Hello Guest it is March 28, 2024, 04:11:05 PM

Author Topic: Glad |I asked first  (Read 8131 times)

0 Members and 1 Guest are viewing this topic.

Re: Glad |I asked first
« Reply #20 on: February 03, 2018, 08:08:46 PM »
Hi,
I found the quote on the MONTENC Lite page. I see the price for the cheaper device is still there at $495 but just above the price list was the note
about the availability of it.

http://www.vitalsystem.com/portal/motion/motenc/motenc_lite.php

Craig

I think you might be a little confused. The 7544 MOTENC-Lite PCI 4-Axis Board (8DAC, 4ENC, 0ADC, 48 I/O) is being dropped but that is a LinuxCNC compatible board that goes into a PCI slot.
The Integra-Hobby Motion Controller 4-Axis 125KHz is #7754 Whic is an Ethernet controller.

Mike
We never have the time or money to do it right the first time, but we somehow manage to do it twice and then spend the money to get it right.
Re: Glad |I asked first
« Reply #21 on: February 08, 2018, 03:40:23 PM »
Wow, thanks for saving me the expense and disappointment by creating this thread, this should be a sticky! I'm really glad I checked/clicked here as I was just about to order the same exact board with a Mach4 license.

When I saw the vendors/manufacturers listed on the Mach4 page I understandably assumed that they were the recommended boards that were fully compatible (and tested) with Mach4's features. In fact, the ONLY reason I was going the Mach4 route over the GRBL setup was because I wanted the backlash compensation. The search continues...

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Glad |I asked first
« Reply #22 on: February 08, 2018, 03:58:08 PM »
Quote
In fact, the ONLY reason I was going the Mach4 route over the GRBL setup was because I wanted the backlash compensation.

This has been debated to death and in some cases backlash comp. can work. The problem is the 99.9% of the cases it wont work in. Backlash compensation on a control truly is an oxymoron. And in those cases it wont work in, this is not just my opinion it's a simple fact. Until someone finds a bit of software that can rewrite the laws of physics the only fix for backlash is removing the mechanical backlash.
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!

Offline thosj

*
  •  532 532
    • View Profile
Re: Glad |I asked first
« Reply #23 on: February 08, 2018, 04:24:45 PM »
Absolutely right. Thinking you're going to take a Bridgeport with .015" backlash in the acme screws and comp it out with software is a fallacy. If that worked machine tools wouldn't need ground ways with scrapped Turcite, ball screws, pressure fed oiling systems. Just whack the machine together out of cold rolled, put a bunch of sensors on it, and comp the whole thing in software as it teeters along.

Take circular interpolation, the X is slowing down as it gets to the right or left side at which point the Y is going full tilt, when X hits the turn around point, what does the X do, jerk the .015 out? Even if some algorithm "eases" it out over time, you still don't have a circle!! And even if it jerks it out, it can't jerk it all out in 0 time while the Y is going full feed rate. 015" is just an exaggeration to make the point, but the same thing happens trying to comp .0015" out. It DOES sound good in the literature, tho'.

All these motion controller sellers spending all this time on Backlash Compensation when they could be spending time on real stuff.
--
Tom
Re: Glad |I asked first
« Reply #24 on: February 08, 2018, 04:26:55 PM »
Thanks for the responses, however, I guess as a programmer and a novice machinist I'm not following the reasoning, and the fact that Mach3/4 has backlash compensation (as well as LinuxCNC and others) means it's something that can be adjusted to work. I've also seen folks who've upgraded from grbl to linuxcnc for this reason and were able to dial it in. Thinking through this for the very first movement you can drive the motor in the opposite direction by the compensation amount, then move it in the correct direction (again in the backlash compensation amount) at this point that axis is good to move until it changes direction (so it does the compensation routine again), no?

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Glad |I asked first
« Reply #25 on: February 08, 2018, 05:41:12 PM »
Wish I had time to explain it all but I don't. It has been explained very well many, many times already. Here (on this forum) and other places.

But I can save you some time. Just reread my previous reply. It's been my job for 25 years to make machines produce good parts. thosj gave a real good real world example. If you have code that rewrites the laws of physics then please ignore my input. Otherwise backlash comp is a selling point most software and hardware manufacturers are pressured into implementing by customers that are certain it is an acceptable solution.
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!
Re: Glad |I asked first
« Reply #26 on: February 09, 2018, 12:47:54 AM »
Hi stoic,
have you tried 'conventional milling verses climb milling' in a machine with significant backlash?

The old Bridgeport copy at work has 0.2 inch backlash in X. If you try climb milling you will break something....I've done it myself and watched and tried to warn
a workmate from the same trap to no avail...he broke a new 16mm High Speed Steel tool, and that takes quite a bit of doing!

The only software that is going to help with backlash is the stuff that starts: 'Dear God'.... and ends in 'Amen'

Craig
'I enjoy sex at 73.....I live at 71 so its not too far to walk.'

Offline thosj

*
  •  532 532
    • View Profile
Re: Glad |I asked first
« Reply #27 on: February 09, 2018, 08:33:05 AM »
Hi stoic,
have you tried 'conventional milling verses climb milling' in a machine with significant backlash?

The old Bridgeport copy at work has 0.2 inch backlash in X. If you try climb milling you will break something....I've done it myself and watched and tried to warn
a workmate from the same trap to no avail...he broke a new 16mm High Speed Steel tool, and that takes quite a bit of doing!

The only software that is going to help with backlash is the stuff that starts: 'Dear God'.... and ends in 'Amen'

Craig

Nice one, Craig.

.2" of backlash, INCHES? Tell me that's a typo!  :o Now THAT would be some software to comp THAT out, that's a full turn of the screw!!

Tom
--
Tom

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Glad |I asked first
« Reply #28 on: February 09, 2018, 08:53:02 AM »
Exactly Craig. Programming strategy (when it can be used) is a far better approach to easing the pains of backlash than backlash compensation.

Lol, we had a bed mill once that the table had so much backlash in it you could see it wobble when changing directions (wish I had a video). We would tweak the program as necessary to comp for it and ran it several months till the new machine could get there. It made a lot of good parts in that time but what a pain it was to pull it off.
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!
Re: Glad |I asked first
« Reply #29 on: February 09, 2018, 02:34:14 PM »
So basically what you guys are saying is that Mach 3/4's as well as the few controller that have implemented backlash compensation is all smoke and mirrors to get the customers to buy into their hype? Glad I learned something new.