PDA

View Full Version : Using current as cutting force indicator?



mitchejc
12-01-2015, 09:04 AM
It would be handy to have some sort of indicator to show how hard a CNC router is working. On my old machine I just play it by ear but sometimes that does not work and all sounds fine an the next moment the bit is broken. I once saw a video of some fancy VMC that had a little electronic dial meter showing cutting force. It did not show an actual force reading but more like a green, yellow and red band, if that makes sense. Do you perhaps have any idea how something like works? Strain gauges or maybe load cells come to mind but I can't see a feasible way of fitting those to a DIY router?


On my build the spindle power is likely to be the weak spot so I'm thinking if I can see how hard that is working under certain conditions it would give me a more scientific reference point. On my VFD I have an analog output that I can set to output current (virtual) and then hook that to an old school analog type 0-12V volt meter. Not sure what current "virtual" means as the in this context but I'm assuming the current the motor is pulling will give me a good indication of how hard its working. Does this sound like a workable plan or I'm just wasting my time?

EddyCurrent
12-01-2015, 09:49 AM
Maybe the vfd has a screen ? yoiu can display current on that.


There was this thread a while back;
http://www.mycncuk.com/threads/7482-Use-spindle-torque-to-control-feedrate?highlight=blade

Robin Hewitt
12-01-2015, 10:59 AM
A current sensor on the power rail perhaps? I think they are usually Hall effect and dirt cheap.

mitchejc
12-01-2015, 11:57 AM
Thanks for the replies, gents and interesting link. My VFD is one of the cheapies (I think Sunfar E300) so does not have a torque output, only volts,amps and frequency as far as I know. Yep, I can display current on the VFD's display, but maybe to rephrase my question: Do you think that the Current reading as produced by the VFD is a good indicator of cutting force or is there a better way to measure or derive cutting force? Sorry I know little about how these motors actually work and I'm assuming it will use more current under higher load?

Jonathan
12-01-2015, 12:14 PM
Do you think that the Current reading as produced by the VFD is a good indicator of cutting force or is there a better way to measure or derive cutting force? Sorry I know little about how these motors actually work and I'm assuming it will use more current under higher load?

As a first approximation, yes the spindle will draw more current as the load increases. The problem is the motor efficiency changes with load and speed, so it's tricky to estimate the actual torque output, though not impossible. If you subtract the no-load current from the current whilst cutting, that's a good start. But if you're more interested in not overloading the spindle, then just looking at the phase current as a proportion of the rated current is sensible.

I'd almost forgotten about that thread discussing linking torque to feedrate, but I've got a bit further with my motor drive I will use in it so it wont be too long now.

mitchejc
12-01-2015, 01:39 PM
Thanks Jonathan. Call it cheap thrills but I'm basically looking into some interesting little future enhancements to make to my router once I have it working to keep me occupied. I'll play around a little to see what the no load current looks like at different frequencies and take it from there. Until my spindle is actually cutting something, is there a safe way to put some load on to see what happens to that current output? The though had crossed my mind to put a thick glove on and gently apply pressure to the collet nut but I'm way too scared to be the first guy to try that :-)

Your plan of automatically varying the feed rate or spindle speed sounds like a very cool idea or maybe even just to do a estop when it figures out its about to break a certain size cutter or stall the spindle but I'll have to learn quite a bit more before I can join in on that fun :-)

Jonathan
12-01-2015, 01:46 PM
I'll play around a little to see what the no load current looks like at different frequencies and take it from there.

If you're keen you could take several readings, graph them then use that as a look-up table or function to subtract from the on-load current in software, on an Arduino (or similar), then connect a display... In fact you could do it quite easily with an Arduino, current sensor and display:

http://www.ebay.co.uk/itm/New-Blue-IIC-I2C-TWI-2004-20x4-Serial-LCD-Module-Display-for-Arduino-/271634672688?pt=UK_BOI_Electrical_Components_Suppl ies_ET&hash=item3f3eb02c30

http://www.ebay.co.uk/itm/Range-ACS712-Current-Sensor-Module-New-20A-/370906424836?pt=UK_BOI_Electrical_Components_Suppl ies_ET&hash=item565bbee604




The though had crossed my mind to put a thick glove on and gently apply pressure to the collet nut but I'm way too scared to be the first guy to try that :-)

You wouldn't be the first ;)

mitchejc
12-01-2015, 02:49 PM
Arduino. Yep those are very high on my list of things to toy with, just wasn't sure what I'd use it for but now I have a plan!

Sorry last stupid question. Using this approach with that sensor would I measure current on then single phase input side of the vfd or on the output side or does it even matter?

Jonathan
12-01-2015, 06:28 PM
Output side - you want the line current. You could use the current sensor to measure the frequency of the current, then scale that to read spindle speed... just for fun.

EddyCurrent
12-01-2015, 08:13 PM
The though had crossed my mind to put a thick glove on and gently apply pressure to the collet nut but I'm way too scared to be the first guy to try that :-)
)

With motors the general method was to jam a piece of 3 x 2 against the rotating shaft to act as a brake, you want it so the shaft is trying to pull the wood from you.

mitchejc
13-01-2015, 04:09 AM
@Jonathan: Ok, thanks, that makes sense.

@Eddie: Thanks, I'll give that a go. My fingers is getting older so they so they don't grow back as quick as they used to :-)

Not sure but I think I read somewhere the peak torque of these 2.2kw watercooled spindles is something like 0.8nm. Does anybody have a rough idea what the torque curve for these looks like?