[vonnieda]

Hi all,

I ran into a problem yesterday that I hope someone can give me some tips on. I have a Taig 2016, conversion done by Dan Mauch of Camtronics that I bought second hand. The controller is by Dan, has a 36v power supply and three Gecko 201s. I'm running it all with Mach 3.

The problem I am having is that if I try to jog with more than one axis at over 15 IPM one of the axes will chatter/grind/skip steps. It basically makes an awful noise and stops moving. It seems to be random as to which one will do it. The interesting thing is that if I just jog one axis I can jog as fast as I want. I've tested up to 40 IPM.

Can anyone give me some suggestions as to where to look for this problem? I'm electronics savvy but not very CNC savvy yet.

Thanks,
Jason von Nieda

[ataxy]

what are the computer spec and kernel speed that you use in mach3 and did you run the driver test to see if you have smooth signal

[vonnieda]

what are the computer spec and kernel speed that you use in mach3 and did you run the driver test to see if you have smooth signal

Kernel speed is default: 25kHz, I believe? I'll need to check it when I get home. I'll also need to check the computer specs when I get home. It's a Dell OptiPlex Gx150, which I think is 1GHz and it has 512 MB of RAM. I'll get the specifics in a few hours.

I have not run the driver test. Will do that as well and report back.

[jalessi]

Vonnieda,


If the driver test passes at 25,000 bump the kernel speed up to 45K and re run the driver test.


Jeff...

[ataxy]

at 1ghz the computer is at the minimum requirement for mach3 so the problem might be there, also does it use a onboard gpu if so you should use a seperate card

[Fixittt]

Sherline mode..............

[vonnieda]

Okay, the computer is a 1GHz P3, 512 MB RAM. It does have onboard video, which is what I'm using.

I ran the Drivertest.exe and it showed Excellent from 25k all the way up to 100k.

I didn't see the option for Sherline mode in my settings, so maybe I have an older version of Mach 3. I'll try that tonight. Unfortunately we're having a bit of a heat wave in Seattle and it was too miserable to stay in the garage for very long last night to try things out.

Any other ideas, though? I would think that if the drivertest passes at 100k it's not a hardware speed issue. Could it be a port voltage issue? Power supply?

[xy_eezy]

First of all, you must make sure that your power supply is easily capable of producing all the current that your drivers may require. If I'm not mistaken, depending on how you have them setup, the drivers that you have, could use as much as 7 amps apiece. And secondly, even if your P.S. is not under any type of strain, you cannot underestimate the effects of improper isolation. Each of the drives need their own power feed, all the way back to the power supply. Also, the power for your controller, should not be derived from the power supply, which feeds your drivers. I have had to troubleshoot issues exactly like the one you are describing, with each axis running flawlessly, until another axis is asked to join in. An o'scope can confirm all that I've said, and that's exactly how I ended up running down the problems that I faced. Just try and avoid pulling out most of your hair in the mean time! LOL

[vonnieda]

Okay, it's finally cool enough in Seattle to spend some time in the garage!

I just did a bunch of testing and I've narrowed the problem down to backlash compensation in Mach 3. With BLC (backlash compensation) turned off I can do 50 IPM in all three axes no problems at all. When I turn BLC on one of the axes will sieze when I start to move another. To reproduce, this is exactly what I do:

Turn BLC on and restart Mach 3.
Move Z down an inch or so using page up key.
Hold down Shift (for full speed jog) and press left arrow to start moving X.
Now while continuing to hold down Shift and left arrow, press page up to start moving Z.
At this point X will seize while Z moves.

This only happens with BLC on, and it only happens when BLC activates, i.e. switching an axis' direction or moving it the opposite direction of the direction it was moving before it last stopped.

So, I think this rules out any electrical problems. I did check on the power supply. It's rated for 20 amps and my drivers are set to 3.5A each; so well within limits there.

Any ideas?

[vonnieda]

Sorry to reply to myself, but I just found this is in the Mach 3 manual:
Mach3 is not able to fully honor the axis acceleration parameters when compensating for backlash,
so stepper systems will generally have to be detuned to avoid risk of lost steps.

Sounds like that's my problem right there?

[ataxy]

check hoss video on backlash

[vonnieda]

check hoss video on backlash

Watched the video and tried the settings. Didn't seem to help. I think there is some difference between what Hoss was getting and what I am. His issue was mainly that without the settings the machine sounded weird and moved too slow. Mine is that the axes actually seize up and miss steps, causing the milling operation to fail.

I just ran some more tests. I hooked up my scope and power looks good. No dips in voltage as different axes run. Parallel port is putting out a steady 4.6v and no dips there.

Finally, I tried cutting a circular pocket similar to what Hoss did. With BLC off it runs fast and smooth... except the backlash causes it to turn out slightly egg shaped. Same as Hoss. With BLC on the machine seizes a bit after the first few arcs and after that the entire operation is out of whack. These were both at 24 IPM. I also tried the same test with BLC on and 15 IPM and this works perfectly fine.

So, I still have some problem with anything higher than 15 IPM when BLC is on.

Any ideas?

Thanks,
Jason

[vonnieda]

I thought maybe a video would help here.
These two videos are running the exact same Gcode. The only thing I changed between the first and second is I unticked the backlash compensation checkbox in the Backlash settings. If you listen in the first video the grinding/seizing sound is really clear.

Backlash on: [ame="http://www.youtube.com/watch?v=PMa6wtscvog"]YouTube - Mill w/ Backlash Compension ON[/ame]
Backlash off: [ame="http://www.youtube.com/watch?v=vrF1-n6oWqU"]YouTube - Mill w/ Backlash Compensation OFF[/ame]

I also found this thread: http://www.cnczone.com/forums/showthread.php?t=17103

This seems to describe exactly the problem I am having but I tried switching to Exact Stop mode and it does not help.

[xy_eezy]

Parallel port is putting out a steady 4.6v and no dips there.

If you are using a quality meter, then 4.6v is nowhere near acceptable for your computers 5v supply. Your computer P.S. could very easily be the source of your somewhat odd issues. Could be as simple as replacing a few leaky/swollen caps. or so, but definitely should be addressed before your controller or something else is damaged.

[HackMax]

What do you have your acceleration set to in the motor tuning section? What version of Mach3 are you using?

If your power supply and driver are working correctly it really looks like your table or Z axis is on the verge of binding up somewhere - either too tight gibs or lead/ball screw binding. I went out and tried the same thing on my X3 mill and it works fine. It looks as though the Z movement is taking most of the voltage/current and stalling the rest of the system.

Rick

[TOOL MQR]

I have an X3 running Mach3. When BLC is on my programs take 1 1/2 times longer to run. It has a long hesitation when changing directions. This, of course, is hard on cutters. I have .003 in X & Y and .007 in Z. I hope to counterwieght Z to help it out. Do I have something set wrong? Or is this the way it is? I was used to a Fanuc control, Mach3 is new to me.

[vonnieda]

If you are using a quality meter, then 4.6v is nowhere near acceptable for your computers 5v supply. Your computer P.S. could very easily be the source of your somewhat odd issues. Could be as simple as replacing a few leaky/swollen caps. or so, but definitely should be addressed before your controller or something else is damaged.

My understanding was that 4.6v is pretty well within tolerances on modern parallel ports due to the power saving measures that go into newer motherboards. I know the computer power supply is putting out 5v cause I get 5.02v on the USB port.

This may be a path worth looking in to, but it seems to me that if the drives and everything else work fine at the same speed with BLC off it's not an electrical issue?

[vonnieda]

What do you have your acceleration set to in the motor tuning section? What version of Mach3 are you using?

If your power supply and driver are working correctly it really looks like your table or Z axis is on the verge of binding up somewhere - either too tight gibs or lead/ball screw binding. I went out and tried the same thing on my X3 mill and it works fine. It looks as though the Z movement is taking most of the voltage/current and stalling the rest of the system.

Rick

Accel is set to 2 for all axes.

Version is whatever the lockdown version was in February. I will check the exact as soon as I get home.

I had my steppers all off not long ago doing some major cleaning and all axes moved nice and smooth, even after I put it all back together. Also, as I said before, with BLC off I can run 50 IPM no problem at all and have complex moves in all three axes. If I was near binding, wouldn't I see it here as well?

Do you use BLC on your X3? And are you running steppers or servos?

[xy_eezy]

I just took a quick look at the 201's manual(actually it said 201A), to familiarize myself a little more with your hardware. I'm not familiar at all with the Dan M. hardware you have. Maybe you could post a few pics(both inside and out, would be helpful) of your controller/P.S. in order to get a more exact idea of what you are working with. But anyway, where is the 5v at pin 10(201's) derived from, and what does this voltage read? Since you have a scope you should be able to look at the dir., step, and 5v inputs at the driver itself(be sure and use the ground there as well). look for proper waveshape(square), and amplitude, with the 36v main disconnected from your drives(compare with and without blc). Then check it all again with the 36v connected along with your motors now running, both properly, and also exhibiting their faulty behavior. Pay particular attention to any level of noise that may now be present on those 3 lines. Also, do you have easy access to another computer? That would be a fairly straightforward way to eliminate about half of the possibilities for the source of your problems.

Clint

[HackMax]

Accel is set to 2 for all axes.

Version is whatever the lockdown version was in February. I will check the exact as soon as I get home.

I had my steppers all off not long ago doing some major cleaning and all axes moved nice and smooth, even after I put it all back together. Also, as I said before, with BLC off I can run 50 IPM no problem at all and have complex moves in all three axes. If I was near binding, wouldn't I see it here as well?

Do you use BLC on your X3? And are you running steppers or servos?

I have mine sety up with steppers. Yes I use BLC when I need to do drilling or squaring metal. I found it leaves a stepped surface when using ot on radius cutting.

The reason I ask these questions when you have the BLC working it almost instantly (at least on my setup) compensates for backlash and doesn't use the acceleration until the backlash it made up for and the table on the verge of binding would surely bind if your backlash settings are off a bit. So if all is ok there I can't help you any further since this is the only thing I've experienced that is somewhat similar to your problem. Good luck I hope you get it going soon. Keep us informed.

Rick