[margni74]

I thought I read in the emc2 docs that emc2 can output a maximum of ~100,000 pulses / second depending on hardware (trying to find that number now for my machine)

I have a 5 turns/inch ballscrew attached to my gantry. driving this gantry is a motor with a 250 cpr encoder, 250*4= 1000 pulses per ballscrew revolution. So, at best I can expect my system to achieve 100,000/1000 = 100 rev/sec = 20 ipm. This seems pretty slow.... What am i missing? This assumes a direct coupling to the ballscrew.

As it stands I guess I'm seeing a min movement/max resolution of 1 inch/1000 = .001". There is most likely a ton of other error and there is a small chance I could actually take advantage of being able to step a thousandth at a time, so perhaps I should gear the ballscrew up a bit? maybe 1:5 or 1:10? This would give me a speed of 100ipm or 200ipm... but then I'd be losing torque at the same factor.

I'd like to be able to get some decent speed out of this. Im not exactly sure what to expect from an emc2 driven machine.

[dannystooblue]

rpm at screw not rev/second. you forgot the to multiply by 60. You will want to gear this thing down a little to get the max torque from your motor while still keeping up the rpm's to the screw that you want. Also remember that you have a max rpm on the motor that you have to stay under, this is probably the limiting factor. if the motor is 3500rpm max and a 1000 count/rev 59k is the max pulse rate you can feed it.


Danny

[margni74]

rpm at screw not rev/second. you forgot the to multiply by 60. You will want to gear this thing down a little to get the max torque from your motor while still keeping up the rpm's to the screw that you want. Also remember that you have a max rpm on the motor that you have to stay under, this is probably the limiting factor. if the motor is 3500rpm max and a 1000 count/rev 59k is the max pulse rate you can feed it.


Danny

Nice catch, oops. so 20 inch/sec = 1200ipm. wow, that would be pretty fast, I think the motors I want to use are something around 3200-3500, like you stated. definately will be gearing down instead of needing to gear up for speed.

Thanks,
Scott

[dannystooblue]

I converted a medium sized benchtop square column mill to cnc before I had 1/2 a clue what I was doing. I ended up talking to Marriss at gecko drive and he opened my eyes a little to what I was looking for. First off was the motor limitations and the fact that peak and rated torque were two very different creatures. I had geared 2:1 on my y and z but 3:1 on x, and I did this because it made since to me at the time. After the conversation I realized that running my ballscrew at 1/2*3500 or 1750rpm would not be good for it, the second thing I realized that the motors I had decided to use were a little on the small side. The solution was to gear them out a little farther. For z 3:1, x 4:1, and y 3:1, 4:1 was the max without making new motor mounts, and it gets the screw down to a speed at which it should live well while giving me the torque to do a few things. I have a flash of it running in the gallery with the old configuration, I haven't redone the video yet but expect the accel/decel to be a good amount better. Oh... when I converted to the GIF format it sped up the thing a little, not 50% more but maybe 20% faster than actual.

[margni74]

I converted a medium sized benchtop square column mill to cnc before I had 1/2 a clue what I was doing. I ended up talking to Marriss at gecko drive and he opened my eyes a little to what I was looking for. First off was the motor limitations and the fact that peak and rated torque were two very different creatures. I had geared 2:1 on my y and z but 3:1 on x, and I did this because it made since to me at the time. After the conversation I realized that running my ballscrew at 1/2*3500 or 1750rpm would not be good for it, the second thing I realized that the motors I had decided to use were a little on the small side. The solution was to gear them out a little farther. For z 3:1, x 4:1, and y 3:1, 4:1 was the max without making new motor mounts, and it gets the screw down to a speed at which it should live well while giving me the torque to do a few things. I have a flash of it running in the gallery with the old configuration, I haven't redone the video yet but expect the accel/decel to be a good amount better. Oh... when I converted to the GIF format it sped up the thing a little, not 50% more but maybe 20% faster than actual.

Thanks for the info... How do I see your pics?

How many ipm is your machine running now in the x-y plane?

I had wanted to be able to get 500+ ipm for rapids on my x-y axis, but maybe that is too ambitious.

my current motors are dinky little Clifton precision motors that have a continuous torque rating of 27 oz-in! I was going to upgrade to 3 keling motors which are anywhere from 120-220 oz-in continuous.

[dannystooblue]

my x-5 speed... 3500rpm, 4:1ratio, 10mm pitch, gives me the posibility of 340ipm. I have it limited much lower than that still since it only has dovetail ways, I think I held it back under 180 or 200.

http://www.cnczone.com/gallery/showg...0/ppuser/17174

or you should be able to right click on my name see a link to them and one to all of my post.

Danny

[chester88]

Just to be clear are you talking steppers or servo motors?

[dannystooblue]

I am talking about servos.

Danny