Re: [CAD_CAM_EDM_DRO] Re: Thanks - was: Things to check? is now: RUNNING~!

Weyland wrote:

>On Sunday 21 September 2003 12:26 pm, Jon Elson spake:
>
>
>
>>>I seem to get my best motion (not ragged, not stalling)
>>>with the axis speed set to 25, and the feed override set to 35.
>>>So, using your example, 25IPM multiplied by 35% = 8.75.
>>>And being set at 4000 on the input scale, comes out to 35000.
>>>Divided by 60, it would seem I'm getting 583.333 pulses per second.
>>>
>>>This is only for jogging the axis', though....
>>>
>>>
>>Well, that seems pretty slow. You should be able to do much better.
>>By the way, why did you use a 2:1 belt reduction. This is almost always
>>counterproductive with steppers. You'd double your speed immediately
>>by going to a 1:1 ratio.
>>
>>
>
>Agreed, and understood, but I was planning on having ballscrews by now. (:>)
>Thus, the reduction in screw pitch will leave me where I'd want to be with
>them.
>
>
>
>>But, let me give one data point, the closest I've got. I have Gecko 201A
>>drives on my minimill, which I got mostly for demos at NAMES. I have
>>size 34 motors on it, with a 4 A rating. I demoed it running at 90 IPM
>>at NAMES. (I had some binding in the non ball-bearing Y bracket.
>>Why the minimill has 2 ball bearings in the X axis, and a sleeve bearing
>>with the hand-wheel as the thrist bearing for Y, I can't fathom. After
>>fixing this with ball bearings, it runs again at 90 IPM.) This is
>>running the
>>motors at 1440 RPM! Obviously, this is pushing old-style 2-phase
>>steppers about as fast as they can go. That is feeding 45,000 steps/second
>>to the G201A, which doesn't have the step multiplier. You probably can't
>>do this with software step generation, but 25,000 steps/sec is not
>>generally a problem. You are quite far from there at the moment.
>>
>>So, what happens when you go faster? Does it rattle, or just stall out?
>>
>>
>
>..depends... and I'm not quite sure on *WHAT* it depends upon yet...
>
>If I'm jogging the axis', and increase the axis speed or feed override too
>much. they will move, but will be jagged and sometimes need to be told to jog
>more than once for movement to occur.
>Sometimes (often) in this scenario, they growl as well.
>Growling being a low frequency (to my ear) sound that rumbles.
>

What is your MAX_ACCELERATION parameter set to in the .ini file?

>
>Another tidbit is that if I set the axis speed or feed override too high, and
>actually get an axis to jog, that it will continue to move after releasing
>the key or mouse click, and most often in the opposite direction.
>
>

I've never seen this related to speed. I do know it is possible by
bobbling on
the arrow keys to get it moving without holding a key down. Pressing a
key for that axis and releasing it will reset the key press reding code in
the GUI. (The problem is the way keyboard press and release messages are
sent by the keyboards when you hit several keys at once.)

Jon