[hax0red]

I have a Novakon NM-135 that I purchased as "not functional". What I received had stepper motors but no power supply or drivers. I believe the original power supply was likely 36 or 48v DC. I have since purchased drivers and they are capable of 110VDC. I have a 48v switching supply here but I want to get an Antec linear toroidal supply so I can eliminate any concerns with a nice clean power supply. My question is should I go for a 48v DC supply or 80v DC assuming they are both 800 watts. I'm a little concerned the stock stepper cannot take the extra voltage. I'm pretty sure they can and only low volage=higher amps should be my only concern?

[joeavaerage]

Hi,
go for the highest voltage your drivers can handle.

The high voltage is applied to the stepper windings at the moment of switching only, thereafter the PWM kicks in to limit the current to a safe level.
Having the high voltage at the moment switching however 'forces' current through the winding in shortest possible time resulting in the fastest possible step
repetition rate and therefore highest stepper speed.

High voltage it what allows a stepper to rotate fast WITHOUT losing steps.

Craig

[hax0red]

Hi,
go for the highest voltage your drivers can handle.

The high voltage is applied to the stepper windings at the moment of switching only, thereafter the PWM kicks in to limit the current to a safe level.
Having the high voltage at the moment switching however 'forces' current through the winding in shortest possible time resulting in the fastest possible step
repetition rate and therefore highest stepper speed.

High voltage it what allows a stepper to rotate fast WITHOUT losing steps.

Craig

Thanks for the reply. Yeah, I went ahead with the 80VDC PS-8N80. I am upgrading my X and Y axis steppers to 12nm closed loop nema 34s that appear to have drivers limited to 100VDC and the next step up from Antec is 105v. Gonna run the stock Nema 34 on the Z with a clone DM860 for now.

[joeavaerage]

Hi,

I am upgrading my X and Y axis steppers to 12nm closed loop nema 34s that appear to have drivers limited to 100VD

I would do more research on those units BEFORE you commit to buy. They have lots of torque but they have high inductance and have limited speed as a result.
There have been a number of posts on this forum of people whom have bought these things only to swap them out for lower torque units and had a MUCH better
outcome....go figure?

Craig

[hax0red]

Hi,



I would do more research on those units BEFORE you commit to buy. They have lots of torque but they have high inductance and have limited speed as a result.
There have been a number of posts on this forum of people whom have bought these things only to swap them out for lower torque units and had a MUCH better
outcome....go figure?

Craig

Yeah I saw those who had better luck with Nema 23/4's as well. My thinking on these was not speed but ability to run the gibs tighter without stalling. Franco on youtube did some comparisons of raw stall torque and they looked quite a bit better. I figure with the extra voltage I should be able to hit regular Nema 34 speeds with increased torque and accuracy.

If I run into trouble I may try to tune them, worst case send them back and go with a couple servos for twice the price. Besides, no more rapids in Fusion 360 without forking over 25+ a month.

[joeavaerage]

Hi,
the extra torque is seductive but in practice it does not seem to materialize. The reduction of torque with increasing speed is so severe that its entirely
probable that a smaller (and much lower inductance motor) will outperform its bigger brother over several hundred rpm.

I chose 750W Delta B2 AC servos and drives for my new mill....and they eat any stepper ever made. They cost $438USD (servo/drive/cables) plus shipping...so
yes they are more expensive, but do the sums. By the time you buy a DC supply, the stepper drives and the steppers themselves my guess is that the price difference
is less than you might imagine. Should point out that the Delta servo drives are 230VAC input and do not require a separate supply.

Craig

[peteeng]

Hi Craig & others - std modulus carbon fibre in a quadaxial laminate (quasi isotropic) will get to 60GPa and 70 if very high volume fraction. In a bending optimised laminate it gets to 80-85GPa. The fibre itself is 200GPa so even a 60% volume fraction laminate with all fibres in one direction can get to 120GPa but this laminate is very tender and will fracture easily if the load is off axis. The ideal CF for machines would be an intermediate modulus fibre and a quasi isotropic laminate. This would give you a modulus of 110Gpa and a very good shear modulus. So it would be the same stiffness as cast iron yet much much lighter.. 1500kg/m3 density and be very damp, much damper then CI.

Some granites are very stiff and can be as stiff as cast iron. But its hard to buy granite by stiffness being a natural material it is variable. Every material can build a successful machine just need to understand the pros and cons of each. Horses for courses. Peter

[joeavaerage]

Hi Peteeng,
this reinforces my contention rather than weakens it.

I have used carbon fibre, particularly on model areoplanes and on the basis of strength to weight carbon fibre is great. For a CNC mill the mass/layup requirement, and therefore cost
of carbon fibre, is just out of the question, even if it is light.

With a CNC mill rigidity is everything, damping is nice but still less important than rigidity. While granite and epoxy granite can and have been used successfully,
but only at the marked increase in cross section....and I ask again why? I'd much rather cut, grind and machine cast iron than granite or epoxy granite. I bet there are a lot
more foundries out there than there are stone masons with the capability to cut and grind granite ultra accurately for CNC purposes.

There is a reason that cast iron has been so ubiquitous when it comes to building machines and it still remains the material of choice. I know you are going to point
to high end manufacturers whom use epoxy granite.......you go and buy one!!! They are eye-wateringly expensive.

Craig

[peteeng]

Hi Craig - There is no argument that CI is the material of choice for mass produced machines, unless very high accelerations are needed then CF takes over due to its lightness. CI offers the correct blend of stiffness, dampness and manufacturability in the current manufacturing paradigm for serial production. In the future with large scale 3D printing this may change. The US army has commissioned a metal printer large enough to print metal hummer chassis' and armoured personnel carriers in one go. That sort of technology will change the way we make things...

Small addition - aluminium may take up some opportunities as well now Megacasters are around. Someone with a megacaster and some spare capacity would make excellent large machine tool parts - Peter

[joeavaerage]

Hi Peteeng,

Small addition - aluminium may take up some opportunities as well now Megacasters are around. Someone with a megacaster and some spare capacity would make excellent large machine tool parts - Peter

I watched a YouTube video of a tour of the Kern factory in Germany.....top end to beat all top end!. They use epoxy granite in the base and cast aluminum elsewhere. The contention is that heat dissipates
throughout aluminum so rapidly that Kerns extensive machine cooling system can achieve a tighter temperature tolerance.

The one piece of my new mill about which I am not happy is the Z axis extension arm from the Z axis saddle to the spindle. I made it out of 100 x100 x 9 RHS and welded two 20mm plates either end.
Works fine with my wee (800W 24000rpm) spindle but starts to vibrate with my 1.8 kW 3500rpm spindle. I thought it would be stiff enough....but its not. I should FEA model it and see.
The upshot is that it will have to be replaced, sooner rather than later. My preference would be cast iron....but maybe cast aluminum might be a better bet.

Notwithstanding the extension arm flexure the machine is operating nicely and allows me and my business to continue as is.

Craig