electronic lead screw

[GreatOldOne]

72 & 24 tooth HTD M3 x 15mm, so the same 3:1 ratio as James. Seemed to be the easiest way forward as he’s proven that it works with the hybrid stepper / servo

 

[GreatOldOne]

Busy day in the shop. Both HTD pullies have been bored to the correct size for their respective shafts, and the larger of the two has had a 4mm key way broached in it. This was my first time broaching!

Checking the fit with a 16mm endmill, prior to broaching. Nice close sliding fit

of course my arbor press didnt have enough daylight between the ram and the broach. I had to do a double boost and add a lower floor.

Both pulleys still need to be drilled and tapped for set screws, but that’s for another day.

Ive laid out some art work mirroring what James has engraved into his enclosure. I’ve 3d printed my enclosure lid, so needed some way of putting the text on there. Heavy duty laser labels do the job nicely.

I laid the label over the whole lid with the intention of neatly cutting out all of the apertures... but it’s a real pain to get a nice neat edge and the aluminium foil labels dull craft knives real quick. So I just used the majority of the label for alignment, and then removed what wasn’t needed. Not as elegent as a CNC engraved lid, but functional.

 

[GreatOldOne]

My OCD got the better of me, and I re-did the decal. I think it looks better now, and it sort of looks like an old C60 cassette. Awesome Mix ELS Edition?

I've put the decal file on my drop box, just in case anyone else wants to use it. Link:

ELS Lid Decals.svg

Shared with Dropbox

www.dropbox.com

Print it out, use a light behind it to line it up with the lid if you've printed one - and one you've got it stuck down use a craft / xacto knife to cut around the dotted lines and the perimeter.

 

[Qtron]

My shop looks tidy only because of careful framing. It's a wreck, and I barely have room to turn around, which is why I planned to build a detached shop this year about three times the size, but it's looking more like a spring thing all the time.

I'm sticking with the "programmable gearbox" concept for the foreseeable future. No CNC for me, thank you. I still might look at a micro with quadrature decoding at some point. I have no need for floating point the way I'm doing things. I have interpolation error but no accumulation of rounding error, and I can stop at a limit on the same count all day long, better than I can measure with a tenth indicator.

I gave very serious consideration to the Clearpath servos, but cheaped out. The "hiccups" I'm seeing are momentary stalls on heavy cuts that aren't quite big enough to cause my microstepper driver to error out. I think I can fix it by increasing the gear ratio between the motor and lead screw. I'm currently running 4:1, but have ordered timing pulleys and belts that will allow me to test 5-, 6- and 8-to-one. If that's still not enough, I might start thinking about Clearpath again. They have one that ought to be a drop-in, and would double the torque. I just balk at the price tag.

I've been reasonably happy with my TouchDRO setups. I'm using Shars scales with added decoupling caps in the heads to make them more stable. I have the cross feed scale on the lathe mounted above the screw, covered and completely out of the way, and I would have a hell of a time doing that with any other kind of scale. About the only possibility would be a DroPros magnetic scale. I wish Yuriy would get fired up and fix a few things, but he seems to have moved on.

Take care.

just saw your Youtube torque test Jon, Wow , the forces involved on a leadscrew = vicious!! no wonder 1/2 nut backlash eventuates
Jon Bryan
8 hours ago, youtube.
"Down the rabbit hole I go. I'm trying to score a force gauge and actually measure it. Rough T = KDP calculation gives me a number between 550 and 1100 pounds of force applied to the carriage depending on the friction loss. I can tell you that even before I geared it down I could push on it my hardest without slowing it down, and I weigh about 270, so I'm inclined to believe the rough calculation. Gives you an appreciation for the kind of cutting forces that are involved when cutting an 8tpi thread."
So,
Not being a mech engineer or F & T, i need clarification here: T = KDP?? (so something related to diametral pitch?)..
(wikipedia = Kurdistan Democratic Party, lol.) its ok, i'll dig for it..
Why not simply do lots of light cuts, (- nicer job? )
to ease up the forces on the 1/2 nuts, leadscrew, bearings etc? After all U don't cut an 8 TPI thread very oft, do you?
Well i don't anyway, & if needed, i wouldn't be in too great a hurry.

- i Just got an answer from Clough42 re his ELS, man is he gunna be busy, new breakout boards for his TI piccolo sold out in hrs!!
I don't quite get the rush, after all he said the board isn't crucial to the project, but nice to have pretty lights to indicate activity, i suppose.
James advised me re your latest Vid., Thanks to him, & certainly, thanks to you - the torque wrench was a profoundly simple way to indicate static stepper torque or holding force. [some drives can be set to 50% idle current, or standstill current ('Leadshine' terminology!) reducing heat build up in the stepper windings].
Q: U seem to be operating a closed loop stepper, it cant be a system reaction to the spindle encoder movement, as there's no direct link from the spindle cog & fwd/rev idlers to the gearbox input anymore, being the point of the ELS!
i mean, i get the holding control of the stepper drive, resisting forced steps, but after 1 step overcome mechanically, i would have thought it would lock into that new step, not go back as depicted, which seemed to be a heap of steps. (unless its a closed loop). Puzzled.
Also, about those idlers & the encoder drive train, wouldn't there be too much backlash / slop between spindle motion & the encoder rotation, given they are only spur gears? I get the need to sense fwd/rev. but that can be done direct off the spindle, with virtually no missed pulse edges.
Maybe i missed something there, too

I do hope U find that suitable force gauge..
cheers, Qtron.
P.S. A blank Mega awaits, Tongue hanging out for that magic code now Sir Jon!!

 

[kb58]

...the first batch of 100 sold out in about 4 hours...

Two hours, and I was the first to lose out. Went to Ebay and it said he had two left. Ordered two and it said, not enough in stock. Ordered one, hit Buy, and "sorry, all out." I think I'm one of the few who actually has the entire thing running; I have the discrete boards but I just like how his board cleans up the wiring. I'm pushing forward on getting it entirely finished such that his board is the only holdup.

To be honest, I'm not quite sure why I decided to put the assembly on the lathe itself instead of the wall. I guess I just like knowing that if I have to move the lathe, there aren't any "appendages" other than the power cord. And yes, there will be a cover over the assembly, because with 0.020 pitch on a couple of the ICs on the development board, it wouldn't take much for a chip to short out things...

Separately, I have no idea if the little belt I'm using will be sufficient for driving the carriage, because there's probably a dozen variables in figuring out the torque. Speed of cut, depth of cut, cutter type, cutter sharpness, material being cut, spindle speed, gear reduction of lead screw, gear reduction of QCGB, gear reduction of belt pullies, torque capacity of the servo motor, and probably more I'm forgetting.

 

[GreatOldOne]

Two hours

Insane. I must have been very fortunate to get one of the first batch then.

 

[RJSakowski]

Two hours, and I was the first to lose out. Went to Ebay and it said he had two left. Ordered two and it said, not enough in stock. Ordered one, hit Buy, and "sorry, all out." I think I'm one of the few who actually has the entire thing running; I have the discrete boards but I just like how his board cleans up the wiring. I'm pushing forward on getting it entirely finished such that his board is the only holdup.

To be honest, I'm not quite sure why I decided to put the assembly on the lathe itself instead of the wall. I guess I just like knowing that if I have to move the lathe, there aren't any "appendages" other than the power cord. And yes, there will be a cover over the assembly, because with 0.020 pitch on a couple of the ICs on the development board, it wouldn't take much for a chip to short out things...

Separately, I have no idea if the little belt I'm using will be sufficient for driving the carriage, because there's probably a dozen variables in figuring out the torque. Speed of cut, depth of cut, cutter type, cutter sharpness, material being cut, spindle speed, gear reduction of lead screw, gear reduction of QCGB, gear reduction of belt pullies, torque capacity of the servo motor, and probably more I'm forgetting.

I ordered a board yesterday and had a message from Paypal this morning to say it's been shipped.

 

[ttabbal]

I ordered a board yesterday and had a message from Paypal this morning to say it's been shipped.

I did as well. I think he put a few more up at some point. I just checked here and there and saw 2 available, so I went for it.

It's not too difficult to build one from the parts he lists. I just wanted to be able to skip that.

His lathe looks similar in size to my PM1127, so I'm going to try 3:1. My lathe gearbox also has a selectable 2:1 I can add in if needed.

 

[RJSakowski]

I've been following James for some time now. Originally, I was planning on waiting until all the dust settled before diving in but given the popularity of this project, I decided that I had best get going on it. Over the weekend, I ordered all the electronics save the stepper/servo driver. Everything except the I/O interface has been shipped.

I am still waiting to see the final verdict on the choice of motors and pulley size. I have found several candidates for motors, a conventional stepper, and a hybrid stepper/ servo. I expect that the maximum lead screw RPM will be in the neighborhood of 600. At 600 RPM, that would be moving the carriage at 50 ipm which is crazy fast. If I were cutting a 4 tpi thread for some unknown reason, threading at the 150 RPM spindle speed would mean the lead screw would be turning at 450 RPM The torque/speed curves for steppers vary quite a bit from motor to motor but it looks reasonable to expect a 50% drop in torque at 600 RPM.

I modified the opening at the bottom of the lathe bed to accommodate either a NEMA 23 or NEMA 24 frame motor.

 

[Briney Eye]

just saw your Youtube torque test Jon, Wow , the forces involved on a leadscrew = vicious!! no wonder 1/2 nut backlash eventuates
Jon Bryan
8 hours ago, youtube.
"Down the rabbit hole I go. I'm trying to score a force gauge and actually measure it. Rough T = KDP calculation gives me a number between 550 and 1100 pounds of force applied to the carriage depending on the friction loss. I can tell you that even before I geared it down I could push on it my hardest without slowing it down, and I weigh about 270, so I'm inclined to believe the rough calculation. Gives you an appreciation for the kind of cutting forces that are involved when cutting an 8tpi thread."
So,
Not being a mech engineer or F & T, i need clarification here: T = KDP?? (so something related to diametral pitch?)..
(wikipedia = Kurdistan Democratic Party, lol.) its ok, i'll dig for it..
Why not simply do lots of light cuts, (- nicer job? )
to ease up the forces on the 1/2 nuts, leadscrew, bearings etc? After all U don't cut an 8 TPI thread very oft, do you?
Well i don't anyway, & if needed, i wouldn't be in too great a hurry.

- i Just got an answer from Clough42 re his ELS, man is he gunna be busy, new breakout boards for his TI piccolo sold out in hrs!!
I don't quite get the rush, after all he said the board isn't crucial to the project, but nice to have pretty lights to indicate activity, i suppose.
James advised me re your latest Vid., Thanks to him, & certainly, thanks to you - the torque wrench was a profoundly simple way to indicate static stepper torque or holding force. [some drives can be set to 50% idle current, or standstill current ('Leadshine' terminology!) reducing heat build up in the stepper windings].
Q: U seem to be operating a closed loop stepper, it cant be a system reaction to the spindle encoder movement, as there's no direct link from the spindle cog & fwd/rev idlers to the gearbox input anymore, being the point of the ELS!
i mean, i get the holding control of the stepper drive, resisting forced steps, but after 1 step overcome mechanically, i would have thought it would lock into that new step, not go back as depicted, which seemed to be a heap of steps. (unless its a closed loop). Puzzled.
Also, about those idlers & the encoder drive train, wouldn't there be too much backlash / slop between spindle motion & the encoder rotation, given they are only spur gears? I get the need to sense fwd/rev. but that can be done direct off the spindle, with virtually no missed pulse edges.
Maybe i missed something there, too

I do hope U find that suitable force gauge..
cheers, Qtron.
P.S. A blank Mega awaits, Tongue hanging out for that magic code now Sir Jon!!

The equation appears in slightly different forms. Of course it's a simplification. The terms are:

T = Torque (in lb)
K = Coefficient of Friction (dimensionless, ranges typically between 0.1 - 0.2)
D = Major diameter of the screw (inches)
P = Force (lbs)

If you rearrange the terms to solve for Force, input 88.5 in-lbs of torque (10Nm), and assume the friction coefficient for a 3/4" screw is the worst-case 0.2, you get:
P = 88.5 / (0.2 * 0.75) = 590 pounds of force applied by the screw.

Caveat: I'm not an engineer, I just spent a lot of years working for them.

Yes, it's a closed-loop microstepper. In the absence of steps it tries to stay where it is. What you're describing is precisely the behavior of an open-loop stepper.

I drive the encoder from the stud gear for simple expedience. It was the easiest and closest. As long as things are in motion backlash can be ignored. When you change direction the backlash is what it is. But that's the normal state of affairs anyway. CNC machines go to an enormous amount of trouble to minimize the backlash because they're changing direction all the time and the less they have to compensate for it the better.