Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Popeye on July 30, 2007, 04:42:05 PM
-
Hello,
I have a problem on my CNC machine that I couldn't solve for the moment.
With a certain combination of X and Y movements, I sometimes loose steps on the Y axis. It occures during a G0 movement and always on the same displacement (same combination of X and Y axis). The G0 speed is set lower than the maximum speed capacity of my machine and there is no mechanical problem.
It seems to have link with the frequency generated by this particular movement combination.
Does anybody have a clue how to solve this problem ?
Thanks
Jean-Louis
-
Make sure enhanced pulseing is enabled
Hood
-
Hello Hood,
Thanks for your aswer.
I checked the enhanced pulsing and it is already enabled.
Jean-Louis
-
slow down that Max speed of the machine and see if that helps... it could be that the power supply is browning out
-
Hello Brian,
I actually didn't think about the power supply. Could you please explain what you mean by "power supply is browning out" ? (sorry, English is not my mother language :P)
Thanks
Jean-Louis
-
Check the voltage of the power supply and see if the volts are dropping as it is moving... I think it may be the problem.
Thanks
Brian
-
You're a very clever guy Brian. I think you pointed out the problem.
I made some test using a lamp to load the power supply at about 3 A. When I connect the lamp, the voltage drops down to 15 V for a while before recovering its rated value (27 V). And when I disconnect the lamp, the voltage then rises (I don't know how much, the power supply voltmeter stops at 30 V) and the overvoltage protection trips. I probably have a voltage regulator that is failing.
Thanks a lot for your help
Jean-Louis
-
The power unit for my stepper motors is a pair of batteries (leisure batteries). Mine are 120 amp hour, because I use them to power a miniature railway train. 75 amp hour ones would be adequate. The batteries can supply the current (amps) without the voltage dropping and mine can power all three stepper motors simultaneously without a problem. I have a 24 volt battery charger that charges them up. Maybe this approach is cheaper than a mains power system.
-
Thanks Jim, it could be an interesting alternative. I don't know the cost for 2 batteries + 24 V charger compared to a new power supply but I'll have a look.
Jean-Louis
-
http://www.kelinginc.net/SwitchingPowerSupply.html
That is where I get my Power from ;)
http://www.kelinginc.net/SwitchingPowerSupply.html
-
I don't know a lot about power supplies, but the reason I use batteries is that the voltage stays constant, both when the motors are on ( and if all three are on together they could pull 7.5 amps) but also when they are off, the voltage does not rise above24 volts ( as I have found with a lot of transformer based power supplies, unless properly regulated).
My stepper motor boards by Stepmaster says specifically that supply voltage should not rise above 30 volts (against a working voltage of 24)
-
http://www.kelinginc.net/SwitchingPowerSupply.html
That is where I get my Power from ;)
http://www.kelinginc.net/SwitchingPowerSupply.html
The prices are actually very good. Unfortunately 120 Vac imput voltage is not suitable for French network.
I checked different solutions and I will finally do my own supply. It will cost about 60 Euros.
The output voltage will be rated about 25 Vdc while the stepper motor board can accept up to 35 Vdc so if the voltage is slightly higher when the motors are not running, it should remain within the driver board limits.
Thanks everybody for your help
Jean-Louis
-
I think they will also take in 220V (The last one that I worked one would).
Thanks
Brian
-
I've ordered the components to make a new power supply but I'll keep think link and contact them if necessarry
I'll confirm you if the problem has been definitely solved by changing the power supply
Thanks
Jean-Louis
-
Hi everybody,
I made new test today with my new power supply and despite a voltage drop that is now limited to 1 V maximum, the problem still occurred.
The "sound" of the machine on this particular segment (about 45° angle, positive X move and positive Y move) was really like a vibration so I slightly modified the fixing of the Y motor (I actually added one fixing point) and since I made this modification, I ran about 10 or 15 times the same program without any problem (I previously had it about once every 3 attempts). The sound is now different and I think that it is definitely linked with a resonance frequency, and this frequency has probably been changed by modifying the motor assembly. I have to run several other programs, hoping the resonance frequency has been moved out of the range I meet on my machine.
Jean-Louis
-
I still think this might be a power supply problem.
If you are running say G0 X10 Y10 from a position of 0,0 then as you say this is 45 degrees. Yours motors will be taking the maximum current exactly simultaneously. THEY ARE INSTANTANEOUS MACHINES AND REQUIRE MAXIMUM CURRENT IMMEDIATELY FOR A SHORT PULSE THEN STOP, then the start again, and stop etc etc.
I assume they are both set up for the same number of steps per rev or per inch, or millimeter, whatever, so the will perform simultaneously, and (of course) since they are controlled via a parrallel port it will be simultaneous. Even if you can only detect a 1 volt drop on a multimeter, your actual voltage drop will be greater.
Try your power supply running a large motor or something, and try stalling the spindle and see if the voltage drops when you get to say 10 amps. It must a a big **** power supply to give 7.5 to 10 amps without loosing voltage.
Jim
-
Hi Jim,
You're probably right when you say that the voltage may drop more than what I see on the voltmeter (even though I used an analog voltmeter that is better than a digital one for such a measurement), but I clearly get a more stable voltage since I made this new power supply (the previous one dropped down to 15 V when running the same program).
However everytime I had these step losses, it occured after having reached the rated G0 speed. It means that the motors were not accelerating anymore (you actually made me realise that :)), and that's why I finally don't think that it was due to a power supply problem, even if I actually really had one.
Jean-Louis
-
Can we see a pic of the axis that is giving you trouble ?
-
DRstein, have you tried to swap the yand x axis after the drives. You will need to rotate the workpiece 90 deg but this should tell you if it is a mechanical issue or a drive/computer related issue. If the error still ocurs on the Y axis then it is mechanical related, if it moves to the x then swap the drives themselves and retest. At least you will eliminated things one at a time as you test. (;-) TP
-
I have no clue what your issues are but when I read about vibration it made me think of this thread. Check out the videos.
http://www.cnczone.com/forums/showthread.php?t=32284
-
What I have found while battling that damn monster. Was that at certain speeds. Resonance would kill the system. So with the dampener installed I am able to reliably run at ALL speeds.
They are a very wise decision and very easy to make.
When you changed out your motors you moved your Resonance level to another speed for the most part. A good test I found is to run a g code that moves a certain axis back and forth for a long period of time and use the feed over ride and slowly step up for speed. Around 22 IPM was my "DEAD" spot. Installed the dampener and it was all good!!!
Cheers
-
Thank's a lot for this very interesting thread !
I can't say that the sound I witness is the same as the machine is mechanically different but I'm quite sure that the phenomenon is identical.
The difference with and without the damper is amazing ! I actually thought about a kind of fly wheel but I wasn't sure if it would really be efficient.
I think I will study this solution as, as mentioned, what I've done may have just moved the resonance frequency to an other rotating speed.
Nb : I also use Xylotex control board and the motors I have are the reference 23LC100-025-8W (details here : http://www.stappenmotor.nl/Stappenmotor/Stappenmotoren/Stappenmotoren_globaal_overzicht.htm)
Thanks
Jean-Louis
-
Dr Stein, did you say you are running nema 34s on the same drivers? What is the amp draw of the 34 motors compared to the 23 motors?? Is the driver rated for a 34s? (;-) TP
When testing don't try to do it in what may look like the easy way. There "is" a method to the madness.
-
So I take it this machine used to have 23s on it?
Did you have the problem then?
If the Amp rating is say 5 amps and you are only driving it at 3 amps the 34 motor may be a little whimpy on the accellaration curve.
Don't know if that driver has current limiting capability or not, but that high of a amp load from the 34s could also drag the performance of the driver down.
If it did not do it with the 23s then I would start digging into the motor/drive specs a little deeper.
That is of course if it is not a mechanical issue.
Just a Thought (;-) TP
-
Does the driver have heat sinks. You may want to increase the size of the sinks and install fans to help cool them, could be the culpret. (;-) TP
-
You said your having problems when the machine changes directions, right? Check the backlash settings, disable them if they are turned on. I had a problem when I started my machine up. I'm running servo's, so they won't stall, if they go out of position by 128 encoder counts, they go into fault, or shut down. Initially, my backlash settings were set to 20% of set speed. This gave me problems in the same place every time in the same programs. At certain points, not many, but the same place every time, the servo's would fault where they were supposed to make a direction change. I now have the backlash speed set to 1% of set speed and have no problems, and I don't notice any problems as far as machining quality either, my initial worry is that the endmill would dwell. BTW, I rapid at 96ipm and cut with a ballmill at 90ipm using constant velocity, no problems since.
Michael
-
I can dial-indicate and inch back and fourth, and it appears to accurately locate back to zero each time.
This is not checking the backlash, only repeatability. You will only see backlash when it changes directions. Put the indicator on something solid. Jog your axis until you move the indicator to a reading of about half of its travel. Write down the reading on the indicator. Jog the axis in the other direction by a set amount, say .010" and see how much difference the indicator shows. If the indicator says you moved .005 and Mach says it has moved .010 then you have .005" backlash.
Brett
-
I used the Logitrol "Low Cost CNC" system. And I don't reccomend them to anyone. Basically, it's a 50V DC power supply, Gecko 320 drivers, US Digital encoders, and Pacific Scientific motors, supposed to be 1800oz/in.
When I recieved the system, the Pacific Scientific motors had encoders pre-installed that were drawing too much current, 150 mA, which is 3 times what the Gecko driver reccomends since the encoder get's it's power right from a onboard regulator in the G320's. I blew 2 of the Gecko drivers immediatly.
I emailed back and forth with Steven Kennef from Logitrol to try to get the system working. He claimed that the Gecko drivers were under warranty and that it must be a problem with the G320's as they were having some issues with "mosfet suppliers", or some other bogus crap like that. So I took the system apart and sent the drivers back to Gecko for evaluation.
After recieving and evaluating the Gecko 320's, I talked to thier engineer on the phone to discuss the problem. He asked a lot of questions and informed me that the encoders that were installed at the time, "remco's", were drawing too much current causing the servo drivers to fail. He also told me that if the drives were repairable, they would repair them for free of charge, even with Gecko paying return shipping! But all 3 drivers were beyond repair...... Gecko did replace 1 servo drive for free under thier "one time stuff happens policy", which they didn't have to do and I was very greatfull for! The other 2 I had to pay for, as well as 3 new encoders that the Gecko engineer reccomended from US Digital Encoder.
Steven Kennef refuses to admit fault to this day. He did send a $140 refund towards my costs of repairing the system that was supposed to be plug and play that his company built. He claims that he can't afford to refund me at this time because his wife is sick in the hospital and he's having financial issues. I apologize, I hope the best for everyone is health and happyness, but not being able to seperate business from personal issues and affairs is poor business. I'm not even going to discuss that further, we'll just say my life isn't perfect at the moment but I don't use that as an excuse, and definatly no to customers.
Sorry for the long story, just pissy about the Logitrol deal and I want to warn others of my dealings with them. Basically, it's a 50V power supply, Gecko 320 drivers, an isolated break out board, sevo's and encoders that are wired up in neat looking boxes. My personal opinion is if your at all capable or atleast willing to learn, just do it yourself.... I will if I do it again. Plus the knowlege gained will be worth 10 fold of what you save as far as money when you can diagnose and repair your own system. Logitrol didn't design or wire anything special and all of the components are available from diffrent suppliers and I'm sure that there's people here in the Mach forum that would be more then happy to help you locate and source. Plus you can probably save a lot of money, I paid Logitrol $2600, plus $600 for replacing damaged drivers and new encoders, and a $400 phone bill for calling Steven in Australia.... Lesson learned.
And I highly reccomend Gecko drivers. Thier customer service is amazing! And this is also why I love the Mach controller! This is an amazing support group and I'm very impressed with the knowlege and helpfulness of some of the members!
Michael