Messages

2081

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 05:25:22 PM »

Hi,
just as an illustration:

Leadshine 86CM80 (4mH,0.63 Ohm) time constant=2.5ms
Longs 34HST9805 (15mH,3.2 Ohm) time constant=  48ms

Thus for the same input voltage the Leadshine motor would run 19.2 (48/2.5) times faster at the same
percentage torque reduction.

That's nearly 20 times faster!!!!

Craig

2082

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 05:11:44 PM »

Hi,

So you you believe 312.5 rpm is far to fast for these motors and at that speed there will be no torque?

Without some published data from the manufacturer I cannot tell for sure.

You stated that you tried to increase the max velocity and they started stalling......doesn't that tell you exactly that?
They work OK at 125 rpm but stall out at 312.5 rpm.

Longs Motors make stepper motors with really high torque but also very high inductance. They sell to first time buyers
who think the higher the torque the better but what they don't realize is that high inductance kills a motor at speed.
Savvy buyers want to know both the torque AND the inductance, or to be technically correct, time constant.

Until I mentioned 'time constant' in this thread had you ever heard of it? Would you consider making a buying decision on
the basis of a time constant?

Now that you have heard of it and seen torque/speed curves that illustrate what a 'time constant' means in the real world
what would be your decision today?

May I suggest that if a company/supplier cannot provide either a torque/speed curve, or a time constant or an inductance
then walk away....they are selling to first time buyers.

Craig

2083

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 04:51:35 PM »

Hi,

Looking at both of those motor graphs the torque is very good at rpm's below 500, so surely running at 312rpm will be fine or am I missing something obvious

Yes, you are correct, they both have excellent torque at 312.5 rpm, but even the larger of the two motors still has only
2/3 the torque of your existing motors. You will probably require q belt/gear reduction to have either of these motors match
the low speed torque of your existing motors.

You could of course go for higher torque motors and not need a belt reduction. The only trouble with that approach is that
higher torque motors also have higher inductance.

The 86CM80 motor by Leadshine has a rated torque of 1133 oz.in and an inductance of 4mH with a time constant of 2.5ms.
This would be a superb replacement for your existing motors, no reduction required and 1/4 the inductance.....a huge step
in the right direction.

Craig

2084

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 04:40:33 PM »

Hi,

I'm a little lost with what your saying here, your first paragraph says to get 5000 rpmmy motor would spin at 312.5 rpm which your implying i will have next to no torque with these motors at that speed. The next paragraph says with gearing i would be at 625rpm to get the same speed surely there would be no torque left at all then?

I understand you confusion.....you are right..... I believe your existing motors have vanishingly small torque at even low speeds.
No amount of gear reduction or change in mirostepping or increase in driving voltage is going to materially improve that.

What I'm planing to do in the morning is change the stepping to 1/8th 1600 on the drivers and reduce the acceleration from 650 to 250 and see what happens.

Try it and see. I suspect changing the microstepping from 2000 to 1600 will be scarcely noticeable, in fact changing microstepping
to any setting will have no effect on the motors torque at speed.

Reducing acceleration may help. High acceleration promotes stalling. However you want the highest acceleration you can for
accurate toolplath following and acceleration is vital to fast cycle times.

What effects will increasing the supply voltage make as it's been mentioned my motors could take 120v rather than the 60v from the current supplies?

Increasing the voltage is the classic means of overcoming high inductance steppers. The highest voltage drivers I have seen
(Leadshine and Gecko) are 80V, so a 25% increase on what you have already. Higher voltage drivers are always an advantage
but I think you will get much MUCH more benefit from replacing those high inductance steppers.

Craig

2085

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 04:28:10 PM »

Hi,
the next step is genuine-honest-to God AC servos like:

https://store.dmm-tech.com/products/dst-0-4kw-ac-servo-motor?variant=23270271046

Its rated torque is 1.27Nm and short term overload rated at 3.82Nm. With a 3:1 reduction that equates to
3.81/11.46  Nm at the leadscrew or 500/1500 oz.in at 1/3 of rated rpm or 1000rpm. On 16mm pitch
leadscrews that is 16000mm/min....it that fast enough???

Craig

2086

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 04:07:54 PM »

Hi,
to flesh out the answer I have taken these published torque/speed curves from the Leadshine website for two of their
34 size steppers.

The smaller of the two (86CM35) has a rated torque of 496 oz.in with a very respectable inductance of 2.67mH. Note also the
low resistance of 0.42 Ohm. The time constant, a figure of merit, is 0.42 X 2.67=1.12ms.

The slighter longer(86CM45) motor has a rated torque of 638 oz.in and a still respectable 2.95mH inductance with a resistance
of 0.43 Ohm for a time constant of 1.27ms.

On the face of it you would say the bigger motor is better....right......but the smaller motor does very well at speed. The somewhat
smaller inductance, and therefore time constant, means that it retains 50% of its torque at 1000 rpm whereas its bigger
sibling retains 30%. In fact at 1000 rpm both motors produce the same torque (1.6Nm).

This illustrates the point that sometimes smaller motors but with low inductance can outperform larger motors but with
commensurately higher inductance.

Leadshine is one of the few companies to publish torque/speed curves but they are very useful. Both of these motors are
'low inductance designs' and are both good but note how even a small reduction in inductance can have a large bearing on the
motors performance at speed.

Your 15mH units will fall well WELL short of these, and note Longs Motors don't publish curves. I think you should
consider replacement.

Your 16mm pitch leadscrews are aggressive and you will require high torque to have reasonable acceleration and thrust to
accommodate cutting forces but really only modest speeds.

For instance at 2000 mm/min with a 16mm pitch the motor rpm need only be 125 rpm.
To get 5000 mm/min with the same leadscrew the motor rpm is 5000/16=312.5 rpm.

In both cases the required rpms are not high. Would you consider putting a belt or gear reduction or 2:1 in place?
That would effectively make your steppers twice the torque and yet still not require high rpms, 625rpm at 5000mm/min.

For example the 86CN35 of 496 oz.in rated will have nearly 1000 oz.in with a 2:1 reduction, even more than your existing
motors and yet according to the torque/speed curve still have 2.2Nm at 625 rpm or 2.2/3.1=71% of its rated torque.

Craig

2087

General Mach Discussion / Re: Motor tuning, please take it easy on me I'm new to this :)

« on: August 30, 2019, 02:54:16 PM »

Hi,
I suspect with those steppers that 2000 mm/min is about as good as you are going to get.

All steppers lose torque the faster they go. The principle determinant in how badly they fare at speed is
the winding inductance, the higher the inductance the quicker the torque degrades with speed.

For 34 size motors you should be looking at units of 2-4mH, such motors will have about 30% of their holding
torque at 1000 rpm.

Your 15mH units are likely to have less than 5% of its holding torque at 1000 rpm. That describes why you
cannot increase the axis speed without loosing steps.

Craig

2088

General Mach Discussion / Re: how to access puls info of axes from within brains

« on: August 28, 2019, 02:39:57 PM »

Hi,

https://www.machsupport.com/forum/index.php?topic=41692.msg272651#new

Craig

2089

Brains Development / Re: is there a way to access puls counts of axes from within brains

« on: August 28, 2019, 05:53:07 AM »

Hi,
Mach does not work that way.

Mach3 (or Mach4) interprets Gcode and plans a trajectory. That trajectory is communicated to the motion controller
as a series of P(osition)V(elocity) over T(ime) data packets in 1 millisecond times slices.
The time slices are stacked up in a buffer, up to 250 millsecond or so long.

The motion controller, and for this purpose Machs parallel port driver is a motion controller, converts those time slices
into pulses. Mach at no time has variables that have pulse data, that is the job of the motion controller.
You do not have access to the motion controller data. In addition MODBUS is a serial communication protocol
and any sense of simultaneous pulses is lost, aside from the fact that MODBUS is way WAY WAY too slow
to communicate pulse timing.

If you wish to make your own motion controller you will need to use USB or Ethernet to get the PVT data packets and then
have your own motion control processor work out the pulse timing and generate the pulses. Its a very big undertaking.

All development of Mach3 ceased six years ago and it seems pointless to invest the sort of effort that would be required
for a software which is in its 'sunset years'. Mach4 is current and developing. If you were prepared to sign a N(on) D(isclosure)
A(greement) with NFS you could get sufficient information about Machs internal operation and the format of its PVT data
and make your own controller. Still a huge undertaking.

Is there a particular reason you want to make your own controller rather than use an off-the-shelf unit from any one of
a dozen or more different companies?

Craig

2090

General Mach Discussion / Re: Reference not stopping

« on: August 26, 2019, 03:01:20 PM »

Hi,
what version of Windows are you using and importantly is it 32 bit or 64 bit?

What make and model is your motion controller?

Craig