[saxa]

Hello, I hope I'm posting this in the correct forum.

Ok my concern is about servo motor makers. Since I'm doing a retrofitting on a industrial gantry CNC plasma/oxy fuel cutting machine, I have opted for a DMM Tech servo motors with absolute encoders. Those would be 2 servos for the Y axis movement and one servo for the X axis movement. The machine is quite big, as it have tracks for travel longitudinally 10000mm (Y axis) and it is 5000mm (X axis) large. The overall weight is about 800 kgs.

Now my idea is to use 2 servos of 1 kW for Y axis, and 1 servo of 750W for the X axis. The X axis will have to move about 3 torches alltogether. (2 separate torches for oxy fuel, and 1 plasma + 1 oxy fuel torch).

DMM Tech have servo motors up to 1,8kW which can fit perfectly for this job, but in the second place I want to extend this machine to have also a drilling/milling device on it. Since we work with big and thick steel plates , with high hardness (ranges from 300HB up to 600HB) we will need big forces for the drilling milling unit. This unit is planned to have a bigger servo motor for the spindle, with power in a range of 7,5kW most probably. And another servo motor for the Z axis (up and down movement of the spindle).

My question is, is there any problems to mix the servo motors from different producers, for example we have some DMM Tech servos but then in the second place we will add some
Panasonic/Yaskawa/Delta or other producer servo ? We plan to use Linux CNC (EMC2) to drive this.

Thanks in advance for the suggestions.

[Al_The_Man]

Normally there should be no problem mix and matching servo's as long as the drive is compatible with the motor and suitable for the controller.
If gearing one servo off another as in dual axis, I would tend to use matching servo/drives in this case.
Al.

[saxa]

Dear Al, many thanks. Just reading on the internet few things and I saw there is many possibilities. Yes, I see that would be no problem in using some DMM servos with its drives and then adding some other brand servo and its drive. Now I'm looking for some other issues with getting the right gearboxes to move that thing.

[Al_The_Man]

One source of a graphic motor sizing program is one by Kollmorgen, which should be available out there.
For economical motor and drive size, you can use the highest ratio of reduction that will still give you the required FPM.
I have generally used planetary gear boxes for this.
Al,

[saxa]

Ok, for planetary gearboxes, I will use those, probably a ratio of 5:1 or maybe even more, since we will work with thick plates and machine will mostly move slowly. I noticed that some people have troubles when machines go slow, with not regular lines and so on.

Another doubt coming up to my mind is, the breakout board. Is there a way to connect different breakout board to the computer ? There is one paralel port usually, ok I can add more of them probably.
Or will the DMM Tech breakout board of 6 axis been able to drive the Toshiba or Yaskwa or other producers servos ?

[Al_The_Man]

It also depends on what the output of the G.B. drives, for example I retrofitted a large oxy/gas cutting machine where the gantry weighed over 300lbs, for this I used a rack and pinion method to move the gantry in conjunction with a BLDC servo and 100:1 planetary G.B.
I have never used Linux CNC so you may get more info on that from their site.
There a couple of Linux CNC experts here also so a post in that forum may bring results
Al.

[saxa]

Al, many thanks, for sure I will open a new thread for the planetary gearbox and linux cnc stuff. For now I can say that the gantry we are retrofitting here is about 800 kg, and it has rack and pinion. I was even thinking to change that rack and pinion from the straight one it uses right now to the inclined one.

[Al_The_Man]

Couple of shots.
Al.

[saxa]

Nice work Al, if you want to see more about the machine I'm working on, you can find the pictures here: Index of /files/cnc
These pictures were taken at the day the machine came in here, and today is not much different, but my plans are to unmount everything and repaint it all, then make it go.

I am doing also a 3D drawing of the machine so it will help me plan few things.

[Al_The_Man]

Certainly needs a good clean up and also new hoses etc .
The old Linde machine I did had a trace cutting head on it which was removed when CNC'd.
Al.

[saxa]

Yeah, I plan to basically remove the color by sandblasting it, then re-paint and mount everything. I am still not sure if I will leave all 3 oxy torches or I will just leave one oxy and one plasma torch. Later my plan was to put on a drilling/milling unit unit. Now the first step is to clean up all the rust, unmount it, and repaint. Meanwhile I will do all the cabling and hoses connections.

I just have few doubts, which are:

- what king of gearbox ratio use for it.
- if I put the CNC computer and controls on the gantry or do I leave them on the side
- Do I put the plasma source on the gantry to travel with it or do I put it on the side

Anybody can comment and advise benefits/downsides for this ?

Rgds
Saxa

[Al_The_Man]

What I did was take the maximum rpm of the motor and decide on a maximum feed rate required and work out the maximum reduction ratio that would still give me this feed rate, the R&P lead obviously has to be build in to this, (distance/one revolution of the pinion.)
Most Plasma power supplies are very heavy, the norm here has been to place it separately from the gantry, the down side is a very long water cooled conductor.
Al.

[saxa]

Thanks Al, what we have here as a source is about 50kg I think, Hypertherm Powermax 1650. Therefore put on the gantry some 50kgs more probably would not make difference. But yeah, maybe its
better from the safety point of view to have it on the ground, and buy the long cables needed. Problem is that we will have about 10m of tracks therefore I estimate I would need about 20m of cables.
Would that work ? another problem I can think of can be the interference of the plasma with the servos, I read about in a lot of sources that the plasma is way noisy with high frequencies and that can damage the servos. At least I think that our plasma source here does not need water cooling, therefore one hose less, but we for sure will have the oxy fuel hoses and the compressed air ones there for sure.

[saxa]

I forgot, for the speed, as far as I know plasma cutters usually go up to 8m/min so we have a rack already there, but I need to find the correct pinion gear diameter I will go with. The motoer speed should be
in a range of 2500 - 3000 RPM which is quite standard. If I get a gear of about 25mm diameter, probably I will need a reduction ratio of 30:1 to get a speed of 8m/min.

Am I correct ?

[Al_The_Man]

Some of the larger Plasma tables here travel at 25mm/min in rapid move between cuts.
They are typically 6m of track.
The BLDC motors I used were 3000rpm but the Linde was gas only not plasma.
Al.

[saxa]

Ok, got it, we will probably also travel way slower than the estimation I made, as the use will be for thick plates cutting of 20 up to 50mm , so probably its good to make a way lower estimation too. I would like to be able to cut al so thiner plates, probably of about 3-4mm as minimum. But maybe is an overkill to try to put a beast like this go so fast.

[Solidworks4u]

I am a new to forum and think it is obvious that there are a lot of people that have a lot going on at this forum. I offer my many thanks! I don't want to hijack this thread, but may I interject a related question about what is the relationship between servo rpm and feet per min. on a belt driven style of feed system? The plan is to squeeze a pvc extrusion between a pair of servo driven timing belts, this custom machine will have rollers to control the other sides, with a tool moving back and forth to cut simple repeating pattern down the length of extrusion. Will a timing belt ratio of 10 to 1 work, keeping the diameter of the timing pulley as small as possible (as in 1 revolution equal as small of distance as possible) or will a gear system you talk about, be the only way to go? Naturally cost is an issue. This is a proof the system will work, will not run every day, and if possible allow for both, suggest brand or type of gearbox to use.

many thanks for your input

Edit: The drives are from a Bridgeport mill, Baldor 5/8" shaft; 500 in-oz

[saxa]

I am a new to forum and think it is obvious that there are a lot of people that have a lot going on at this forum. I offer my many thanks! I don't want to hijack this thread, but may I interject a related question about what is the relationship between servo rpm and feet per min. on a belt driven style of feed system? The plan is to squeeze a pvc extrusion between a pair of servo driven timing belts, this custom machine will have rollers to control the other sides, with a tool moving back and forth to cut simple repeating pattern down the length of extrusion. Will a timing belt ratio of 10 to 1 work, keeping the diameter of the timing pulley as small as possible (as in 1 revolution equal as small of distance as possible) or will a gear system you talk about, be the only way to go? Naturally cost is an issue. This is a proof the system will work, will not run every day, and if possible allow for both, suggest brand or type of gearbox to use.

many thanks for your input

Edit: The drives are from a Bridgeport mill, Baldor 5/8" shaft; 500 in-oz

Ok, I am not sure if I understood it correctly. The relationship between the servo rotations er minute and feet per minute is a conversion of rotation into the linear movement. So to get it you willhave to see exactly few things:
- how fast you will roatate
- what is the diameter of the driving pulley
- what is the diameter of the driven pulley
And then you can calculate how fast the belt will move.

For example: If you r motors does 1000 RPM and your pulley ratio is 1:10 means that you will have the following calculation:
Morot pulley diameter 20mm
Belt pulley diameter 200mm
Belt pulley circumference is 2*3,141*100 = 628mm
Each rotation of motor will do 1/10 rotation of the belt pulley. So, if you rotate 1000 rotations in one minute, you will linearly make 62800mm/minute
Now how many feet this is I think you are able to figure it out.
If you want to go faster with the belt or with the motor, it is just a matter of change the diameters or simply the ratio.

I hope I did not understand it wrongly.

[Solidworks4u]

Many thanks saxa for your response and interest.

My thinking the drive needs to be up in the power curve vs. just barely moving to have more torque. For example I would not use a coupling direct to the timing pulley and belt assembly. (just to clarify, the timing belt will have considerable length to have firm grip on extrusion, I am just not shore just how far is enough). I realize it will reverse direction so at some point it will need to come to complete stop. I read that servos are better than steppers as with having more power from a complete stop.

[saxa]

Ok , I think if you need more power depending on how much for sure servos in my opinion are a way to go, but maybe again, depending on the exact construction and use you can use also a simple AC motor if you need then a feedback of the position, which I think you do not need I would go with a normal ac motor.

About the forces, I have no idea how much it would be needed, as it depends on a lot of factors. Extruded material, friction, diameter , etc. etc....