors are an Industry standard. They are lockable isolators.....obviously you would isolate after you have "stopped" the motor:)

[A_Camera]

Connecting/dissconnecting a motor from a VFD or stepper controller when it is powered is generally considered a risky business due to the voltage spikes that can occur. Power transistors have been described as 'the fastest fuse on three legs' for good reason. If your switch has a back contact that can be used to prevent the VFD from trying to start when the motor is not connected that might be a good move.

Perhaps I was not clear, it is NOT meant to be on/off switch, just safety switch to make sure the spindle will NEVER be started by accident or error while there is a tool change in progress. Of course, the spindle is not spinning when the switch is flipped to off, and the VFD is not feeding the motor power outputs when the switch is flipped to on. Starting and stopping the motor is of course done by the CNC software, in my case through Modbus commands.

Power transistors provide ZERO safety, quite the opposite, they give you a false sense of safety and may not operate when needed. A mechanical switch will always work because it is just what it is, simple and 100% independent from electronics and software/firmware.

[magicniner]

This is like a health and safety thread on a model engineering site! ;-)

Commercial CNC machines rely on the spindle motor drive, with a suitable signal from the controller, to prevent the spindle starting up whilst changing tools, that works, failure to understand why will lead you to think you need breakers and stuff between the motor & VFD when you don't.

If you have a decent VFD it has the facility for an E-Stop data input, this will suffice on equipment of a reasonable quality.

[Kitwn]

A-Camera,
I appreciate that a proper procedure for stopping and isolation protects the VFD. My concern would be the occasional operator error, possibly starting the VFD before closing the contactor which is why I suggested the back-contact interlock to prevent the VFD from operating with the contactor open. This ties in with my earlier comments about the 8 second delay etc. These spindles are a big investment for a hobbyist and anything you can reasonably do to prevent one of those potentially expensive "Oooops!" moments is worth considering.