Router losing steps. Solution needed. Will this help?

Hi All,
I have a typical Chinese R1325 router which after a great deal of work to complete the wiring of the control panel, and truing/squaring up the axes is now working. The control system is Nc Studio driving Nema 34 stepper motors. I mainly use it for cutting profiles in 18mm melamine faced chipboard for kitchen units. I use a 12.7mm dia PCD cutter and cut the full depth in one pass. For the most part this is producing acceptable results. However if I try anything more ambitious where multiple tool paths are required I get lost steps all over the place. I don't think it's due to any tightness in the machine as it occurs anywhere in the tool path. It's quite obvious when it happens as the motors can be heard to "knock" quite loudly. This can even be heard when the machine is sent "back to zero". My question is this, would fitting a hybrid closed loop step servo drive system in place of the existing drives and motors solve the problem? I've seen a few around and like the look of this http://www.ebay.co.uk/itm/CNC-Hybrid...item417b121221

What does the collected wisdom think? am I barking up the wrong tree? should I maybe be thinking of changing the control system to Mach 3 instead or as well?

Many thanks for any advice you can offer, this is driving me quite mad.
Bob.

Hi Bob,

Woo Lad hang fire before buying anything lets find the problem first.? It won't be the motors that's 99.95 sure.

Few questions.

At what feed rates and point in G-code does it seem to lose steps most.?
What Size Nema 34 motors.?
What Voltage are the drives running on.?
Make and model of Drives and settings ie: amps, micro stepping.
Is it using ballscrews or Rack & pinion.?

Double check all the axis for binding that's a must.! . . . Check for lose motor couplings etc while at it.

Changing to closed loop won't help anything if you have a binding issues. Equally they will still stall if you haven't got enough Voltage for the speed your trying to run at. They only keep position within the tuning parameters and if your overtuning for system capabiltys then they won't do shit for you.!!

I'd tip that your trying to run too fast for the amount of volts your providing the motors with. Nema 34 motors, even small sizes, need high voltage to get any decent feed rates from them and it's common for people to under power them and then get stalling and missed steps. Throw into the mix running on a parallel port and then this sort of thing happens.
It's down to the fact larger motors have more inductance so need much more voltage, Nema 34 motors are only good to about 900Rpm before hitting the corner speed and then torque drops away quickly. What they do over smaller nema23 motors is hold higher torque up the speed curve and drop off quickly nearing the corner speed.

Give so specs of your system we can see what might be causeing the problem but DONT buy anything yet as I'd tip High voltage drives will cure your problem without new motors.

Thanks for the reply Jazz, I will get back with all the specs on Monday when I get back to work. Meanwhile I think it unlikely it's due to asking for too much speed from the motors, or low voltage drive, the problem is less evident with a feed rate of 3600mm/min and never happens at rapids of 10000mm/min. The drives are fed from 4 separate windings on the transformer at 70 volts so should be high enough. Drives are Leadshine MA860H. More to come next week.

70V is only just enough if they are smaller 4nm 34 motors with low inductance but if they are larger motors then it's not really.

It sounds like this could be few things working together.? Possibly some binding and overtuned motors but without knowing more regards screw pitch and size etc then difficult to say. But with 10mtr/min rapids and guesiing 10mm pitch screws your all ready tuned over the corner speed of the motors and running the motors with low voltage so torque less higher up the speed curve.

I'd tip these steps are lost on directional changes thru inertia pushing with low torque and guess your using 25mm screws with a 10mm pitch.? (Things only get worse if your using R & P due to low efficency)

Give full SP on machine ie: size/pitch/ratio etc of linear motion. Motor tuning setup Velocity/accel etc. Drive setup amps and Micro step setting.

Oh and forgot to say in last post.! . . . . YES Dump NC Studio and go with Mach it's much better IMO. I'd also recommend you think about an external motion control card to get the best from your machine and setup. They really do blow the parallel port away giving much better pulses which translates into smoother motor performance and higher tuning capabiltys. They also make Mach much more reliable as they take all the strain of generating pulses leaving mach to get on with other tasks. Also PC doesn't need to be anything special and even allow laptops to be used.

Hi Jazz,
A few details to update you

The machine has rack & pinion drive to the x and y axes and 10x25 ballscrew to Z Y axis has 2 motors with the drivers connected to the same pulse/dir circuit. All motors have 5:1 reduction drive to the screw/pinion via toothed belt. Motors for X and Y are set to 1600 steps per rev and machine output set to 0.01042mm/step, the rack pitch is approx 4.15mm which is confirmed by the drive ratio and machine setting. I therefore calculate the maximum speed of the motor in rapid at 10000mm per min is 600RPM. Drive current is 4.86A. The motors are Nema34 type about 115mm body length They have 86BYGF 450B on them but a google search reveals only ads for machines no specs. The machine manufacturer reckon they have no details they can let me have. Hope this all helps.

I take your point about using Mach3, I'll be getting around to making the change asap. NcStudio offers no motor tuning facilities at all, it's not possible to alter the pulse duration.

I do have a USB driven breakout board for use with Mach 3 which has on board a PWM spindle control circuit and differential drive outputs for step/dir. I would like to make use of the PWM spindle control but have also been looking at the smoothstepper board which may mean better control. Any thoughts?

Many thanks again for whatever you can do.

Bob.

Quote:

---

Originally Posted by cncbobuk

The machine has rack & pinion drive to the x and y axes and 10x25 ballscrew to Z Y axis has 2 motors with the drivers connected to the same pulse/dir circuit.

---

This needs explaining better.? Do you mean your just using 1 Step and 1 Dir signal to drive borh drives.? This is a Big NO NO. One reason is the drives have resonance compensation circuitry built into them and doing this just confuses the drives so they don't know which drive to apply compensation too. The net result is that one drive/motor can get thrown into resonance. Then you have timing issues thrown into the mix and this just results in all sorts of random and strange happenings.!!

Not saying this is the problem but it certainly won't be helping any if it is setup this way.

Regards the R&P setup then what size are the pinions, ie diameter not No teeth.?
Reason I ask is because from the 0.01042 per step then that translates to 16.672 pitch with 1600MS and with a 5:1 ratio this means a pinion diameter of approx 83.5mm and this means a small pinion size.!
With Module size then pinion size for same number of teeth increases which means your module size would need to be small to have diameter of only 83mm And I'd be surprised if it's less than module 1 and more likely to be Mod 1.5 or even imperial.? . . . All this makes me wonder if your actually set correctly.?

Regards the Smooth stepper then it doesn't have any spindle speed control built into it or really any real I/O capabilty's that you would use directly so this means using with a BOB. It's main function is Motion control which it does very well.
Regards USB then I'd steer clear of using USB to power the BOB as it's unreliable and if it as the option to use external power then do so.

If you want the Best Motion control card for Mach3 then look at the Cslabs Csmio-IP-M or IP-S products which blow the others away. They also have Spindle speed control built into them. DONT use BOB so take away the biggest hassle in any control box and provide 24V I/O which connects directly to the Unit thru nice DIN rail mounted connectors so very neat and industry standard. (See pic.)
Attachment 13546

Quote:

---

Originally Posted by JAZZCNC

This needs explaining better.? Do you mean your just using 1 Step and 1 Dir signal to drive borh drives.?

---

Sorry I probably didn't make this clear. Y axis has one motor at each end of the gantry each driving their own rack. Each motor has it's own MA860H driver but the pulse and direction inputs are connected to the same output from the controller.

Quote:

---

Originally Posted by JAZZCNC

Regards the R&P setup then what size are the pinions, ie diameter not No teeth.?
Reason I ask is because from the 0.01042 per step then that translates to 16.672 pitch with 1600MS and with a 5:1 ratio this means a pinion diameter of approx 83.5mm and this means a small pinion size.!
With Module size then pinion size for same number of teeth increases which means your module size would need to be small to have diameter of only 83mm And I'd be surprised if it's less than module 1 and more likely to be Mod 1.5 or even imperial.? . . . All this makes me wonder if your actually set correctly.?

---

I assume you mean 83mm circumference??? if so you are correct, the pinion diameter is approximately 26mm I can only conclude the basic setup is right as when I do a simple full depth cut the produced dimensions are essentially correct.

Quote:

---

Originally Posted by JAZZCNC

Regards USB then I'd steer clear of using USB to power the BOB as it's unreliable and if it as the option to use external power then do so.

---

The board does have external 12-36v supply connection, it connects to Mach 3 via USB not parallel port.

Quote:

---

Originally Posted by JAZZCNC

If you want the Best Motion control card for Mach3 then look at the Cslabs Csmio-IP-M or IP-S products which blow the others away. They also have Spindle speed control built into them. DONT use BOB so take away the biggest hassle in any control box and provide 24V I/O which connects directly to the Unit thru nice DIN rail mounted connectors so very neat and industry standard.

---

I'll check them out, cheers.

Quote:

---

Originally Posted by cncbobuk

Sorry I probably didn't make this clear. Y axis has one motor at each end of the gantry each driving their own rack. Each motor has it's own MA860H driver but the pulse and direction inputs are connected to the same output from the controller.

---

Ok well this is BAD news and needs sorting. Chances are this has been done because NC studio cannot use slaved motors. Mach3 does allow slaved motors and is what you need. This allows each drive to have it's own set of pulses which are controlled as one axis. Effectively you are slaving 2 axis together Ie: X with B or whch ever axis you designate to be slaved.

The USB device you have isn't just a BOB if it connects to the USB and doesn't have Parallel port. It's Motion control device that provides I/O.
So in this case then you could just connect this and use with Mach. It will need a Plug-in to work with Mach. How good it will be will remain to be seen but will get you going.? Also use the external supply don't rely on USB to power.
Do you have details of it.?

Yes sorry ment circumference and if this is the case then that's ok.! . . estimate pinion about 18T if it's Mod 1.5.? 20T if Mod 1.25

Many thanks again for your assistance, Ill try the Mach 3 option as soon as I can get round to it. The board is freely available on ebay a version of it is here http://tinyurl.com/kpwfrel

Cheers,
Bob.

I randomly found this thread on google while trying to search for why my CNC is losing steps. After reading through this thread, I just want to say that JAZZCNC is an amazing human being for providing so much useful information!! You are my new CNC god! Have to give respect where it’s deserved!!! Cheers!