Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Katoh on September 18, 2008, 01:44:54 AM
-
Hi To All
This is my first post so please excuse me if its in the wrong area.
If have built a 6' x 10, 3 axis router based on the design by Tom (cadcut). The trouble I'm running into is I have run twin steppers on both X and Y axis, The Xaxis is rack and pinion style drive both motors are connected to a drive shaft via bully and belts, the Y axis is a ballscrew and again both motors are connected to either side of the ballsrew by pulley and belts again, Both X and Y axis are configured in mach3 as Master and Slave All Fine! When the machine works it tends to get a lot of motor lockup and tuning and speed is way down to what it should be. If I disconnect one of the belts from either motor and run that axis I can achieve 5x the speed without loosing steps, I have come to the conclusion that somewere the motors are fighting each other or they just are not getting the same signal.
Has anyone outhere got any suggestions?
Specs on the Machine.
Steppers Nema 34 900oz/in
Gecko 201 drives (one per motor)
Power 48v 20 amp
On both X and Y axis one motor CW the other CCW
-
Get rid of the connecting shaft between the X axis motors. Let them each drive one side.
Also ditch the second motor on the Y axis. One should be fine.
Well you did ask for suggestions. ;D
Greg
-
Thanks for the suggestion I dont realy want to go to that extreme yet.
-
I'd call putting two motors on one screw going to extremes. :) I don't think I've ever seen anyone mount two motors on one screw like that. But I have read posts advising against it, for the reasons your seeing. The twio steppers will always be fighting each other. One thing you can try is loosening the pulleys, power up the steppers, then tighten the pulleys. That will get them aligned as close as possible, but it still might not work correctly.
-
Hello Katoh,
You didn't mention if there was a reduction ratio with your pulleys. Either way, you could reduce the current ratio by 2:1 and use 1 motor. It would then have the same torque as two motors at 1:1 but of course the speed would be cut by half. You may then be able to increase the speed in motor tuning, upping the kernel speed or getting yourself a SmoothStepper.
Or go to a single larger motor......like a COOL AC servo.
RC
-
Thanks for the Info
for Overloaded my ballscrew only has a 5mm movement per turn so my ratio is actually 1:1.3 this gives me about 7mm per turn.
I suppose I could live with having one motor on the Y. My problem now lays with the X if I break the driveshaft as Greg suggested whats going to stop the whole thing from binding. Just a Thought could it be I am using two two breakout boards C10 from CNC4PC that Ive taken the master from BOB1 and the Slave from BOB2?
-
One thing you could try is have the two drives share one step and direction signal.
Not suggesting as a permanent fix but would be easy to try and interesting to see the result.
If needed to reverse one motor then reverse one winding connection.
Greg
-
Hi Greg
Tried Doing that run probably a little better, tried Even taken both signals from the one board really made no difference. all I can see is I'm going to have to run the Y on one motor. The X its a good idea to cut the shaft in halve but I still have my reservations about racking and binding. Looks to me that if one motor looses a step the the other motor jams up is it ties to correct it. I wonder if it would be any different if both motors were turning the same way?
-
"Looks to me that if one motor looses a step the the other motor jams up is it ties to correct it"
That concern is shared by a lot of builders but not those who have tried it.
Many machines run independent slaved drives. Mine included. Mach handles it well.
Stepper driven machines need to operate within their performance envelope.
If steppers are loosing steps then you need to retune or redesign.
My way of seeing it is one motor driving both pinions via a linking shaft is a legitimate way of doing it.
Independent slaved drives definitely works. Two motors hard linked is not the way to go. IMO of course.
Greg
-
Katoh,
If you are using 2 BoB's, you should probably have your X Master and X Slave on ONE BoB, and the Y Master and Y Slave on the OTHER BoB.
The way it is, it sounds like you have them mixed between ports....Probably not good.
RC
-
Thanks again Greg
I'm really starting to lean towards loosing one motor on the Y What I am probably going to do now is cut the X shaft in two using each motor to run one side independent to the other. If I find the machine Racks and binds I can always put a coupling back in the shaft and connect up as one again. Just to back track a bit The main reason I designed the machine to use twin motors is all my linears are Igus W200 slides not bearings, and in my thought I always believed that ill need double the power to make the machine move at any decent speed, From what Ive seen now on the Y axis with one motor 6000mm/min is not a problem but with the 2 I could only achieve 2000mm/min. The X axis with the linked motors runs at 2500mm/min but the belts are slightly longer an must just have a little more give in them.
Any More comments , suggestions or Recommendations are GREATLY APPRECIATED.
-
What I meant was to at least have each slave on the same board and port as its master.
-
Hi Overloaded
Have Tried that used BOB 1 to run X master and Slave and used BOB 2 to run Y master and Slave, Made no difference to performance. When I tried Greg's suggestion of just using one Step and Direction Signals to run both motors there was a slight improvement Not Much. with the motors hard mechanically connected I can not get to run nowere near what I can with only one motor. What really gets me is I have seen Two steppers coupled together with a steel coupling joining the shafts that was used in a commercially made CNC embroidery machine. Just tells me it can be done somehow.
-
I can't see 2 steppers fixed together working unless they are syncronised.
I've been told that steppers can be slightly out on micro steps but will always be correct on a step
If 1 is only 1/2 degree out of step from the other then they will fight each other all the time, they would have to be completly in step.
I think someone already said to try unlocking one pulley and power both steppers up and then lock the pulley back up.
Steve
-
Hi Steve
I tried that as well, The problem I came across is it works fine till one looses a step then it seems to escalate, I can get it going with the two motors at a decent speed until the first lock up, Then after that you find for some reason you cant get motors back into sync, at a low speed they work fine, but for a timber router a bit of speed is not a bad thing. Ive being giving it some thought and I agree with the Philosophy that you really cant run twin motors in the way I have been. The X axis should work fine if split one motor controlling either side, the y axis the only way to run twin motors is to have a second ballscrew both driven individually from their own motor. I may have the opportunity in the morning to cut my X axis shaft in half, I will keep everyone posted on what happens after that. Thanks again to all your suggestions it has taken me a long way to resolution of my troubles.
-
One last Thought before the X drive shaft meets its fate with angle grinder.
Because the steppers work OK at low speed could Power be the problem here,
My 48v 4amp per motor needs to be a lot more to have the top end speed.
Or will the extra power just mask for a while the main problem?
-
What's the current and voltage ratings of the motors?
-
HI Gerry
The motors are rated in Series 4 amps 3.0 volts cant run the motors in Parallel as they require 8 amps and geckos only go up to 7.
-
If you run them bipolar parallel at 7 amps, you'll probably get more top speed out of them than running them in series.
And for best performance, power supply voltage should be 32 times the square root of the motors inductance in mH. The 48V might be a little high for parallel, but the 20 amps should still be plenty, even when parallel wired.
-
Gerry
I have thought about doing that, But wont I run out of power even by loosing one motor on the Y axis still leaves me with 4 motors each wanting 7 amps thats 32 amps in total.
-
Geckos will never draw more than 2/3 of the motors rated current, and usually will draw much less. So, first of all, 7x4=28, not 32 ;) 2/3 of 28 is 18.66amps is the max the Gecko's will draw, if all 4 motors are running at full power at the same time.
-
sorry about my math for some reason I still was thinking each motor wanted to draw 8amps, Ill try that first before I start hacking into drive shafts and making major changes, its easier swapping a few wires then altering machinical parts. On the G201 just to make sure I'm doing it right you connect a straight wire between terminals 11 and 12 for max amps? Would I also be advised to use Standby enable setting on the jumper block to save a few amps or just use the default setting that I'm running now?
-
For Max amps you dont connect anything to the resistor terminals, leave them open.
Hood
-
Thanks Hood
Ill keep everone posted on what happens when I try a few different things, wont be for another day or so now.
-
Ive just a chance to change the motors from a serial to parallel connection, i changed both x motors and both y and had a little play with it. Well heres one for my books. the X axis in my Original Config one shaft twin motors x motor from BOB 1 X slave from BOB 2 top speed 5000mm/min without loosing steps or lock ups. The y axis with the twin motors running y motor from BOB 1 y slave from BOB 2 top speed of 5000mm/min without loosing steps or lock ups. Actually Mach wont let me drive the Y any faster its hit the mach limit. The whole machine is very comfortable running at 3000mm/min. Put through a basic not to complex programme, run the programme 10 times in a repeat and was happy to see that it returned back to zero.
Thanks to all how helped!
-
Hi All Again
A problem has arisen were I can run a test program through Mach Eg x400 y400 then x0 y0 and have the machine return to zero time after time using speeds anywere from 2 to 7000mm/min, but as soon as I run a programme through mach (some previous project) the machine will not return to x0. I have tried motor tuning changing the pulse lengths, using exact stop using constant velocity nothing seems to fix this error. Doesn't matter wich programme I load its always out by a constant 3mm over the initial x0.