LinkBack URL
About LinkBacks
Check out Keling or Marchant Dice for fixed and free end bearing blocks. If you needed to you could probably sell those pretty easily on eBay.
| | |
Situation: I purchased ballscrews for a DIY CNC that I have been planning. It never occurred to me that they could be a wee-bit too big.
Story:
I've been planning my second CNC for a while now. It is (was?) going to be a 4'x3'x1' steel construction machine. Before I bought the two behemoth ballscrews I had already acquired a fair amount of material. This included a stepper driver power supply, stepper drivers, a large stepper motor, all the linear rails I would need, and a Z axis ballscrew.
I then went on to purchase the said ballscrews. I eventually got in contact with a nice fellow that bought up ballscrews at auctions. I arranged to purchase the ballscrews and sent him a check, and he sent the ballscrews in two separate boxes. My face when opening the boxes:I did know how big they were because I had seen pictures and dimensions. However, seeing something in person is always totally different than with an image and numbers. Needless to say, they were not as I had imagined.
Since the purchase of the two ballscrews I have been a little scared. I called about 20 people and finally tracked down an engineering diagram for the ballscrews so I could get a better idea of how everything is supposed to go together. The diagram shows a bushing going onto the ballscrew, bearings going on top of that, and a mounting case going around all of that and the motor coupling.
The Good:
- Ballscrews are heavy-duty and very large.
- They are precision ground.
- They move very easily and appear to be very high quality (USA Made!).
- Good price.
The Bad:
- Ballscrews are very large.
- Few details on dimensions. I've been measuring with a caliper.
Problems:
- I don't know what to use for motors. How in the world am I supposed to drive these two huge ballscrews? I was planning on using a NEMA 34 stepper I bought, but the shaft is much smaller than the engineering drawing for the ballscrew. I want to stick with steppers but I fear that they won't be strong enough and I really can't afford any really nice servo setups (unless someone can advise me of a decent solution).
- I don't know what material, tollerance to make the bushing that goes around the ballscrew with. I have submitted a RFQ for this part but I think I need some details ironed out first before I accept an offer.
- There are more, but I'll just start with these two since they are the most pressing.
Check out Keling or Marchant Dice for fixed and free end bearing blocks. If you needed to you could probably sell those pretty easily on eBay.
How big are the screws (diameter and length)? It's possible you could use them, but if they're too big and have too much inertia, it would limit your acceleration. Usually that is not a big factor - you're just worried about the mass of the parts of the machine you'll be moving and the cutting force required. But if you've vastly oversized the screws from what is typical for a machine like what you're building then the inertia of the screws might need to be taken into account.
As far as practical mounting considerations, you need thrust bearings on one end (typically the end pointing up in your first picture). On big screws like this those are usually fancier and pretty expensive, but you could put your own together using separate thrust bearings and radial bearings if that would work for what you want your machine to do. Then you need a coupler to fit on the end of the shaft that mates to a coupler on your motor. Alternatively you can use timing belts on the screw and on your motor if you need to gear the motor down.
I don't see why not having dimensions is a problem... You have the screws - measure whatever you need... If you need to know diameters more precisely buy a micrometer.
CNC mill build thread: http://www.cnczone.com/forums/vertical_mill_lathe_project_log/110305-gantry_mill.html
jsheerin,
Both are 2.25" in Diameter. One is ~3ft the other is a few inches shorter than 3ft. They are heavy, I'd say ~50 lbs each.
I am going to put together an angular contact bearing assembly as seen in the third picture. The bearings are 7314 bearings as recommended by the ballscrew manufacturer.
Maybe I'm just worried, but it doesn't seem right to mount a small nema 34 motor shaft to such a large (1.373in!) shaft end. See the comparison in the last image. The motor is on the left and the screw, on the right. Any idea on if I will need a bigger motor?
The engineering diagram provided specifies a MAC 12D-0-ED servo motor which can kick around 38 N*m = 28 lbf*ft. That is astounding. A motor like that would cost me $1,000 or more!
Do you see why I am worried?
--------------------------------------------------------------------
I don't think I want to sell them just yet. Nice try at getting them though.Originally Posted by louieatienza
Actually I just bought such a motor used for $10 on ebay (plus $100 shipping). It will do 38N*m continuous, probably something like 250N*m+ intermittent. The big but is the current required to do that - 30A for continuous, ~200A for the intermittent, and that it's a 3 phase ac servo motor and would require much more complicated electronics to run (my current project - sorting out those electronics).
The bearings they're recommending are very nice I'm sure, but you don't have to use those. They're probably suitable for using these screws in a large milling machine which you're probably not doing. The thrust loads you will be applying will likely be much smaller if you're making a diy wood router (I assume that's what you're making as that's what this forum is about).
In any case, if they're 2.5tpi and you went with something like a 960oz*in N34 stepper from Keling, had a 300 pound or less moving mass, and kept acceleration at 100in/s^2 or less, you'd probably be okay running up to 300-350ipm. If the screw pitch is different or you know what your moving mass might be, I can run that again.
CNC mill build thread: http://www.cnczone.com/forums/vertical_mill_lathe_project_log/110305-gantry_mill.html
How expensive are the servo electronics? I'd assume you also need an additional PSU too.
The loads will be less, yes, but I intend on using the CNC to cut Aluminum and hopefully a bit of steel too (nothing serious). I posted here because of the large DIY base and because I couldn't find a better place to post.
Also, I priced the 7314 bearings and they are nearly the cheapest AC bearings that would fit around the ballscrew. The OD of the ballscrew where the bearings go is 1.65in. Browsing VXB, few bearings I find have an ID that large.
My ballscrews are 1.00 inch pitch double start. I'd say the gantry would be 300lbs or less (rough guestimation) and I figure I can get a 1000ozin stepper from e$bay or Keling.
Side note: I will be getting 4000 steps/in via 1/10 microstepping.
You don't have to run the calc again, but I am curious as to what formula you are using.
In theory you could drive ANYTHING with ANY size motor, you just need to use some gear reduction. Granted it wouldn't be ideal, but as you mentioned the alternative might cost you $1000 for a motor.
If you use a high performance stepper with an upper RPM limit of 4000 RPM, and use (just an example) a 10:1 reduction, you would still get 400 RPM at the screw. I know that a lot of the steppers from various internet sources don't give you the full specs, especially as far as recommended RPM range, but I have purchased some "high output" NEMA 23 steppers from Automation Direct. They rated them with an upper RPM limit of 4000, and said running at that speed would be "no problem". They also have torque curve graphs available in their documentation which will give you a clear picture of what the motor will do, and what voltages you can pump into it.
Problem with steppers running at high RPM is the drop in torque. Again, the high output motors have a better torque curve than others, especially once the RPMs get above 500 RPM.
The other issue with running high RPM steppers is getting a good power supply that can put out the voltage (and amps) required to run the steppers at a high RPM. Higher voltage = more torque at higher RPM, but it also = more heat, so you may actually need to cool the motors (fan cooling or even a water cooling block) if running under heavy load continuously.
I believe the motors from Automation direct will handle up to 80VDC (just a guess, I forget the actual value). If you pair a good power supply with a Gecko G203V, you might be able to make something like that work.
I don't know what kind of accuracy you are wanting, but a 10:1 belt drive wouldn't be that hard to build and shouldn't affect accuracy too much. I don't know if you have machine tools at your disposal to build something like the gear reduction. I might be able to help with that if not.
Hope this was useful.
He is more machine now than man.....
Forgot to mention, if you need some help in determining the size motor to use, especially with different gear reductions, there is a free program called VisualSizer that can give you a good estimate of power required. It's got a bit of a learning curve, and the amount of detail you have to provide to get an estimate can be overwhelming if you aren't familiar with the terminology.
Here is the link to the description page for VisualSizer. There is a download link in the right-hand menu.
VisualSizer - Servo Motor Sizing and Selection Sofwtare for Windows
He is more machine now than man.....
cjjonesarmory,
Sadly, I do not own any significant set of tools so a gear reduction system would be a much harder for me to make/setup than I am sure it sounds to you. That does not rule out gear reduction, but it is a significant downside for me. Thank you for the idea. I will think about it. Also, thanks for the link. I will check it out.
The best and easiest solution for me would be a high performance stepper motor directly driving the ballscrew. I just need someone to assure me that I can get away with a 1000+ozin (don't have) stepper on a 50V 20A PSU (have) rather than spend all my cash on a new servo system.
The servo is somewhere I really don't want to go because I can't fully afford it and it would involve new electronics I already bought once to drive steppers.
High-end Servos:
- Guarenteed to work.
- Very very costly.
- Have none of what is needed.
High-end Steppers:
- Not sure to work.
- Cheap(er).
- Already have most of what is needed.
Servos being guaranteed to work is a bit of a stretch... They can be much more complicated to set up. As far as cost, you might be looking at $1500 or so for a 3 axis setup plus a power supply to run big motors, but there's a wide range of options available ranging from very cheap to very expensive. It all depends on what kind of performance you want. I'd agree with you that sticking with steppers is a better idea for your project.
I used an excel spreadsheet I wrote to do the calculations necessary to see what kind of performance you could get from the setup I mentioned. The formulas were originally from a pdf posted on cnczone about how to calculate how much torque is required for a given screw and machine setup, but I've added a bunch of other stuff. It's probably similar to the link someone else posted.
Gearing a motor down can be as simple as ordering two timing pulleys and a timing belt. If you use different size pulleys, you can make the screw turn faster or slower than the motor and change the torque you apply to the screw. Physical gears are actually less than desirable as they'll typically have some backlash, but it's still referred to as 'gearing down'.
CNC mill build thread: http://www.cnczone.com/forums/vertical_mill_lathe_project_log/110305-gantry_mill.html
Wow, those ball scews could have been used for pretty heavy VMC.
To give you a point of reference the conversion on my J-head Bridgeport mill uses 1.25" dia 5 TPI ball screws for the X and Y axis. These are driven by by some 1500 in-oz steppers with a 2:1 timing belt reduction using a Gecko 203V stepper driver. The result is 20,000 steps per inch. I can do 100ipm rapids but usually limit it to 75 and cuts less than 50 ipm.
My knee is driven by a 1810 in-oz stepper and a Gecko 203V with a 3:1 belt drive to a 1.5" dia ball screw at 5 tpi. This will do atleast 75ipm accelerating 750 lbs or knee and table. The knee is assisted with 400# of gas springs to reduce the static load.
I would use a 2:1 belt drive on your system -- There a numerous advantages to using a belt drive system.
Don't expect much accuracy from the micro-stepping it mainly helps out for motor mid band resonances and smooth operation. Just assume the motor's basic steps are you maximum accuracy.
Craig
Posting Permissions
| | |
