Rods "Spaceship" Scratch built Plasma Cutter build

Tommy, after your post yesterday, I did some research. Linear rails are rated for velocities of 3-5 m/sec and we are driving them way beyond that. Thats why those 600 balls make so much noise complaining about being overdriven on a plasma machine!

There was nothing really wrong with the Spacehip before I tackled this upgrade. It was built on some motors and drives I had on hand and a 48 volt switch mode power supply that seemed popular and then some NEMA34's so I could use cheap premade belt reductions. I just got the smallest 34's I could find so my stepper drives could still drive them.

We still want to harness that low down torque Steppers have plus retain their simplicity for DIY builds. With the predictable forces on plasma tables (in the absence of cutting forces), there is not much benefit to be derived from moving to servos provided the whole package is engineered correctly. Thats actually much harder for steppers than it is for servo machines.

I struggle to see why stepper manufacturers have so many products in their lineup and it was hard to make a product selection from our engineering model output. Hardly any motors matched our requirements. The Sanyo Denki's were a bit opportunistic as I got them cheap as chips and they are a perfect match to our model output. Time will tell but one would think replacing $20 motors with $200 ones and doubling the voltage should make a difference!

The servos Bevins uses are an 80ST. If you Google "80ST servo" listings will abound.

Linear rails are rated for velocities of 3-5 m/min

Are you sure about the unit?

CPC rates their linear systems for 10m/s and acceleration beyond 500m/s²

and here an excerpt from Hiwin:

Reinhard wrote:

Linear rails are rated for velocities of 3-5 m/min

Are you sure about the unit?

CPC rates their linear systems for 10m/s and acceleration beyond 500m/s²

and here an excerpt from Hiwin:

Reinhard, my mistake, it should be m/sec when I review the source I was quoting. My bad! See:

www.rollon.com/IN/en/news-item/378-under...ed-and-acceleration/

... we are driving them way beyond that. Thats why those 600 balls make so much noise complaining about being overdriven on a plasma machine!

The noise comes from the balls scouring at each other. On cheap linear cariages the balls push each other when moving. That scratching makes the sound ...

Better systems provide a spacer between the balls (image taken from CPC, Hiwin and others offer similar):

with such ballchain the friction is lower and lifetime is longer.

At work a carriage was broken and needed to be replaced. But cheese-pairing support collegues bought a carriage without ballchain - now the machine screams like hell.

cheers Reinhard

I had some troubles today getting my Lam drivers up and running. For some reason my old crappy Chinese motor on the Z axis refused to play with the LS1073a driver. I tried everything. The in desperation I swapped the drivers around and it worked perfectly on the larger DS1076a driver. Go figure.

Anyway, I still have to replace 3 motors in the coming days but on a quick test, it looks like we'll get away with at least 30 m/min (1200 in/min) and 5 m/sec/sec acceleration (200" / sec/sec) Thats 0.5 G's.

So this is what those figures look like:



And for the record, a good solid table like the spaceship, handles those sort of numbers in its stride. No stripped belts and no screaming balls!

A small-ish issue there, if you were jogging the machine on that video, set the Display and Traj section max_velocity at 25 to 50% more than each individual axis in the ini, that way it will not limit the X and Y axis speeds while jogging both of them at the same time.