Hello Guest it is March 07, 2021, 05:16:06 PM

Author Topic: Motion controllers - Take me to school  (Read 7251 times)

0 Members and 1 Guest are viewing this topic.

Offline ger21

*
  • *
  •  6,289 6,289
    • View Profile
    • The CNC Woodworker
Re: Motion controllers - Take me to school
« Reply #20 on: September 04, 2017, 11:54:19 AM »
Do NOT buy a chinese motion controller, unless you want partial functionality, and no support.

Quote
I paid 1k for a basic Sherline stepper system so as far as cost goes in comparison, its rather insignificant considering the performance gain.

You can probably put together a stepper system for half that, that performs twice as well.


A $2K upgrade on a sherline that does little more than reducing noise is crazy, but that's just my opinion. ;)
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline sn96

*
  •  53 53
    • View Profile
Re: Motion controllers - Take me to school
« Reply #21 on: September 04, 2017, 01:54:42 PM »
In this image, just one piece takes 8 hrs to complete. Much of the time is wasted in position transfers. I'm ok if you call me crazy, nuts, whack-o... for wanting to not buy, but to look into and research the possibilities outside conventional steppers. I've been using steppers for 3 years, and if I'm ready to upgrade I have many options to choose from. If speed, quietness and step accuracy are important factors to me, then the value relates to what is important to the end user. No?  ???

Point taken about Chinese products. We agree there.





« Last Edit: September 04, 2017, 01:57:07 PM by sn96 »
________________
Mike

Sherline 5400 with Sherline steppers
Visual Mill 6.0 Standard
Mach3
Windows XP
Re: Motion controllers - Take me to school
« Reply #22 on: September 04, 2017, 02:31:17 PM »
Hi,
if you want to spend that sort of gold on servos but the save $10 buying a Chinese motion controller?

Some research on the forum will reveal hundreds if not thousands of posts about Chinese controllers and the battle people have had trying to
get them to work. Even if they succeed features we regard as standard fare like probing often don't work. China might as well be a different planet
when it comes manufacturers support.

In the 'value' end of the market
My wife left with my best friend...
     and I miss him!
Re: Motion controllers - Take me to school
« Reply #23 on: September 04, 2017, 02:50:18 PM »
Hi,
to continue...the value end of the market..

https://warp9td.com/   the ESS recommended
https://www.pmdx.com/ the 411 for simplicity and price the 424 for increased functionality
https://www.poscope.com/products/pokeys-devices/ the 57CNCd25 for simplicity and price the E57 or E57CNC for functionality

http://www.cncdrive.com/ have useful product, the simple and cheap UC100 is Mach4 capable, the larger and more functional products are Mach3 capable only

http://www.vitalsystem.com/portal/index.php have a superb reputation but useful solution start at $600.

You may note that all of the recommendations I've made are Mach4 capable which is also my recommendation. The list might be  somewhat longer for good Mach3 only
products like
http://en.cs-lab.eu/
http://dynomotion.com/KFLOP.html

Craig

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

Offline sn96

*
  •  53 53
    • View Profile
Re: Motion controllers - Take me to school
« Reply #24 on: September 04, 2017, 03:02:00 PM »
Hi,
to continue...the value end of the market..

https://warp9td.com/   the ESS recommended
https://www.pmdx.com/ the 411 for simplicity and price the 424 for increased functionality
https://www.poscope.com/products/pokeys-devices/ the 57CNCd25 for simplicity and price the E57 or E57CNC for functionality

http://www.cncdrive.com/ have useful product, the simple and cheap UC100 is Mach4 capable, the larger and more functional products are Mach3 capable only

http://www.vitalsystem.com/portal/index.php have a superb reputation but useful solution start at $600.

You may note that all of the recommendations I've made are Mach4 capable which is also my recommendation. The list might be  somewhat longer for good Mach3 only
products like
http://en.cs-lab.eu/
http://dynomotion.com/KFLOP.html

Craig



Thanks for those links. I like the Model PMDX-340 that was in the second link you provided.
________________
Mike

Sherline 5400 with Sherline steppers
Visual Mill 6.0 Standard
Mach3
Windows XP
Re: Motion controllers - Take me to school
« Reply #25 on: September 04, 2017, 03:13:04 PM »
Hi,
the 340 is NOT a motion controller it is a boxed parallel input breakout board.

You need either the 411 as a single port output controller or the 424 which has two effective ports worth of IO.

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

Offline ger21

*
  • *
  •  6,289 6,289
    • View Profile
    • The CNC Woodworker
Re: Motion controllers - Take me to school
« Reply #26 on: September 04, 2017, 09:53:33 PM »
Quote
http://www.cncdrive.com/ have useful product, the simple and cheap UC100 is Mach4 capable, the larger and more functional products are Mach3 capable only

They now also have a Mach4 plug for the UC300ETH. The UC400ETH plugin will probably follow.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html
Re: Motion controllers - Take me to school
« Reply #27 on: September 04, 2017, 11:05:07 PM »
Hi ger21,
kool, that's good to know. I think we are well served by our long term suppliers with Mach4 ready products.

Craig
My wife left with my best friend...
     and I miss him!
Re: Motion controllers - Take me to school
« Reply #28 on: September 05, 2017, 03:03:05 AM »
Hi Mike,
I notice the servo you are looking at has rated torque of 58 oz.in.  Seems very wimpy to me...don't under do it...you'll regret it each time the servo
faults. When a stepper gets pushed it gets hot, starts making noises, push a bit more its starts losing steps and then stalls. An under specified servo
gets hot, starts making noises, push a bit harder and the drive faults 'following error'. If a servo can't keep up and lags behind where its supposed to
be by 20 (programmable) encoder counts it faults out and usually E-stops the machine.

While cutting forces can prevent a servo from keeping up it much more likely because it can't accelerate fast enuf to follow the trajectory you have set it.
Torque= J(moment of inertia) x dW/dt (angular acceleration).... that is without torque there is little or no acceleration and so it will take forever to get to
any speed.

May I suggest an experiment. You say that the Gcode job to machine the item pictured takes several hours to run. Have you tried the 'Simulate Program
Run' button on the toolpath tab? I use it a bit and find it accurate enuf for most purposes. Here's the experiment....edit your motor tuning as if your steppers
can go twice as fast but leaving everything else the same and <Simulate Program Run>, how much did your cycle time improve? Now go back and edit
your motor tuning with your regular speed but double the acceleration setting on all the steppers and <Simulate Program Run>, how much did your cycle
time improve.

There have been a number of posts that talk about the relative merits of speed verses acceleration and most of us vastly underestimate the acceleration
required to achieve fast cycle times.

While my mill is homemade it is cast iron and steel probably not dissimilar to your machine in terms of weight. With the gearboxes on the steppers I have
rated torque of 705 oz.in. I have settled for 0.3g for acceleration, the steppers can handle better than 1g but the whole machine rocks around and has to be
bolted down. I can't help but think 58 oz.in sounds too low.

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

Offline RICH

*
  • *
  •  7,419 7,419
    • View Profile
Re: Motion controllers - Take me to school
« Reply #29 on: September 05, 2017, 06:58:27 AM »
Quote
If speed, quietness and step accuracy are important factors to me, then the value relates to what is important to the end user. No?  

It's your money so do as you wish with it,BUT, when asking for advice realy consider what is being said.

Step Accuracy - testing we did with steppers counting steps / pulses ( ie;simutaneous recording  of 6 inputs for comaprison with a measurement accuracy of 29 ppm ) showed that actual steps / pulses  sent at various points along the system could for all practical purposes be spot on through the system if the system is capable of it.

Heck my engraving machines smallest actual move is 0.000040 but is limited to 7 ipm!

Yes step accuracy is important as it relates to theoretical resolution but actual accuracy
of the system comes from the mechanical components.

In 2.5 / 3d machining acceleration is very important (more so than velocity) and even if the actual material removal is small, fast  acceleration  ( which requires torque ) provides for reliable movement of small axis moves.

Hmm......, spend your money so you feel good, let us know how all goes!  ;D

RICH