PDA

View Full Version : Unwanted axes movement.



corrie
02-16-2009, 10:40 AM
not sure if this is the place to ask this question but here it goes. What would cause all three axes to move when jogging only one axes , directly after startup. Problem goes away after jogging a different axes.And all seems fine after. Thanks.

trubleshtr
02-16-2009, 11:28 AM
Are they trying to find home position or locate the marker pulse on the encoder of the axis?
what type of machine/controls?

corrie
02-16-2009, 07:49 PM
Hello Shooter, no encoders,running Mach3.The axes, all three are usually left in the center of the table.What happens is this, i,ll start the computer, launch Mach, turn on the power supply, jog the Z with shift button pressed while pageing up or down,depends if anything may be in the way.At this point all three axes move simultaneously. While holding the shift, I,ll move the Y and all seems normal. Keyboard is not USB,nor the mouse.I also noticed that while in a holding pattern I can hear the X-axes creeping/stepping slowly and spuraticaly but the DRO shows no X movement in numbers. I may have to put a seperate P.Port...The P.C is a IBM net vista ,no virus protection and all background saving features are turned off ,including updates.I had someone put a scope on the power supply and some slight noise is visible.By all means I,m not any type of electronics Guru,far from it,..... I,m stumped. Thanks in advance and hope my info can help you think of something,cheers,Corrie.

trubleshtr
02-17-2009, 07:49 PM
Hmm and thisis only at start up you say...I thought maybe you need to tune the drives/adjust any gains going to the axis...but if it is only at start up????
Is it possible you are picking up some "noise" on the feedback and the drive starts trying to compensate?
I am sure there are other guys here at the zone that have seen this before.

corrie
02-17-2009, 08:55 PM
I,m going to look into the breakout board it might be the culprit. It,s almost like cross talk between the axes.If not the B.O.B then it must be the P.Port.This may explain some of the lost steps when doing a 3d carve,not sure though.Where in canada are you?Cheers and thanks.

corrie
02-20-2009, 05:18 PM
Well as stated before I could hear my X axes creeping but giving no DRO movement, and now all three axes are creeping simultaneously. Now scratching my head with both hands instead of one any ideas, anyone.

trubleshtr
02-21-2009, 10:48 AM
I don't know how safe your set-up is...But If you power up the machine only, no computer ,no Mach , does it still do it? (Dont attempt if you know the machine will run away uncontrollable with no controls on it! I dont want you to get hurt or damage the machine)
BUT..Then you can rule out anything computer or software related, that is...
If you are getting power leakage to the drives before the "automation" or control takes over the machine.

When you say no movement is seen on the DRO, Do you mean the MACH screen, or do you have a seperate DRO on the system?
The fact that you can hear/see the x axis moving, (when no control is asking it to move), is probably the reason why you see no movement on the DRO. If you had encoders you would see the movement coming back to the DRO.
If you are also seeing all the axis creep now when no control command is given, I would look at your power set-up/ gains ect... you also mentioned that you are missing steps during a run, which could mean the controls are having a hard time positioning the axis due to the fact that the drive(s) is not set-up 100% or the power going to the drives is too "Dirty".
Hope this makes sense?
I am out in Georgetown...where are you in Ontario?

corrie
02-21-2009, 04:54 PM
Hello Shooter,for starters I,m out in Whitby east of T.O... Yes the machine does move without the aid of P.C. or Mach3..., axes were moving in the plus side, whether that means much....The DRO values we talked about that were not seen was with Mach not a seperate controller.I supose that the power setup and gains controls would be directly on the drives,is this correct?What would be the reason the power to the drives would be dirty? This is the part of the CNC world I,m weakest at.The drives came from www.durhamrobotics.com. You can see them on his site.Hear is something that happened a long time ago, my power supply in he begining did not have a fuse on the power side and my bridge rectofier came loose due to heat and I ended up melting some wires.The toroidal transformer had a slight bit of sheathing come off but it seamed fine,none of the windings took a hit. I had the bridge recto replaced ,and a fuse put in since then.I,m wondering if there was some underlying damage that was not caught because since then I have also tried a different BOB with a changed chip for noise.I got a feeling I may need to have someone make me up a new controler/power supply. Would it hurt to put in a seperate PPort?Thanks for your help ,much appreciated.

trubleshtr
02-21-2009, 06:28 PM
Hi Corrie,
I had a quick look at the link you provided. I dont see any provisions for gain/voltage adjustments on these drives. He does however have a jumper to limit/set the amount of current on the "polar drive" he sells. Is that the drive you purchased?
The "Critical motion" drive he offers is listed as un-available due to "Performace test" not meeting his criteria...Did you buy one of these possibly before they realised they had issues?
I dont see any way of tuning the drives unless he provides you with documentation after purchase re-garding how to adjust gains...ect???

I would suspect ( as an armchair mechanic here) that if you are seeing movement with just the machine power on...no pc, no MACH 3 turned on, and you had the power supply problem in the past, that you might have damaged part of the drive? Or if you have one of his "Critical motion" drives, maybe it is possible they have un-covered a problem with these drives??

Assuming the power supply was rebuit ok, and is putting out the proper rated power to the drive, then, there is potential that you have a bad/poor connection somewhere (power supply or on drive)
The only other thing I can conjur up at the moment is that the drive can't handle the "rush in" current when first turned on...In which case you could argue the drive needs protection against this, In the way of proper power supply, and if that is the case then again, the drive has potentially been damaged from past "un-protected" use.
He say's there is a low power standby mode, perhaps this area of the drive is not functioning correctly.
It is hard to say for sure without being in front of the machine. Have you e-mailed him and asked if they have seen this before on their drives?

corrie
02-21-2009, 07:51 PM
Hey shooter, I must say you are quick on the replies.I don,t have the critical motion drives, I have the polar drives.He has been over a few times to have a look into the problem but is not sure himself.He put a scope on the controller to see if and where there was any noise in the system.We ran it in sherline mode half step just to see anthing inbetween the pulses because not much was showing up according to him, this all well above my head.

trubleshtr
02-22-2009, 12:32 PM
Well, something in the way of power has to be coming down the line to the motors to make them move.
If you know how...?With a volt meter on the motor connector, measure what you are getting when it is drifting. Work your way back to the drive and then power supply to trace it.

corrie
02-22-2009, 12:50 PM
Hey Shooter, is this probing to be done with only the controller powered up while the axes are moving or should the computer be connected as well.

trubleshtr
02-22-2009, 07:22 PM
Hi Corrie,
I would unplug the break out board from the pc so you know that no step and direction signal is being sent to the drives. Turn your power supply on. If the axes start creeping, measure the votage at the drive output.
The drive should not be allowing anything to go out to the motors until the step and direction signals are sent to the drive in a normal situation.
Check also the voltage coming out of the power supply. If the output from the power supply is within the range that you built it to, and does not exceed the rated voltage for the drive, then you should be ok.
If you are seeing voltage coming out the drive to the motors then your drives are not "holding back" the voltage and or current until commanded by the BOB to go. The other test would be to check the currrent coming out of the power supply, but that will be a little more tricky.....

Torchhead
02-22-2009, 07:39 PM
Motor drives move for one reason. They are getting a signal into the Step & Dir inputs that "tells" them to move. Noise on the power supply will not make them move. With no signals to the drives (from a lower impedance source like a buffered parallel port) they may be picking up noise but should never move when you have everything hooked up and no signals being sent.

See, chopper drives use a method called PWM (pulse width modulation) it turns the windings on/off at some higher frequency (16KHZ for the drive you have). It is at high current. Move a signal across a wire at high current and turn it off and on at a frequency and it forms (drum roll) a transmitter. The signal can feedback and effect parts of the logic. I have seen the signal get back into the PC and parallel port. The fact the DRO's in MACH do not move shows MACH is not sending the signals. Noise is funny stuff. It will travel down grounds and cross talk between conductors running in parallel.

Part of the design process in power electronics is to understand the paths for noise transmission and deal with it with proper layout. The input circuits need a level of noise immunity.

I can't tell if the inputs on the drive are opto isolated. If they are not the drive motors, drive components and low level logic all share the same ground path. That can be problematic if grounding is not done correctly.

Your BOB could be at fault but it would be unusual for it to produce crosstalk at logic levels. If you have your BOB mounted close (in the same case) as the motor drives try taking the BOB outside the case and running it that way. The PWM from the motors may be getting into the low level inputs of the BOB. Make sure your BOB is a power version and has bufferes on the step & dir lines. Better yet that it has opto isolation on all of the INPUTS.

TOM CAUDLE
www.CandCNC.com
Totally Modular CNC Electronics.

trubleshtr
02-22-2009, 08:18 PM
Ahh finally someone with some electronic background....thank you Tom.

corrie
02-22-2009, 08:22 PM
Thanks for the replies gentlemen. While no P.C was hooked up I turned on the controller briefly but left the P.P cable attached to the B.O.B and all axes moved.I did this agian without the P.P cable attached and the axes still moved but there was a five second delay before they moved. The B.O.B is in the same case,yes.I may have to get a larger case and reconfigure the internals.I,ll post a pick as I can giving a closer view of part placement.Thanks to you both for your replies.

corrie
02-27-2009, 08:25 PM
Hello folks, thought I would chime in with the latest news. The owner of Durham robotics came over today and installed a revamped B.O.B. Also, we tried out some Gecko 203v,s and I gotta tell ya what a difference. The Geckos were mounted to a large piece if 3/4 aluminum with heat sinks on opposite side and mounted on edge, will post pics later. I,ll run some code tommorrow and let you all know how it turned out.Thanks for the help Gary, I appreciate it. Also thanks to the good people here that chimed in with their thoughts,thanks.

corrie
03-01-2009, 08:15 AM
Good day all, thought I would post a pic of a carving I did yesterday with the new setup. Eagle dimensions are 3x1.125x .093. I used a 30 degree v-bit .015 and I think it was around 8% stepover or less at 100 I.P.M. Compared to many other carvings the new setup proved out well.

trubleshtr
03-01-2009, 08:46 AM
Hey look at that, good job! Thanks for posting your results.
So in the end it was a new drive set-up and a new BOB?...I suspect your old drives were damaged?

corrie
03-01-2009, 10:10 AM
Hello Shooter, not sure if it was the drives or not ,but I would put my money on the BOB. Garry will fill me in on that. But none the less the gecko,s realy rock.garry must have spent a better part of that day helping me out,good guy.Maybe if Garry gets any info on the culprit he may chime in on this thread.I,m still running some test files and see how things progress. Will take a picture of my setup latter,cheers.

corrie
03-15-2009, 09:57 PM
Here are some pics of what kind of accuracy my machine is up to after installing the Gecko,s and rebuilt BOB,cheers.