Advanced Search

SCALEBELTS.COMP
Calculate spindle speed based on belt positoin for a machine with 2 belts and 4 positions
Use with a matching QTdragon custom screen which was included in the qdragon release
linuxcnc.org/docs/stable/html/gui/qtdrag...ed_custom_vcp_panels

STEPBOOST.COMP
For Lam Technologies Stepper drivers with Boost function
Enabling the stepper driver boost pin when at constant velocity reduces the stepper amperage allowing the motor to cool when not working hard.
This allows masssive improvement on performance from the stepper motor which only works when accellerationg at start and end of motion segment. Set the boost reduction to be say 30% of the default amps.
Certified tested safe for production
 

i seen his video it's a pretty nice build but it is a manual self-centering if i remember right. the idea is to have the overhang so you can have both the flat table for larger stuff and the rotary off to one side.
what would be different about what i want to do is have the automated clamping and a second liner axis as a bar feeder and drive the rotary from a second in the rear in line with the fixed end. the fixed end is basically a steady rest with automated clamping system to hold the place while the other can be released to move back and grab another bite on the material and keep feeding material if you want to cut something long,

you basically have infinite travel on a small footprint machine.

How many amps are the motors, and what voltage?  Got a pic of the motor plates?

Because if it was orignally a 400v machine, you'll need a transformer and high(er) voltage drives than the typical 220/240vac 1p/3p offerings unless you want the motors to run at reduced max speed.

There are a number of 'agnostic' servo drives out there that have resolver inputs.  Copley and Advance Motor Controls (AMC) come to mind, and are generally available on the used market.

Another option is to replace the resolvers with encoders.

It's not difficult for me to lay out one of these configurations.
I tried to simplify it to make it easier to model and understand the essence of the problem.
I hope this configuration helps.

PCW wrote:

Do you have more than one encoder enabled in the hal file?
 

That was the problem.
I simply changed the count and now I can see the new encoder.

I think I'm away to the races.

Big Thanks

Do you have more than one encoder enabled in the hal file?

I think converting this to a table addon, I would extend th x axis past the table edge and suspend the chuck below it and just have a lazy unpowered chuck at the end of the table aligned to the roller chuck. It would be the roller chuck moving then.

PCW wrote:

The first P1 encoder is  encoder 1 so the position would be:

hm2_7i96s.0.encoder.01.position

and the most important parameter for that encoder would be:

hm2_7i96s.0.encoder.01.scale

 

Update, I received and populated my isolating board that plugs into the P1 connector of the 7i96s, I connected the Z scale and when I move Z, I can see the A & B quadrature data changing in PI-1 (io34), and P1-2 (io35), so looks like the data is making it to the FPGA.

However I can't see anything P1 related in HAL, and when I try to connect net "hm2_7i96s.0.encoder.01.position" to a net I get
postgui.hal:21: Pin 'hm2_7i96s.0.encoder.01.position' does not exist

So I'm doing something wrong.

Help please.
 

Travis MItchell has an excellent series on buidling a rotary tube cutter for a laser head for circular, square and rectangular tube.


You should review his chuck design befor going to far into this. The material is held between 2 seperate self centring roller assemblies so it can move in and out while driven by a powered chuck at the other end on a tailstock on the Y axis. I don't get why he powers two chucks. I think only one powered chuck is required.

Excellent!

Stanislavz wrote:

No official rtai kernal with mint yet ? 

No and there is not going to be any, but there is the one for Debian that might work, linux-image from here
www.linuxcnc.org/dists/bookworm/base/binary-amd64/
and LinuxCNC probably from here
www.linuxcnc.org/dists/bookworm/2.9-rt/b...xcnc_2.9.3_amd64.deb
Might also need linux-headers from the first link and rtai-modules

langdons wrote:

I don't understand why it does not work with just the stepconf config.

Why is LinuxCNC so stupid?


Why does it move away from zero and toward 20 while homing????!!!

1. because you told it so.
2. it is not.
3. because you told it so.
-
Try changing "home latch direction" in StepConf for each axis that needs it.

If you are unwilling to share you setup files, hal files etc, then please stop posting asking for help./

So this is pretty weird. Most of the encoder wiring is pretty straightforward but the 5VDC common is not wired. See attached schematic. I looked at the plug that is associated with the x-axis servo and sure enough there was no pin. I checked at the terminal block and found that the 5VDC common was wired to case ground. The same is true of the servo tachometers so that was the manufacturer's original intention. Kinda leery of wiring my MESA card's 5vdc PS to case ground.