electronic lead screw

[RJSakowski]

No cus

Wow u must have a lot of customers for $14 k spend. Slick machine!
BTW, after relisten to vid 2 I suddenly realised he has no accel/decel built in. Not good.

No customers. Strictly hobby. It was a self administered present for my immanent retirement. Actually, I had a kitchen lighting project that could only be done with a CNC mill. I could have farmed it out at a cost of several thousand $ but what fun is that?

As to the lack of accel/decel, I wouldn't be concerned about it. The stepper is always tracking the spindle and the spindle's natural momentum will limit acceleration and deceleration.

 

[Qtron]

No cus

No customers. Strictly hobby. It was a self administered present for my immanent retirement. Actually, I had a kitchen lighting project that could only be done with a CNC mill. I could have farmed it out at a cost of several thousand $ but what fun is that?

As to the lack of accel/decel, I wouldn't be concerned about it. The stepper is always tracking the spindle and the spindle's natural momentum will limit acceleration and deceleration.

Agree re fun too.
Good point re accel. I was thinking purely of the mass of my LS considering its length. But yes spindle dictates.
I suppose you will be watching for that potential flaw / setup sensitivity KB58 described.
BTW, u seem very studious! Thanks for info re 50 oz in friction of the 770, a good observation!

 

[RJSakowski]

I completed the ELS build this afternoon. The last task was to make the faceplate for the display module. I used a piece of .063" sheet aluminum for the panel. I used my spindle mounted microscope in the Tormach CNC is an optical comparator to scan the display pcb to determine the position of the components and laid out the panel in SolidWorks. Since I had no way to make the flexible keypad, I opted for simple round buttons. For a bezel for the seven segment display, I used a piece of 1/8" red acrylic and I milled a shoulder around the perimeter so the bezel would be flush with the front of the panel but still retained by the seven segment display. This required .330" standoffs for the display pcb mount. For the push button switches, I made Delrin top hat button extensions,

I replaced the original LED's with new so I could take advantage of the longer leads to bring the LED's through the panel. The panel lettering was set up in SolidWorks. In engraving fonts, it is desireable to use a single line font. Unfortunately, Truetype fons are outline fonts. I chose the SWGDT font as the spacing is close enough that for all intents and purposes, it is a single line. For the switch lables, I used a 6 pt. font to avoid crowding due to the close spacing of the switches. For the LED's, the spacing was more relaxed so I used a 9 pt. font there. The G code was done in SprutCam. Engraving was .008" deep, using a 30º engraving tool. I chose .187" for the push button diameter so I could use my 3/16" 5C collet in the collet chuck.

To make the seven buttons, I used 1/4" black Delrin rod with the first op done with the three jaw. I
used my carbide parting tool for all machining. To machine the button, I set up my carriage stop plus an 1/8" spacer. With this setting, I faced the top of the button. Then I pulled the spacer and cut the shoulder. Then I moved the carriage to slightly past the facing position and move the stock so that I could part a slightly over length part. The spacer was reinserted and the second part was made. Once all the buttons were made, I switched to the 5C collet chuck with a 3/16" collet. The part was inserted in the collet and the carriage set up to cut a .152" length. When using a collet in this manner, especially on a material like Delrin, it is important to use consistent torque on the chuck key as the more torque you use, the deeper the collet is seated and the longer the part will be. I tightened the key until I could just feel some resistance.

The lettering was filled in with Rustoleum flat black spray on a plastic sheet and applied with a swab. The excess was sanded off with 360 grit wet and the faceplate was coated with automotive clear coat.

The enclosure is a phenolic box acquired many years ago. It is mounted on a second box made from a piece of 3" aluminum channel and a cover plate. This box houses the two power switches, one for the driver and the other for the controller, plus a fuse. That box is mounted on a repurposed residential mains fuse box which contains the 42 volt power supply for the driver, a 5 volt power supply for the Clough42 board and Launchpad. and the stepper driver.

ELS Faceplate back

ELS Faceplate front

ELS Control and Power enclosures, mounted

ELS mechanical

 

[RJSakowski]

Agree re fun too.
Good point re accel. I was thinking purely of the mass of my LS considering its length. But yes spindle dictates.
I suppose you will be watching for that potential flaw / setup sensitivity KB58 described.
BTW, u seem very studious! Thanks for info re 50 oz in friction of the 770, a good observation!

Thanks for the compliment.

I already found that the stepper doesn't like 5mm/rev @ 588 rpm. It will handle 2.5mm/rev and possibly higher. 5mm/rev is 130 ipm at that spindle rpm. However, so far no other issues. Lots more testing to do.

The friction torque was measured as a better way to set preload on the angular contact thrust bearings. Tormach's recommendation was to loosen the gib to remove any friction associated with it and then loosen the thrust bearing. In theory, the remaining friction is due to the ball nut. Then slowly increase the bearing preload as you rotate the ball screw, stopping as you start to feel some resistance. That is where I devised the torque meter to measure the increase. My initial torque was 21 oz.-in. and I tightened the preload to 37 oz. in. Then I tightened the gib to the final 50 oz.-in. I did a thread here. https://www.hobby-machinist.com/threads/measuring-ball-and-lead-screw-torque.37492/ John Saunders, NYC CNC, picked up on it and did a You Tube video on it as well.

 

[kb58]

Clough42 recommends against bundling the encoder and servo cables, for EMI reasons. I bundled mine and had it hiccup (change settings) on me, so he may be right.

 

[Qtron]

Clough42 recommends against bundling the encoder and servo cables, for EMI reasons. I bundled mine and had it hiccup (change settings) on me, so he may be right.

Encoder input on the piccolo would be impacted by EMR from the stepper/ servo but I fail to see how that would cause it to jump from feed to thread mode or vv.
But the MCU should be in a metal box with earthing strategy well thought thru.!

 

[GreatOldOne]

I fail to see how that would cause it to jump from feed to thread mode or vv.

At a guess, it’s probably down to the coms to and from the keypad to the launchpad. It’s serialish IIRC... noise on the data line could be misinterpreted by the launchpad, and it does the switching thinking it’s a legitimate command from the pad - sending the change of state back to the pad.

 

[Qtron]

At a guess, it’s probably down to the coms to and from the keypad to the launchpad. It’s serialish IIRC... noise on the data line could be misinterpreted by the launchpad, and it does the switching thinking it’s a legitimate command from the pad - sending the change of state back to the pad.

Ur right, of course, was regretting that statement overnight, lol.
I guess its best to have short leads from piccolo to the ''led&key''. (Maybe place both in the same metaI box?).
I read somewhere that in some situations a ferrite bead on the earth wire/lead to a sensitive device on the end of a chain provides better noise immunity..
But maybe not relevant if cables/wires are short.
I do think a ferrite bead on the encoder cable right at the piccolo would be beneficial.
(Think of better type of USB Leads for PC's etc.)

 

[Qtron]

Thanks for the compliment.

I already found that the stepper doesn't like 5mm/rev @ 588 rpm. It will handle 2.5mm/rev and possibly higher. 5mm/rev is 130 ipm at that spindle rpm. However, so far no other issues. Lots more testing to do.

The friction torque was measured as a better way to set preload on the angular contact thrust bearings. Tormach's recommendation was to loosen the gib to remove any friction associated with it and then loosen the thrust bearing. In theory, the remaining friction is due to the ball nut. Then slowly increase the bearing preload as you rotate the ball screw, stopping as you start to feel some resistance. That is where I devised the torque meter to measure the increase. My initial torque was 21 oz.-in. and I tightened the preload to 37 oz. in. Then I tightened the gib to the final 50 oz.-in. I did a thread here. https://www.hobby-machinist.com/threads/measuring-ball-and-lead-screw-torque.37492/ John Saunders, NYC CNC, picked up on it and did a You Tube video on it as well.

Will check out vid. Very good, lots of fiddle for U.
Try the damper plate technique on your stepper.
May raise the revs. Whats ur microstepping rate set to? (U may have already stated).

 

[Qtron]

I completed the ELS build this afternoon. The last task was to make the faceplate for the display module. I used a piece of .063" sheet aluminum for the panel. I used my spindle mounted microscope in the Tormach CNC is an optical comparator to scan the display pcb to determine the position of the components and laid out the panel in SolidWorks. Since I had no way to make the flexible keypad, I opted for simple round buttons. For a bezel for the seven segment display, I used a piece of 1/8" red acrylic and I milled a shoulder around the perimeter so the bezel would be flush with the front of the panel but still retained by the seven segment display. This required .330" standoffs for the display pcb mount. For the push button switches, I made Delrin top hat button extensions,
View attachment 308143
I replaced the original LED's with new so I could take advantage of the longer leads to bring the LED's through the panel. The panel lettering was set up in SolidWorks. In engraving fonts, it is desireable to use a single line font. Unfortunately, Truetype fons are outline fonts. I chose the SWGDT font as the spacing is close enough that for all intents and purposes, it is a single line. For the switch lables, I used a 6 pt. font to avoid crowding due to the close spacing of the switches. For the LED's, the spacing was more relaxed so I used a 9 pt. font there. The G code was done in SprutCam. Engraving was .008" deep, using a 30º engraving tool. I chose .187" for the push button diameter so I could use my 3/16" 5C collet in the collet chuck.

To make the seven buttons, I used 1/4" black Delrin rod with the first op done with the three jaw. I
used my carbide parting tool for all machining. To machine the button, I set up my carriage stop plus an 1/8" spacer. With this setting, I faced the top of the button. Then I pulled the spacer and cut the shoulder. Then I moved the carriage to slightly past the facing position and move the stock so that I could part a slightly over length part. The spacer was reinserted and the second part was made. Once all the buttons were made, I switched to the 5C collet chuck with a 3/16" collet. The part was inserted in the collet and the carriage set up to cut a .152" length. When using a collet in this manner, especially on a material like Delrin, it is important to use consistent torque on the chuck key as the more torque you use, the deeper the collet is seated and the longer the part will be. I tightened the key until I could just feel some resistance.

The lettering was filled in with Rustoleum flat black spray on a plastic sheet and applied with a swab. The excess was sanded off with 360 grit wet and the faceplate was coated with automotive clear coat.

The enclosure is a phenolic box acquired many years ago. It is mounted on a second box made from a piece of 3" aluminum channel and a cover plate. This box houses the two power switches, one for the driver and the other for the controller, plus a fuse. That box is mounted on a repurposed residential mains fuse box which contains the 42 volt power supply for the driver, a 5 volt power supply for the Clough42 board and Launchpad. and the stepper driver.
View attachment 308149
ELS Faceplate back

View attachment 308150
ELS Faceplate front

View attachment 308148
ELS Control and Power enclosures, mounted

View attachment 308151
ELS mechanical

Very nice RJ.
Any chance of dxf files for the front panel?