Re: [Emc-users] crazy

[Gene Heskett]

On Monday 30 November 2020 17:23:47 Peter C. Wallace wrote:

> On Mon, 30 Nov 2020, Gene Heskett wrote:
> > Date: Mon, 30 Nov 2020 17:07:04 -0500
> > From: Gene Heskett 
> > Reply-To: "Enhanced Machine Controller (EMC)"
> > 
> > To: "Enhanced Machine Controller (EMC)"
> >  Subject: [Emc-users] crazy
> >
> > Greetings all the pid experts;
> >
> > I changed out the bob for one w/o the input capacitors, looks 1
> > million times better on the scope
> >
> > Playing with what little servo I have configured so far, trying to
> > get it to actually act like a servo. I don't have a valid SCALE yet
> > but expect it will be large so I set it to 1150. Hooked the encoder
> > feed back back up in hal. Couldn't make it move, pwmgen limited to
> > about a 2% duty cycle. Not enough to move the motor. looking at the
> > pid_a with halconfig, I saw that pid_a.maxerror was set to .0005. 
> > Curious, since it was at that point sitting around 1000 out of
> > balance, I did a setp to set it to 1000. Motor almost jumped off the
> > stepladder and ran toward balance at a good clip but eventually
> > stopped when encoder feedback into pid_a.feedback was about 2 from
> > equalling pid_a.cmd.  That .0005 was an old setting, but worked when
> > a stepper was doing the shaft twisting. I certainly don't remember
> > adding it 5+ years ago.
> >
> > That was set in the .hal file but I've no clue why.  Whats the
> > default if its not set? It is not in the .ini file
>
> Default if not set is 0 which means the error is unbounded.

Cleared out of the file now.  Thanks Peter.

> Maxerror is bound on stepgen loops since the stepgen hardware is
> an almost perfect velocity mode device (so FF1 does 99.99%
> of the work). The advantage of the bound is that servo thread
> jitter creates bogus position feedback errors and by bounding
> these errors you reduce their effect on motion. This has largely
> been made obsolete by the DPLL that reduces the stepgen and
> encoder position sampling jitter to a small fraction of a
> microsecond.

[...]

Stay safe and well now.

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] crazy

[Peter C. Wallace]

On Mon, 30 Nov 2020, Gene Heskett wrote:


Date: Mon, 30 Nov 2020 17:07:04 -0500
From: Gene Heskett 
Reply-To: "Enhanced Machine Controller (EMC)"

To: "Enhanced Machine Controller (EMC)" 
Subject: [Emc-users] crazy

Greetings all the pid experts;

I changed out the bob for one w/o the input capacitors, looks 1 million
times better on the scope

Playing with what little servo I have configured so far, trying to get it
to actually act like a servo. I don't have a valid SCALE yet but expect
it will be large so I set it to 1150. Hooked the encoder feed back back
up in hal. Couldn't make it move, pwmgen limited to about a 2% duty
cycle. Not enough to move the motor. looking at the pid_a with
halconfig, I saw that pid_a.maxerror was set to .0005.  Curious, since
it was at that point sitting around 1000 out of balance, I did a setp to
set it to 1000. Motor almost jumped off the stepladder and ran toward
balance at a good clip but eventually stopped when encoder feedback into
pid_a.feedback was about 2 from equalling pid_a.cmd.  That .0005 was an
old setting, but worked when a stepper was doing the shaft twisting. I
certainly don't remember adding it 5+ years ago.

That was set in the .hal file but I've no clue why.  Whats the default if
its not set? It is not in the .ini file


Default if not set is 0 which means the error is unbounded.

Maxerror is bound on stepgen loops since the stepgen hardware is
an almost perfect velocity mode device (so FF1 does 99.99%
of the work). The advantage of the bound is that servo thread
jitter creates bogus position feedback errors and by bounding
these errors you reduce their effect on motion. This has largely
been made obsolete by the DPLL that reduces the stepgen and
encoder position sampling jitter to a small fraction of a
microsecond.



Right now I can run it either way, but it takes a long time to settle
like it needs a lot more Pgain yet, taking at least 30 seconds, to over
a minute to reach a quiet sorta balance as the motor stops before
hitting exact balance, but it may take many seconds of bumping along
before it stops for good, leaving the pwmgen outputting 10 to 15
microsecond pulses forever. But its progress, and I'm learning. And
wondering if I need to put a low level dither in it somehow.

And still no bolts for the motor mounts. So I'm twiddling my thumbs.


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



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] crazy

[Gene Heskett]

Greetings all the pid experts;

I changed out the bob for one w/o the input capacitors, looks 1 million 
times better on the scope

Playing with what little servo I have configured so far, trying to get it 
to actually act like a servo. I don't have a valid SCALE yet but expect 
it will be large so I set it to 1150. Hooked the encoder feed back back 
up in hal. Couldn't make it move, pwmgen limited to about a 2% duty 
cycle. Not enough to move the motor. looking at the pid_a with 
halconfig, I saw that pid_a.maxerror was set to .0005.  Curious, since 
it was at that point sitting around 1000 out of balance, I did a setp to 
set it to 1000. Motor almost jumped off the stepladder and ran toward 
balance at a good clip but eventually stopped when encoder feedback into 
pid_a.feedback was about 2 from equalling pid_a.cmd.  That .0005 was an 
old setting, but worked when a stepper was doing the shaft twisting. I 
certainly don't remember adding it 5+ years ago.
 
That was set in the .hal file but I've no clue why.  Whats the default if 
its not set? It is not in the .ini file

Right now I can run it either way, but it takes a long time to settle 
like it needs a lot more Pgain yet, taking at least 30 seconds, to over 
a minute to reach a quiet sorta balance as the motor stops before 
hitting exact balance, but it may take many seconds of bumping along 
before it stops for good, leaving the pwmgen outputting 10 to 15 
microsecond pulses forever. But its progress, and I'm learning. And 
wondering if I need to put a low level dither in it somehow.

And still no bolts for the motor mounts. So I'm twiddling my thumbs.


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] Faulty THCAD board? 7i96 error? Advice needed! Was: Plasma config with Telwin plasma source

[Peter C. Wallace]

On Mon, 30 Nov 2020, Viesturs L?cis wrote:


Date: Mon, 30 Nov 2020 22:35:21 +0200
From: "[UTF-8] Viesturs L?cis" 
Reply-To: "Enhanced Machine Controller (EMC)"

To: "Enhanced Machine Controller (EMC)" 
Subject: [Emc-users] Faulty THCAD board? 7i96 error? Advice needed! Was:
Plasma config with Telwin plasma source

Hello!

I am stuck with the plasma machine that has 7i96 board and also THCAD board.
My biggest issue is getting THC module to work. I have managed to get
the encoder module to receive pulses from THCAD board and to count
them.
Here are the encoder settings:
https://www.picpasteplus.com/v.php?i=b5f8366a2e
The problem is that THCAD board is not sending any pulses with the
frequency division jumper set at 1/1. In that case the rawcounts pin
of encoder module remains at 1. If I move the jumper to 1/32,
rawcounts pin immediately starts to increase and the encoder velocity
is approximately -120.


Thats correct, with the default encoder filter settings you can not
count the ~1/2 usec pulses from the THCAD. That said, for noise 
immunity its suggested to use the /32 setting and the default encoder

filtering.



The thing is that I was expecting the value to be 120. But why is it
NEGATIVE?!? I tried setting encoder scale value to be negative. It
fixes the issue of velocity being negative.

Its negative because the counter mode is up/down and the FPGA direction pin
is low if the B input is unconnected (which means down)


Nex problem with THCAD card is that if I fire the plasma and it
creates the arc, encoder velocity decreases. Yes, it gets smaller.
Approximate value is 96 for that particular torch height I have set. I
watched both velocity and position pins of the encoder module.
Position pin was increasing, so the pulses were still coming and it
was not a glitch of a "frozen" value at velocity pin.
My way of circumventing this is negative value for thc.scale_offset
parameter. Then with negative encoder velocity value, which is
decreasing as the plasma starts and negative offset I get positive
voltage value in THC module.


Not sure what you are saying, I would expect the plasma voltage to
decrease once the arc is established (compared to the higher open circuit 
voltage)




And then there is a minor problem with 7i96 board:
https://www.picpasteplus.com/v.php?i=f609e19c95


This indicates either poor latency in the host or lost communication with the 
7I96. You might go through the standard latency reducing steps on the PC

(disabling power management/speed switching,EIST,C states >C1 etc etc)
Also if you have a Intel Ethernet chip on your PC you need to disable
IRQ coalescing as described in the hm2_eth manual page



This error was triggered a minute or two after I launched the
LinuxCNC. Is that subject to poor config of the system? PC has
mini-ITX board with dualcore Celeron CPU and 4 gab of RAM. Should be
plenty of power for LinuxCNC but apparently I am wrong.
The thing is that after this error HAL pin values were "frozen" and
encoder velocity pin was showing approximately 120 while the position
pin was also frozen and did not change.


Second thing is that I have to use an existing plasma source that they
already have - "Telwin enterprise plasma 160 hf", which is meant for
manual cutting. So there are 2 challenges - get the plasma voltage and
arc-ok signal.

Does anybody happen to know, if the controller board of this machine
has dedicated outputs for these 2 things?

I have connected 2 wires to plasma ground and torch cables
respectively and connected them to THCAD board. I hope that it would
work and not cause too much of EMF noise. Cable is shielded, shield is
connected to ground at a single point with all the other cable
shields.

How do I obtain arc-ok signal? Can I do it in HAL by watching plasma
voltage and trying to assume that it is ok, when voltage drops below
some preset treshold? Or is there a doable way to take a signal that
would drive a relay to trigger input of 7i96?

Viesturs


___
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] Faulty THCAD board? 7i96 error? Advice needed! Was: Plasma config with Telwin plasma source

[Viesturs Lācis]

Hello!

I am stuck with the plasma machine that has 7i96 board and also THCAD board.
My biggest issue is getting THC module to work. I have managed to get
the encoder module to receive pulses from THCAD board and to count
them.
Here are the encoder settings:
https://www.picpasteplus.com/v.php?i=b5f8366a2e
The problem is that THCAD board is not sending any pulses with the
frequency division jumper set at 1/1. In that case the rawcounts pin
of encoder module remains at 1. If I move the jumper to 1/32,
rawcounts pin immediately starts to increase and the encoder velocity
is approximately -120.
The thing is that I was expecting the value to be 120. But why is it
NEGATIVE?!? I tried setting encoder scale value to be negative. It
fixes the issue of velocity being negative.
Nex problem with THCAD card is that if I fire the plasma and it
creates the arc, encoder velocity decreases. Yes, it gets smaller.
Approximate value is 96 for that particular torch height I have set. I
watched both velocity and position pins of the encoder module.
Position pin was increasing, so the pulses were still coming and it
was not a glitch of a "frozen" value at velocity pin.
My way of circumventing this is negative value for thc.scale_offset
parameter. Then with negative encoder velocity value, which is
decreasing as the plasma starts and negative offset I get positive
voltage value in THC module.

And then there is a minor problem with 7i96 board:
https://www.picpasteplus.com/v.php?i=f609e19c95

This error was triggered a minute or two after I launched the
LinuxCNC. Is that subject to poor config of the system? PC has
mini-ITX board with dualcore Celeron CPU and 4 gab of RAM. Should be
plenty of power for LinuxCNC but apparently I am wrong.
The thing is that after this error HAL pin values were "frozen" and
encoder velocity pin was showing approximately 120 while the position
pin was also frozen and did not change.


Second thing is that I have to use an existing plasma source that they
already have - "Telwin enterprise plasma 160 hf", which is meant for
manual cutting. So there are 2 challenges - get the plasma voltage and
arc-ok signal.

Does anybody happen to know, if the controller board of this machine
has dedicated outputs for these 2 things?

I have connected 2 wires to plasma ground and torch cables
respectively and connected them to THCAD board. I hope that it would
work and not cause too much of EMF noise. Cable is shielded, shield is
connected to ground at a single point with all the other cable
shields.

How do I obtain arc-ok signal? Can I do it in HAL by watching plasma
voltage and trying to assume that it is ok, when voltage drops below
some preset treshold? Or is there a doable way to take a signal that
would drive a relay to trigger input of 7i96?

Viesturs


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

Re: [Emc-users] Changed mind on motor, now for sale

[Bruce Layne]

On 11/30/20 1:56 PM, Gene Heskett wrote:
> ...it did a fine job, very good finish with a 4 flute 1/8" uncoated carbide 
> tool 
> plowing out the outline of these motor mounts

When using these high RPM spindle motors on the router, I try to use
single flute small diameter end mills.  That allows a higher spindle RPM
where the motor makes a bit of power without having a cut depth that is
friction burning rather than cutting.  Managing friction is very
important when milling plastics, and the Onsrud O flute carbide bits do
a good job, although they're expensive so do some slow air cuts when
writing your own G code.

I run the water cooled spindle motor between 9000 and 24000 RPM. 
Bogging it down with low RPM makes more heat than power.


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

Re: [Emc-users] Changed mind on motor, now for sale

[Gene Heskett]

On Monday 30 November 2020 09:54:13 Todd Zuercher wrote:

> The main problem is that you keep trying to run these cheapo 24000rpm
> spindles at speeds less than 9000rpm for extended time periods.  Of
> course you're going to have trouble cooling it, (or have much usable
> torque.)  They are made for wood carving, not milling metal.

You are right about the cheapo's, I don't think I'll get a 100 hours out 
of this one, at 75 or thereabouts hours, it sounds like its grinding up 
the bearing cages already. Not built to carve 1/2" alu although it did a 
fine job, very good finish with a 4 flute 1/8" uncoated carbide tool 
plowing out the outline of these motor mounts, with the mister working, 
Digging .75mm deep per pass at 350mm a minute and 14 to 15 thousand 
revs.

But I was hearing more than tool noise at the end of the second mount. 
Then my stepson came down from Maryland to spend Sat afternoon with me, 
and I spun it up to about 8 grand for him and was not impressed by the 
noise, its close to done. But I have a good 4 bearing 250 volt version 
and a higher voltage vfd to match coming. Cooling didn't seem to be a 
problem as it didn't heat more than 10C over room during those two 24 
minute jobs. The new one s/b here by Pearl Harbor day. Making a home 
switch mount for the BS-1 can wait, or can be done on the GO704 as 
theres some room left on the end of its table if I absolutely must be 
doing something.

So the short life of this one, even if it had been carving wood, is not 
exactly news. This machine was a prime example of BBLB as the first 
thing I did was throw away ALL the electronics including the box. With 
decent drivers, those itty bitty motors can do 190 ipm for XY but the z 
is still puny at about 35 going up. I'm hoping it can lift the longer, 
heavier new motor. Out of the box? About 7 to 10. Plug in the A and it 
dropped like a rock. The supply, supposedly 24 volts, was folded back to 
14 with 3 motors plugged, and about 6 with the A plugged in.  With Mach3 
running it, then with linuxcnc, the vfd ran the spindle in a random 
direction. I asked the guy to refund the cost of the electronics ($500) 
and got told off.  So I thought I'd make a decent machine out of the raw 
material.  This motor is just part of that effort. Next is thicker X 
rods as I expect someone has that spot on the shelf filled since this 
mill is a bit famous for needing that. It will get done if I don't stop 
using air. At my age? All things are possible.

Thanks Todd.
> Todd Zuercher
> P. Graham Dunn Inc.
> 630 Henry Street 
> Dalton, Ohio 44618
> Phone:  (330)828-2105ext. 2031
>
> -Original Message-
> From: Gene Heskett 
> Sent: Wednesday, November 25, 2020 9:38 PM
> To: Enhanced Machine Controller (EMC)
>  Subject: [Emc-users] Changed mind on
> motor, now for sale
>
> [EXTERNAL EMAIL] Be sure links are safe.
>
> Greetings all;
>
> I bought a good 4 bearing motor to replace the partially blocked water
> cooled 1.5KW motor that came on the 6040 mill. But thinking on the 110
> volt version's performance at the lower revs, and the flickering of
> the lights when its running slow, I've decided the 2.2KW version is a
> bit much for the 120 volt wall plug wiring it my garage.
>
> I am adding a 250 volt run to the machine so I can use a 220 vfd and
> motor, and ordered that same motor and its matching vfd for 220 volt
> service.
>
> So these two motors will be for sale along with a 1.5kw vfd, all for
> 110 volt service.  The OEM motor has about 20 hours service on it, the
> good one is some what unpacked, repacked and brand new.  The vfd has
> been reprogrammed to properly drive the OEM motor. It does that very
> well down to about 1000 revs.
>
> I've got $160 in the new 4 bearing 2.2kw & the $125 vfd has a few
> hours on its capacitors, estimated 5 year life, so make me an offer if
> you can use them.
>
> Thanks everybody.
>
> 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
>
>
> ___
> 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

Re: [Emc-users] Changed mind on motor, now for sale

[Todd Zuercher]

The main problem is that you keep trying to run these cheapo 24000rpm spindles 
at speeds less than 9000rpm for extended time periods.  Of course you're going 
to have trouble cooling it, (or have much usable torque.)  They are made for 
wood carving, not milling metal.

Todd Zuercher
P. Graham Dunn Inc.
630 Henry Street 
Dalton, Ohio 44618
Phone:  (330)828-2105ext. 2031

-Original Message-
From: Gene Heskett  
Sent: Wednesday, November 25, 2020 9:38 PM
To: Enhanced Machine Controller (EMC) 
Subject: [Emc-users] Changed mind on motor, now for sale

[EXTERNAL EMAIL] Be sure links are safe.

Greetings all;

I bought a good 4 bearing motor to replace the partially blocked water cooled 
1.5KW motor that came on the 6040 mill. But thinking on the 110 volt version's 
performance at the lower revs, and the flickering of the lights when its 
running slow, I've decided the 2.2KW version is a bit much for the 120 volt 
wall plug wiring it my garage.

I am adding a 250 volt run to the machine so I can use a 220 vfd and motor, and 
ordered that same motor and its matching vfd for 220 volt service.

So these two motors will be for sale along with a 1.5kw vfd, all for 110 volt 
service.  The OEM motor has about 20 hours service on it, the good one is some 
what unpacked, repacked and brand new.  The vfd has been reprogrammed to 
properly drive the OEM motor. It does that very well down to about 1000 revs.

I've got $160 in the new 4 bearing 2.2kw & the $125 vfd has a few hours on its 
capacitors, estimated 5 year life, so make me an offer if you can use them.

Thanks everybody.

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


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