Hello Guest it is October 23, 2021, 04:39:19 AM

Author Topic: mach 4 learning the machine drive ratios  (Read 1210 times)

0 Members and 1 Guest are viewing this topic.

Re: mach 4 learning the machine drive ratios
« Reply #20 on: December 17, 2020, 09:32:28 PM »
I was referring to the quill as the z axis. I am thinking that the knee will be "set" for a given operation. Once set, than the tool heights, fixturing and work can be programmed. My understanding of the machine is that all the dynamic functions of all the motors will be
x/y and the quill I'm referring to as z. I will focus on getting this machine functioning as well as I can get it with my limited skill set. I don't want my personal skill limits to compromise the build, but its not going into production either. Our discussions have already corrected a huge amount of mistakes not yet made. If I can get this machine to run well, maybe it will enable me to run well. That's the goal. I'm looking for the "goldilocks" outcome for my little situation.  8) That said, I would like to be conservative, but not cheap. I seem to want to overbuild everything, but I have in the past underbuilt, and have reaped the consequences.
Thanks again, Chris
« Last Edit: December 17, 2020, 09:35:03 PM by Retriever »
Re: mach 4 learning the machine drive ratios
« Reply #21 on: December 18, 2020, 04:16:48 AM »
not sure what to make of the quill, I don't know how it works and the forces involved. The X and Y axes are a little more straight forward.

Lets assume that you wish to get close to the performance that I posted earlier, ie 10m/min rapids and 0.5g.

If the ballscrews have a pitch of 6mm, per your previous post, then to achieve 10m/min the ballscrews must rotate at
10,000/6 =1666 rpm.

There is your first hurdle, steppers lose torque badly with speed, and only very few of them can maintain any useful torque above 1000rpm.
You could use an increasing gear reduction but that would require even more torque from the steppers, I suspect its unlikely to happen.

Per my previously posted calculation you can get about 0.2g with 1Nm, so to get 0.5g will require 0.5/0.2= 2.5Nm. Note also that this makes no allowance
for friction nor ballscrew efficiency, so I would suggest that around 2.5Nm should be considered the minimum torque. Note that this torque is required
at top speed.

A very low inductance (1mH) stepper may retain 40-50% of its torque at 1000 rpm, which would suggest that the stepper needs at least 5-6Nm at low
speed. Steppers that have that sort of torque and low (sub 2mH) inductance are rear beasts. A more normal 34size stepper may well have adequate torque
at low speed but 34 size steppers have inductances of 4mH and more. Such a stepper is going to run out of 'puff' at 500-750rpm.

Side note: 1Nm= 130oz.in

All in all I think steppers may well provide enough torque at low speed but I suspect that they will fail to satisfy at speed. High speed operation
is always steppers Achilles Heel.
I think 34 size steppers of around 800oz.in and 4mh (or less is preferred) would work OK, but maybe fail to get to 10m/min, but more likely 5m/min.

The other alternative is servos. Servos come in a wide variety of specs, but the readily available brands likely to be affordable for hobbyists are:
200W 3000rpm =0.65Nm cont, 1.9Nm peak
400W 3000rpm = 1.27Nm cont, 3.81Nm peak
750W 3000rpm = 2.4Nm cont, 7.2Nm peak

If you used any of these three servos with a 2:1 reduction then you would get very close to your desired 10m/min rapids.

The 200W servo would, after the 2:1 reduction, have barely enough torque without dipping into overload all the time.

The 400W servo would after the 2:1 reduction would have plenty of torque, (1.27 x 2 =2.54Nm) without using overload.

The 750W servo has enough torque without reduction. Without reduction you could have rapids of 20m/min, alternately
with reduction you could still get near 10m/min but have a surfeit of torque for even higher accelerations.

So servos will certainly provide the best performance but they cost more than steppers, and steppers will do OK except
at high speed.

For my new build mill have bought three (one braked) 750W Delta B2 series servos. They cost (excluding the braked one) $460USD each, which
includes the drive and cables, plus shipping. I found that the 750w models were only $40 more than 400W ones, so I got the 750W's.

I notice that many US buyers favor Clearpath (US) over DMM (Canadian made in China) or Delta (Taiwanese made in China) but they are
more expensive, about $480USD for 350W. For various reasons, cost primary among them I went with Delta, almost twice the power for less money!

Ultimately your budget will determine what you do. I would recommend 750W servos if you can afford them.

« Last Edit: December 18, 2020, 04:21:41 AM by joeaverage »
My wife left with my best friend...
     and I miss him!
Re: mach 4 learning the machine drive ratios
« Reply #22 on: December 18, 2020, 02:54:58 PM »
Your counsel is super immensely appreciated. I think that long after I have forgotten how much they cost, I will be living with the results. Right once beats twice cheap every time. Therefore, I will be focusing on servos. Its tempting to go with steppers, but I want to continue to like/love my machine instead of the alternative. Success breads success.
 I already feel like you have steered me around big trouble. I think I will still do a 2:1 drive, and the 750w servos. I think I will be very happy with those results.
Again, much appreciated, and thanks, Chris.
BTW, what profile and size are your servo drive shafts?
Re: mach 4 learning the machine drive ratios
« Reply #23 on: December 18, 2020, 06:43:44 PM »
my 750W Delta B2 series are 80mm across the flange and 19mm shaft with keyway and seal.

I think you are right to go with servos despite the initial cost. If you get good quality and capable units they will give
twenty years untroubled service and even hold resale value whereas underperforming units will not.

My wife left with my best friend...
     and I miss him!
Re: mach 4 learning the machine drive ratios
« Reply #24 on: December 19, 2020, 03:26:31 PM »
That makes perfect sense. The  ballscrews are the same. Also, as I understand, servo hold is determined by position. How does the "brake" function work on your braked servo?
Is there a possibility for a conflict between programmed stop, and applied brake stopping the servo short of its programmed satisfied position? In other words, could the brake apply causing the servo to continuously keep attempting to reach perfect position?
Thanks again for your generosity, and understanding dealing with someone with entry level , if that, skills. Chris
« Last Edit: December 19, 2020, 03:28:30 PM by Retriever »
Re: mach 4 learning the machine drive ratios
« Reply #25 on: December 19, 2020, 04:08:07 PM »
a braked servo has an electromagnetic brake that comes on when its depowered.

With a normal servo when its depowered the armature may turn with only modest amounts of torque to overcome
friction and bearing drag. With such a servo if and when the Z axis is depowered gravity will act on the axis and potentially
the ballscrew will turn and the axis will sag.  With a brake however, when the servo is depowered the brake comes on to prevent
unintentional movement.

The brake could also be used to lock an axis say. If you drive a servo to a given position the servo drive will use its available power
and control to hold its position. There will always be a small amount of deflection about the controlled position however. It may be
that an electromagnetic brake is better in that circumstance.

I am not planning to use the brake it that manner....but who knows...things may change. My intention is to use the braked servo
on the Z axis so that on intentional, or otherwise, de-power the Z axis does not sag into the work zone. The spindles (two)
I have and use on my existing mini-mill will be used on my newbuild mill and they are light enough that I do not expect the Z axis
to sag under their weight.

I do, at a future time, plan to build a third spindle based on a largish AC servo of 2.7kW at torques of 14Nm. That servo and associated spindle,
which I envisage will be ATC, would be heavy enough that Z axis sag would be an issue. I decided therefore to bite the bullet and buy
a suitably equipped servo for that future plan.

I paid  $460 (excl shipping) for a 750W un-braked servo kit (servo, drive and cables) and the braked one cost $600....so you pay a premium for them.

In the case of my 750W Delta servo, and I imagine other brands would be similar, the brake is 24V at 250mA. So I need an auxiliary
power supply of 24V for the purpose. The brake is switched on or off by a electronic switch within the drive and that switch is in turn controlled
by the motion controller. There are a number of programmable parameters that govern how and when the brake operates within the drive,
but the typical brake-on and brake-off delays are 20ms.

My wife left with my best friend...
     and I miss him!