Hello Guest it is March 29, 2024, 05:06:19 AM

Author Topic: Linear encoders and steppers (newbie ?)  (Read 5843 times)

0 Members and 1 Guest are viewing this topic.

Linear encoders and steppers (newbie ?)
« on: August 10, 2009, 02:16:52 PM »
Well, I got stuck with this project at work, and I am green when it comes to CNC setup and conversion, so please bear with me.

I'm currently researching what I need to retrofit a "Handyman" knee mill. I'll be replacing everything. I've been reading in the posts about position loss using stepper motors, and a question came to mind.

Can I use feedback from the glass scales currently installed on the mill with the new stepper motors to create a closed loop system for positioning and running? If I can, I'll use a double parallel port so that I have enough input. (Anybody know of a good double BOB?)

I am hoping this is the case as it would eliminate problems with both loss of position, and backlash.

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #1 on: August 10, 2009, 02:24:43 PM »
No, you can not get closed loop in Mach itself. You could use encoders to fault out Mach if a certain amount of error is reached but for that you would need to write some sort of macropump up to look at the encoders and the commanded position.
Having said that if you have sized and tuned your steppers correctly then you will not lose position unless you crash.
 With a servo system you will have closed loop back to the drives, Mach is still only presuming the axis are where they are meant to be but the drives will do their best to make sure they are and if not will fault and tell Mach (if you have that set up) Again properly sized motors are needed or you will constantly be faulting with following errors.

Hood
Re: Linear encoders and steppers (newbie ?)
« Reply #2 on: August 10, 2009, 03:01:34 PM »
Thanks, Hood.

The only reason I asked was because there are encoder inputs, and I haven't seen any where in the software where it differentiates between steppers and servos.

I'll be using 1260 in/oz motors utilizing Gecko G203V drives. Should that be sufficient?

Will I be able to use the DRO ccrdinants in manual mode?

I appreciate your help and input.

Keith

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #3 on: August 10, 2009, 03:09:15 PM »
 Not sure what a Handyman mill is like, presume similar to a Bridgeport series 1? If yes then these motors should be fine, maybe even direct drive but certainly with 2:1.

Even with servos the encoders dont go back to Mach, they just go back to the drives. You could have them going to Mach if you wished but Mach would not act on them unles you had a special macropump made up to monitor them and then it would only act as an error checking feature and not a closed loop.

Yes you can input the encoders and use them as confirmation of position or for  DROs in manual positioning.
Hood

Offline simpson36

*
  •  1,369 1,369
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #4 on: August 12, 2009, 07:49:59 PM »
Gecko 204V is a great stepper drive. I just sold the last of mine and they did a great job for me while I had them.

You are handing them a tall order with those big steppers. The Geckos top out at 7A

Why have you decided on steppers over servos?
Re: Linear encoders and steppers (newbie ?)
« Reply #5 on: August 13, 2009, 09:04:28 AM »
Gecko 204V is a great stepper drive. I just sold the last of mine and they did a great job for me while I had them.

You are handing them a tall order with those big steppers. The Geckos top out at 7A

Why have you decided on steppers over servos?

Quite frankly, cost and inexperience are the 2 major factors. If I screw up, it won't cost too much to fix my mistake, and maybe I won't get fired. (j/k)

The steppers I've spec'd are rated @ 4.0 A/phase. I'm hoping that they will do the trick. The parts I'll be making (right now) are all fairly small and speed isn't as much of a factor as accuracy is. If everything works out this time, I might be able to convince the powers that be to cough up the funds to go all the way. We have 6 Bridgeport type knee mills, and this is the 1st I'll be refitting, but ti probably won't be the last.

I appreciate your input. I've relied a great deal on info that I've gleaned from the posts on here. You (Simpson) and Hood seem to have the best handle on some of the idiosyncrasies of a project like this.

Thanks for your input

Keith

Offline simpson36

*
  •  1,369 1,369
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #6 on: August 13, 2009, 12:28:21 PM »
Steppers can be set up to run in series or parallel. From the size motor and only 4A, sounds like series. That would be a mistake in my opinion. By way of example, here is a quick discussion using round numbers:

If those are 8 wire motors, running them in series reduces the current, but doubles the voltage. leaving you with only 10X possibility with the Geckos 80V limit. Parallel would put you at 8A but give you 20X voltage. Take a look at the chart on this similar motor:

http://www.kelinginc.net/KL34H2160_62_8A.pdf

When I first decided to plunge into a retrofit, my initial research quickly introduced me to the never ending stepper vs servo debate. Some of the 'common knowledge' is true and some is not.  I am of the opinion that one cannot  reasonably compare an animal know as 'servos' to steppers, because the term 'servos' is completely inadequate to describe the species. Within the various servo types there are significant differences in cost and complexity that often are not mentioned. AC servos get VERY expensive in a hurry, whereas DC brush servos (arguably the logical first tier upgrade from steppers) are actually very competitive with steppers on a total installed cost basis.

True: DC brush servos are more complex to select and tune for a given application . . UNLESS . .  you have a reliable resource to copy from. In which case, the complexity moves to nearly zero.
        Gecko 203V drives require tuning, albeit very simple compared to servo drives.
        Small and medium sized DC brush servo installs can be nearly the same installed cost as steppers

False: Blanket statements like: 'Servos' cost 3 or 5 or 10 times what steppers cost.
         Blanket statements like: Steppers are 'just as good' as servos. This is equivalent to saying a Corvette is 'just as good' as a pickup truck. Neither is 'better', it depends on what you are trying to move and how fast.
         Steppers can be compared to servos by their torque ratings.




Offline RICH

*
  • *
  •  7,427 7,427
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #7 on: August 13, 2009, 12:50:01 PM »
HMM....
Good stuff for sure,yep there are choices / decisions to made. But Sometimes you need to pull back and
say, what do i have, what do i want to do, what are the limitations.....
So the "system" should be designed to meet a need. Each part of the system requires a lot of knowledge along
with experience and practical considerations.
I like to think of it all, like a stereo system, as what is heard is only as good as all of the components of the system.

Just a few thoughts,
RICH

Offline simpson36

*
  •  1,369 1,369
    • View Profile
Re: Linear encoders and steppers (newbie ?)
« Reply #8 on: August 13, 2009, 01:03:33 PM »
"If I screw up, it won't cost too much to fix my mistake, and maybe I won't get fired."

Time for a bedtime story "Why bosses actually fire people"  >:D.

I had a designer working for me many years ago who came into my office and proudly announced he had saved my department money by repairing a tool. Turns out the clogged hot melt glue gun was brought to him by one of the model makers who then proceeded to spend the next hour watching the senior designer fix the glue gun. To his surprise, rather than kudos, I informed him that his pay would be docked for any time he spent in the future fixing tools. Employees have a tendency to not regard their own salaries as 'cost'. In effect, I now had a $150 'Refurbished' hot melt glue gun charged against my billable man-hours account instead of a new $20 tool charged to the general shop equipment budget. More importantly, I was now two man-hours in the hole on the project they were supposed to be working on. I explained to the designer the priorities at issue and that neither the CEO nor the client is going to ask me in a project meeting why I spent $20 on a new glue gun, but they are going to be keenly interested in why my design group held up a critical path item.

I'm just glad there was not a shop crew at $hundreds per hour sitting idle waiting for a drawing while the designer was fixing a $20 glue gun? The shop forman would certainly have been asking for one of my budgets to charge the time to. Those are the realities that bosses deal with.
  
In my opinion, if one is concerned about the success of an assigned project, steppers are a bigger risk due to some of their peculiarities. Employee salaries are going to be a very large portion of the total cost to implement the capability. Mo' better to do it right the first time, even if the cost is initially higher, and be in a defensible position if things get messy. Things to consider:

However much the hardware and man-hour cost to implement the new CNC capability would be appropriated up front and since it a first endeavor, could probably go over budget without too much repercussion. Lost production and man hours lost to endless diagnostics trying to figure out why the steppers are suddenly loosing register and ruining parts is going to be visible to the degree of being a neon sign on the bosses office.

Servos will not eventually be found inadequate and need to be upgraded to steppers, thereby converting all hardware and labor costs (from the original conversion) from investment to loss. Bosses do not like loss.
Servos are not disposed to mysteriously 'learning' new bad behaviors that were absent yesterday and will take three days to track down and fix. Numerous active posts attest to that.
Servos do not stealthily and with no warning or cause (seemingly) ruin parts by happily cutting along many invisible silent steps from where they should be.

Your boss will never come to you and say 'why didn't you go with steppers in the first place?"

 
FWIW . . .
« Last Edit: August 13, 2009, 01:34:33 PM by simpson36 »