What's new
What's new

SB16 electric lead screw options. Need a hand taking measurements

Rockmonton

Plastic
Joined
Dec 2, 2020
SB16 power lead screw ideas - Album on Imgur

Looking to fit a brushed DC treadmill motor for am electric leadscrew, but unfortunately have the lead screw and carriage ripped off already.
Just thinking at this point, nowhere near committing to it.

Can someone tram their carriage all the way to the headstock and grab a measurement and picture for me? It LOOKS like there's enough room with a bit of trimming on the motor housing, I've already found room for encoders on the headstock and gearbox (surplus 4096 SSI from work that are no longer used).

Not planning to go CNC, but would like to be able to cut metric, and have the benefit of super slow feed rates.

Anyways, just looking at options. I plan to thread/key my replacement shafts for the gearbox to take encoders and drive from treadmill motor.

Anyways, if anyone can give me a hand it would be appreciated. Working in -30 can't keep the frost off in the shop.. this will be slow.
 
I'm afraid I can't help you with the measurements, but it may be worth reconsidering the treadmill motor if it's one of the usual 3500rpm suspects. It'll be hard to step down the speed far enough for an ELS (50-250rpm? just a guess) and at low speeds they don't have much torque.

Something like a 24V 250W DC motor with a belt step down or even a DC gear motor of appropriate size/ speed might be a better fit. Plus the speed control options for those low volt/power DC motors are alot cheaper.
 
It's a 5650rpm 2hp treadmill duty, so even worse on the torque curve.

I was planning on running the secondary shaft with the 14/28 ratio on the RH tumbler with the LH disengaged....

In previous experience as long as you stay above 250rpm or so the motor will spit out torque all day long.

Since I have a vfd (powerflex 525) and will have encoders on both ends I can vary the speed on the lathe roughly 400-800rpm (have not put a tach on yet). It looks like I'd need to gear roughly 16:1 plus the secondary in the 2:1 position. That would give revs at 8tpi of 4800rpm and 64tpi at 600tpi. Add in the vfd control down to 400rpm that gives the 4tpi range.

PID control between lead screw and VFD is a trivial task with some hefty drivers to the motor, now just to figure out where to mount it. Looking like under the QCGB may be the best option.

Thanks!
 
Just a thought, Most wheelchair motors are 24V, right/Left angle drive. They also have a lever to disconnect the motor so the chair can be easily pushed. Don't know what the RPM would be, but it should be close. Torque should not be a problem. Mounting should be fairly easy as well.
You might want to check eBay.
 
My calculations were wrong. A 5:1 reduction works nicely and by varying spindle speed from 400-800rpm it's possible to keep the motor within the power peak (roughly half the motor speed) easily.

128tpi would require ~300rpm on the treadmill motor at 800rpm spindle speed. Anything slower would still be done by the gearbox.

Give most of what I thread is 10tpi-64tpi and 0.5mm-2.0mm, that gives me motor speeds at the 10:1 total reduction of 4800rpm at 10tpi and 750 at 64tpi. Well within the operating parameters.

A 3/4" htd 5mm belt should be sufficient to turn this without breaking. That leaves an OD on the secondary input pulley of approx 3.75" which will fit nicely on the secondary shaft.

Unfortunately I have ZERO experience with timing belts so I have to make sure it can handle the speed and torque

The reason I went against wheelchair motors is that they top out approx 150rpm, usually closer to 100. To thread 10tpi at 800rpm spindle speed that would require a leadscrew speed of approx 480rpm (800rpm spindle x 6 tpi leadscrew /10tpi [someone check my math please]). Slowing the spindle speed down to 400rpm gets things in the range of a wheelchair motor, but that's pretty slow to cut coarse pitches nicely on this lathe.
 
Turns out the calculations are easy. Htp5 is WELL within the 2.6n-m of torque the motor can put out. With a 3/8" bore there is negligible shaft deflection (~0.03 degrees) and I'll have to hog it out to 11/16 anyways on the input side so there's no worry about that.
 
You want really accurate index pulses.
An optical encoder is the best choice.
To get good steady cuts at the right point in time and rev/turn/cut.

You want realy fast pullout at the end of the thread.
And it must be really repeatable.
To 1 micron repeatable as a goal, or 10x better on electronics, under speed, with mechanical retraction issues, and bend and flex.
This is hard to do.

You want really heavy torque and fast response at the spindle end.
This makes the thread steady and not wander.


My advice is to aim for vastly more spindle torque than You think you need.
And aim for vastly more resolution and stiffness than You think You need.
Rigidity and repeatability is everything.

It´s hard to do a ELS.
Or a good CNC lathe.
 
The encoders I have are 4096 optical SSI with a third output as an Electronic adjustable index pulse (embedded pic microcontroller for those curious that also outputs a 360ppr conversion). Piece of kit I had designed for work that has been replaced with A newer style. Plenty of encoder. I was looking to put the encoder on the second gear from the spindle to keep the thru hole on the spindle.
Tightening up calculations on shaft torque flex, etc, gets me down to 0.004 degrees of angular deflection on the motor shaft, everything else is beefy enough not to worry.

Trying to figure out how to compensate for gear lash on both the last of the RH tumbler and the leadscrew, I’m thinking it’s trivial if moving in a consistent direction, but would still need a thread dial or electronic limit switch to get back to a starting position and keep the lash accounted for in the same direction. Once I get my broken casting brazed back together And bushed I’ll fit the new secondary shaft for the timing pulleys and encoder.

Total cost is about $260 for pulleys, estimated belt length, electronics, and control. Plus probably 40 hours of design and another 40 of fabrication and making mistakes, followed by 20 hours of tuning and turning scrap into chips. Mounting bracket looks easy, but figuring out forces, belt tension, and stiffness will take some effort.
 
I started watching these videos. I was impressed. But I am impressionable. ;)

He is thorough that is for sure. I did watch the whole series as I got hooked and wanted to see the project completed. He did a great job detailing the entire process.


Sent from my iPhone using Tapatalk
 








 
Back
Top