Author Topic: Underpowered Y axis  (Read 3909 times)

Offline Dragonfly

  • CNC Jedi
  • *****
  • Posts: 2380
    • View Profile
Re: Underpowered Y axis
« Reply #15 on: October 23, 2019, 14:52:10 pm »
Yes Dave, fully agree.
That's why when using open loop drive (ordinary steppers) they should be selected to have some excess of torque in reserve.
Besides true servo drives there are (and became more on offer) hybrid servos using stepper motor with encoder. Price wise they are somewhere between stepper + driver combo and true servo drive sets. If, however, the stepper stalls they provide a 'fault' signal to the machine control.

Offline Garyhlucas

  • CNC Jedi
  • *****
  • Posts: 1378
    • View Profile
Re: Underpowered Y axis
« Reply #16 on: October 23, 2019, 19:08:32 pm »
Take that accuracy thing a little further.  On my 0.2” pitch ball screws that have no backlash that 5% step error on my 200 steps per revolution motor is a position uncertainty of 0.00005”.  Now consider that over thousands of moves where I have not lost any steps I am certain that I am within that tolerance of true position along the path.

 Then consider that servos require error of some magnitude in order to respond and they don’t just stop if the error increases until they through a alarm for servo lag.  On high performance machines you have to be aware of servo lag as it causes rounded corners and other issues. When roughing that may be acceptable, but you may need to tell the CNC what the acceptable deviation is for finishing.

So when you compare servo VS stepper you need to be aware that in order for the servo to be substantially better it is actually going to cost way more. In fact on small machines where bigger steppers are available spending the same dollars as servos is likely to get you very similar performance.  The place where this isn’t really true is when you need really high speed, which is mostly about acceleration not velocity.
Gary H. Lucas

Have you read my blog?
 http://a-little-business.blogspot.com/

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1353
    • View Profile
Re: Underpowered Y axis
« Reply #17 on: October 24, 2019, 01:01:18 am »
Yes Fly.
I have had a look at those Hybrid stepper systems and agree they are a half way house.
They do operate in servo mode as opposed to open loop that is say they have a PID loop
and an external, rotary encoder for feedback.

They do offer a performance advantage over steppers and are cheaper than Servo systems.
Here's a random torque sheet from a Hybrid set-up from the internet. eBay
They say that it can do 1000 rpm but the torque is 40 percent less than at 200 rpm.

If I were given one,(A Hybrid Drive and Motor) would I use it, Hell Yeah!, But would I replace my nema 24 570 oz\in  motors and Leadshine drivers (real ones). 
Well No there's not enough 'Advantage to gain' for me.

What I will do though is when each drive expires I will replace it with the Hybrid Drives and motors.
I do note that they have come down in price quite a bit since the last time looked.

Garry it's plus or minus 5 % so it's 10 %, it's no co-incidence that the step angle error is 10 % and you should step no more than X10 to ensure accuracy.

If you are interested, the subject is called Control Theory.
I'm subbed to this guy he's good.
https://www.youtube.com/user/ControlLectures
Dave

Offline Dragonfly

  • CNC Jedi
  • *****
  • Posts: 2380
    • View Profile
Re: Underpowered Y axis
« Reply #18 on: October 24, 2019, 07:22:41 am »
From my practical experience less than 1/10 microsteps bring worse results. And it's not only the motor mechanical quality but also the driver electronics ability to maintain very small discrete voltage steps to the coils.

Offline Garyhlucas

  • CNC Jedi
  • *****
  • Posts: 1378
    • View Profile
Re: Underpowered Y axis
« Reply #19 on: October 24, 2019, 18:11:58 pm »
Dave,
Yes it is plus or minus 5%, as in +-0.00005” of true position! Good enough for what I do!
Gary H. Lucas

Have you read my blog?
 http://a-little-business.blogspot.com/

Offline Garyhlucas

  • CNC Jedi
  • *****
  • Posts: 1378
    • View Profile
Re: Underpowered Y axis
« Reply #20 on: October 24, 2019, 18:22:24 pm »
You’ll notice that virtually all the good stepper drivers offer microstepping up to 256 steps per full step. I think there are two uses for this.  The first is getting extremely smooth motion at very low speeds, say driving a telescope mount.  The second is combined with a high resolution encoder holding a very exact position or following a very exact path again at low speeds.

Genuine Leadshine drives were mentioned. In the world of counterfeits on Ebay, Amazon, etc. where are you gauranteed to get the real thing?  Personally I think Ebay should be forced to clean house because all the benefit there is going to cheating customers and real manufacturers.
Gary H. Lucas

Have you read my blog?
 http://a-little-business.blogspot.com/

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1353
    • View Profile
Re: Underpowered Y axis
« Reply #21 on: October 25, 2019, 04:03:34 am »
Garry you have confused the property of resolution and the property of Accuracy (with includes the error terms).
To cut to the chase, by the time you take into account the Step angle error and Drive electronics error and the Lead Screw error.
And more importantly the rigidity of the machine itself (which on home made machine can be a much greater magnitude than a built for purpose machine).
Then your Repeat-ability  and Accuracy figures (the ability to move your tool bit around and to actually know where it is) little lone put a cutting load on it are drastically reduced.

Here is the  standard test you could to use to really see what your machine is doing.
https://www.iso.org/standard/36688.html

Here is a list of the expected tolerance of parts you would expect from each process, claiming much more performance this means you have made an error in you calculations.

Quote
The second is combined with a high resolution encoder holding a very exact position or following a very exact path again at low speeds.

Then it is acting in servo mode meaning no Step Angle Error term in the calcs.

Quote
Genuine Leadshine drives were mentioned. In the world of counterfeits on Ebay, Amazon, etc. where are you gauranteed to get the real thing?  Personally I think Ebay should be forced to clean house because all the benefit there is going to cheating customers and real manufacturers.

Agreed, you're never going to stop though, Honestly if people new just how much better the Gecko\Leadshine drives work as compared to say a TB 6600 they would be shocked, I do have some, to use for  my sons projects (He has a girlfriend now and has lost interest) and have variety of spare servo's about.

I have directly compared the bench a TB6600 to a Leadshine DM556  and the difference was night and day in favor of the DM556.

I have directly compared (on the machine so under typical load) the Gecko 80v 20 servo drive to a DCS810  and the Gecko won hands down.
 
If I won the lotto I would Buy a set of Clearpath Nema 23 Servos they have the very latest software with RAS.
I have downloaded the C++ software that they provide with which you can write your own motion control.
I have non-cnc applications where this would be fantastic.

Dave

Offline chuckeroo

  • Wookie
  • ****
  • Posts: 465
    • View Profile
Re: Underpowered Y axis
« Reply #22 on: October 26, 2019, 13:47:47 pm »
The Gecko 201 x's is what's on the mill with 48 volt power supply's. Manufacturer's can recommend Motor size's for moving loads,got that . But they can't determine the cutting load. A few factor's come into play, depth of cut, sharpness of bit,type of material among quite a few other's, no doubt.The user has to have enough insight look at and measure what he has cut and make MOp adjustment's to compensate for short comings of the mill. Now where is that winning lotto ticket anyway, daddy needs a new set of CLEARPATH'S.  
Thanks for all the food for thought.
« Last Edit: October 26, 2019, 13:55:41 pm by chuckeroo »

Offline tshawkins

  • Ewok
  • *
  • Posts: 11
    • View Profile
Re: Underpowered Y axis
« Reply #23 on: November 09, 2019, 02:23:33 am »
I bought a couple of of nema 23 kl23h2100-50-4bm. Using them on the y&a axis for the gantry movement. Best speed with a 3/8 upcut spiral is around 50 ipm. I really need to bring this speed up to 100 ipm. Thinking about using 2 clearpath servos on this axis. Used an online calculator and came up with 38 watts for a 200 pound gantry. Is this correct?The cheapest servo they sell is 38 watts and thats per unit . point being if i have 2 on this axis it should be 76 watts right? Which should take care of my problem plus quite a bit of wiggle room? Lastly how do i verify how many watts the kelling is putting out. Thanks .Thing is i have spare 48 volt power supplies and the motor mounts are in place as we speak. Hmm just occured to me i didnt factor in the load of the bit cutting the wood unless this is already factored in?

Have you tried cutting tbe number of microsteps per step, i had a simular probblem with a much smaller rig where it was having problems with both the x and y axis, cutting tbe number of microsteps  from 16/step to 4/step had a significant effect on available torque, making tbe machine viable again for cutting aluminium. It did not seem to impact percison to much, i still have 400 microsteps/mm, down from 1600.