Nomad with deep DOC, shallow WOC

Thanks, that’s good to know! In that case I’ll avoid rigid couplers for the Z-axis.

I was also thinking about putting the steppers outside the machine. It would require a bit of drilling but that isn’t too bad.

Results don’t lie. Increased rigidity and power go hand in hand with cut depth and width :slightly_smiling_face:

1 Like

But the “Devils in the [disappointingly missing] details”! :wink:

What machine and spindle? Did you measure power/spindle current? Why only 15000 RPM?

Modded Nomad, 1.5kw air-cooled 80mm.

100 ipm due to max feed limits at the speed (400 steps per mm 1204s) also why the rpm is at 15k. Cutting super small chiploads with 3F has never been reliable for me.

Thanks,
So, deep and narrow works on ruggedized Nomad. Assuming the Nomad 3 is also limited to 100 IPM, it would likely be spindle power limited (24kRPM 150W) to the following with that endmill assuming Millalyzer is accurate.


Do you think the following would work with a stock (10kRPM 70W) Nomad PRO?

Better to stick with a good single flute imo. 45 degree 3f are “touchy” especially running dry. 30 degree, more forgiving for sure.

Hard to say about the Nomad 3…better anti-backlash nuts should make a huge improvement. Ide personally love to get my hands on one to see how it would handle.

That 883 pro cut…Definitely not. I mean it might cut, but its not going to be good for your tools, finish, or machine. Would predict “angry” noises.

I’m with Vince, absolutely not. That cut requires 65W of power at the endmill and the Nomad can’t get anywhere near that, even with high radial engagement cuts. With higher axial engagement, the jelly-like Nomad 883 Pro turns into a chattery mess.

Speaking of/to @Vince.Fab, Vince, what kind of linear rails and blocks did you end up using? I’m still not settled on whether to use the short blocks or make the carriage a bit bigger for the longer blocks.

So much for trying to match flute count, helix angle, and depth of cut to minimize cutter impact forces! Kind of makes me wonder how abrasive cutters would perform.

Oh no,

Now I have a mental image of somebody magnetic mounting an angle grinder onto their CNC spindle…

2 Likes

Regardless of forces, you need to get the chips out of the way somehow. With 3-flutes and without flooding, that’s not so easy, especially with large helix angles.

1 Like

HGH15s
Overkill, but better stronger than weaker. Plus they are thick enough to allow for easy enough mounting of the ball nut. The Z plate only needed a small pocket for the mount to clear the rear.

Oh those things still matter. The tricky part for desktop machines is having enough rigidity to take a big enough depth of cut for good enough flute engagement to smooth out everything. Its been very interesting trying to learn how to use @spargeltarzan calc and seeing how the data relates to real world usage and cuts.

This is that cut with 30 degree 3F, smooooooth

3 Likes

You’re such a tease!
I’m guessing it’s something like this? (Even deeper and shallower).

How many? Just one on each rail, or two?

In other news, I think I’ve just about finalized the Z-axis design:

I’d appreciate any feedback anyone has. In particular, the “bridge” that the ballscrew goes under is only 4.5mm thick. I figure that should be okay because most of the force should go straight into the rails but does anyone have an opinion?

And I think most importantly, can anyone see any fatal flaws in the design that should prevent me from purchasing the parts?

The parts I ended up selecting are:

  • Ballscrew: TBI XSUR1204T3D-FC5-231-P1 (note: C5 rolled ballscrew, not C7)
  • Rails: TBI TR15-166-H
  • Rail carriages: TBI TRS15VS, precision H, Z1 preload
  • Fixed support: SYK FK10
  • Floating support: SYX FF10
  • Motor bracket: SYK MBA10-C
  • Stepper motor: Leadshine 57CM06 (NEMA 23)
  • Coupler: HZ HZ8-D20L27-8*6
  • Ballscrew nut bracket: DSG12H

Total cost is looking to be ~$280 ex. shipping (which is ~$130).

@Julien, I’d like to talk about and link to the suppliers I went through to get to my final choice of parts and price, is that okay? I thought it might be interesting to compare them.

Yes, it’s ok to provide purchase links (as long as you don’t make any personal profit by doing so, I’m sure it’s not the case but prefer stating it anyway)

1 Like

Marvellous, I of course don’t get anything from this.

I went through a loooong process of finding a supplier. I looked at:

TBI Motion distributors:

  • TBI Motion directly, they told me to talk to their distributors.
  • Tuli, the TBI distributor I talked about earlier: They’re relatively expensive and have a 3 week lead time on ballscrew machining. They do however also sell some of the really nice linear rails if you want them. Service was knowledgeable and helpful, so I feel a bit bad about not ordering from them.
  • MEW, one of the distributors TBI recommended. They’re also quite expensive and like Tuli, take 3 weeks. Service was fairly quick and knowledgeable.
  • TOPPER, an AliExpress store which was recommended over on the CNCWerk forum. This is the one I’m certain to go with. Their service was super quick and responsive, they answered all my questions on AliExpress’s chat function, usually within seconds to minutes each time. At this point I already knew what I wanted so it was just a matter of asking them what they had. Like Tuli and MEW, they’ll take 3 weeks to get the product to me but they do it at a drastically lower price.

Chinese ballscrew sellers:

  • Dold Mechatronik: Reasonably priced, responsive and seems to ship fast. Their linear rails are okay (they’re CPC rails, just with the branding slightly obscured) but their ballscrews are rated with 0.05mm of backlash while TBI claims no axial play at all.
  • Kamp & Kötter, a German reseller of Chinese ballscrews. They are cheeeeeap and seem to have really fast processing because all their screws are pre-machined. I found the lack of specs worrying though so decided against them. Didn’t try talking to them.

And Hiwin: They apparently accept orders directly from private customers and their delivery was fast but their quoted prices were sky high. Had to bug them about the quote but the process was painless.

So on price, which I talked about a bit, I’ll compare the cost of the 231mm ballscrew (with machining but not shipping or tax):

  • Tuli: 159.93€ for TBI C7
  • MEW: Won’t quote the price directly but it was a decent chunk more expensive than Tuli
  • Topper: $77 for TBI C7, $91 for TBI C5
  • Dold: 59€ for C7
  • Kamp & Kötter: 67€ for C7
  • Hiwin: >300 CHF for C7

So I think TOPPER stands out here. TBI ballscrews seem fairly reputable and yet I can get them for a price more comparable to Chinese ballscrews.

Another thing I liked about TOPPER is that they stocked every part I needed. They have the screw, nut, rails, supports, coupling, motor mount, mounting block and the motor. The only other store that had everything was Dold.

3 Likes

The HDZ also has a surprisingly slim bridge to it, so long as you have the ballnut well attached it seems OK.

Here’s an approximately correct HDZ model in Fusion for comparison;

https://drive.google.com/file/d/16-EZjk4Jk4NpZqhM7MJeRXvCLIPw7Skt/view?usp=sharing

1 Like

Well I ordered all the parts finally. Here’s the final total:

I’ll finally have covers for the guide rails!

And I found a bit of a solution to the “bridge” problem: SYK has a lower-profile motor mount that gives me an extra 3mm of height back and I can also carve out a little indent in the mounting plate to get everything a little lower still.

And regarding the flat surface for the rails, I can get:

For my ~80 mm rail separation, assuming flatness can be roughly extrapolated, I’d be getting 20µm difference with the 7XXX series plates or 10µm with the 5083, so either is plenty good.

The 7XXX series plates are mechanically superior and make my life a bit easier regarding mounting (I can tap threads into them directly) so I’m thinking I’ll go with them, probably the 7021 specifically because it’s a bit easier to obtain (less insane shipping cost).

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.