Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Cartierusm on August 22, 2009, 08:14:31 PM
-
OK, ever since I converted my Bridgeport to CNC the X and Y axis lock up when trying to make shallow moves at the same time. So let's say when trying to go from Y0.2 X0.1 to Y8 X.2. I think it's because the machine is trying to move that shallow sliver of a move of X over such a long distance. I noticed it during drilling cycles originally. And before you ask I've tried everything slowing down the motor in motor turning and changing the accerlation down. The motors are more than strong enough 1200 oz. steppers at 150 IPM it can pull me across the floor without missing a step, but even then I tuned them down to 70 IPM as my top speed and of course it still happens.
Some things I noticed months and months ago when trouble shooting it, I readjusted my Y gibs, they were a little loose. I eneded up just forgetting about it and just altering my Gcode to make each move separately until it gets to where it's going. Instead of X2Y2 on one line I would say X2, [new line] Y2, just to be safe. The weird parts is I use this mill all the time doing very complex 3 and 4 axis stuff and it never messes up until I get to shallow moves on a drill cycle; that is until yesterday I was doing some pocketing where there were some extremely shallow moves and it messed up, luckily I was there to Feed Hold and resetup.
Any help would be appreciated. I'm just worried that it's going to mess up somthing real important.
-
How about power supply, is it powerful enough to power all 3 axes at the same time?
-
As I said above I do 3 axis 3D all the time and no problem, the problem is only when doing a shallow X and Y move together. Anyway, I'm using a 72VDC 20AMP power supply.
-
OK, ever since I converted my Bridgeport to CNC the X and Y axis lock up when trying to make shallow moves at the same time. So let's say when trying to go from Y0.2 X0.1 to Y8 X.2. I think it's because the machine is trying to move that shallow sliver of a move of X over such a long distance. I noticed it during drilling cycles originally. And before you ask I've tried everything slowing down the motor in motor turning and changing the accerlation down. The motors are more than strong enough 1200 oz. steppers at 150 IPM it can pull me across the floor without missing a step, but even then I tuned them down to 70 IPM as my top speed and of course it still happens.
Some things I noticed months and months ago when trouble shooting it, I readjusted my Y gibs, they were a little loose. I eneded up just forgetting about it and just altering my Gcode to make each move separately until it gets to where it's going. Instead of X2Y2 on one line I would say X2, [new line] Y2, just to be safe. The weird parts is I use this mill all the time doing very complex 3 and 4 axis stuff and it never messes up until I get to shallow moves on a drill cycle; that is until yesterday I was doing some pocketing where there were some extremely shallow moves and it messed up, luckily I was there to Feed Hold and resetup.
Any help would be appreciated. I'm just worried that it's going to mess up somthing real important.
"...the X and Y axis lock up..." - What does that mean. You mean neither axis moves? Or only one moves? Or one axis may, or may not move the correct distance?
Regards,
Ray L.
-
Hi, Cartierusm
Try setting Sherline 1/2 Step Mode in ports & pins, or this xml below and see if it helps, It's one you posted on your 4 th axis topic eariller, Let us know, It should give your cont board's have a little more time to see Mach3's pulses.
What axis controls are your using.
Chip
-
Thanks for the help.
Himy, it seems that the one making the short move is crawling along and the axis that's supposed to do the long move locks up, but I think I've seen it reversed. Not sure if the one axis that's is moving if it's actually moving at the correct speed.
Chip, is the only alteration to the XML the Sherline part as I'd rather change by hand so I know what's going on. Also for axis control do you mean what stepper drives I'm using? I'm using Keling 8078 Drives, here's a link to their specs:
http://www.kelinginc.net/KL-8078.pdf
-
Hi, Cartierusm
Changing it on your own is fine, Sherline 1/2 Step Mode just set's pulse width's to 40 ms, I did notice that your step & dir Pulses were set low, dir being set at "0" in motor tuning.
I'm PM messaging you my phone #, If you'd like to talk about this problem.
Chip
-
Thanks for all the help and a BIG THANKS to CHIP for spending time with me to solve my problem.
I changed the settings to the Sherline 1/2 Pulse Mode and reran the program and it worked, I was ecstatic, then I tried again from the beginning of the program...then it didn't work DAM. So I went and changed the G00 move that it was stalling on to G01, which at that point in the code was 20 IPM and that worked. I remember months ago trying different velocities, but I guess I didn't go low enough for the problem to stop. So something weird is going on for those shallow moves as it doesn't happen during other moves with high rapids.
OH well I'm not a complainer, I'm happy the problem got solved and it's not like I'm make 40" moves to where I'd notice the speed difference.
-
I have the big brother to these drives on the coil winder, I now have a SmoothStepper on it but during testing I was using the parallel port and it needed 5 for pulse widths. Have you tried that? Maybe they are not happy with the extra wide pulse of Sherline.
Hood
-
Hood, I tried the Sherline and it worked, then per Chip's suggestions unchecked the Sherline mode and went into motor tuning and changed Step Pulse and Dir Pulse to 10 saved the axis and restarted Mach. Tried it didn't work, retried Sherline didn't work, then tried Step Pulse and Dir Pulse at 5 and it didn't work. So I tried changing the G00 to G01 in the actual Gcode and that worked. Chip said that the Step Pulse and Dir Pulse should be the same, so just for posterity I kept the Step and Dir Pulse at 5. So it seems something electronic or mechanical is happening, but the problem got solved and I changed alot of my regular Gcode files in Rhino.
Are your Keling drives set at active low? Chip said something about that but I didn't bother trying to change them, Keling documents say to have them at Active Low so I figured I'd leave them.
-
My drives are not badged Keling, they were from ArcEuro in the UK but no matter what name is on them they all look like they are made by Leadshine in China.
I think mine are set active high for both step and direction.
Hood
-
Do you think that chaning the active position would make a difference?
-
It may do, I cant recall whether the docs I got with my drives said Hi or Low but it seems to work fine that way for me. Then again even if the docs did say it I am not sure I would believe them as the latest drive I got was slightly different and had a dip switch to select between CW/CCW mode and Step/Dir mode and it was wrong.
Hood
-
If I switched it to Active high and it didn't work what would I see the machine doing differently?
-
Would probably lose the odd step here or there.
Hood
-
hmm, ok, well the problem gets solved by slowing down the rapids, so I think I'll just stick with that, thanks for the help.
-
hmm, ok, well the problem gets solved by slowing down the rapids, so I think I'll just stick with that, thanks for the help.
That would really worry me. You may have solved the immediate problem, but there is something wrong somewhere, and it will undoubtedly come back and bite you in the ass at the worst possible time. I don't believe you've seen the last of this problem....
Regards,
Ray L.
-
Well if you have any suggestions? I mean it only does it when it's doing a rapid move where one axis is long and one axis is super short. I do 3D all the time with no problems.
-
OK Himy was kind of correct. This problem still exists, kind of. So I don't do any rapid moves during machining anymore, I stay at 20 IPM. But I noticed recently that the problem would present itself at 30 IPM if both the X and Y were moving at the same time, such as going around a corner at full chip load, so 1/2" endmill 30 IPM and 1/4" DOC. I believe the gibs to be adjusted accurately and both axis' are solid and move smoothly. I can't move the tables by pulling more than .001 - .0005 and it's certainly not binding. When the machine is off I can turn the ballscrews by hand and they are very smooth. The only mechanical fault I would see is maybe the steppers are not strong enough for this mill under full chip load.
So PLEASE HELP ME solve this problem. Specs: Keling 1200 oz./in. Steppers -- Direct Drive to Ball Screws -- Keling 8078 Drives -- 72 VDC @ 20 Amps Power Supply (The Steppers are set to 5.7 amps on the drives). At 150 IPM it can drag me across the floor without losing steps, but that's not a very technical test and a full load cutter might be more. Is there any way this could be an electronic problem with the drives or something? If it's not an electrical thing would I get more power from a 1810 oz./in. or 2830 oz./in. Stepper with the same 72VDC Power Supply? I'm asking cause I don't know how much more power at 20-50 IPM I would get from the other motors.
-
I would put an oscilloscope on the power supply. My guess is it's not holding up under peak loads. You can try looking at it with a voltmeter first. If that's not it, then I'd look for possible ground problems, or noise affecting the Step/Dir lines. Does one axis typically stall, or both?
Regards,
Ray L.
-
Cartiersium,
Unless someone has experience with equivalent steppers and in that voltage and amp setting range you can't know what the 1810 or 2830 in oz would do in terms of power. You need to take a look at the motor curves to get a flavor. Problem is you may not get a curve at that voltage ( which is usualy the case ). So you have an idea but nothing tested. You can always ask the vendor.
If you are finding the problem at a particular feed rate, then maybe it's a motor resonance problem at that rpm. Take a look at the Gecko site for info on it. Just a guess on my part as i have never experienced the resonance stuff.
Sorry i couldn't be of more help,
RICH
-
Cartiersium,
Unless someone has experience with equivalent steppers and in that voltage and amp setting range you can't know what the 1810 or 2830 in oz would do in terms of power. You need to take a look at the motor curves to get a flavor. Problem is you may not get a curve at that voltage ( which is usualy the case ). So you have an idea but nothing tested. You can always ask the vendor.
If you are finding the problem at a particular feed rate, then maybe it's a motor resonance problem at that rpm. Take a look at the Gecko site for info on it. Just a guess on my part as i have never experienced the resonance stuff.
Sorry i couldn't be of more help,
RICH
Rich,
Actually, voltage has nothing to do with torque. Torque is determined by current. The high voltage is used simply to allow the drive to "overcome" the motor inductance and ramp the current up much faster. The peak torque rating of a stepper gives you the low-speed torque capability, so he should be seeing pretty close to that peak torque at his 30IPM feedrate.
Regards,
Ray L.
-
What kind of power supply are you using here? Linear regulated, linear unregulated, switcher? Make/model?
Based on the somewhat flaky behavior, I'd bet you've got either a noise or ground problem, and not a motor/drive problem. What BOB are you using? How do you have things wired? Some pictures of your electronics might help a lot.
Regards,
Ray L.
-
I'm using a Keling Power Supply KL- 7220, Unregulated Power Supply 1440W, 72VDC/20A, I'm using it with 120VAC. It's on this page http://www.kelinginc.net/SwitchingPowerSupply.html. I'm using a CNC4PC BOB C10 http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=45. I'm using all shielded wire. I don't have the shield part terminated anywhere, but I don't get any noise problem otherwise. The Problem is not flaky it's pretty much static, if I run it too fast X and Y at the same time where the one of the axis is shallow and no load I get the problem at about 65 IPM, if I'm hogging out material then 30 IPM. Seems pretty consistent to me.
Himy I don't have a scope, but I might be able to borrow one, if I use a multimeter what would I be looking for and what would I run on my machine to test it?
-
Here are some pics. I use CPC Circular Plastic Connectors to hook up the Steppers to the electronics as I use the electronics for more than one CNC machine. I rewired it a few months ago and I had the problem before and after the rewire and all the other machines work fine.
-
Ray,
Agree, but it does affect the motor curve. And what your after is power and that is where my thought was.
Off thread and nothing to add, so off i go,
LOL Cartiersum, your in good hands,
RICH
-
I can see several things that are less than ideal in your wiring. First, you really MUST have a very solid single-point ground. Normally, this would be a metal plate, to which everything is mounted. At a minimum, it should be a solid buss-bar to which ALL grounds are connected. Never daisy-chain grounds - they should ALL connect directly to a single point. It appears your power supply has two separate output? If so, the grounds on those MUST be tied solidly together. Never, ever connect a signal ground point to a power ground point (e.g. - connect an encoder power- signal directly to motor power- on the drive, to save a wire). Run two separate wires all the way back to the single ground point.
What you're seeing "smells" like a bad ground, since it gets worse under load - that suggests that different parts of the system "see" different grounds at different times, which can cause an active input to be "seen" as inactive, or vice-versa. Basically, the idea is to make sure that for each signal, there is one, and only one, path back to ground, and that path must NOT include shields. Otherwise, the return current will be split among the available return paths, and things may not behave as expected. In your case, I would add a buss bar alongside your power supply, tie BOTH power supply grounds to that, along with the power line ground, and make ALL ground connections directly to that buss bar. My guess is this alone will go a long way towards resolving your problem.
Next, the shielded cables - If you don't have the shields connected, then they're effectively unshielded cables. The shields should be connected to ground at the SOURCE end only. This means the signal cables (STEP, DIR, etc.) should have their shields connected ONLY at the BOB end.
There is no reason for this stuff to be particularly sensitive, if wired up properly. As an example, my E-box has the PC, power supply, two BOBs, SmoothStepper, home/limit/relay I/F board, spindle speed I/F, VFD, Geckos, Modbus board, etc. all in a single enclosure, and works flawlessly. The mounting plates and the enclosure itself are all securely tied to ground. You can barely see in the pictures that the grounds of both power supply filters are tied directly to the mounting plate, and all ground connections branch off from there.
Regards,
Ray L.
-
Thanks Himy, I appreciate all the help. I understand what you're saying but might have a few questions, but right now I'm running the machine and actually making some money for once, but later today I'll study what you said.
-
Ok Himy, I understand what you're saying but help me clarify some things. I want the shield on each cable going to one lug on the metal body of the control box I have everything in. Now can I slide a copper wire up into the shielded cable as most of the shields are a foil, I think, then zip tie it so it has contact? If they have a drain wire for the shield then I'll just run off of those off course.
Now for the grounding of the power. Here are all my devices what wire should go to a ground location.
Main Power Supply -- It's a transformer so there are a ton of wires. Do I want to ground the actual ground coming off of the plug coming out of the wall? Then you mention bridging the grounds on the power supply, but there is only one 120V ground, or do you mean I should run the NEGATIVE from the DC side to a ground point on the metal cabinet?
Next I have 4 Stepper Drivers what wire do I use on those to ground? And which on the VFD Speed Control and BOB? And I have a 5VDC power supply coming in for the power on the BOB and for the Steppers, nothing is daisy chained. And I have a 12VDC Power Supply for the VFD Speed Control.
Thanks. Oh you said have all electrical grounds together, do you mean the NEGATIVE side of the DC goes to the same spot on the metal cabinet as the 120V GROUND?
-
See below in []:
Ok Himy, I understand what you're saying but help me clarify some things. I want the shield on each cable going to one lug on the metal body of the control box I have everything in. Now can I slide a copper wire up into the shielded cable as most of the shields are a foil, I think, then zip tie it so it has contact? If they have a drain wire for the shield then I'll just run off of those off course.
[That should work OK. The important thing is that the shield be connected only at the source end.]
Now for the grounding of the power. Here are all my devices what wire should go to a ground location.
Main Power Supply -- It's a transformer so there are a ton of wires. Do I want to ground the actual ground coming off of the plug coming out of the wall? Then you mention bridging the grounds on the power supply, but there is only one 120V ground, or do you mean I should run the NEGATIVE from the DC side to a ground point on the metal cabinet?
[Yes, the AC ground should absolutely be tied solidly to the machine itself, and to your E-box. The wire used for making these connections should be as large as the largest wire used in the system. This is for safety as much as anything else. The - ouput(s) of your power supplies also should be tied to the metal case of your E-box, and to a buss-bar for making all the ground connections to the devices.]
Next I have 4 Stepper Drivers what wire do I use on those to ground? And which on the VFD Speed Control and BOB? And I have a 5VDC power supply coming in for the power on the BOB and for the Steppers, nothing is daisy chained. And I have a 12VDC Power Supply for the VFD Speed Control.
Each stepper driver should have a motor power - connection. This should be tied directly back to the main ground buss bar from each driver. If there's a logic ground as well (e.g. - for the step/dir signals), that should be a separate wire, even though it goes to the same place. The VFD will also have a ground connection, that needs to tie back to the ground buss bar. The control signals for the VFD should be run through a shielded cable. If your input is single-phase AC, twist the two wires together with a couple of turns per inch - shielding should not be necessary. If that 12V power supply for the speed control is powering the computer-side of the circuit, then it's - output should also be grounded. If it's powering the VFD side of the circuit, then its - output should be connected to the - connection on the VFD analog input terminals.]
Thanks. Oh you said have all electrical grounds together, do you mean the NEGATIVE side of the DC goes to the same spot on the metal cabinet as the 120V GROUND?
[Yes.]
Regards,
Ray L.
-
Thanks. I'll do that this weekend.
-
Himy, I just reread your posts to me and I've still got some questions, I'll try and make this simple. I haven't done any grounding yet, I've kept the machine under 10 IPM and these days that's too slow so I'm going to do the wiring today.
120V Ground = Lug on Metal Housing Box
12VDC Negative From PS Side One = Lug on Metal Housing Box
12VDC Negative From PS Side Two = Lug on Metal Housing Box
Lug on Mill = Lug on Metal Housing Box
220V Ground on OUTPUT of VFD = Lug on Metal Housing Box
12VDC Negative Power Supply for VFD from it's circuit board = Lug on Metal Housing Box
5VDC Negative Power Supply from BOB = Lug on Metal Housing Box
Shield Wire from Signal Cables for Limits Switches and Accessory Wires = Lug on Metal Housing Box
Shield Wire from Each Stepper Cable = Lug on Metal Housing Box
If the above is correct then the only thing I don't get is where you say to ground the Stepper Drivers, the Negative 72V going into it is from the Powersupply and I mention that above as grounded, do I need to tie in the A- and B- of the drivers that go to the steppers? Thanks.
-
Do NOT ground any signal wires coming from the VFD - that may blow up the VFD. That supply should be floating.
Cable shields should be grounded to the device driving the signa wires. So, cables coming from the BOB should have their shields grounded to the BOB ground terminals.
Not sure I'd bother shielding stepper cables. The power supply wires to the stepper controllers should tie directly to the power supply terminals. The - connections can go to your common ground point if that is very close (no more than a few inches) to the power supply.
Regards,
Ray L.