Re: [Emc-users] Help with stepgen joint error using Mesa 5i25 7i76 cards

[andrew beck]

Hey guys

PCW.  I tried changing the setp pid.XXX.maxerror .0005  to .01 and that
fixed the error I think.  I don't get a joint error now.  So thanks so much
for the help.

Gene Haskin

Just a bit of info as you were asking about it.

This is a commerical vmc that I am retrofitting..  The ballscrews are
single start 16 mm pitch.  40 mm dia.  But this machine used to run like
that and was a nice accurate machine.  I have 131072 ppr encoders so heaps
of resolution.  But then I am running step and direction lol so I probably
loose so much of that accuracy anyway.  and as I understand it the actual
accuracy is only as good as the tuning of the motors.  The encoders are
actually pretty interesting.  if I connect up the batteries they are
absolute encoders and don't loose their position values.  But If I don't
use a battery and change a parameter they work as normal incremental
encoders I think.  though under the hood they may be serial anyway I am
picking.  They were a extra 30 usd each above the normal 2500 ppr encoders
so I decided to get them.  And they come out of either germany or taiwan
depending on what resolution I need.

anyway here are the ini files requested.  you guys might be able to find
some more silly stuff that I have done if you want to scan it..  But PCWs
fix worked so the machine is currently working.

*INI file*
# Generated by PNCconf at Sun Jul  7 12:14:14 2019
# If you make changes to this file, they will be
# overwritten when you run PNCconf again

[EMC]
MACHINE = test_2
DEBUG = 0

[DISPLAY]
DISPLAY = axis
POSITION_OFFSET = RELATIVE
POSITION_FEEDBACK = ACTUAL
MAX_FEED_OVERRIDE = 2.00
MAX_SPINDLE_OVERRIDE = 1.50
MIN_SPINDLE_OVERRIDE = 0.50
INTRO_GRAPHIC = linuxcnc.gif
INTRO_TIME = 5
PROGRAM_PREFIX = /home/andrew/linuxcnc/nc_files
INCREMENTS = 5mm 1mm .5mm .1mm .05mm .01mm .005mm
POSITION_FEEDBACK = ACTUAL
DEFAULT_LINEAR_VELOCITY = 0.25
MAX_LINEAR_VELOCITY = 100.00
MIN_LINEAR_VELOCITY = 0.01
DEFAULT_ANGULAR_VELOCITY = 0.25
MAX_ANGULAR_VELOCITY = 1.00
MIN_ANGULAR_VELOCITY = 0.01
EDITOR = gedit
GEOMETRY = xyz

[FILTER]
PROGRAM_EXTENSION = .png,.gif,.jpg Greyscale Depth Image
PROGRAM_EXTENSION = .py Python Script
png = image-to-gcode
gif = image-to-gcode
jpg = image-to-gcode
py = python

[TASK]
TASK = milltask
CYCLE_TIME = 0.010

[RS274NGC]
PARAMETER_FILE = linuxcnc.var

[EMCMOT]
EMCMOT = motmod
COMM_TIMEOUT = 1.0
COMM_WAIT = 0.010
SERVO_PERIOD = 10

[HOSTMOT2]
#  This is for info only 
# DRIVER0=hm2_pci
# BOARD0=5i25

[HAL]
HALUI = halui
HALFILE = test_2.hal
HALFILE = custom.hal
POSTGUI_HALFILE = postgui_call_list.hal
SHUTDOWN = shutdown.hal

[HALUI]

[TRAJ]
AXES = 3
COORDINATES = X Y Z
LINEAR_UNITS = mm
ANGULAR_UNITS = degree
CYCLE_TIME = 0.010
DEFAULT_VELOCITY = 20.00
MAX_LINEAR_VELOCITY = 200.00
NO_FORCE_HOMING = 1

[EMCIO]
EMCIO = io
CYCLE_TIME = 0.100
TOOL_TABLE = tool.tbl

#
# Axis X
#
[AXIS_0]
TYPE = LINEAR
HOME = 0.0
FERROR = 5.0
MIN_FERROR = 6
MAX_VELOCITY = 200.0
MAX_ACCELERATION = 1000.0
# The values below should be 25% larger than MAX_VELOCITY and
MAX_ACCELERATION
# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 250.00
STEPGEN_MAXACCEL = 1250.00
P = 50.0
I = 0.0
D = 0.0
FF0 = 0.0
FF1 = 1.0
FF2 = 0.0
BIAS = 0.0
DEADBAND = 0.0
MAX_OUTPUT = 250.0
# these are in nanoseconds
DIRSETUP   = 1000
DIRHOLD= 1000
STEPLEN= 1000
STEPSPACE  = 1000
STEP_SCALE = 468.75
MIN_LIMIT = -0.01
MAX_LIMIT = 1000.0
HOME_OFFSET = 0.0

#
# Axis Y
#
[AXIS_1]
TYPE = LINEAR
HOME = 0.0
FERROR = 5.0
MIN_FERROR = 6
MAX_VELOCITY = 200.0
MAX_ACCELERATION = 750.0
# The values below should be 25% larger than MAX_VELOCITY and
MAX_ACCELERATION
# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 250.00
STEPGEN_MAXACCEL = 937.50
P = 50.0
I = 0.0
D = 0.0
FF0 = 0.0
FF1 = 1.0
FF2 = 0.0
BIAS = 0.0
DEADBAND = 0.0
MAX_OUTPUT = 250.0
# these are in nanoseconds
DIRSETUP   = 200
DIRHOLD= 200
STEPLEN= 200
STEPSPACE  = 200
STEP_SCALE = 1500.0
MIN_LIMIT = -0.01
MAX_LIMIT = 500.0
HOME_OFFSET = 0.0

#
# Axis Z
#
[AXIS_2]
TYPE = LINEAR
HOME = 0.0
FERROR = 10.0
MIN_FERROR = 1.0
MAX_VELOCITY = 200.0
MAX_ACCELERATION = 750.0
# The values below should be 25% larger than MAX_VELOCITY and
MAX_ACCELERATION
# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 250.00
STEPGEN_MAXACCEL = 937.50
P = 50.0
I = 0.0
D = 0.0
FF0 = 0.0
FF1 = 1.0
FF2 = 0.0
BIAS = 0.0
DEADBAND = 0.0
MAX_OUTPUT = 250.0
# these are in nanoseconds
DIRSETUP   = 200
DIRHOLD= 200
STEPLEN= 200
STEPSPACE  = 200
STEP_SCALE = 1000.0
MIN_LIMIT = -500.0
MAX_LIMIT = 0.01
HOME_OFFSET = 0.0

#
# Spindle
#
[SPINDLE_9]
P = 0
I = 0
D = 0
FF0 = 1
FF1 = 0
FF2 = 0
BIAS = 0
DEADBAND = 0
MAX_OUTPUT = 2000



*Hal file:*

# Generated by PNCconf at Sun 

Re: [Emc-users] 8255 type cards

[bari]

It doesn't matter what does the counting, fpga, pic, chickens, etc as
long as the count is accurate and often enough for the rate of movement.
Servo loops using Mesa are 1KHz and faster.

On 7/6/19 12:04 PM, Roland Jollivet wrote:
> Hi  all
>
> This might expose my ignorance ;-) and I'm not wanting anyone to do
> in-depth calculations for me..
>
> In the days of ISA cards, the 8255 was quite popular, but now we're onto
> PCI and PCIe
>
> So, if one looks at the Mesa 5i25, it can read encoders because presumably
> because the FPGA is doing the counting.
>
> Now, if I used one pci-bridge 8255 type card and used all the I/O (A,B,C)
> to support 3 encoders, could I use, say a Pic processor to read the encoder
> and convert to 8bit data, and then;
> These 8bits are presented to the 8255 ports so Lcnc can do a snapshot read
> of the encoder values to close a servo loop.
>
> With 8bit counters on the encoders, would Lcnc read the encoders fast
> enough to run a servo based machine comparable to an open loop stepper
> system? Obviously it has to read before roll-over or all is lost
>
> So, lets say the encoder provides 5um linear resolution on an axis, moving
> at 10m/min   (32ft/min or 0.05Furlong/min  ;-)
> = 10 000 000um/5um /min
> = 2 000 000 counts/min
> = 33 333Hz counter rate
>
> With a  /256 pic counter, roll-over will be at 130Hz or every 7.7ms
> Could Lcnc cope with that?
>
> Roland
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] 8255 type cards

[Roland Jollivet]

Hi  all

This might expose my ignorance ;-) and I'm not wanting anyone to do
in-depth calculations for me..

In the days of ISA cards, the 8255 was quite popular, but now we're onto
PCI and PCIe

So, if one looks at the Mesa 5i25, it can read encoders because presumably
because the FPGA is doing the counting.

Now, if I used one pci-bridge 8255 type card and used all the I/O (A,B,C)
to support 3 encoders, could I use, say a Pic processor to read the encoder
and convert to 8bit data, and then;
These 8bits are presented to the 8255 ports so Lcnc can do a snapshot read
of the encoder values to close a servo loop.

With 8bit counters on the encoders, would Lcnc read the encoders fast
enough to run a servo based machine comparable to an open loop stepper
system? Obviously it has to read before roll-over or all is lost

So, lets say the encoder provides 5um linear resolution on an axis, moving
at 10m/min   (32ft/min or 0.05Furlong/min  ;-)
= 10 000 000um/5um /min
= 2 000 000 counts/min
= 33 333Hz counter rate

With a  /256 pic counter, roll-over will be at 130Hz or every 7.7ms
Could Lcnc cope with that?

Roland

___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help with stepgen joint error using Mesa 5i25 7i76 cards

[Gene Heskett]

On Saturday 06 July 2019 08:55:54 andrew beck wrote:

> Hey guys.
>
> I will post this on the forum also...
>
> So I'm just setting up my cnc mill with Mesa cards and step/dir
>
> Using 5i25 and 7i76.
>
I'm using that combo, twice. One is driving a 6040 quite well, the other 
is moving a Grizzly G0704

> Chinese servos are working well.
>
> But I just spent most of my Saturday chasing this funny stepgen joint
> error.

Copy/paste it here please.

> I am wanting a resolution of at least.  0.005 mm.  And I want at least
> 12 m per min rapid speed as I'm retrofitting a high speed linear rail
> machine. The ballscrews are 16 mm pitch.

16mm pitch and .005mm resolution?  Multistart screws I assume. Thats 
asking for vanishingly small backlash, maybe even dual nuts loaded 
against each other with preload springs.  And better screws than any 
I've used which are usually C7, the cheapest.

> Does anyone know why I get a stepgen joint error?  And also when
> moving at high speed Linux cnc rapids to about 3 mm of the end and
> then it slowly inches up the last little bit.
>
That sounds an awful lot like a way to low 'accel' someplace in that 
control chain, so it thinks it will take a long time to get stopped.
But if thats the case it should also be accompanied by a long rather 
leasurely accel to top speed too. OTOH from 12m a minute, 3mm doesn't 
sound like too long a distance to get stopped.  Your creep could be a 
PID windup too, too much IGain?

Something you may have to observe with the halscope while driving it back 
and forth with a siggen module.  That, despite the very poor bandwidth 
of the halscope, can be quite educational. 

Search back thru the archive for a mention of disconnecting one of the 
axis feedback signals, ISTR there was a problem that sounded like that 
which popped up about the time the joint/axis merge was done around a 
year ago.

> I don't think its a big problem and I'll go ahead and bolt the motors
> to the machine.  But would appreciate some help from people.

Also, if thats a gantry style machine, the y moves the whole gantry, 
which is the most massive part, and will often need a few percent less 
in the accel dept.

HTH. May not, it seems my wet ram is getting less attentive. Too many 
birthdays or something like that. :(

> Regards
>
> Andrew
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help with stepgen joint error using Mesa 5i25 7i76 cards

[Peter C. Wallace]

On Sun, 7 Jul 2019, andrew beck wrote:


Date: Sun, 7 Jul 2019 00:55:54 +1200
From: andrew beck 
Reply-To: "Enhanced Machine Controller (EMC)"

To: Emc-users@lists.sourceforge.net
Subject: [Emc-users] Help with stepgen joint error using Mesa 5i25 7i76 cards

Hey guys.

I will post this on the forum also...

So I'm just setting up my cnc mill with Mesa cards and step/dir

Using 5i25 and 7i76.

Chinese servos are working well.

But I just spent most of my Saturday chasing this funny stepgen joint error.

I am wanting a resolution of at least.  0.005 mm.  And I want at least 12 m
per min rapid speed as I'm retrofitting a high speed linear rail machine.
The ballscrews are 16 mm pitch.

Does anyone know why I get a stepgen joint error?  And also when moving at
high speed Linux cnc rapids to about 3 mm of the end and then it slowly
inches up the last little bit.

I don't think its a big problem and I'll go ahead and bolt the motors to
the machine.  But would appreciate some help from people.



This sounds like an error caused by a scale problem in pncconf on metric 
configurations.


Check your hal file are remove any lines like:

setp pid.XXX.maxerror .0005

(this number makes sense for inch machines but not mm machines)



Regards

Andrew

___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users



Peter Wallace
Mesa Electronics

(\__/)
(='.'=) This is Bunny. Copy and paste bunny into your
(")_(") signature to help him gain world domination.



___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Help with stepgen joint error using Mesa 5i25 7i76 cards

[andrew beck]

Hey guys.

I will post this on the forum also...

So I'm just setting up my cnc mill with Mesa cards and step/dir

Using 5i25 and 7i76.

Chinese servos are working well.

But I just spent most of my Saturday chasing this funny stepgen joint error.

I am wanting a resolution of at least.  0.005 mm.  And I want at least 12 m
per min rapid speed as I'm retrofitting a high speed linear rail machine.
The ballscrews are 16 mm pitch.

Does anyone know why I get a stepgen joint error?  And also when moving at
high speed Linux cnc rapids to about 3 mm of the end and then it slowly
inches up the last little bit.

I don't think its a big problem and I'll go ahead and bolt the motors to
the machine.  But would appreciate some help from people.

Regards

Andrew

___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users