IF you have the choice then run both and tune them seperately as each does do a specific job. It also may be that I remember it wrong that was a long  time ago.

(;-) TP

I asked the question on the CandCNC forum, I might get an answer one day...

Thinking about designing a better plasma z axis...


Anybody used anything like this for a z axis leadscrew?

High helix leadscrew, 10mm dia, 12mm lead, with a delrin nut (22mm of)

http://www.igus.co.uk/wpck/7001/DryLin_Steilgewindemuttern

Was looking at 1610 or 1616 ballscrew as a faster response zaxis for a plasma as well.

Why the 10mm leadscrew you may ask, smaller, lighter, easier to couple, no need to have ends machined for bearings or couplings, especially when I would be only after a short length (150mm)

Open to comment, thanks

Edit, drive by stepper, with programmable digital drive, hence choice of microstep resolution.

Thinking about designing a better plasma z axis...


Anybody used anything like this for a z axis leadscrew?

 
Open to comment, thanks

Edit, drive by stepper, with programmable digital drive, hence choice of microstep resolution.

Hi Robert,
  I picked some of these up some time ago (when they were cheap  ) to use in a p&p gantry robot.

Teflon coated steel screw, Delrin ? Acetal ? zero backlash nut (self compensating).
KERK ... I think.
3/8" +/- OD x 1/5" +/- lead. VERY slick.

I will contribute to the discussion by donating one of these to you If you'd like.
Guess it depends where in the world you are to be practical as far as shipping goes.

Kind regards,
Russ

Thanks Russ for the offer, much appreciated.

1/5" lead is about a 5mm pitch, and 5/8 is about 16mm dia... So unless I got it wrong they are approximately eqivelent to a 1605 ballscrew although using delrin/acetal etc nut.

I'm presently running a 2mm pitch 10mm trapezoidal leadscrew.

I'm in the UK and may take you up on the offer if it's still available after a bit more advice on leadscrew + plasma + z axis.

Oooops, my mistake.
I crossed 1/2 " with .5"

They are about one half inch lead.

  ..... and 3/8" diameter, not 5/8".

Russ
 

For a direct drive stepper you need between  2-5 tpi on the feed screw.  AND always keep this in mind That screw will be in the NASTIEST, abrasive place you can ever place a screw in period, bar none.  Either count on replacing it often or use bellows to protect the screw and nut.

Play on the screw will drive your THC nuts and give terrible cuts.

You only need about 6" of travel max.

(;-) TP

Just got a reply from Denis (MiniTHC guy) regarding the frequency of the THC, not had time to really study yet but looks like he is talking about 20KHz if my quick calcs are right.
Anyway here it is.



Sorry for delayed reply.
I had to talk to the engineers to ask your question.

Internal feedback loop is about 0.05-0.06 milliseconds (0.00005 s).
(Oscillator clock is 8MHz, and ADC accumulates about 400 values in row, to apply the filter. So 1/(8E6 Hz) * 400 = 0.05 ms)

If you connecting THC to computer, response time (with LPT) will be anyway greater, due to signal need to be transferred to computer and back to motor, which is 1/25000 * 2 = 0.08 milliseconds.

25 kHz seems to be enough to respond even for curve sheets.
For example, with max possible cutting speed (10 m/min), machine travels 170 mm/sec.
That mean, for every 1 mm, Mach3 is able to receive 75 signals up/down


Hood

Have they ever said if they use a spanvolts (deadband ) or PID control for the thc ?  The numbers look very good to me . BUT I am not sure that Mach3 can keep up that fast but it is better to be quicker than slower in this case.

Thanks for the Numbers Hood, Those are VERY hard to come by (;-)

(;-) TP

No TP they never said but then again I never asked.
I presume that is what the 400 samples bit is, ie they take 400 samples before they issue an up/down command. How they sort these samples I have no idea.

There was another part to the email that talked about PID but that was in relation to driving the motor directly from the THC, I presume without a servo drive. I had mentioned I was thinking of driving the motor independent of Mach and I presume they were thinking DC motor.

Anyway here is that part.

As far as I can tell, there is two ways to implement direct motor control:
H-bridge and PID controller (like the servo motor drivers).
I can't say if you really want to implement your own PID, but we tried both ways, and I will be happy to share its pros and cons.


If I do decide to drive the motor direct from the THC it will be via the Preset Inputs on the servo drive so that doesnt really matter to me.

Hood