My DIY CNC, "CNC2011"; build log

[koaDave]

another idea for balancing Z: Quoting Romanlini in "5-axis build full steam ahead" in this forum;

"To balance the Z without adding an equal mass counterbalance I would use a strut or thin piston connected by a hose to a air resevoir, and keep the air resevoir at X PSI to dial in the exact amount of lift. Provided the air resevoir is significantly larger than the piston's total displacement then the pressure will remain pretty constant (within a few percent) and the lift at the Z axis will also remain constant so you can dial it in very nicely..."

A little more work, but elegant, light weight and easy to adjust.

[RomanLini]

Hey! I said that!

Actually I was just browsing through the thread and trying to think of a good way to balance the weight.

A more expensive style of gas-lift strut has a resevoir attached that is significantly larger than the piston displacement, so they provide a relatively constant lift force. The cheaper gas struts like they use of car hatchbacks don't have the resevoir and they have a big difference in lift depending on the stroke position. Of course if your total stroke is only small (3 inches?) then the non-linearity won't matter that much.

Once suggestion I would make is to ditch the cast iron motor and get a much lighter router motor. There are many good routers on the market and most have variable speed control which is another bonus.

[hub]

Thanks for the ideas!

Yes we have a quite a few hardware stores here in Finland, but none with springs the size required. Online stores have some though. Been having a nice easter & not thinking much about this the last few days

The Z travel is about 500mm ( 19.5 inches)... Springs would be available that size I think. Gas springs are a bit hard to find with that long a stroke. And those that might have a total length of.. (too much)
Same with compressed air cylinders. I haven't tried to find any, but it seems like a lot of stroke & total length.. Plus my current compressor makes lot of noise, would like to avoid using air in the machine if possible.
I'm still leaning towards the springs so far..

The motor, yes a nice light spindle motor would be nice, but the ones I have found are rated 5.000-30.000 RPM and are quite expensive (I'm on a budget). Also my VFD's max freq is 200Hz, they seem to want 400Hz..
I really don't want anything faster than 10.000 RPM. Why? Because if I have too much RPM I would have to have a very fast feed with very powerful steppers/servos.. That would work well with "softish" wood, foam etc. only I think. ?

I'm not planning to do any high volume production. So, if the machine is slow that's fine with me.
I would like to have decent accuracy though, it's more important to me.

And WOW Dave! That's some serious stuff! You're very lucky!! Keep me updated what you do with them please, would like to see what you make of them
If I'm lucky, hope so..., I'm getting 7 big servos & drives from an old robot being updated. I have to pay something, but practically free. Keeping my thumbs up!

[hub]

Got the machine assembled to a point it can do some cutting (test cuts only, it's not adjusted properly/precise yet except for the X axis).
Works fine so far
Lots of temporary solutions though.. which will be changed one at a time. Like the nice wooden bearing blocks for example! lol

The Z axis works very well without any springs or anything to fight the weight. No problems there. I'm worried that the ball screw or nut would wear quickly if it has that much weight to lift all the time?? If that's the case, I'll put in some springs to help lift it.

Also, I think I have made a design error. With the current design, I'm not able to adjust squareness/parallelism of Y and Z separately. If I loosen the bolts in the gantry it will loosen both Y & Z axis. (Trying to adjust Y square to X & the table; and also Z square to the table).
At the moment it's "ok", but not accurate at all.

And some pics of the current situation.

[hub]

To save money, I'm going for this design, check the picture.., (one less ball screw, ball nut, belt, 2 pulleys, bearings, blocks etc).
This would work I think? Has anyone tried this? If so, is there problems in accuracy and would it be rigid enough?
The steel wire can be tightened pretty hard and will hold a lot of force/weight.

Or is this a bad idea? Should I just go for a second ball screw and connect both with a belt? Or some other idea?

[koaDave]

Wow, Hub, great progress. Also glad the Z weight is not too great for the mechanism. (Yet) As for the steel cable idea, I had a table saw once that used that to keep the fence square. It was rock solid and reliable. Of course, who is measuring thousandths of and inch in woodworking? If the pulleys and mounts were sturdy, I think it would work, but since I have no experience. . .

Dave

[hub]

Thanks Dave
I'm definitely going to try the wire "thing" to see how well it works, unless someone tells it's a no-go

[hub]

I have done some adjustments and further tests. Runs pretty good so far.
Started to mount the wires to fix both sides of the gantry. I have only had the time to start it. I still need to design and machine the four wire rollers.
Will do some more work on it tomorrow I hope.

Some pics to keep it more interesting

[tommylight]

I have used the wire system on two of my builds and to a certain point it works perfectly ( on a 50x35 cm engraver and a 2.5x1.5m plasma cutter), i have also had this temporarily on a 2.5x1.5m router with a 1mm steel wire but it would play a bit ( maybe 1mm at full load cutting 10mm thick plastic with a 8mm end mill doing 700mm/min in one pass. I think for your size the 1mm wire would suffice and if it doesn't try thicker wire ( more load on the motor).
You can find some of my builds using that method here just do a search for tommylight.
Regards,
Tom

[hub]

Thanks Tom
I still haven't got to a point to install the wires. (Need to machine the rollers & put some feet on the machine because the wire must be underneath)

1mm play would be way too much, I'm hoping to achieve much less..
Tight enough it should theoretically be close to zero play? And I don't understand why thicker wire would require more power? Using good bearings there should not be much friction?
Thanks again for your help

[tommylight]

For rollers you can use bedroom closet door rollers and most of them have ball bearings, and the power thing is the wire is not elastic enough for small radius turns so it will need a bit more. Trust me it is noticeable.
Regards,
Tom

[hub]

Thanks for the tip and advice!
The rollers I was planning are pretty small, check the picture. Their diameter is 28mm (1.1inch). They would require machining a groove for the wire. I'll check the rollers you suggested.

The wire I have is pretty thin, 1.6mm (0.065inch) but I don't think it should be elastic at all, even that thin? It's steel wire. Even if it would be a tiny amount elastic, wouldn't tightening it enough resolve any issues? It'll be tightened "against itself" if you know what I mean.