how to choose a Servo motor

[rceebuilder]

I used to use stepmotors for my designs so far but my new project requires servomotors . My new project will be a big router with the workable area of 3 m x 2 m and with the jogging speed of 1 m / sec . Probably I will use kremayer - pinion for the axis movements to achieve this speed .I will cut 18 mm MDF sheets for furniture industry. For optimum servo selection I was studying on data sheets of some motors. But I 'm confused a lot . I hope someone will help me to find optimum solution. At the beginning I have started with some considerations ; for example the cutting force will be 30 kg ,the jogging speed will be 1 m/sec . if I use 40 mm pinion I need 60 kg-cm( 6Nm) tork and I need 480 rpm at the output shaft of the motor ( or gear drive) . I thought if I use a servo without gear reduction very big motor must be used but I never use it at 3000 rpm where the eff. is high as 70 % and as the speed reduced the tork also reduces ( please see the data sheet) So I thought if I use a 5/1 reduction I will get 500 rpm at shaft when the motor at 2500 rpm and I will get very high tork with a small 1.5 Nm motor.This is the first step which I need your confirmation. Second step is the selection of the motors . In my opinion there is two kind of DC motors brushed and brushless . I think Brushless motors are better since they don't require maintenance . This is the second subject that I can not decide . For the drivers there is no big price difference between brushed and brushless drivers . I suppose I will get all the motors and drivers from here http://www.transmotec.com/ since the prices are very affordable.

[ger21]

How fast will you be cutting? You may end up with higher than 30Kg cutting force if your cutting near 8-10meters/minute. You better have a powerful spindle as well.

[Geof]

I think you might have to reconsider your possible motor size. The chart you posted showed a maximum input power of 838 watts. I cannot comment on what torque you might need but I think you need much more power than 838 watts. I have a gantry machine that rapids the Y axis at 2100 ipm (53 m/min), it is ballscrew and uses a servo motor with a peak input power of just under 10kw and it runs up to this during acceleration on a rapid move.

[rceebuilder]

Originally Posted by ger21 How fast will you be cutting? You may end up with higher than 30Kg cutting force if your cutting near 8-10meters/minute. You better have a powerful spindle as well.

Gerry this is exactly what I want .According to my calculations a 500 watt DCbrushservo with 1.6 Nm @ 3000 rpm will be fit perfectly or 660 watt DC brushless servo can be used. What do you recommend? These motors data sheets are attached.

Another point that I don't understand is the sychronization of all 3 axis as servo motors are used. Let's say we use 3 servo and 3 drivers .Okay ,each servo works as closed loop but if one of the servos is jammed how the others know this In my opinion when this occurs machine must stop and must give an alarm. Is it possible to to it with these hobby servodrivers ?

Murat.

[Mariss Freimanis]

My calculations show 294W are needed for 1 meter/sec and 30 kg of thrust.

Mariss

[Geof]

Originally Posted by Mariss Freimanis My calculations show 294W are needed for 1 meter/sec and 30 kg of thrust. Mariss

At 100% efficiency this is correct but a bit optimistic for the real world.

[Mariss Freimanis]

Well, I don't know the transmission efficiency; 30% for the crappiest ACME thread, over 95% for the best ball-screws. I would assume a rack and pinion with reduction gearing falls between those two limits.

Mariss

[Geof]

You also need to take into account motor efficiency (probably not much better than 85% and considerably less if the motor is operating away from its optimum rpm) and frictional drag on the whole moving assembly at that speed. The design intent is to get 30kg of thrust above the frictional drag so if it takes 5kg of thrust just to move the assembly at that speed then the total thrust needed is 35kg. If there is a gear train in the drive allow for an efficiency here of around 85% maximum (a belt would be closer to 65%), rack and pinion will be comparable to a gear train so the calculation comes out as:

85% x 85% x 85% = 61.41%

The total thrust of 35kg at 1m/sec gives 343 divided by 0.6141 = 558watts

But this does not give any over capacity for acceleration so it is still an underestimate.

[Mariss Freimanis]

It's not quite that bad.:-)

1) Motor efficiency is not taken into account. The calculated power is Watts mechanical, which is what servo motors are rated at. Motor efficiency only comes into play when you need to compute Watts electrical going in.

Motor efficiency is a function of torque, specifically: eff = 1 - (load torque) / (stall torque). A frictionless motor would have 100% efficiency at an infinitesmal load, falling to 50% at 1/2 of stall torque load, dropping to zero % at stall torque. There's always friction around so efficiency tops out at 90% + for the best motors at very light loads.

2) The calculations are very much on the conservative side; 30 kg was specified as a work load while 1 meter/sec was mentioned as a 'jog' which I took to mean a rapid during which no work load is required. The calculations were based on a 30 kg work load at the 'rapid' speed. This is an unlikely, worst-case scenario.

Any torque in excess of work load can be invested in acceleration; were the moving part of the mechanism weigh 30 kg, this would allow 1G of acceleration, meaning you could go from zero to 1 m/s in 102 milliseconds.

Mariss

[rceebuilder]

thanks guys It seems my calculations are not so bad. What do you think about brush or brushless motors. I'm gonna choose brushless DC type as they don't need any maintenance . But I still don't know how the servodrivers communicate each other. Or in other words do they need to communicate ?
Murat.

[turmite]

Murat the kind Mr Freimanis did not bring attention to himself, but he is Mr. Gecko. I have no idea if his drives will suffice for your needs or not, but he can tell you. The Gecko's will trip if they get to a pre-set number of pulses out and that in turn will trip an e-stop if you have your machine wired that way. One axis trips and the complete system should shut down.

Mike