Advanced Search

Usually it's a Sauter revolver, but I have a test stand at home. It works the same.

tommylight wrote:

No idea where or how you got those numbers, but utmost respect for effort in getting things right.
Start with 300mm/s/s, do some tests, then go higher, but most probably 750 will be to much for short fast movements like Fusion360 does.
I am refraining from further "muddying the waters" for now as:
1. i have no idea what "HLFB" is
2. -deleted-
 

HLFB "high level feedback" is many things, but it can work like a drive alarm and that's what I will be using it for. There have been a couple of posts on it, but they all go nowhere and the OP never comes back and says "I did this thing exactly like this, and it worked. Here is how to recreate what I did."

For examples of dead ends, here you go:

forum.linuxcnc.org/49-basic-configuration/41375-clearpath-hlfb

and

forum.linuxcnc.org/10-advanced-configura...with-hlfb-mesa-7i76e

and

forum.linuxcnc.org/10-advanced-configura...ervo-with-mesa-7i76e

I just want to know what to enter into what box, and move on. I am at the pncconf step, so any talk about ini files or HAL or anything past pncconf is not yet relevant to me. I need to know what to tell pncconf what to do with the inputs I connected the HLFB circuits to. And I need to know if I have connected it properly in the first place. I don't know what I don't know.
 

tommylight wrote:

Do NOT fix drift!
All analog servo systems will have a bit of drift at 0V input.
Wire the control, it takes care of it when encoders are in use.
 

Well, I wouldn't go that far.  Make sure the velocity DAC is outputting zero, and then adjust the "balance" or velocity offset put to get minimum drift.  You will not be able to get it down to absolute zero, but any drift in the amp will offset your position when not commanding movement.  If you really have 12 - 16 V output to the motor with zero in, then something is QUITE wrong!  That could be an open connection to the tachometers.  If the machine has been sitting a long time, the commutator of the tachs might be oxidized, maybe running the motors a bit might clean them.  Or, it could be a dirty connector contact.
Jon

Send me .esi files for canOpen devices from your config... there are not supported startup SDO and it can not be generated automaticly ...

regards

No idea where or how you got those numbers, but utmost respect for effort in getting things right.
Start with 300mm/s/s, do some tests, then go higher, but most probably 750 will be to much for short fast movements like Fusion360 does.
I am refraining from further "muddying the waters" for now as:
1. i have no idea what "HLFB" is
2. -deleted-

Looks like halcmd was run without running LinuxCNC

mesaflash --device 7I96s --addr 10.10.10.10 --write 7i96s_p1encd.bin

should work

mesaflash --device 7I96s --addr 10.10.10.10 --write 7i96s_p1encd.bin --reload

is probably better

sharp_reaper wrote:

I have an error with the reading of arc voltage (-45 V) after one piercing.
Before piercing I was reading 0 V.

The THCAD-300 is mounted inside of the plasma source connected directly to raw Arc Voltage.
The output from THCAD-300 is connected to the encoder from 7i96s (TB2.7 and TB2.8).

Already switch the THCAD-300 and 7i96s to brand new ones and at start the readings of arc voltage seems to be ok.
But when I put Torch On the Arc Voltage goes to -45 V, after that the value doesn't change. No matter if plasma is on or off.

Should have asked before doing all that...
And do not mount THCAD's inside plasma sources, mount it on the back of it, in a plastic box.
Using new equipment after a failed test, also not good.

PCW wrote:

That's what I call starting with a bang!
 

Understood, and thank you for the reply. I calcluated 49800 mm/sec^2 from the data I had available to me . If 10000 mm/sec^2 is still too high, then what is a known good value for this motor? How does the max. accel. entry factor into the behavior of the software? As a dummy check?

From my limited time on planet Earth, I have learned that if I include more than one thing in a forum question, the one that matters the least will be casually non-answered and the important one will be completely ignored. My dilemma is that I feel the need to include background information to give context, and this context is usually where people latch on. That's what happened here.

Again, I need to reiterate that I am not hung up on one specific data point. I want to enter the correct information in pncconf in its entirety before I power on the axis motors, and there are many other variables to consider, for the motors and beyond.

We already talked about HLFB, but no one has ever explained how to assign the inputs in pncconf so I can use them for this purpose. I currently have the machine wired exactly as shown, based on the limited amount of information here :

To put into perspective the limited information available about this topic, this post is already the top Google search hit, and there aren't many more. So Teknic, please, if you're reading this, I am looking for SPECIFIC settings I need to make this work with LinuxCNC, including HLFB.

halcmd show all hm2 > hm2names.txt does not yield very much, see attached.
  Okay,

---

I have mesaflash v3.4.9
My 7i96s is jumpered and presently setup at 10.10.10.10

Should I change the jumpers to 192.168.1.121 for the upgrade, or just change the upgrade commands as follows?

mesaflash --device 7I96s --addr 10.10.10.10 --write 7i96s_p1encd.bin

Than just power cycle the board.



 

Pncconfs 10000 limit is about 1G. I would consider this super fast for routers
with more typical values in the .5 G range.  49800 mm/sec^2 is about 5G
which is crazy for almost anything except some 3D printers and perhaps laser
cutters (with linear motors)

Higher accelerations will excite vibrations and cause position errors due to
machine flex (especially when you are accelerating something heavy like a spindle)

Lower accelerations will cause slowing on sharp curves. One way to look at this is to
calculate the minimum radius possible before the machine needs to slow on corners.
This is V^2/A so at a acceleration of 10000 mm/s^2 and a max cutting velocity of say
100 mm/sec, you can cut a 1mm radius without slowing down.

Another way to get a feel for this is to calculate the time needed to reach full velocity.
at 10000 mm/s^2 and 100 mm/s full speed, its only 10 ms from zero velocity to full velocity.
That's what I call starting with a bang!


 

LINUXCNC_HOME is not set. I thought that I saw in the "linuxcnc" script where it was set upon execution when starting?

This may be why I had so many problems in the past. I prefer to run RIP but, I may have accidentally/inadvertently installed the uspace version
and perhaps not knowing the difference. I recently removed the uspace version however not totally sure that is what caused DESIGNER to
fail.

Additionally, what's a mystery to me is the use of paths in the INI file. Should be using FULL paths or ABSOLUTE paths or other?.......

Here's what is set:
andy@file0x:~$ env
SHELL=/bin/bash
EMC2VERSION=2.10.0~pre0
GLADE_CATALOG_SEARCH_PATH=/home/andy/linuxcnc-dev/lib/python/gladevcp
GLADE_ICON_THEME_PATH=/home/andy/linuxcnc-dev/share/glade/pixmaps:
PWD=/home/andy
LOGNAME=andy
XDG_SESSION_TYPE=tty
MANPATH=/home/andy/linuxcnc-dev/docs/man:/usr/local/man:/usr/local/share/man:/usr/share/man
TCLLIBPATH=/home/andy/linuxcnc-dev/tcl
HOME=/home/andy
LANG=en_US.UTF-8
LINUXCNCVERSION=2.10.0~pre0
XDG_SESSION_CLASS=user
LINUXCNC_EMCSH=/usr/bin/wish8.6
PYTHONPATH=/home/andy/linuxcnc-dev/lib/python
TERM=xterm
USER=andy
SHLVL=1
XDG_SESSION_ID=9
LD_LIBRARY_PATH=/home/andy/linuxcnc-dev/lib
XDG_RUNTIME_DIR=/run/user/1000
LINUXCNC_NCFILES_DIR=/home/andy/linuxcnc-dev/nc_files
EMC2_HOME=/home/andy/linuxcnc-dev
PATH=/home/andy/.local/bin:/home/andy/bin:/home/andy/linuxcnc-dev/bin:/home/andy/linuxcnc-dev/scripts:/home/andy/linuxcnc-dev/tcl:/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games:/home/andy/.local/bin:/home/andy/.local/bin
DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/1000/bus
SSH_TTY=/dev/pts/2
_=/usr/bin/env
 

Both the 7I96S and the THCAD(if damaged) are likely repairable with a
single chip/card

I don't really get your point of view.
The ethical thing would be to share the GPL code.

I don't have the time nor money to sue Tormach to get it. IMO what they are doing is simply malicious (non) compliance.

Yes it is... at the end is homing subroutine ... have a look carrefully how it works in the whole scale

I am using it in the CSP for less then two years and it is rock solid... but not run CSP over them because it is incremental(60C1) and not SI(607A) units because of Kollmorgen stuff... 

but I defenetly recomend you to do all config and startup SDOs via the TwinCAT3(I can help u)... Try it over that because it is easy and it will generate you a configuration itself... with SDO config too

I will have to create a python script which will did it automaticly but time lack is present..

or reply me over my last mails

regarrds

tommylight wrote:

Unless that is a very small 3D printer, 49800 mm/sec^2 is way, way, way to high.
 

Okay. I am looking for known good values for Clearpath SDSK servos on an Avid router, specifically a Benchtop Pro. There is no way that I am the first person to set up this specific configuration.