One Big, One Smaller ... My 1st builds - TrickyCNC

[TrickyCNC]

So

Here I am all green and new.
Bursting with (over) enthusiasm !!

Me ... I currently am self employed, and 'make/do stuff' for a living.

I find myself loving woodwork (in the proper sense) but hating doing it for a living !!! MDF dust just kills me

So... CNC ... I've been semi interested for a while, but not had any funds to seriously think about it.

I have a product (MDF) that has taken off quite well, and in turn I have a little spare cash to 'play' with. This means I can think about CNC seriously for the 1st time.
I must admit to being totally addicted already and cant think of much else ...

So

I will build a CNC router to work with a full 8' x 4' sheet (probably second). and also a smaller machine that can be more precise. For PCB and Aluminium milling, along with engraving and carving. This one I will build 1st and try and lean from.

So ... My 1st question to you all.
What is the largest optimum size machine for detailed work like PCB isolation, small engraving etc etc.
I also want to use this to machine wood. So as large as practical, without loosing precision.

I was thinking MDF/PLY for the construction, but I am quite happy working with steel/welding too.. I am more set up for wood at the moment. Steel would need more planning of work space etc.

Thanks
Rich

Similar Threads:

• Looking for DIY 5 Axis builds
• New Machine Build- Two new builds at the same time
• New Machine Build- Mastercam X5 Builds
• Newbie- VM builds
• TRON cnc re-builds

[CarveOne]

That's a loaded question with no absolute answer - other than "It depends".

Just to throw you a number, it would be less than 18" square and as stiff as your choice of materials will allow.

CarveOne

[TrickyCNC]

That's a loaded question with no absolute answer - other than "It depends".

Just to throw you a number, it would be less than 18" square and as stiff as your choice of materials will allow.

CarveOne

Hi C1

hmmm... so are you saying your newest build, is not up to the job I want ?
I read between the lines that you were hoping to mill some Ally with it ?

If 18" is the answer, then 18" it will be.

I was hoping for a bit(double) more though ! LOL

24" x 36" seems like a nice size. would this be possible with a good strong build ? I beams, RSJ's ? (do you know 'rsj's' over there ?)

Rich

[CarveOne]

Hi C1

hmmm... so are you saying your newest build, is not up to the job I want ?
I read between the lines that you were hoping to mill some Ally with it ?

If 18" is the answer, then 18" it will be.

I was hoping for a bit(double) more though ! LOL

24" x 36" seems like a nice size. would this be possible with a good strong build ? I beams, RSJ's ? (do you know 'rsj's' over there ?)

Rich

If I were to build a dedicated circuit board/engraving machine I think I would make it the 18" square format and use the same steel box tube.

I won't know my latest machine passes muster for brass nameplate engraving and circuit board cutting until I get the G540 repaired and get some cutters to try it. Things like the letter "O" can be difficult to cut correctly when it is only 0.1" or so tall, due not so much to stiffness of the machine, but due to backlash. The shapes aren't cut true to form and look ugly.

It should have no problem making aluminum router plates of better quality than my big machine. That was the primary goal. If it is precise enough to cut circuit boards and do engraving I will be that much happier with it. I think the box tube frame I ended up with will allow for all of this. Lead nuts, and the Z axis design will be the potential areas of any backlash and weakness. Fine tuning the bearing preloads will be essential.

24x37 working area is a good size. Many solid body guitar and necks are made on this size machine. I ended up with slightly over 28x35 once I made the last modification to the rear X axis lead screw mounting plates. If the machine will do a good job with the fine stuff it will have no issues with the wood and plastic projects.

CarveOne

[TrickyCNC]

sorry, I'm not upto date with my reading. I didn't know your G540 was damaged ?

I'm thinking of a zero (or there abouts) backlash cable system. direct drive, or belt geared to a wheel/drum .

I will design it failure proof, in that I can add a new drive mechanism at any time, if the cable doesn't work.

The 8x4 one would be geard down belt drive to a wheel (large drum) with a cable attacthched to prevent slippage.

I estimate roughly a 30" wheel will give 8' of linear movement per revolution, therfore allowing the cable to be fixed to the wheel. The drive being driven by geared belt system.

I have yet to read up on 'micro stepping' and its advantages / disadvantages. for extra precision.

I will have fun learning, even if I end up going ballscrew or R&P in the end !

Rich

[TrickyCNC]

anyway, I'm off to bed now. It's 11.35PM , and I've been up since 4AM thinking about this stuff !

I've sent out for quotes on skate bearings and M8 threaded rod, so I should be ordering tomorrow

I dont know what rail system I'll go with yet. Round/Square/angle/edge /etc.

Hopefully a good night sleep tonight !

Rich

[CarveOne]

sorry, I'm not upto date with my reading. I didn't know your G540 was damaged ?

I'm thinking of a zero (or there abouts) backlash cable system. direct drive, or belt geared to a wheel/drum .

I will design it failure proof, in that I can add a new drive mechanism at any time, if the cable doesn't work.

The 8x4 one would be geard down belt drive to a wheel (large drum) with a cable attacthched to prevent slippage.

I estimate roughly a 30" wheel will give 8' of linear movement per revolution, therfore allowing the cable to be fixed to the wheel. The drive being driven by geared belt system.

I have yet to read up on 'micro stepping' and its advantages / disadvantages. for extra precision.

I will have fun learning, even if I end up going ballscrew or R&P in the end !

Rich

Yep, did it while I was tuning the X and A motor drivers. Too busy to arrange to send it back for repair this week. Maybe in the coming week.

2/3 of the fun is in designing and building. Read the opinions and go do your own thing with what seems to fit your needs best. I'm doing good so far. Three builds and I've spent less than a trillion dollars total on them.

CarveOne

[stevespo]

I always recommend that people look at the successful designs and take what they can from those. It's certainly great fun to think outside the box, but the proven designs have already solved many of the problems you will encounter.

Focus on the small machine. Design it to do what you need to do. Go back to the large machine when you've got the first one up and running. If done well, the first machine will be a huge help on the 2nd.

A PCB mill can take just about any form imaginable. A mill capable of cutting aluminum will probably be a different beast entirely. You've got to set some realistic parameters and let your needs dictate what to build.

People always say "building a CNC is easy". And it is! Like hitting a baseball. It all depends on what league you're playing in and what you need the machine to accomplish. It's a really fun, educational (and costly) balancing act.

Steve

[TrickyCNC]

So, looking for CNC parts, I came across some timing belt that isn't in a loop. You buy it by the meter. hmmm.... I thought, and instantly 'invented' the "Servo Belt" drive in my head ! I searched on here, to find it is already invented :-) and looks quite feasible for DIY.

So, until I change my mind again , belt drive it is , at least for the X and Y. I'm not sure about the Z yet.

My latest idea for my 1st machine is to make it 3' in the X and 4' in the Y (working area), then I can still load it with full 8 x 4 sheets to save having to break them down. I'd like to hear some feedback on that idea please. with the belt drive ... should I just jump in with 8 x 4 working area ? it will mean a big change to my small workshop layout.

I've ordered 100 skate bearings off ebay for £25 delivered, which I thought was OK. So, I have officially started collecting stuff for the build.
I'm going to use steel angle for the rails and bearing trucks, which I will phone round for localy.

I'm hoping this will work, each side of the bed, for the X rail

/ bearing
> rail
\ bearing

I need to get some software together to start drawing things in CAD. I'll be using Linux and looking for anything that's free !

[ger21]

hmmm.... I thought, and instantly 'invented' the "Servo Belt" drive in my head ! I searched on here, to find it is already invented :-) and looks quite feasible for DIY.

It's not that simple. It took me a long time to come up with my version, and after I built it, I've decided it wouldn't be good enough, so I'm redesigning it and will rebuild it. In the several years that the Servobelt thread's been going, only one person has shown a running DIY version. And it's performance (at least what was shown) was probably much less than 10% of the real thing.

[TrickyCNC]

thanks for the warning Gerry. I'll see what I can do then.

I've decided to build an 8 x 4, as this will be most useful to me from the start.

I'm just typing out my order for 95% of the parts needed to complete the build. It should be with me next week.

I've put my prices up on the stuff I make, so I can temporarily reduce my order book volume, and get building this machine. By the end of the week, I will have cleared my current orders and be free to dedicate some time for the build. I hope to have it built and fired up within a week of starting.

Yes... optimistic ! but that's the way I work.
It doesn't mean I will meet my deadline, it just means I will get a lot done

Rich

[TrickyCNC]

Just ordered steppers , drivers, BOB, PSU, Belt, and a couple of pulleys.

No going back now !

So ... what are the things I will forget to order ? such as wiring ? how much ?, connectors ? which ones ?, etc. etc.

[louieatienza]

It's not that simple. It took me a long time to come up with my version, and after I built it, I've decided it wouldn't be good enough, so I'm redesigning it and will rebuild it. In the several years that the Servobelt thread's been going, only one person has shown a running DIY version. And it's performance (at least what was shown) was probably much less than 10% of the real thing.

In all fairness [to HomerSimpson] he built a "StepperBelt" drive, using surplus helical gear-reduction, as opposed to the "real deal" made with direct-driven servos...

[CarveOne]

- snip -

I need to get some software together to start drawing things in CAD. I'll be using Linux and looking for anything that's free !

There are a number of free Linux CAD programs out there. Download and try out HeeksCAD. You can still use a Windows CAD program (I use TurboCAD) and Vcarve Pro to make gcode that will run in EMC2. CamBam can make the gcode also. CamBam and VcarvePro are UK companies.

CarveOne

[ger21]

In all fairness [to HomerSimpson] he built a "StepperBelt" drive, using surplus helical gear-reduction, as opposed to the "real deal" made with direct-driven servos...

I did say that "from what I saw" it was slow. For all I know it could be lighting fast. But my point was mainly that it's the only functioning DIY implementation that I've seen. If it was easy, there'd be more.

And, I fully expect my "Stepperbelt" to perform similarly to the real thing. But we'll have to wait and see.

[louieatienza]

I did say that "from what I saw" it was slow. For all I know it could be lighting fast. But my point was mainly that it's the only functioning DIY implementation that I've seen. If it was easy, there'd be more.

And, I fully expect my "Stepperbelt" to perform similarly to the real thing. But we'll have to wait and see.

True, though I don't know if his design goals were top rapids speed. My thinking is that there'd be more if it were more practical to implement, for the average DIYer using non-CNC tools.

I was hoping to see your y axis going at 1g, until you abandoned your first design....

[TrickyCNC]

to me, I am not looking to build the fastest machine on my 1st attempt. Just that it works faster than I would without it, and it is as accurate as I would be by hand. The fact that I can press 'GO' and then get on with something else would much improve my workshop output.

I expect to learn heaps from the 1st build and start a new/revamp design from my learning experience

[TrickyCNC]

So ... what are the things I will forget to order ? such as wiring ? how much ?, connectors ? which ones ?, etc. etc.

Anyone ?

Is there a list of parts needed anywhere ? ... emergency stop, limit/home switches, Zeroing ? manual positioning ??? I've never used one so I am sure to miss something out !

Rich

[Joe S.]

The fact that I can press 'GO' and then get on with something else would much improve my workshop output.

That might be a while in coming, it takes a while to get over standing and watching it run and thinking . . . . cool!

Is there a list of parts needed anywhere ? ... emergency stop, limit/home switches, Zeroing ? manual positioning ??? I've never used one so I am sure to miss something out !

Dust collection ? What manner of limit switches I. E. magnetic, mechanical, optical ?

[TrickyCNC]

Dust collection can come after I get it working. I have a box of microwsitches and probably some reed switches too, so that's OK.

As an added bonus...

I took my kids to see my Dad today, before the summer holidays are all used up. He used to have a metal fabrication shop, so I thought I'd ask him what kind of metal I should be buying for the rails n stuff.
Off we went on a tangent, that he had a few bits he'd took out of a factory they were closing down ... linear rails and bearings, maybe some motors too ! So we went out to his workshop and started rummaging around.

some rail, 4 sets of rail and bearings, some Ally extrusion, 2 ballswres with bearings and steppers - 5 TPI 430 Oz 29" long. and a larger stepper without a rating plate.

So, that's my other machine sorted, if I can get it all to work.