Re: [CAD_CAM_EDM_DRO] (API) CMD-260 Controllers.
Posted by
art
on 2002-06-07 14:01:44 UTC
Hi All:
I recently purchased a couple of API drives from Tim Goldstein's web site.
Normally I'd use Gecko's, but in this case I needed an extreme of 50,000
pulses per rev of the motor due to a low drive ratio. (1 rev = 3.5" ). Its
hard to get a 100uStep driver at a reasonable price. I had reports of
missing steps on these controllers from others, and when hooked up found I
did indeed have many missing steps. I eventually found the opto-couplers are
biased to expect a buffering resistor inline to their inputs. Hooking
directly to a printer port could cause loss of activation just from the
small currents induced by capacitance on my cables. (Looking at them wrong
would cause activation of the signals). I hooked a 1Kohm resistor in-line to
the wire from my printer port and they now work great with no missing steps.
I suspect these drives won't be around long as they are a manufacturer
clearout, but if anyone else has a problem with them, this might save you
alot of trouble. At 100-200uSteps per step, they are an incredibly quiet
ride. Mach1 drives them at 25Khz with no apparent problems so far. So my 1
rotation for 3.5" system now runs at .000175 resolution at about 4 IPS. (I
realise I don't really get anywhere near that resolution as larger than
10uSteps really doesn't add anything, but it solves my ratio problem
nicely.)
Hope it helps, if you have a very low drive ratio it is one solution,
Art
ArtSoft CNC Software Company
http://users.andara.com/~fenerty/master.htm
I recently purchased a couple of API drives from Tim Goldstein's web site.
Normally I'd use Gecko's, but in this case I needed an extreme of 50,000
pulses per rev of the motor due to a low drive ratio. (1 rev = 3.5" ). Its
hard to get a 100uStep driver at a reasonable price. I had reports of
missing steps on these controllers from others, and when hooked up found I
did indeed have many missing steps. I eventually found the opto-couplers are
biased to expect a buffering resistor inline to their inputs. Hooking
directly to a printer port could cause loss of activation just from the
small currents induced by capacitance on my cables. (Looking at them wrong
would cause activation of the signals). I hooked a 1Kohm resistor in-line to
the wire from my printer port and they now work great with no missing steps.
I suspect these drives won't be around long as they are a manufacturer
clearout, but if anyone else has a problem with them, this might save you
alot of trouble. At 100-200uSteps per step, they are an incredibly quiet
ride. Mach1 drives them at 25Khz with no apparent problems so far. So my 1
rotation for 3.5" system now runs at .000175 resolution at about 4 IPS. (I
realise I don't really get anywhere near that resolution as larger than
10uSteps really doesn't add anything, but it solves my ratio problem
nicely.)
Hope it helps, if you have a very low drive ratio it is one solution,
Art
ArtSoft CNC Software Company
http://users.andara.com/~fenerty/master.htm