Voron Cascade CNC

[kenny]

4 hours ago, PFarm said:

the pdf link dont work..

 

[PFarm]

2 hours ago, kenny said:

7 hours ago, PFarm said:

the pdf link dont work..

Try it now, since that was your first post, it needed approval.

[NikosSprocket]

You guys have built and using Milo's so you are very familiar with desktop CNC machines.

What is your opinion of the Cascade compared to the Milo?

I did a quick assessment between the Milo 2.1 and the current Cascade CAD files and my thoughts are as follows:

• Belt driving the Y screw. I think this will create some inaccurate cuts as the belt will “give” when forces are applied on the Y-axis or that's good enough for these machines' intended loads? I understand this was done for compact design, but I rather have the motor directly connected to the screw. Honestly, it would not take that much more to extend the screw through the 3060 extrusion and directly mount the motor from behind. As for the frame rigidity losses due to hole through the rear 3060, then can be re-enforced with sandwiched 3060s or other extrusions, which will support the Y-axis motor too. 
• The plexiglass case is nice, but a secondary will be required for stopping savings from flying off the top opening.
• The Milo Casa is huge, actually that's the main reason I cannot build one given the space limitations, should make cleaning easier compared to the Cascade. Also, liquid cooling and circulation will be interesting to see.

As you know I love my Voron printers and with today's offerings I will still build another 2.4r2 Voron; maybe with INDX toolhead soon ; but I think the Cascade needs more work.

[PFarm]

1 hour ago, NikosSprocket said:

You guys have built and using Milo's so you are very familiar with desktop CNC machines.

What is your opinion of the Cascade compared to the Milo?

I did a quick assessment between the Milo 2.1 and the current Cascade CAD files and my thoughts are as follows:

• Belt driving the Y screw. I think this will create some inaccurate cuts as the belt will “give” when forces are applied on the Y-axis or that's good enough for these machines' intended loads? I understand this was done for compact design, but I rather have the motor directly connected to the screw. Honestly, it would not take that much more to extend the screw through the 3060 extrusion and directly mount the motor from behind. As for the frame rigidity losses due to hole through the rear 3060, then can be re-enforced with sandwiched 3060s or other extrusions, which will support the Y-axis motor too. 
• The plexiglass case is nice, but a secondary will be required for stopping savings from flying off the top opening.
• The Milo Casa is huge, actually that's the main reason I cannot build one given the space limitations, should make cleaning easier compared to the Cascade. Also, liquid cooling and circulation will be interesting to see.

As you know I love my Voron printers and with today's offerings I will still build another 2.4r2 Voron; maybe with INDX toolhead soon ; but I think the Cascade needs more work.

• On Miley, a V2 version, the belt drive is on the X axis. Milo V1.5 has the Z axis driven by a belt and has no issues with accuracy. Inaccuracy with the Milo bench mills has more to do with leadscrews and backlash. The V2 Milo/Miley and V1.6 now have a direct drive stepper on the Z axis, and drilling with the direct drive has been affected by the loss of torque, so the belt drive does have it's adventage.
• On Casade, I do like that they're using a ballscrew, but it's going to struggle with spindle stalling with the 1204 ballscrew, which only has a 4mm lead. Milo uses a T8-8 leadscrew which has a 8mm lead. The lead is the amount of mm of travel one rotation of the stepper motor that the axis will travel. With high-speed spindles and less rigid frames, increasing the cutting speed is the way to compensate for that. A 1210 ballscrew would have been a better choice.
• The self-contained, smaller footprint will help with smaller home shops.
• I do like the toolsetter they chose, I have one, and they work very well. These do have an overtravel microswitch that can be wired to halt the operation.
• The two control board options are nice, Scylla and Expatria Flexihal, but if you're going to use the Flexihal board, a Pi5 is a better choice to reduce latency. With the Flexihal board, there is an additional cost of stepper drivers compared to the Scylla drive, which has onboard drivers.  Both excellent controllers.
• The CAD file doesn't show how the plumbing and associated hardware will be handled.
• Not sure how the 500W DC spindle is going to perform with the lower ballscrew lead.

I agree I think Casacde can be a good bench mill with the correct components installed.

[NikosSprocket]

1 hour ago, PFarm said:

• On Miley, a V2 version, the belt drive is on the X axis. Milo V1.5 has the Z axis driven by a belt and has no issues with accuracy. Inaccuracy with the Milo bench mills has more to do with leadscrews and backlash. The V2 Milo/Miley and V1.6 now have a direct drive stepper on the Z axis, and drilling with the direct drive has been affected by the loss of torque, so the belt drive does have it's adventage.
• On Casade, I do like that they're using a ballscrew, but it's going to struggle with spindle stalling with the 1204 ballscrew, which only has a 4mm lead. Milo uses a T8-8 leadscrew which has a 8mm lead. The lead is the amount of mm of travel one rotation of the stepper motor that the axis will travel. With high-speed spindles and less rigid frames, increasing the cutting speed is the way to compensate for that. A 1210 ballscrew would have been a better choice.
• The self-contained, smaller footprint will help with smaller home shops.
• I do like the toolsetter they chose, I have one, and they work very well. These do have an overtravel microswitch that can be wired to halt the operation.
• The two control board options are nice, Scylla and Expatria Flexihal, but if you're going to use the Flexihal board, a Pi5 is a better choice to reduce latency. With the Flexihal board, there is an additional cost of stepper drivers compared to the Scylla drive, which has onboard drivers.  Both excellent controllers.
• The CAD file doesn't show how the plumbing and associated hardware will be handled.
• Not sure how the 500W DC spindle is going to perform with the lower ballscrew lead.

I agree I think Casacde can be a good bench mill with the correct components installed.

Thank you for the great comments/info. Good to know about the belt driven screw and the electronics.

The plumbing in virtual and driven by AI of course