Designing a new router called Brevis-HD

Hi All - I have been whittling away at a heavy duty benchtop router. This thread is picking up close to the rough design completion. I'm about to compare a belt drive to a ballscrew drive and make the decision which way to go. The machine is intended for hobbyists and makers. It is a 1/4 sheet machine and will take a full sheet overhanging. So in theory if you register the setups you could cut a full sheet. Its for hobbyists who have a benchtop or small garage space. It's bigger then chinese machines and uses 20mm square rails for its motion control. I looked at 15mm miniature rails, 15mm std rails and the price was similar so went with the bigger rails and cars. This thread will be more about the build then the engineering. Its bench is a made from 19mm hardwood ply or birch ply. It's a high rail design so its stiff and contains dust well. It's a good size for guitar makers. I may extend the outer sheets to create a box to contain dust even better...All comments appreciated :) Peter

Hi Pete.

I guess the first thing that comes to mind is how do you define a bench? The short X axis and long Y bother me, this especially with the thin gantry beam. Of course if the goal is to fit on a narrow bench then this arrangement May be required however that would likely mean giving up your full sheet capacity.

As for hobbiests the needs vary so much there that I’m not sure targeting half sheet capacity makes sense. For example somebody doing Guitars would likely not need that capacity. He might need a long axis to do necks (or whatever they are called) but to achieve quality results a stiff machine is required.

I guess my idea of a bench top machine is generally much smaller. Unless someone has an unused bench someplace suitable for a large machine a machine this large should be self supporting in my mind. A self supported machine adds mass where it does a lot of good.

Hi Wizard - Yes What is a bench? 700x1800? 1200x2400? If parked against a wall then yes hard to feed a sheet through.... I have been fending inquiries for Brevis-M for some time. Brevis-M is 800mm deep and 1100mm wide. Fits on most benches easily. Brevis-HD is 800mm deep and 1600mm wide and would fit on any of the benches I have. I lose inquiries to large chinese extrusion machines because they are ~$1000 less. New users don't get the value of rails and cars vs extrusions and wheels. Brevis-HD is a step up in envelope aimed at people that want to use sheet goods or make instruments, plus have the extra Z height for jigs etc. The first one is for a person just like that and I think there are others. So the size will investigate the market that the chinese machines don't cover. Being bigger the cost of rails seems to fit better as well. Most people who have looked at my machines indicate they would build their own bench so you are correct in that area.

I suppose "Benchtop " to me means that 2 people can pick it up and move it from bench to bench.

I have done a few necks for guitar makers and there are advantages to doing it on the Y axis. There is only one drive system on the Y so the long 3D cuts are done slightly smoother than if you do them on the X axis as the dual drive always has a small difference in it due to the gantry being controlled by two systems. The "thin" gantry will be addressed when I do some FEA work on it but its currently stiffer then my 1/2 sheet machine which does quite well, plus I think it will be a ballscrew drive so even better. Peter

I have to agree with Peter.

I actually like the fact that it is "widthwise". I believe it is easier to arrange a standard depth long bench along a garage wall/small shack instead of a deep bench which sticks out into the space. Makes also material loading easier than feeding into the narrow side of the machine in a small/crowded space. The possibility of full sheet feed-through is not possible if oriented along a wall but somebody who often cuts full sheets would probably get at least a 4x4 anyway. For occasional use, one could temporarily turn the machine 90 degrees into the space to feed through full sheets.

Half sheet is possible by flipping. The question of gantry rigidity remains.

I will have the same orientation on my Brygga desktop line of machines once all sourcing, revisions and long-term testing is finished.

Put another guide rail upside-down on the gantry. You could use a smaller rail to offset the cost but I think it would decrease deflection a lot. I like where you’re going with this design.


Sent from my iPhone using Tapatalk

Hi Yoyo - I'm not sure what you mean by another rail upside down? Does it include a car? If its just a rail for extra stiffness that's an expensive addition, can use a steel bar. But the required size of the gantry is the same as any full sheet machine. Its about 120x120x3 mm at the moment which is very stiff for a kit machine. But when I get to the FEA I can change this. Currently I'm working on clean simple solutions for each structural loop which I think I'm close to. The gantry is also sized by the spindle.

I originally was going to spec a 48V DC ER17 600W as the top spindle size but people invloved kept talking about 1.5kW ER11 spindle so that's seems to be the area needed. But the 1.5kW is heavy so this contributes to gantry sag. Again once I get to FEA I'll resolve that question. Peter

I don’t have any engineering experience but it looks to me like your rail is going to twist with a tool deflection load. What I’m suggesting is a second rail and second bearing. If your calcs bear out the results you want then mine is a moot point!


Sent from my iPhone using Tapatalk

Hi Yoyo - The gantry (Brevis-HD) has two 20mm rails with 2 cars on the front and one at the rear. (All axes have 2 rails) Square rails support moments so it will actually work with one rail. My first machine used one rail for all axes. Worked fine for a while but developed a slight wobble over time as the cars wore. The rear single car is for stability and transmitting the torque of the gantry evenly across the faces.

I think you mean the machine in the photos which is Brevis-M. Yes it has one rail per axis and one car. It cuts plastic and timber very well with a 400W DC spindle. Brevis-M was made as a demo machine so it can be carted around to trade shows etc. Its been used as a knife cutter, plotter, laser engraver, mechanical engraver and carver. I wouldn't go as far as to say its a router as yes it does flex a bit. But I can pick it up and put it in my truck and take it somewhere to show people. That's its intent and its doing great at that at the moment. Its current purpose is to trial geared steppers and drivers that I intend to use on a large router called Maximus. Do a search for it here if you like engineering.. Peter

can anyone check for cnc information...https://cnchomedecor.com/

Hi Peter,
looks good.
A few years ago I helped a customer with a design for a plasma table. He cuts small parts as a rule so he wanted to
be able to fit a 1200 wide sheet but the y axis only had 600mm of travel so the ends of the sheet overhung.

He was adamant that it had to have zero lash ballscrews rather the rack and pinion of his existing table and therefore
he limited the y axis movement to save money on the long screws and rails which would otherwise be required.

Funny thing is that while he did save money by constructing that way but then went out and bought 4 Omron servos and drives,
they are the same company and identical to Yaskawa, and from the local (New Zealand) importers for $3000NZD each, without
considering other brands or sources of supply!! One step forward, two steps back in my book.

Craig

Hi All- had a look at costing of the drives. Belts vs ballscrew. Ballscrew won by about half. The precision belt system would cost about $120AUD per metre and a C7 ballscrew will be $70AUD per metre. So now need to address that in the model. I costed the ballscrews from BST Motion... Any comment on these for a hobby, maker level machine? Good/Bad Peter

Hi Peter,
quality is commensurate with the price. Whatever else may be said about Chinese manufacturers they certainly have bought bargain prices
for ballscrews to the hobby market. Not so many years ago these would have been very VERY much more costly.

Clearly a major goal of your design is to keep the price down.....these ballscrews will help that goal.

Craig

Hi Criag - Yes this is an entry level machine and I think the C7's match that at +/-0.05mm per 300mm. Plus the target users are timber workers so better then +/-0.1mm is fine. Cheers Peter s

Hi Peter,
thanks for the attachment, very succinct.

Rolled ballscrews tend to have a high cyclic error, or fluctuation/ 2PI, in THK's terminology, I have seen it specified elsewhere and 20um/2PI is typical.
This is still very much academic for an entry level woodworking machine but for a metal cutting mill would be problematic.

Is it your intention to direct couple the ballscrews to the steppers? I imagine direct coupling would save dollars both in parts and assembly time offsetting
the extra cost of high(er) torque steppers.


Craig

Hi Craig - I'm in that design space now. If I direct couple then the motors stick out the front of the bench like many routers do. Being a bench top machine this maybe a problem or not. If I use a belt drive then I can gear the motor and reflex it above the drive screw so it does not stick out and potentially use a smaller motor. But brackets/pulleys/belts will increase cost and the belt will add some compliance to the drive. The front of the bench could be an apron so the rails will need to come fwd a bit so the bit can get there. But the immediate issue is to connect the drive nut to the gantry. The stack heights of parts don't agree with bearing heights so some fangdangelry will have to happen.I dislike spacers just for spacers must be a clean way to do this. Peter

Quote:

---

Originally Posted by peteeng

Hi Yoyo - The gantry (Brevis-HD) has two 20mm rails with 2 cars on the front and one at the rear. (All axes have 2 rails) Square rails support moments so it will actually work with one rail. My first machine used one rail for all axes. Worked fine for a while but developed a slight wobble over time as the cars wore. The rear single car is for stability and transmitting the torque of the gantry evenly across the faces.

I think you mean the machine in the photos which is Brevis-M. Yes it has one rail per axis and one car. It cuts plastic and timber very well with a 400W DC spindle. Brevis-M was made as a demo machine so it can be carted around to trade shows etc. Its been used as a knife cutter, plotter, laser engraver, mechanical engraver and carver. I wouldn't go as far as to say its a router as yes it does flex a bit. But I can pick it up and put it in my truck and take it somewhere to show people. That's its intent and its doing great at that at the moment. Its current purpose is to trial geared steppers and drivers that I intend to use on a large router called Maximus. Do a search for it here if you like engineering.. Peter

---

Quote:

---

Originally Posted by joeavaerage

Hi Peter,
thanks for the attachment, very succinct.

Rolled ballscrews tend to have a high cyclic error, or fluctuation/ 2PI, in THK's terminology, I have seen it specified elsewhere and 20um/2PI is typical.
This is still very much academic for an entry level woodworking machine but for a metal cutting mill would be problematic.

Is it your intention to direct couple the ballscrews to the steppers? I imagine direct coupling would save dollars both in parts and assembly time offsetting
the extra cost of high(er) torque steppers.


Craig

---

i have never bothered with screw mapping in the past because what i usually ended up making was small in size. but this is probably something to consider with error in lead accuracy over a large distance. i wound imagine once you figure out exactly how much you have over a known value this is a pattern that repeats and could be controlled somewhat in the control software? might not be a perfect solution but i would think would make the over all accuracy acceptable to some degree.

I'm probably in your target "market". Have considered designing and building my machine but have far too many other things on my plate to do the full monty. Building from a well thought out design is definitely possible. I have been looking at an Avid Pro4824 class machine - in Metric, I'd think 1200x600 mm is pretty close. 12060 in Chinese. Even though the Pro4824 is not a "benchtop" machine, I'm not hung up on labels. The real issue to me is overall machine footprint as I don't have a huge amount of space and the Pro4824 is pushing the boundaries of my available shop space. Getting smaller than the Avid machine's complete envelope (including electronics and VFD cases) but still maintaining that working envelope would be a very good thing. I could opt for a smaller machine but have a number of projects that will need at least a full meter of working space in X or Y. That leads me to think the extra cost and small amount of compliance loss is a worthwhile trade off.

So, if this moves forward fairly soon, it's quite likely that I will "play along".

Hi all - my landline has been down for 10 days and it's a PITA doing this on my phone. I have been looking at the screw drives and 1204 will give lots of force but will top out at maybe 2000mm/min so I'm looking at 1610 screws. OTHERWISE DESIGN IS NEARLY. Oops complete. Will have quotes for parts in next 2 weeks. Peter

Hi All- Back on line at last. 16 days no connection thanks Telstra. Systems seem to be normal now. So update... I'm going with 1610 screws, now I have to size motors. I've looked at whirling of the screws and my FE and other equations seem to be in conflict. So that's an interesting area to sort. My development model has developed a few bad habits so its time to rebuild into correct sub assemblies from scratch and build a production grade model. Been sorting some wiring and I have a meeting with the sheet metal company to sort some preferred metal thicknesses and bending data. Then its onward to the build. I have decided to solder some of the bracing on the machine vs welding or bolting... minimal distortion, very compact and easy to do...Peter

10 mm lead seems like a very good happy medium on these machines. i used a 5 mm on mine and wished i would have went with a 10 mm lead.