Messages

221

Building or Buying a Wood routing table.. Beginnners guide.. / Re: Servo ratios and drives..

« on: October 15, 2006, 02:57:15 AM »

"Id like about 12,700mm's per minute. Or about 500 inches per minute as my rapid. It may be overreaching, but we'll crunch some numbers and see. That means on a 4" pulley, Id need about
2*3.14*2 or about 1 foot per rotation on a 4" pulley with  reversed belt, so an end speed of about 500/12 = 41rpm , so a reduction of about 75:1 woudl seem great under that scheme.
Gonna have to look around to see what I can find and make a compromise I suspect. "

Hi Art, This is pretty close to what I have set up on my mill, and in retrospect I would have done it differently.
My motors are a little slower than yours and I am using a 50:1 planetary gear reduction to a belt drive. It is all very heavy duty but i still have a little backlash ~.025 in the x (my long axis) .01 in the y and .002 in the z (ball screw). My gantry with the z weighs around 1300 lbs. There is lots of mass there, and as a result to avoid oscillation (i am suspecting belt spring) i have to have the accel to around 20. full out the machine is capable of 550 ipm on x and y and 160 on the z.

You are indicating that you would like to have around 72:1 for a reduction. To accomplish this you basically have two options. First a $1600 low backlash gear head or a big mess of belts and pulleys. for 72 to 1 it is going to end up probably being a 4 stage belt pulley assembly. that turns out to 8 or nine not cheap pulleys 4 belts lots of bearings and shafts and a bunch of machining for the transmission case. Belts do stretch a little and do have a little backlash and the more you have the worse it gets. Especially if you have a bunch of mass to deal with at high speeds. Don't forget you have to drive the long axis from both sides otherwise you will get racking in the gantry, that adds to the complexity and cumulative backlash.
I opted for option one with the gear drives. My plans this week are to definitively find the backlash in my x. I have a feeling it is a combination of belt stretch 6 arc min of backlash in the gear head and a twisting of the drive 2" shaft.

Any way I said if i had it all over to do again it would be different. What i think would be a better route is to go bigger on the servo and do a one stage 4 or 5 to one with a 2" belt. This gets rid of the backlash in the already expensive gear head, gives more speed and makes for a simple transmission. We are planning on converting the entire thing over to ac servo and the plan is to put around 3hp on the x,y,z and use as little reduction as we can get away with.  We recently got a Haas tl1 and after carefully digging around it and getting ideas i realized that they aren't doing any reduction with the exception of the ball screw pitch. The servos are 750 watt and are directly coupled to the screws. This seems to provide plenty of power and i have a feeling they have limited the rapid electronically to 200 ipm as it should be capable of way faster rapids ( not that that would get you anything other than shaking your machine apart, at least on this one).

To summarize i would suggest saving the money on fancy planetary gear heads or tricky pulley belt creations and put it into a bigger motor and a simple reduction..

Just my $.02

Chad

222

General Mach Discussion / Re: inserting pics in post

« on: October 13, 2006, 05:31:11 PM »

Hi, when you are writing the message look toward the bottom of the text box on the left there is a link" Additional Options" click on that. It will then ask you for attachments. select your jpeg and click post.


Don't forget to scale them.

Chad

223

Building or Buying a Wood routing table.. Beginnners guide.. / Re: Servo ratios and drives..

« on: October 13, 2006, 12:49:54 AM »

HI Art,

Is permitted to discuss things here or is it only a log for you? I have a few comments based on my own design that I would like to share. And some problems that I have run into in this area.

If not please feel free to nuke this post.



Chad

224

General Mach Discussion / Re: Zoom in Mach 3

« on: October 13, 2006, 12:33:41 AM »

I second that about the zoom, I think it is a bug that art needs to look at. It was frustrating me just today.  It seems to zoom incrementally to a point then it jumps to like 200%. It basically makes the tool-path display useless, at least to spin around and take a close look at something.


Chad

225

General Mach Discussion / Re: Printer port

« on: October 13, 2006, 12:30:46 AM »

Yes, the "printer port" is usually a separate chip on the board. Sometimes it can combine things like, serial, usb, lpt.

It is possible to fry the chip and not be able to use the printer port but the rest of the machine would still be fine. I always like to install a cheap lpt card. If it gets fried then is is no big deal to replace. 

Chad
 

226

General Mach Discussion / Re: Newbie servo question

« on: October 10, 2006, 02:43:20 PM »

Hi Yoda!   Ok you have to take a couple of steps back before moving forward. The first thing you have to figure out is your steps per unit. The 2000 you entered is just an arbitrary number until you know: how many counts your encoders are, what pitch your ball screw is, your belt reduction. Until you know these things that number won't mean anything, and neither will your velocity or accel.  Look at the top of this screen under tutorials and watch the videos they will help fill in some gaps.

Now if you don't have any idea the encoder count or the pitch of the screws there is a nifty utility to get you in the correct ball park.

Get a ruler and a calipers or better yet a dial indicator and go to the settings page. get your machine in the middle of it's travel and click on set steps per unit button. This will bring up a window that will ask for an axis to set. pick the x.
Next it will ask you to enter an amount to move enter something small to start with like 1". The machine will move what it thinks 1" should be. it will then ask you how far it actually moved, use the ruler to get a rough amount.
So if you entered 1" and the machine moved .25" then enter this number. Click on and do it again. this time try to get more accurate. Put one inch in again and measure the result and enter it when asked, this time it should be a lot closer than the first time.
Once you get in the ball park you can use a dial indicator to get really accurate. repeat this process for all of the other axis. I have a feeling that your 2000 steps per unit (in this case inch) is going to be WAY too low for a servo system with a ball screw, encoders and a 4 to 1 belt reduction.
Once you get this axis calibration down to the point that you enter 1 inch and the machine actually travels one inch you can use that number with a measurement of the pitch of the ball screw, and the 4 to 1 reduction to determine the encoder count.   

Now that your machine is set NOW you can start playing with the velocity and the accel and try to tune your motors for good results.

Chad

227

Show"N"Tell ( Your Machines) / Re: My Frankenstein

« on: October 10, 2006, 01:32:32 PM »

I have a similar machine, I used retrofit components from Elrod machine. He makes a VERY nice quill drive. It is cool in the fact that it can be manually disengaged so you can still use the quill handle when you need to. It is a great company to work with. This would really help with the speed problem in the z. I can get over 300 ipm on all three axis but I have it limited to about half of that as it is just scary at those speeds.

You can see some pictures of our machine here (towards the bottom) http://www.machsupport.com/forum/index.php?topic=627.0

Here is the link to the elrod site. http://www.elrodmachine.com/CNC_Z_Axis_Quill_Drive.htm

Chad

228

General Mach Discussion / Re: Z axis moves on its own in program/jumps steps

« on: October 07, 2006, 01:29:26 PM »

Hi, this could be a bunch of things.
First make sure there is no binding in your z axis travel.
Second slow down your accl and velocity and see if that helps.
Third your z could be just to heavy for the size stepper you are using.
Fourth just for fun try increasing you step and Dir pulse width to 5us and see if that makes any diff.

The dro does not know if the motor stalled so they will always read the correct location. This tells me that mach is doing it's thing and the problem is in the hardware..

Chad

229

General Mach Discussion / Kudos for the excellent support.

« on: October 03, 2006, 07:29:25 PM »

Just thought I would relate a little story about what happened today.

I have a converted Bridgeport style mill with a g100. I had been holding off on updating mach and the plug in because things were mostly working and I have some stuff to get done by this weekend. I was having a problem with mach and the g100 just stopping mid job. I know art has been working on this and it was probably time to update.
I updated to the latest 2.0g and the latest g100 plug in. I loaded gcode and did a dry cut and the thing would hang just after start, no jog no Nada. Not good.
I sent a couple of panicked emails to art and an hour later I had a new g100 plug in and have been cutting without problems sense.

Art and the artsoft team offer astonishing customer service, especially for the price.

Thanks!

Chad 

230

General Mach Discussion / Re: Mach 3 SDK ?

« on: October 01, 2006, 08:23:36 PM »

C++ , aaah ok. For some reason i thought you guys were using c#.

Chad