MechMate CNC Router Forum

Go Back   MechMate CNC Router Forum > Common Assemblies & Parts > Motors & their mountings
Register Options Profile Last 1 | 3 | 7 Days Search Today's Posts Mark Forums Read

Reply
 
Thread Tools
  #1  
Old Thu 30 November 2006, 06:58
Manjeet Singh
Just call me:
 
Hi,

I have reading about MechMate and every one was after stepper motors and controllers. As steppers run in a open loop systems there are chances of getting lost steps which can somtimes ruin the part we are engraving

There is DIY UHU servo motor controller project By Mr. Ulrich Hubber. What a great controller for hobbiests. I have made one and tested on my small engraving machine and had very good results

For more details on UHU controller Please visit http://gsst.wikispaces.com/UHUget

Manjeet
Reply With Quote
  #2  
Old Thu 30 November 2006, 07:15
Gerald_D
Just call me:
 
Sorry Manjeet, for your very first post, I think this is advertising. You are welcome to start a discussion on servo motors in general. Servo systems are a lot more complicated subject than simply to push Mr Hubber's controller. (I notice that you are a commercial seller of kits to build that controller.....)
Reply With Quote
  #3  
Old Thu 30 November 2006, 10:04
Manjeet Singh
Just call me:
 
First I would like to explain that I am purely a hobbiest and I had no interest in selling Kits of UHU controllers. As I had to make one set for my own use I made the PCBs the cost of making three pcbs was quite high so I made 50 instead. for the balance I offered to other cnczone forum members at no profit no loss only.

Then came the components story the same thing happened again so I ordered for more quantity componets and offered others complete as well as hard to find components.

Above all I am by profession a mechenical engineer running my own manufacturing business of printing and packaging related machinery with a skilled workforce of 36

Manjeet
Reply With Quote
  #4  
Old Thu 30 November 2006, 11:30
Gerald_D
Just call me:
 
Manjeet, I mean you no harm, and I believe you could be an asset to this group. However, it was not a good start for your first post to divert us from the plans to a site where you are listed as a supplier.

Why don't you, as a mechanical engineer, start off by telling us how we could mount servo motors (with their gearing/belting systems) onto a MechMate? Show us some examples? (Remember we are talking of a fairly big machine here).
Reply With Quote
  #5  
Old Thu 30 November 2006, 21:57
Manjeet Singh
Just call me:
 
Hi Gerald,

Thanks for understanding me and my thaughts, Previosly I had built a small engraver with drawer slides and steppers it was not as accurate I wanted. I tested it with servos but again he backlash in the slides caused the problems. Now I am on a halfway completeing 30"X30" table engraver. I hope It will be completed within december 06, as I had very little time for my hobby.

More over I had discussed MechMate with my younger brother he is also a mech engineer and shrtly we will start buiding this. I have all the facilities of machining and fabricating in my workshop. For slides I can arrange for grinding V upto 4.5 meters.

I will post pics of my engraver as well as "The MechMate Beast" when I finished up.

Regards....Manjeet
Reply With Quote
  #6  
Old Thu 30 November 2006, 22:58
Gerald_D
Just call me:
 
4.5m V-grinding! you sure are a lucky person! Here in Cape Town we can do 3m max, while Johannesburg does a lot longer. You can weld your rails to the y-gantry before grinding and then take the whole gantry to be ground as one unit.

Tell us a little more of the technical aspects of the UHU controller and then I will put this topic back in the right place.
Reply With Quote
  #7  
Old Thu 30 November 2006, 23:58
Manjeet Singh
Just call me:
 
Please do read the following pages and then decide.

http://www.uhu-servo.de ( Authors Site)
http://gsst.wikispaces.com/ (Wikipage for servo controllers)
http://www.cnczone.com/forums/showthread.php?t=14217 (A complete discussion on open source DIY servo Controller)

The Main feature of this controllers you can modify it according to your power needs and also if any parts smokes out you can replace it thus making easy to repair thus saving your $$$ for replacing drives.

More over ordinary DC motors with encoders can be used as servomotors another cost saving feature for DIY hobbiests. No need to buy costly servomotors.

Manjeet
Reply With Quote
  #8  
Old Fri 01 December 2006, 01:41
Manjeet Singh
Just call me:
 
Gerald,

One more point to use UHU servo controller is being its Open Source it does match your open source DIY MechMate. So why use dedicated controllers. Who ever the person on this forum here must put his source/ideas to faclilitate others.

Manjeet
Reply With Quote
  #9  
Old Sun 03 December 2006, 07:34
Gerald_D
Just call me:
 
Manjeet, some of the things about servo's that it make it difficult for a DIY builder:

1. They have low torque and therefore there must be fitted with "torque-multipliers" (gearboxes/belts/screws).

2. They are expensive if you want to buy ready-packaged NEMA-34 solutions - otherwise you need to find individual cheap solutions on E-bay and in scrapyards.

3. They need a lot of wiring (many connections) - a mistake in the wiring normally makes the motor run out of control. A mistake with a stepper causes no motion.
Reply With Quote
  #10  
Old Sun 03 December 2006, 10:15
Mike Richards
Just call me:
 
Gerald,
I agree with everything except point #2. The ID3300x brushed servos from MCG cost about the same as the Oriental Motors PK299 steppers. The ID3300x are in the $150 price range, depending on the exact model. Adding an encoder from US Digital adds another $60, again depending on the model. The Gecko drivers for servos start at $114, or about $20 less than the G202 for steppers.

BUT, as you clearly pointed out, you'll have to gear them down at least 10X, maybe more before they're usable. That is a big issue, since a simple belt-driven gear box is in the 3:1 range. Of course, using a gear box also multiplies the motor's torque.

One big advantage of using servo motors over steppers is that they hold full torque when stopped - without drawing much power (again, depending on how much 'pull' is being generated by the load). A stepper normally goes into 'power-down-mode' to reduce heating problems.

If I can ever find a practical solution for the gear-box problem, I'll mount servos on my Alpha.
Reply With Quote
  #11  
Old Sun 03 December 2006, 10:44
Gerald_D
Just call me:
 
Mike, by "ready-packaged" in point 2 I meant that to mean inclusive of gearing. With steppers, we can take a $200 package off the shelf and screw it to a MechMate and it will immediately cut wood reliably. Right now there is no $200 servo option straight off the shelf.

Manjeet's point of comparing open loop steppers to closed loop servo's is not valid since steppers can also be fitted with feedback. And we have seen with the SB Alpha that closed loop can give a false sense of security.

Servo's with 3:1 belt-drives are practical, but not with rack & pinion. 3:1 belt drives work with ball-screws, but long ball-screws needed by big routers is a whole big problem area by itself. Hobbyists get away with servo's because they typically have short ball-screws and a fairly clean environment.

The appeal of servo's over steppers is the smooth (non-step) motion.
Reply With Quote
  #12  
Old Sun 03 December 2006, 12:29
Greg Waggy
Just call me:
 
Gerald, You forgot to mention that servo motors also fail to get lost should a blip happen in the communications. Here's another place that has servo motors for under $200.00: http://www.cadcamcadcam.com/index.as...TS&Category=11

I don't know how good they are or if they will fill the bill for power.
Reply With Quote
  #13  
Old Sun 03 December 2006, 12:48
Gerald_D
Just call me:
 
I didn't forget to mention that "fact" because it isn't true. This is that false sense of security I was talking about.....a servo will recover from a lost position, but for the duration of the time that it is lost it will leave a mark on the cutting job.

None of those "servo-motors" can be bolted straight to a MechMate and start cutting.

The name "servo-motor" has become to be applied to most any DC motor. Sure, one can find DC motors at all prices and qualities, while stepper motors are more select.
Reply With Quote
  #14  
Old Sun 03 December 2006, 12:50
Greg Waggy
Just call me:
 
Hmmm, I didn't know that. Thanks for that information because it sure is passed around that they won't get lost.
Reply With Quote
  #15  
Old Sun 03 December 2006, 13:36
Gerald_D
Just call me:
 
Greg, there is a lot of popular half-truths out there....

The broad belief is that servo-motors have encoders to give positive feedback for closed loop control and that this makes them better than stepper motors. So, what if an encoder is fitted to a stepper motor, etc., etc.??

Point is that the type of motor has nothing to do with the closed-loop issue - any type of motor can be fitted with an encoder. Most people fail to recognise this.

When a motor (servo/stepper/whatever) is slightly under-powered for the application, an encoder feedback system allows the motor to play catch-up when the load is lighter. Open-loop steppers don't let you get away with an underpowered motor.

Regarding torque, the servo guys like quoting "peak" torque loudly and "continuous" (holding) torque softly. With a stepper motor, the peak & continuous are the same.
Reply With Quote
  #16  
Old Sun 03 December 2006, 14:40
Greg Waggy
Just call me:
 
I am glad you are willing to enlighten us. Me, anyway. I'm new to the CNC world and every little bit of information I can learn from is most appreciated. I wonder why encoders aren't being used on stepper motors.

Maybe, if I can find out how to do it, I will outfit my stepper motors with encoders and see how that works. You do seem to have a lot more power with stepper motors than with servo motors. And it's a lot easier to figure out just what power you have.
Reply With Quote
  #17  
Old Sun 03 December 2006, 19:37
Patrick Toomey
Just call me:
 
Gerald,
That answers a question I have long had which is basically, wouldn't over-powered steppers be the best solution since you can count on them not losing steps? It seems like bigger steppers are cheaper than all the other issues with servos, encoders, more advanced controllers, etc... I know servos have their place but I've been cutting with relatively small servos (steppers?-Gerald) for awhile now and short of crashing into a clamp (or three or four) I don't think I've ever lost steps. I would like to go with bigger steppers so I could push it harder but it seems to me that for machines of this size, servos are a solution in search of a problem.
Reply With Quote
  #18  
Old Sun 03 December 2006, 22:51
Gerald_D
Just call me:
 
Patrick, it appears that going bigger on steppers is not ideal either - I guess this is because they become too "steppy" (rough?) for smooth light cuts. Introducing gearing/belting and making a small stepper run faster seems like the best way to go. But, as you have noticed, you only lose steps when things go wrong - "lost steps" is not as big an issue as the servo guys want to make out.
Reply With Quote
  #19  
Old Mon 04 December 2006, 09:44
Mike Richards
Just call me:
 
Servo motors can be hard to work with. (That's why my servos are still being tested on my test bench instead of running my Alpha.) Gerald has made some very valid comments. Steppers are ideal in every way except holding torque when stopped - assuming that an auto current-reduction driver is being used. When that happens, a big, beefy 34 size stepper with 600 oz*in torque suddenly turns into a puny 23 size stepper with 100 oz*in torque. Granted, as soon as a step command is given to that motor, it immediately transforms itself into a 600 oz*in powerhouse again. BUT, the problem, for me, is that if that motor stays powered down, while a motor on a perpendicular axis starts to move, the inactive axis can get pushed around - 'Chatter'.

On the other hand, a 34 size servo ID3300x series has about 100 oz*in of continuous torque that has to be multiplied by a gearbox, giving 1,000 oz*in or more when it is geared to give about 15-ips feed speeds on rack/pinion machine. That 1,000 oz*in torque is still holding the motor in position when it is at rest. Unless something is trying to push the motor out of position, it acually draws very little current when resting.

Gerald's point about closed-loop servos with encoders giving a false sense of security is perfectly valid. It's the same problem that I've had with my Alpha, and probably worse. If I remember correctly, the Gecko G3xx series of servo drivers can send up to 127 pulses before 'faulting' if the encoder doesn't respond. The distance that the servo would move depends on the encoder being used, but it would certainly be enough to ruin the part being cut. So, whether using a stepper or a servo, moving too fast is going to ruin the work.

In a conversation that I once had with Mariss at Gecko, he suggested that most size 34 steppers should be run at 500 - 600 RPM to get the best compromise between speed and torque. On my Alpha, 200 RPM = 628 ipm or 10.5 ips, so to use Mariss's recommendation, I would have to gear the motor down about 3:1. That's exactly what I did and it works great. Holding torque is about 300 oz*in in reduced current mode. Torque at low speeds is over 1,500 oz*in. Torque at normal cutting speeds (5-8 ips) is still over 600 oz*in.

I would still like to switch to servos to get the highest possible holding torque, but spending close to $1,000 per motor for a good no-backlash gearbox makes it impractical at this point.
Reply With Quote
  #20  
Old Mon 04 December 2006, 10:27
Gerald_D
Just call me:
 
Mike, that holding torque thing is interesting....

If our ShopBot or MechMate is switched on, it takes about 70 lbs of force to move the gantry (and probably about 35 lbs to move the y-car), "breaking" the holding force of the static motors. These figures are estimates, I've never had reason to measure them.

But the time I did wish I could measure a holding force was when I pushed an Alpha car easily out of position. Your point about the current reduction causing a big loss in holding force seems to be peculiar to Alphas only? I don't think we are suffering that extent of problem with G201 drives straight into open-loop steppers. Our gantries/cars feel pretty solidly locked up even in current reduction mode. Would you be interested in a definite measurement?

We must accept that the lockup is to the nearest full step in open loop systems. Your Alpha may have been hunting in the microsteps?
Reply With Quote
  #21  
Old Mon 04 December 2006, 11:07
Mike Richards
Just call me:
 
Gerald, you might have solved the mystery for me. I've used stepper motors in various non-CNC machines for years, and my Alpha was the first machine to make me worry about reduced current mode. In total desperation, I added the 3:1 gearbox. When it worked, I thought about it over and over, trying to figure out why it so completely solved my problem. What you wrote makes total sense to me.
Reply With Quote
  #22  
Old Mon 04 December 2006, 13:08
Greg Waggy
Just call me:
 
Gerald, What do you think of this motor? NEMA 23 BIPOLAR STEPPER MOTOR 282 oz-in, ? Diameter Shaft with flat
Reply With Quote
  #23  
Old Mon 04 December 2006, 14:44
vadeem
Just call me:
 
I may add in Manjeet's defence, that the prices on components he offers in the kit are less than I can buy them anywhere here in the States. Many of them are 1/3 the cost here, even from surplus places like Hosfelt.

That being said, I'm still going with the Gecko 203V and steppers. I want to stick with the group for easy troubleshooting.
Reply With Quote
  #24  
Old Mon 04 December 2006, 15:12
ralph hampton
Just call me:
 
Gerald and Mike,
Very interesting...Is the "reduced current" mode when not stepping something that can be switched off, or is it essential for high performance systems? It is obviously used with alphas, is it also used with gecko drivers, (when driven by the alpha processor - g4 - as well as mach3?)? I get the feeling (though don't really know) that it isn't used with the standard prt.
Any enlightenment most welcome.
The more I get to know, the more I realise there is yet to learn...
Reply With Quote
  #25  
Old Mon 04 December 2006, 16:28
Mike Richards
Just call me:
 
Greg, I wouldn't use a 1/4-inch shaft motor for a CNC router.

Ralph, As far as I know, the 'reduced current' mode can NOT be switched off, nor would it normally be a good idea, because it would lead to the motors and drivers overheating. (Full current would be passing through the motors. That current would be converted to heat.)

I know absolutely nothing about the standard PRT. However, Oriental Motors supplies a driver in their entry-level CSK driver that allows you to turn off current reduction. (They strongly advise against turning it off.)

I'm in the middle of reading the driver manual for the AS911AA stepper motor. There is one reference in the manual for a current adjustment switch, but, so far, I haven't found detailed information about that switch.

STOPPED motor, FULL current = HEAT and full torque. STOPPED motor, REDUCED current = less heat and less torque. The trick in using stepper motors is selecting the correct components so that everything works as expected. (When someone finds a formula that always works to determine those components, please let me know.)
Reply With Quote
  #26  
Old Mon 04 December 2006, 17:43
Greg Waggy
Just call me:
 
Mike, I'm only using these for a proto type machine. Then they'll be put on my foam cutting machine. http://www.kelinginc.net/ has some husky such as NEMA 34 HIGH TORQUE STEPPER MOTOR 1810 oz-in, ? Single shaft with flat.
Reply With Quote
  #27  
Old Mon 04 December 2006, 21:45
Mike Richards
Just call me:
 
Greg, I hope that I didn't offend you by suggesting that 23 size motors were useless. As a matter of fact, I still have at least six Oriental Motor PK268-02A motors laying around from old projects when I used them exclusively for roll paper advancement in Kodak-S printer upgrades. They always performed flawlessly, but back then the loads were very light, speeds were low, and simple 8051 type micro-controllers were more than sufficient to create pulse trains for the motors.

Every few weeks, I catch myself thinking about building a small CNC-router with ball screws and linear rails for making small plaques and signs. Reality usually hits when I remind myself that I'd be spending several thousand dollars and only end up with a machine whose functionality would be duplicated on my current Alpha. Maybe someday the whimisical in me will win out because I would really like to see those little motors do their thing.
Reply With Quote
  #28  
Old Mon 04 December 2006, 23:14
Gerald_D
Just call me:
 
Greg asked: "Gerald, What do you think of this motor? NEMA 23 BIPOLAR STEPPER MOTOR 282 oz-in, ? Diameter Shaft with flat" Much too small for a MechMate, as Mike said.

Ralph asked about "reduced current mode"....
This is standard practice in non-Alpha's, MechMates, Gecko's and everywhere where a stepper gets locked into a stationary position. Everybody agrees it the good and right thing to do, and there is seldom any discussion about it. Mike is a bit suspicious of the reduced current's effects on his Alpha, but the Alpha is not a typical stepper motor application. As Mike said, it is not something that you want to switch off, because then the motors will cook.

GeckoDrive have published this excellent paper:
http://www.geckodrive.com/photos/Step_motor_basics.pdf from which this section is taken...


The reduced current mode drops the torque lower than the horizontal "rated torque" line, but still leaves you with more torque than you would have had at the higher speeds to the right of the curves. Ralph, I bet you that you have a reduced current system sitting in your workshop right now - are you having a problem that an axis cannot hold a stationary position? If not, then don't worry about "reduced current mode".
Reply With Quote
  #29  
Old Tue 05 December 2006, 01:06
ralph hampton
Just call me:
 
Gerald, thanks - that document looks invaluable. I have no problems in the position hold department. So prt's, gecko's etc lock when they stop, while alpha's depend on feedback repositioning. I have always thought that feedback systems were a bit cookie, and must involve an endless series of messages along the lines of:
Where are you?
Move left.
Where are you now?
Move right.
Where are you?
left a teeny bit...

etc etc
But then I know nothing..
Reply With Quote
  #30  
Old Tue 05 December 2006, 01:47
Gerald_D
Just call me:
 
Ralph, you have described the "hunting" perfectly!

But don't knock "feedback" and "closed-loop" too hard - they do work much smoother than the steps of stepper motors under the right conditions.
Reply With Quote
Reply

Register Options Profile Last 1 | 3 | 7 Days Search Today's Posts Mark Forums Read

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -6. The time now is 01:46.


Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.