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
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