Newbie- Help with EMC2 timing for CW230 drivers?

[kazoochips]

Hi all, new guy here in sunny Kalamazoo, Michigan. Dad bought the book from buildyourcnc.com and we've finished building a modified version of that machine with a 3x3' x/y table and about 8" of z.

Setting up the electronics now and I've got the wiring figured out all right, but I'm not sure about what numbers to feed EMC2.

As part of the software setup, the stepconf wizard wants to know the following timing values for the driver board: Step time, Step space, Direction hold, Direction Setup, and whether it steps on the rising or falling edge of the pulse.

See, for example, the table at the EMC2 wiki:

http://wiki.linuxcnc.org/cgi-bin/emc...r_Drive_Timing

and this page from the EMC2 docs:

Configuration Wizard for Stepper-Based Machines

We have CW230 drivers (like this one here: Step Motor Driver, 3A Max, CW230-MPJA, Inc. ) Manufacturers like Gecko and Xylotex apparently have all of the timing values published in their datasheets, but I can't find the similar numbers for the CW230. I've been googling for like two days now and I'm starting to get frustrated. Also I don't have handy access to a oscilliscope, so unless somebody close by has one I can borrow, I'm running out of ideas.

Can anyone help me out with the right values so I can feed them to EMC2? Or, alternately, if someone else has EMC2 working with the CW230 drivers, could you tell me what numbers you decided to use?

I'm cross-posting this in the stepper motor / driver forum in case someone there can help.

Thanks in advance, and can't wait to start making chips!

[doorknob]

This is only a guess, because I don't have a CW230 and also because while I have installed EMC2 on a test machine, I haven't actually used it, so keep that in mind as I think out loud here:

If you want to be able to tune your system to get the best mix of ultimate performance and reliability, then you will want to get those numbers right on target.

However, if your immediate goal is to get your system up and running, then you can use some guesswork to get started. You should be conservative in choosing the numbers (meaning, use "larger" numbers instead of "smaller" numbers, which translates into specifying "slower" signals rather than "faster" signals).

In fact, the link that you provided advises as much:

If your unsure about your drive timing start high like 10000 for each and test. remember that signal conditioning and opto-isolation can increase timing requirements.

My best guess is that the CW230 uses optoisolated inputs, which are typically "slow" devices. So, I would approach this task by taking one of the "slower" configurations shown in the list, and either use that for your starting values or else slow them down even further (by increasing the numbers), just to see if you can get your system to work. Afterwards you can work on better tuning.

[kazoochips]

Hey doorknob, thanks for your thoughts.

I'm pretty sure I can set the numbers high just to get things working initially. I guess I was just thinking that since there seem to be a lot of folks selling (and presumably buying) those CW230 driver units, someone might know what the timing values should be. I hate guessing, you know? Especially when the information must be out there somewhere.

Anyhow- it looks like Friday is going to be the day that we finally power it up and see what happens. So if I don't find an answer before then, I'll probably just set everything way up high like 10k or 20k and see if the machine loses any steps...

[cncbasher]

their quite standard in the timings you should find that
3.0us step and 5.0us direction will be fine

[kazoochips]

OK, cool, I'll try 3 and 5 and see if the motors complain...

[ryckebusch]

How did your timing work out? I have a similar setup and am having troubles getting it all working. When testing the axis while in the setup I get really jerky motion and sometimes it reverses direction.

Thanks,

Jules

[jbmcc]

I am helping a friend with an opto-isolated breakout board and the opto-isolated CW230 (too much protection?). It seems reliable setting all values to 20,000 (20 uS). We will try adjusting down and I'll report back.
EMC-2 10-04, Dell D810 laptop

[doorknob]

Originally Posted by jbmcc I am helping a friend with an opto-isolated breakout board and the opto-isolated CW230 (too much protection?). It seems reliable setting all values to 20,000 (20 uS). We will try adjusting down and I'll report back. EMC-2 10-04, Dell D810 laptop

Using an optoisolated output from the breakout board to drive the optoisolated input of the motor driver may not be the best approach - you run the risk of degrading your signals, which at the extreme could possibly lead to problems such as missed steps or incorrect timing of some step pulses with respect to the direction pulse (although EMC2 does give you some flexibility in adjusting those parameters, so if it occurs you may be able to overcome it).

[kazoochips]

sorry for the lag in my response, things have been busy here. New baby in the house and all that means less time to play in the shop.

I've been running at 5000 step and 20000 dir for a few months now without issue, at least not any that I know about. I picked those numbers somewhat arbitrarily and I don't have any reason to suspect that they are "good" or "bad." I do know the machine works with those values. I have not tried the 3 and 5 values that were suggested above.

The most significant performance tweak we made was to stick a couple of hockey pucks onto the non-driven ends of the stepper motors. google "hockey puck dampers" or "stepper motor damper hockey puck" and you'll find lots of stuff about it. The upshot is you find the center of the flat face of an actual $2.00 hockey puck, and you drill a hole in it that is just slightly smaller than the axle sticking out the back end of your stepper. Then you slip that puppy on and re-test your max speeds and accelerations.

Before the pucks, I got stuck in nasty resonance around 30 IPM on both X and Y axis and couldn't run the machine any faster without freezing up the stepper and/or dropping lots of steps. After adding the pucks I can comfortably cut at 70+ IPM on both axes which is plenty fast enough for my purposes.