[StrawberryBoi]

Absolutely makes sense to use a Left/Right screw when the motor is a significant expense. In automotive traction motor design work I'm so very often asked about hub motors or one motor per wheel with CV shaft, paired with a gearbox typically (because wheels are very low speed and high torque). Sprung mass discussions entirely aside, splitting motors is great when you need independent control at high fidelity, but when you are looking for a net effect and not a specific effect (ie moving the car and not managing specific torque at an individual wheel) then the cost of separate motors and drivers can be significant.

Quite often my clients are shocked to learn the motor controller is often the largest and most expensive component in high performance traction applications. Plus when they are split you often need to oversize the individual motors.

Now this isn't a traction application, there isn't a slipping wheel that's not putting traction to the ground that another wheel now must compensate for, but there is a variation in the inertia and potentially the drag depending on the spindle location and build variations, wear, etc. This means that one motor may end up holding more load than the other sometimes and your total net performance is less than the two motors combined would seem to have on paper.

[henkfromkullsta]

I have now spent quite some time to look through this thread, really inspiring.
I have had my CNC plans on hold for 10-15 years. But now my life situation makes it possible to start a project. I have been thinking about a central screw lifting the gantry. I add a sketch. (The ideas a really early).

Edit, Cannot manage to rotate the picture on the phone.. And it was quite confusing already from its intended viewpoint.

[peteeng]

Hi Henk - If I'm interpreting correctly - you have a moving table, and a moving Z axis on a gantry and the gantry is lifting up and down. For this to work you will need a bridge above the gantry to hold the drive system (say its a ballscrew) Then the bridge is the reaction point for the central screw. If you drive in the middle of the gantry the bearings need to be well spaced to stop the gantry "walking". All doable you need to develop the idea a bit more. Peter

[henkfromkullsta]

Hi Henk - If I'm interpreting correctly - you have a moving table, and a moving Z axis on a gantry and the gantry is lifting up and down. For this to work you will need a bridge above the gantry to hold the drive system (say its a ballscrew) Then the bridge is the reaction point for the central screw. If you drive in the middle of the gantry the bearings need to be well spaced to stop the gantry "walking". All doable you need to develop the idea a bit more. Peter

Yes correct! The bridge is the beam with the four holes. Thanks for the feedback. I will try to do a CAD later in the autumn.

[Zorbit]

I should of explained it better a Left and Right hand Ballscrew is need when using ( 1 ) motor, 2 Ballscrews, here is an engineering post on this subject so nothing silly about it at all, Helical racks also have to have left and right hand which a lot of builder's are now using

Depending on the drive arrangement this could still be done with ( 2 ) right hand Ballscrews with ( 1 ) motor being used like you are saying but with any gearing being used ( 1 ) Ballscrew would be reversed so then would need a Right and Left hand Ballscrew

Why in the world would anyone need a ball screw with a left-handed thread? Why not simply reverse the program and drive the device in the opposite direction? Unfortunately, this may not always be the best solution. Several industrial applications require a left-handed ball screw to help resolve high-precision programming requirements and reduce the number of needed expensive components.


For example, a system designed with two ball screws; one right-handed and the other left-handed, with each unit positioned on either side of one drive motor moving in opposite directions. This configuration is less expensive and much more practical because only one motor is required and there is no need to synchronize two separate right-handed ball screw systems driven with two servo or stepper motors.

Can you link to any machines that use a left and right hand ballscrew ? The only reason I can think for doing it is if the two screws were very close to each other and gear driven one from the other, if both screws were driven from one gear on the motor there's no need to reverse one screw thread.

[Zorbit]

OK, I found the article you cut and pasted from, the picture shows a "self centering" arrangement. Nothing relevant to routers or mills as far as I can see.

Left-Handed Ball Screws Level the Playing Field (linearmotiontips.com)

[StrawberryBoi]

Nice catch on that Zorbit. Good point, the one motor two screw situation for the machines actually in question would have the same motion direction on both screws and so reversing the thread hand makes no sense.

[peteeng]

Hi All - something like this. Peter

https://www.mabi-robotic.com/home/in...asp?IDinfo=254

[peteeng]

Hi9 Guys - Don't get bogged down in details LH and RH screws exist so they do have applications. Peter

[peteeng]

Morning all - I have been chatting briefly to RoboDK in Canada and they can write a post for the robot. So Milli-IRB maybe a good moonshot project... Stay tuned. Peter

[peteeng]

Evening all - I have found two more packages. That's really good, gives me confidence there will be a controller solution out there.

look up Octopuz.com rhinorobot and RoboDK. all look good... So get Frankenrouter sorted and move onto Milli-IRB.. Peter

[Zorbit]

Morning all - I have been chatting briefly to RoboDK in Canada and they can write a post for the robot. So Milli-IRB maybe a good moonshot project... Stay tuned. Peter

I look forward to it.

Might I suggest a pick-and-place machine rather than a full-strength milling robot ? It would allow you to experiment with the kinematics on cheaper hardware, and play to your strengths in carbon composites - lightweight and fast machines.

Mirobot | 6-axis Mini-industrial Robot Arm - YouTube

[peteeng]

Hi Zorbit - I'm a go hard or go home sort of designer. Everyone I deal with wants routers on robots. Pick and place can be left to the others. An economy robo router is the go... I can buy a mirobot for about $1500AUD if I want to play with them. But that won't tell me anything. I used to play with welding robots as a job with ABB. Time for play is over.... robots are kool. Peter

A software engineer from RoboDK just sent me a note asking for more info so they are keen! I said I'll send them a CAD file when I get it moving...

Hi Zorbit - I used to take half steps and general opinion is that's a good plan but I've converted to taking really big steps and then falling back. Takes same time and you gain ground faster and generally end up further forward. Time is slippery and precious, got to max out some stuff otherwise it doesn't happen...

The first thing to do is a bit of a trade study to determine the geometry that gives me a reasonable envelope.

[mactec54]

Hi All - something like this. Peter

https://www.mabi-robotic.com/home/in...asp?IDinfo=254

I have seen these machining Robots just like this at the IMTS industrial show that normally is every 2 years in Chicago, they have lots of working Robots doing all kind of things machining being one of them, this may be the cam side of it which you may already have

https://www.autodesk.com/solutions/r...mming-software

[peteeng]

Hi Mactec - thks for that., Some years ago I was working for a boat builder and did research on robot trimmers. I decided Powermill was the best package on a Kuka robot. That's mentioned in your link. Ta Peter

[ardenum]

What I'm wondering is how you're gonna manufacture the frame, seems the shapes are something that need to be cast or printed, the latter not really being an option due to $$$

[peteeng]

Hi Ard - Earlier in the thread I investigated ultra high performance concrete UHPC. So this can be used to cold cast parts. The masters or plugs can be machined on my router in MDF then fibreglass or urethane moulds can be made. I can also make parts in carbon fibre. I have some short fibre CF here for test. Agreed metal 3D printing is too expensive but maybe some day the cost will be OK. Would love to get a project that justifies buying a metal printer. The best material for the parts is infused CF, it will be the stiffest and lightest material. I spent a bit of time on this today and identified that the bearings will be a hurdle. Slewing bearings are very expensive and thin bearings are $$$. I'd like to give air bearings a go but I don't think they will be stiff enough. Peter

[peteeng]

Hi All - I have been studying the Kawasaki robot geometry, very interesting. I'll do the bottom drive quite differently though. My main thoughts are about the lower big bearing... Peter

[ardenum]

Slewing rings up to 200mm dia are somewhere around $200-400 per piece on aliexpress. XU series from INA, RU from THK and these are cross roller types. There's a third one available at the same price range, 3 roller type, super stiff, the kind they use in rotary tables, this one is YRT series.

[peteeng]

Hi Ard - Like this? My sisters name is Susan and these things have always irritated her. Peter

but this asian harmonic drive company maybe useful. My current plan for the bearings is to use Hub-2 or Hub-3 bearings from 4WD's. I have used these on a scara robot before. But maybe slewing rings are the go.