Re: [Emc-users] 7i90 not listed in pnc

[Gene Heskett]

On Tuesday 22 May 2018 23:10:24 jeremy youngs wrote:

> On Tue, May 22, 2018, 20:53 Gene Heskett  wrote:
> > Ouch, unless its a timing belt, otherwise there will be slippage,
> > making the scale more of a SWAG than accurate.
>
> Yes, 1 inch wide xl 037
>
>
> That, and the fact that I wasn't aware the encoders
>
> > output was differential, meant I had to use some under a buck a
> > copy, rs485 gismo's setup as receivers only to convert the
> > differential to TTL for the 5i25 inputs. But to get to the 5i25, I
> > had to rip out the opto's in the BoB's input for the A/B inputs.
>
> My encoders are completely unknown to me, and I cannot find much
>
> > information on them.
>
> I have 2 Bob's that are just vias directly to db 25. I may have to
> make a 50 pin adapter for them .

You won't need a BoB, the 7i42TA is both a BoB and 7i90 protector. 24 
pins at a time. Just make sure that all things being controlled are 
active low. The 7i90 can sink lots more currant than it can source.

> I don't know yet if I will have to buffer the signal, how did you know
> you needed to ?

If you are refering to the encoder, the spindle servo went wide open at 
about 250 revs on the gui's tach, while the tach went to zero. The 
opto's simply weren't fast enough to get the A/B signals into the 5i25 
inputs. So the PID thought the spindle was stalled and fed the motor the 
whole 127 volts the supply is capable of.  Its a 90 volt motor. In my 
case, the BoB's are Sainsmart BoB's, no opto's in any output.

But this isn't germane to your setup, I just want you to be aware it 
might be needed if the encoder is spun too fast for the opto's. The 
rs485 gizmo's are under a buck each on ebay. Get a 10 pack, they are 
handy as can be for fixing bad signals, giving you a clean rail to rail 
output.

They'll need a ground from the 7i42TA's and some 5 volts, and a solder 
blob IIRC to convert them from bidirectional to just a receiver of the 
diff signals from the encoder. Tested here on the bench, works well at 
more than my cheap function generator can manage, which is 2 megabaud.

> > > is there any reason I should choose a higher count?
> >
> >  Way more than that is well tolerated now.  If you are 1/1, then I
> > wouldn't consider an encoder under 1000 lines, your PID tuning will
> > be much much easier.
> >
> >
> >
> > 1/2" 13 tpi, is a tall order. How much torque is available?
>
> 620 oz inches , 3.2 kw at this time.

Thats not enough. A 10x geardown might be enough.
 
> I have my eye on a bldc at 37.5 n/m at 5kw if I can scratch up the
> funds. Rigid tapping isn't an immediate issue but would be nice.

Fixing that later is spending the money twice. And once you've seen it 
work, you'll wonder why you ever considered it to just be "nice". Even 
if the power to drive it all the way in one pass isn't there, my case 
with this G0704, and you have to write peck loops, you won't be willing 
to go back to turning T handles and breaking taps.

You'll have to init some vars, but the peck loop looks like this:

o300 WHILE [#<_ztmp> gt #<_tapbot>]
g33.1 z[#<_ztmp> + #<_tapdec>] k#<_tpmm>
#<_ztmp> = [#<_ztmp> + #<_tapdec>]
o300 ENDWHILE

That taps one of the 4 grub screw holes in this tap hat I'm making a 
bunch of. 76 L.O.C. drills all 4 holes, change from a shank sized drill 
to a tap, and rerun after setting the if triggers to tap, and it then 
taps them 12mm deep. Put the tap into the "hat", install the grub screws 
to grip the square butt of the tap, tightening to about 1/8th turn from 
broke/stripped, done.

> > Gripping the tap in a collet, R8 or a TTS ER-20 without the positive
> > coupling, tightly enough to prevent slippage.
>
> I have a bt30 spindle in the works.

Now I'm jealous. :) But I'll have to settle for a metal gear kit and some 
better bearings in the head of this G0704 at some point. The thing has 
plastic gears!

> I hope soon to have this thing reassembled, it's been a couple months
> since it made chips. Thank you for your advice.

NP Jeremy.

-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] No Charge Pump signal

[John Dammeyer]

I have a solution.  Part of the problem with the LinuxCNC system was the 
behaviour of the PMDX-126 Break Out Board with respect to the Charge Pump 
Signal.  

I've posted the results to:
https://www.pmdx.com/PMDX-Forums/index.php?topic=521.0

The stepconf generated hal file is  incorrect when using the charge pump signal 
assigned to PIN 17. It didn't help that the PMDX-126 also had an odd behaviour 
requiring removal of the charge pump in order to allow the ESTOP switch input 
to be reflected on the parallel port connection.

I don't know if an update should be posted here too since this is specific to 
the PMDX-126
https://github.com/LinuxCNC/linuxcnc/issues/315

John

> -Original Message-
> From: andy pugh [mailto:bodge...@gmail.com]
> Sent: May-22-18 5:06 AM
> To: Enhanced Machine Controller (EMC)
> Subject: Re: [Emc-users] No Charge Pump signal
> 
> On 22 May 2018 at 07:29, John Dammeyer  wrote:
> 
> > If I understand the hal file correctly the charge pump code is loaded, the
> > estop_out pin and charge-pump.enable are dependent on the user-
> enable-out
> > which is controlled by the ENABLE button on the top left part of the
> screen.
> 
> If you look at charge-pump.enable with Halmeter does it change state?
> You could also Halscope parport pin 17, that should indicate if HAL
> _thinks_ it is twiddling the pin.
> 
> It is possible that the scope might not see anything without pull-up
> or pull-down on the pin. P-port pins are a bit funny that way I seem
> to recall (I haven't used one for ages)
> 
> --
> atp
> "A motorcycle is a bicycle with a pandemonium attachment and is
> designed for the especial use of mechanical geniuses, daredevils and
> lunatics."
> � George Fitch, Atlanta Constitution Newspaper, 1916
> 
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 7i90 not listed in pnc

[jeremy youngs]

On Tue, May 22, 2018, 20:53 Gene Heskett  wrote:

>
> Ouch, unless its a timing belt, otherwise there will be slippage, making
> the scale more of a SWAG than accurate.

Yes, 1 inch wide xl 037


That, and the fact that I wasn't aware the encoders

> output was differential, meant I had to use some under a buck a copy,
> rs485 gismo's setup as receivers only to convert the differential to TTL
> for the 5i25 inputs. But to get to the 5i25, I had to rip out the opto's
> in the BoB's input for the A/B inputs.
>


My encoders are completely unknown to me, and I cannot find much
> information on them.
>

I have 2 Bob's that are just vias directly to db 25. I may have to make a
50 pin adapter for them .
I don't know yet if I will have to buffer the signal, how did you know you
needed to ?

>
>
>
> > is there any reason I should choose a higher count?
>  Way more than that is well tolerated now.  If you are 1/1, then I
> wouldn't consider an encoder under 1000 lines, your PID tuning will be
> much much easier.
>
>
>
> 1/2" 13 tpi, is a tall order. How much torque is available?

620 oz inches , 3.2 kw at this time.
I have my eye on a bldc at 37.5 n/m at 5kw if I can scratch up the funds.
Rigid tapping isn't an immediate issue but would be nice.

>
>
> Gripping the tap in a collet, R8 or a TTS ER-20 without the positive
> coupling, tightly enough to prevent slippage.
>


I have a bt30 spindle in the works.

I hope soon to have this thing reassembled, it's been a couple months since
it made chips. Thank you for your advice.
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 7i90 not listed in pnc

[Gene Heskett]

On Tuesday 22 May 2018 20:36:45 jeremy youngs wrote:

> On Tue, May 22, 2018, 18:05 andy pugh  wrote:
> > halrun
> >
> > loadrt hostmot2
> > loadrt hm2_7i90 ... and the rest of the line ...
> >
> > loadrt threads
> > addf hm2_7i90.read thread1
> > addf hm2_7i90.write thread1
> > start
> >
> > setp hm2_7i90.0.pwmgen.00.value 1 (use tab completion, I am guessing
> > the pin names)
>
> Pretty sure I can follow that, and match the pins to my printed off
> configuration
>
> > > I also need to determine encoder count
> >
> > loadusr halmeter pin hm2_7i90.0.encoder.00.rawcounts
> >
> > Then twirl the shaft and look at the counts.
> >
> > Thank you very much , I hope to get this far along in the next day
> > or two.
>
> Once the counts are known my next question will be how to edit the Hal
> file to match what I have. I will refer to the dead tree copy of Hal
> tutorial and try to work through it first though.
>
> I have to purchase an encoder for my spindle motor, I'm thinking a 256
> count should suffice , spindle is 1:1 belt driven,

Ouch, unless its a timing belt, otherwise there will be slippage, making 
the scale more of a SWAG than accurate. Your situation is different from 
mine as there is a gearset for the heads gear shift between the motor 
and the spindle. Because of the gear ratio, and the encoder being a 1000 
line encoder, my high gear "scale" is 7161.61, and low gear is even more 
outragious at 14095. That, and the fact that I wasn't aware the encoders 
output was differential, meant I had to use some under a buck a copy, 
rs485 gismo's setup as receivers only to convert the differential to TTL 
for the 5i25 inputs. But to get to the 5i25, I had to rip out the opto's 
in the BoB's input for the A/B inputs.

> is there any reason I should choose a higher count?

Yes, I found the way high counts also gave me the ability to regulate the 
spindle speed so well that I hear zero slowdown as the load goes up, my 
only clue being the PMDC motors squeak as the current limit in the 
pwm-servo amp kicks in at about 16 amps, around 160% of full scale for 
that motor. My previous encoder, on the spindle itself, had 268 for a 
scale, and the coarse granularity drove the PID unstable at Pgains above 
3. Way more than that is well tolerated now.  If you are 1/1, then I 
wouldn't consider an encoder under 1000 lines, your PID tuning will be 
much much easier.

> I would like to rigid tap 
> eventually , mostly to put 1/2 13 thread into my subplate as I really
> don't want to do 400 of them by hand. High performance is not
> necessary here but accuracy is. Thank you

1/2" 13 tpi, is a tall order. How much torque is available?  With only a 
1 hp motor geared nominally 7/1 or 14/1, I'd have to wrap the g33.1 up 
in a peck routine, probably advancing 1/4 turn or less per peck. 
Although I have not yet mounted that big a tap in one of the tap hats 
we've been discussing. So that test has not been made. Yet.

Gripping the tap in a collet, R8 or a TTS ER-20 without the positive 
coupling, tightly enough to prevent slippage, which will ruin the 
thread, isn't possible. These tap hats have a cap screw in the side that 
engages a notch cut in the bottom of an R8, locking the hat to the R8, 
and the hat has 4 grub screws to grip the square flats on the butt end 
of the tap. Its either going to stay in time, rip the grub screws out, 
threads and all, or twist the tap off.

-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 7i90 not listed in pnc

[jeremy youngs]

On Tue, May 22, 2018, 18:05 andy pugh  wrote:

>
> halrun
>
> loadrt hostmot2
> loadrt hm2_7i90 ... and the rest of the line ...
>
> loadrt threads
> addf hm2_7i90.read thread1
> addf hm2_7i90.write thread1
> start
>
> setp hm2_7i90.0.pwmgen.00.value 1 (use tab completion, I am guessing
> the pin names)
>
Pretty sure I can follow that, and match the pins to my printed off
configuration

>
> > I also need to determine encoder count
>
> loadusr halmeter pin hm2_7i90.0.encoder.00.rawcounts
>
> Then twirl the shaft and look at the counts.
>
> Thank you very much , I hope to get this far along in the next day or two.
>
Once the counts are known my next question will be how to edit the Hal file
to match what I have. I will refer to the dead tree copy of Hal tutorial
and try to work through it first though.

I have to purchase an encoder for my spindle motor, I'm thinking a 256
count should suffice , spindle is 1:1 belt driven , is there any reason I
should choose a higher count? I would like to rigid tap eventually , mostly
to put 1/2 13 thread into my subplate as I really don't want to do 400 of
them by hand. High performance is not necessary here but accuracy is.
Thank you
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[Gene Heskett]

On Tuesday 22 May 2018 18:58:52 Chris Albertson wrote:

> Try this one, sorry it moved.
>
> http://www.cds.caltech.edu/~murray/books/AM08/pdf/am08-complete_22Feb0
>9.pdf
>
> -- Chris

That worked. Pretty well covers feedback systems from a cursory read, 
less technical but still complete. It looks like I ought to add it to my 
dead tree library, after I put some fresh ink in my big Brother, its got 
two nearly empty tanks ATM. And pick up another ream of good paper and a 
2" D ring binder.

Thanks Chris.

> On Tue, May 22, 2018 at 3:04 PM, Gene Heskett  
wrote:
> > On Tuesday 22 May 2018 12:10:47 Chris Albertson wrote:
> >> I just got a copy of "Feedback Systems - An Introduction for
> >> Scientists and Engineers, Karl Johan °Aström and Richard M. Murray"
> >>
> >> The book is FREE and pdf format at
> >> http://www.cds.caltech.edu/∼murray/amwiki
> >
> > I, like others are reporting Chris, am getting a not found error,
> > and I suspect its the character in the little box in front of the
> > "murray" string above. Its reproduced as a small "1/4" symbol here
> > in the browsers (palemoon) font, and I don't think it belongs in a
> > url address. Swag: caltecs way of establishing copyright?
> >
> >> It is an actual set book that seem to be exactly what anyone
> >> working in this field needs to know.  I'd say it is not is
> >> mathematically oriented as a real university control theory book
> >> and certainly well above the hobby level.  It assume some know age
> >> of Calculus but not a lot.  Pretty much what the title says.   But
> >> it covers feedback, not just motion control. so things like op-amps
> >> are covered.   But you need to know that to if designing control
> >> circuits
> >>
> >> I'm using stepper motors but already have DRO scales.   It seems
> >> that I should be able to combine open loop counting with closed
> >> loop linear sensors.
> >>
> >> Suggestion above were good, thanks, because they contain a solution
> >> I had not thought of, drive "I" with different data than "PD"
> >>
> >> On Tue, May 22, 2018 at 6:44 AM, John Kasunich
> >> 
> >
> > wrote:
> >> > On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:
> >> >> I looked at the unit.   It seems like a good idea.  It uses
> >> >> inputs from both the motor's shaft encoders and another encoder
> >> >> on the object that is being moved, like a linear encoder on the
> >> >> table.
> >> >>
> >> >> Question:  Let's say I wanted to do this myself.   Is there a
> >> >> method that "everyone" in the machine tool industry uses for
> >> >> combining the reading of multiple encoders?  If not it seems
> >> >> like the perfect application for a Kalman filter.
> >> >>
> >> >> But maybe you don't combine them but use the linear DRO for
> >> >> position loop and the motor shaft encoders for velocity.
> >> >>
> >> >> This is a common problem I think with robot arms.  The joint has
> >> >> an angle sensor but the motor has a shaft encoder.   So the
> >> >> control loops might be nested.
> >> >>
> >> >> It reminds my the old saying the "A man with a watch knows what
> >> >> time it is, a man with two watches is never sure of the time."
> >> >
> >> > We did something like this several years ago at Stuart's shop in
> >> > Wichita, on a big Giddings and Lewis boring mill.
> >> >
> >> > We used two PID loops, with their outputs summed.  The position
> >> > command went to both loops.  The feedback for one loop came from
> >> > the motor encoder, and the feedback for the other loop came from
> >> > the linear scale.
> >> >
> >> > The motor loop was tuned as normal, except that the I-gain was
> >> > kept at zero. The linear scale loop was tuned using ONLY I-gain. 
> >> > So the linear scale loop corrected the fairly small steady-state
> >> > errors due to things like the lead screw heating up (10 foot long
> >> > screw, it adds up).  It also compensates for backlash in the
> >> > screw, and if there is much of that it leads to disturbances on
> >> > direction reversal.  Isn't going to fix a clapped out machine,
> >> > but can improve the accuracy of a tight machine.
> >> >
> >> >
> >> > --
> >> >   John Kasunich
> >> >   jmkasun...@fastmail.fm
> >> >
> >> > -
> >> >--- -- Check out the vibrant tech community on one of the
> >> > world's most engaging tech sites, Slashdot.org!
> >> > http://sdm.link/slashdot
> >> > ___
> >> > 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)
> > Genes Web page 
> >
> > 
> >-- Check out the 

Re: [Emc-users] 7i90 not listed in pnc

[andy pugh]

On 22 May 2018 at 21:54, jeremy youngs  wrote:

> The question is how do I go about bench testing this system? I would like
> to verify pwm outputs before I connect hardware .

You can run HAL from the command line (I do this a lot when testing).
Halrun starts a HAL session, and from then on you can use
tab-completion to help type the commands.

halrun

loadrt hostmot2
loadrt hm2_7i90 ... and the rest of the line ...

loadrt threads
addf hm2_7i90.read thread1
addf hm2_7i90.write thread1
start

setp hm2_7i90.0.pwmgen.00.value 1 (use tab completion, I am guessing
the pin names)

> I also need to determine encoder count

loadusr halmeter pin hm2_7i90.0.encoder.00.rawcounts

Then twirl the shaft and look at the counts.

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[Chris Albertson]

Try this one, sorry it moved.

http://www.cds.caltech.edu/~murray/books/AM08/pdf/am08-complete_22Feb09.pdf

-- Chris



On Tue, May 22, 2018 at 3:04 PM, Gene Heskett  wrote:
> On Tuesday 22 May 2018 12:10:47 Chris Albertson wrote:
>
>> I just got a copy of "Feedback Systems - An Introduction for
>> Scientists and Engineers, Karl Johan °Aström and Richard M. Murray"
>>
>> The book is FREE and pdf format at
>> http://www.cds.caltech.edu/∼murray/amwiki
>>
> I, like others are reporting Chris, am getting a not found error, and I
> suspect its the character in the little box in front of the "murray"
> string above. Its reproduced as a small "1/4" symbol here in the
> browsers (palemoon) font, and I don't think it belongs in a url address.
> Swag: caltecs way of establishing copyright?
>
>> It is an actual set book that seem to be exactly what anyone working
>> in this field needs to know.  I'd say it is not is mathematically
>> oriented as a real university control theory book and certainly well
>> above the hobby level.  It assume some know age of Calculus but not a
>> lot.  Pretty much what the title says.   But it covers feedback, not
>> just motion control. so things like op-amps are covered.   But you
>> need to know that to if designing control circuits
>>
>> I'm using stepper motors but already have DRO scales.   It seems that
>> I should be able to combine open loop counting with closed loop linear
>> sensors.
>>
>> Suggestion above were good, thanks, because they contain a solution I
>> had not thought of, drive "I" with different data than "PD"
>>
>> On Tue, May 22, 2018 at 6:44 AM, John Kasunich 
> wrote:
>> > On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:
>> >> I looked at the unit.   It seems like a good idea.  It uses inputs
>> >> from both the motor's shaft encoders and another encoder on the
>> >> object that is being moved, like a linear encoder on the table.
>> >>
>> >> Question:  Let's say I wanted to do this myself.   Is there a
>> >> method that "everyone" in the machine tool industry uses for
>> >> combining the reading of multiple encoders?  If not it seems like
>> >> the perfect application for a Kalman filter.
>> >>
>> >> But maybe you don't combine them but use the linear DRO for
>> >> position loop and the motor shaft encoders for velocity.
>> >>
>> >> This is a common problem I think with robot arms.  The joint has an
>> >> angle sensor but the motor has a shaft encoder.   So the control
>> >> loops might be nested.
>> >>
>> >> It reminds my the old saying the "A man with a watch knows what
>> >> time it is, a man with two watches is never sure of the time."
>> >
>> > We did something like this several years ago at Stuart's shop in
>> > Wichita, on a big Giddings and Lewis boring mill.
>> >
>> > We used two PID loops, with their outputs summed.  The position
>> > command went to both loops.  The feedback for one loop came from the
>> > motor encoder, and the feedback for the other loop came from the
>> > linear scale.
>> >
>> > The motor loop was tuned as normal, except that the I-gain was kept
>> > at zero. The linear scale loop was tuned using ONLY I-gain.  So the
>> > linear scale loop corrected the fairly small steady-state errors due
>> > to things like the lead screw heating up (10 foot long screw, it
>> > adds up).  It also compensates for backlash in the screw, and if
>> > there is much of that it leads to disturbances on direction
>> > reversal.  Isn't going to fix a clapped out machine, but can improve
>> > the accuracy of a tight machine.
>> >
>> >
>> > --
>> >   John Kasunich
>> >   jmkasun...@fastmail.fm
>> >
>> > 
>> >-- Check out the vibrant tech community on one of the world's
>> > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> > ___
>> > 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)
> Genes Web page 
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



-- 

Chris Albertson
Redondo Beach, California

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing 

Re: [Emc-users] Newall DRO integration

[Gene Heskett]

On Tuesday 22 May 2018 12:10:47 Chris Albertson wrote:

> I just got a copy of "Feedback Systems - An Introduction for
> Scientists and Engineers, Karl Johan °Aström and Richard M. Murray"
>
> The book is FREE and pdf format at
> http://www.cds.caltech.edu/∼murray/amwiki
>
I, like others are reporting Chris, am getting a not found error, and I 
suspect its the character in the little box in front of the "murray" 
string above. Its reproduced as a small "1/4" symbol here in the 
browsers (palemoon) font, and I don't think it belongs in a url address. 
Swag: caltecs way of establishing copyright?

> It is an actual set book that seem to be exactly what anyone working
> in this field needs to know.  I'd say it is not is mathematically
> oriented as a real university control theory book and certainly well
> above the hobby level.  It assume some know age of Calculus but not a
> lot.  Pretty much what the title says.   But it covers feedback, not
> just motion control. so things like op-amps are covered.   But you
> need to know that to if designing control circuits
>
> I'm using stepper motors but already have DRO scales.   It seems that
> I should be able to combine open loop counting with closed loop linear
> sensors.
>
> Suggestion above were good, thanks, because they contain a solution I
> had not thought of, drive "I" with different data than "PD"
>
> On Tue, May 22, 2018 at 6:44 AM, John Kasunich  
wrote:
> > On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:
> >> I looked at the unit.   It seems like a good idea.  It uses inputs
> >> from both the motor's shaft encoders and another encoder on the
> >> object that is being moved, like a linear encoder on the table.
> >>
> >> Question:  Let's say I wanted to do this myself.   Is there a
> >> method that "everyone" in the machine tool industry uses for
> >> combining the reading of multiple encoders?  If not it seems like
> >> the perfect application for a Kalman filter.
> >>
> >> But maybe you don't combine them but use the linear DRO for
> >> position loop and the motor shaft encoders for velocity.
> >>
> >> This is a common problem I think with robot arms.  The joint has an
> >> angle sensor but the motor has a shaft encoder.   So the control
> >> loops might be nested.
> >>
> >> It reminds my the old saying the "A man with a watch knows what
> >> time it is, a man with two watches is never sure of the time."
> >
> > We did something like this several years ago at Stuart's shop in
> > Wichita, on a big Giddings and Lewis boring mill.
> >
> > We used two PID loops, with their outputs summed.  The position
> > command went to both loops.  The feedback for one loop came from the
> > motor encoder, and the feedback for the other loop came from the
> > linear scale.
> >
> > The motor loop was tuned as normal, except that the I-gain was kept
> > at zero. The linear scale loop was tuned using ONLY I-gain.  So the
> > linear scale loop corrected the fairly small steady-state errors due
> > to things like the lead screw heating up (10 foot long screw, it
> > adds up).  It also compensates for backlash in the screw, and if
> > there is much of that it leads to disturbances on direction
> > reversal.  Isn't going to fix a clapped out machine, but can improve
> > the accuracy of a tight machine.
> >
> >
> > --
> >   John Kasunich
> >   jmkasun...@fastmail.fm
> >
> > 
> >-- Check out the vibrant tech community on one of the world's
> > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > ___
> > 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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Water or Air cooled spindles

[Bruce Layne]

I've had two of the Chinese 2.2 KW water cooled spindles for the last 
few years and have had no trouble with them.  I consider them to be a 
good value.  Well worth the plumbing hassle, in my opinion.


Be sure to use good quality very flexible tubing of the correct size.  I 
think I got 8mm outside diameter tubing from McMaster-Carr (red for 
supply and blue for return).  I use pink RV antifreeze as the coolant.  
It's used full strength and not diluted.  I use it in the hope that it's 
less corrosive than water.  Neither machine's coolant has had any rust 
or other issues, although there was a slight film of oil that's flushed 
out of the spindle motor.  I'm not worried about it freezing because one 
of the CNC routers is in an attached garage and the other is in my 
basement, and neither get very cold.  I'd actually be more worried about 
the machine rusting if it was in a condensing environment, and the CNC 
routers are mostly aluminum.


I mounted a thin liquid crystal thermometer on the spindle facing the 
operator so I can tell at a glance if the spindle overheats. These 
liquid crystal strip thermometers are readily available on eBay and are 
sold for reptile terrariums.  I buy a bunch of them and put them in 
electrical panels, etc.


I haven't finished wiring it yet, but the production machine will have a 
130F bimetallic button thermal switch siliconed to the spindle motor 
housing and wired into the e-stop circuit to shut everything down if the 
spindle overheats.


The Huanyang VFD produces a lot of electrical noise, apparently mostly 
radiated.  I used ultra flexible shielded four conductor cable (three 
phases plus ground) to keep the cable from radiating much energy.  The 
only place I had an RFI problem was the VGA monitor and a good quality 
VGA cable fixed that problem.


On the larger router, I tried to place a five gallon coolant tank under 
the router and pump the coolant up and then back down to the spindle 
motor.  I was partly motivated by not wanting a leak that siphoned the 
five gallon coolant tank empty.  After some experimentation, I gave up 
and put the coolant tank on top of the CNC router enclosure.  When it 
was underneath, I needed to use such a large pump to have enough 
pressure to pump the coolant six feet vertically that the submerged 
coolant pump was heating the coolant more than the spindle.  It was a 
spindle heater, not a spindle cooler.


We're finally ramping up production, with some programs running 
unattended for ten hours.  I'm going to need to add a radiator in the 
coolant return line and a couple of muffin fans to keep the coolant 
temperature low enough.  The other alternative might be ten or fifteen 
more gallons of coolant to increase the thermal mass, but that seems to 
only delay the overheating problem with greater risk of a severe coolant 
leak.







On 05/22/2018 03:18 PM, Roland Jollivet wrote:

I'm looking at getting one of those 2.2kW air or water cooled spindles +
VFD kits out there for a router.
I'm not worried about the noise difference between the two types.

Has anyone taken apart a water cooled spindle?
How are they doing the cooling? Is it a water jacket or just some copper
tubing inside?
How likely is it to leak in a few months time?
I can only find one video describing leaks and water related shorts etc.
(so they must be good?)

I actually prefer the idea of using air-cooled and making ducting to take
the exhaust away from the spindle nose. Make a closed loop fan-assisted air
duct.
The irony is that I want to use flood cooling on the work. (composite
material)  So it won't be a dry environment.

Regards
Roland
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users





--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 7i90 not listed in pnc

[jeremy youngs]

7i42 in hand, 7i90 configuration built. Linuxcnc starts, and no real time
error , that's awesome.
50 pin cables and connectors assembled. All hooked up, all power on.

Mesaflash readhm2id gives all pins and their assignment. So I presume that
is all as it should be.
I printed a copy if this report for hookup purposes.

When I start lcnc and move it I get a following error ( obviously this is
because nothing is hooked up) .
But as I cannot use pncconfig to test this setup ( because pnc 5i20 doesn't
match the 7i90 and 5i20 is what I lied and said it was )
The question is how do I go about bench testing this system? I would like
to verify pwm outputs before I connect hardware .
I also need to determine encoder count , which I figured I would hookup
encoders and use Hal scope then put my 18 inch degree wheel on the motor ,
put a marker on it and spin one turn , this should get me fairly close .
All input welcome and thank you for your help.
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Water or Air cooled spindles

[Ken Strauss]

Don't you use a rust inhibitor in the water (or use antifreeze) for spindle
cooling? Freezing could be an issue if your CNC is in a sometimes heated
garage or even in an industrial building.

> -Original Message-
> From: Nicklas Karlsson [mailto:nicklas.karlsso...@gmail.com]
> Sent: Tuesday, May 22, 2018 3:24 PM
> To: Enhanced Machine Controller (EMC)
> Subject: Re: [Emc-users] Water or Air cooled spindles
>
> There is a huge diffence in heat capacity between air and water. Water
make
> me think about rust.
>
> On Tue, 22 May 2018 21:18:34 +0200
> Roland Jollivet  wrote:
>
> > I'm looking at getting one of those 2.2kW air or water cooled spindles +
> > VFD kits out there for a router.
> > I'm not worried about the noise difference between the two types.
> >
> > Has anyone taken apart a water cooled spindle?
> > How are they doing the cooling? Is it a water jacket or just some copper
> > tubing inside?
> > How likely is it to leak in a few months time?
> > I can only find one video describing leaks and water related shorts etc.
> > (so they must be good?)
> >
> > I actually prefer the idea of using air-cooled and making ducting to
take
> > the exhaust away from the spindle nose. Make a closed loop fan-assisted
air
> > duct.
> > The irony is that I want to use flood cooling on the work. (composite
> > material)  So it won't be a dry environment.
> >
> > Regards
> > Roland
> >

--
> > Check out the vibrant tech community on one of the world's most
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > ___
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
>
>

--
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[MC Cason via Emc-users]

On 05/22/2018 02:02 PM, Peter Blodow wrote:
Sorry, Chris, I get a "Not Found"-Error with your caltech link without 
any further comment.

Peter


  It must be a code page error.  It worked after I deleted the tilde, 
and then added it back in.

http://www.cds.caltech.edu/~murray/amwiki/index.php?title=Main_Page

--
MC Cason
Eagle3D - github.com/mcason/Eagle3D
Lathe Electronic Edge Finder - raccoonelectronics.com



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Water or Air cooled spindles

[Nicklas Karlsson]

There is a huge diffence in heat capacity between air and water. Water make me 
think about rust.

On Tue, 22 May 2018 21:18:34 +0200
Roland Jollivet  wrote:

> I'm looking at getting one of those 2.2kW air or water cooled spindles +
> VFD kits out there for a router.
> I'm not worried about the noise difference between the two types.
> 
> Has anyone taken apart a water cooled spindle?
> How are they doing the cooling? Is it a water jacket or just some copper
> tubing inside?
> How likely is it to leak in a few months time?
> I can only find one video describing leaks and water related shorts etc.
> (so they must be good?)
> 
> I actually prefer the idea of using air-cooled and making ducting to take
> the exhaust away from the spindle nose. Make a closed loop fan-assisted air
> duct.
> The irony is that I want to use flood cooling on the work. (composite
> material)  So it won't be a dry environment.
> 
> Regards
> Roland
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Water or Air cooled spindles

[Roland Jollivet]

I'm looking at getting one of those 2.2kW air or water cooled spindles +
VFD kits out there for a router.
I'm not worried about the noise difference between the two types.

Has anyone taken apart a water cooled spindle?
How are they doing the cooling? Is it a water jacket or just some copper
tubing inside?
How likely is it to leak in a few months time?
I can only find one video describing leaks and water related shorts etc.
(so they must be good?)

I actually prefer the idea of using air-cooled and making ducting to take
the exhaust away from the spindle nose. Make a closed loop fan-assisted air
duct.
The irony is that I want to use flood cooling on the work. (composite
material)  So it won't be a dry environment.

Regards
Roland
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[Peter Blodow]

Sorry, Chris, I get a "Not Found"-Error with your caltech link without 
any further comment.

Peter

Am 22.05.2018 um 18:10 schrieb Chris Albertson:

I just got a copy of "Feedback Systems - An Introduction for
Scientists and Engineers, Karl Johan °Aström and Richard M. Murray"

The book is FREE and pdf format at http://www.cds.caltech.edu/∼murray/amwiki

It is an actual set book that seem to be exactly what anyone working
in this field needs to know.  I'd say it is not is mathematically
oriented as a real university control theory book and certainly well
above the hobby level.  It assume some know age of Calculus but not a
lot.  Pretty much what the title says.   But it covers feedback, not
just motion control. so things like op-amps are covered.   But you
need to know that to if designing control circuits

I'm using stepper motors but already have DRO scales.   It seems that
I should be able to combine open loop counting with closed loop linear
sensors.

Suggestion above were good, thanks, because they contain a solution I
had not thought of, drive "I" with different data than "PD"

On Tue, May 22, 2018 at 6:44 AM, John Kasunich  wrote:


On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:

I looked at the unit.   It seems like a good idea.  It uses inputs from
both the motor's shaft encoders and another encoder on the object that is
being moved, like a linear encoder on the table.

Question:  Let's say I wanted to do this myself.   Is there a method that
"everyone" in the machine tool industry uses for combining the reading of
multiple encoders?  If not it seems like the perfect application for a
Kalman filter.

But maybe you don't combine them but use the linear DRO for position loop
and the motor shaft encoders for velocity.

This is a common problem I think with robot arms.  The joint has an angle
sensor but the motor has a shaft encoder.   So the control loops might be
nested.

It reminds my the old saying the "A man with a watch knows what time it is,
a man with two watches is never sure of the time."


We did something like this several years ago at Stuart's shop in Wichita, on a 
big Giddings and Lewis boring mill.

We used two PID loops, with their outputs summed.  The position command went to 
both loops.  The feedback for one loop came from the motor encoder, and the 
feedback for the other loop came from the linear scale.

The motor loop was tuned as normal, except that the I-gain was kept at zero.
The linear scale loop was tuned using ONLY I-gain.  So the linear scale loop 
corrected the fairly small steady-state errors due to things like the lead 
screw heating up (10 foot long screw, it adds up).  It also compensates for 
backlash in the screw, and if there is much of that it leads to disturbances on 
direction reversal.  Isn't going to fix a clapped out machine, but can improve 
the accuracy of a tight machine.


--
   John Kasunich
   jmkasun...@fastmail.fm

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users






--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[Chris Albertson]

I just got a copy of "Feedback Systems - An Introduction for
Scientists and Engineers, Karl Johan °Aström and Richard M. Murray"

The book is FREE and pdf format at http://www.cds.caltech.edu/∼murray/amwiki

It is an actual set book that seem to be exactly what anyone working
in this field needs to know.  I'd say it is not is mathematically
oriented as a real university control theory book and certainly well
above the hobby level.  It assume some know age of Calculus but not a
lot.  Pretty much what the title says.   But it covers feedback, not
just motion control. so things like op-amps are covered.   But you
need to know that to if designing control circuits

I'm using stepper motors but already have DRO scales.   It seems that
I should be able to combine open loop counting with closed loop linear
sensors.

Suggestion above were good, thanks, because they contain a solution I
had not thought of, drive "I" with different data than "PD"

On Tue, May 22, 2018 at 6:44 AM, John Kasunich  wrote:
>
>
> On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:
>> I looked at the unit.   It seems like a good idea.  It uses inputs from
>> both the motor's shaft encoders and another encoder on the object that is
>> being moved, like a linear encoder on the table.
>>
>> Question:  Let's say I wanted to do this myself.   Is there a method that
>> "everyone" in the machine tool industry uses for combining the reading of
>> multiple encoders?  If not it seems like the perfect application for a
>> Kalman filter.
>>
>> But maybe you don't combine them but use the linear DRO for position loop
>> and the motor shaft encoders for velocity.
>>
>> This is a common problem I think with robot arms.  The joint has an angle
>> sensor but the motor has a shaft encoder.   So the control loops might be
>> nested.
>>
>> It reminds my the old saying the "A man with a watch knows what time it is,
>> a man with two watches is never sure of the time."
>>
>
> We did something like this several years ago at Stuart's shop in Wichita, on 
> a big Giddings and Lewis boring mill.
>
> We used two PID loops, with their outputs summed.  The position command went 
> to both loops.  The feedback for one loop came from the motor encoder, and 
> the feedback for the other loop came from the linear scale.
>
> The motor loop was tuned as normal, except that the I-gain was kept at zero.
> The linear scale loop was tuned using ONLY I-gain.  So the linear scale loop 
> corrected the fairly small steady-state errors due to things like the lead 
> screw heating up (10 foot long screw, it adds up).  It also compensates for 
> backlash in the screw, and if there is much of that it leads to disturbances 
> on direction reversal.  Isn't going to fix a clapped out machine, but can 
> improve the accuracy of a tight machine.
>
>
> --
>   John Kasunich
>   jmkasun...@fastmail.fm
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



-- 

Chris Albertson
Redondo Beach, California

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] tap hats again

[Roland Jollivet]

Re. the video, clever, but messy.

You always know where the back of the vice is, so;
Use a caliper to measure the stock, or the vice gap, and move forward half
that amount?




On 22 May 2018 at 15:30, Gene Heskett  wrote:

> On Tuesday 22 May 2018 08:12:51 andy pugh wrote:
>
> > On 22 May 2018 at 09:55, Gene Heskett  wrote:
> > > I don't think I've got one of those seesaws,
> >
> > Steel rule.
> >
> I still don't have one. Micrometer, dial indicator, or digital caliper do
> it here.
>
> > https://www.youtube.com/watch?v=m3WWoQDGO2c
>
> Some error, the body of that chuck he has clamped the gage to, is NOT the
> bore holding the tap. Its motion will be the centerline of the spindle.
> Jaw wear may not be holding the tap dead on.  Not to mention that the
> tip of the inverted tap he used is forged, not precision ground, and
> could easily be off center by more than that. I have some in my
> collection that are at least 20 thou off.  Flea market stuff I have to
> sharpen everytime I use them.
>
> Theres right, good enough for the girls we go with, and wrong.
>
> His chances of being wrong are not quite zero but is good enough.  My use
> of the g38.2 to find locations has worked well for over 15 years as long
> as the screws are good. So I'd have to call it good enough for the girls
> I've gone with. :) Its allowed me to drill halfway thru a pcb, turn the
> pcb over and finish the hole from the other side with no visible offsets
> where they meet in the middle of the boards thickness.
>
> --
> 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)
> Genes Web page 
>
> 
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Newall DRO integration

[John Kasunich]

On Fri, May 11, 2018, at 11:16 PM, Chris Albertson wrote:
> I looked at the unit.   It seems like a good idea.  It uses inputs from
> both the motor's shaft encoders and another encoder on the object that is
> being moved, like a linear encoder on the table.
> 
> Question:  Let's say I wanted to do this myself.   Is there a method that
> "everyone" in the machine tool industry uses for combining the reading of
> multiple encoders?  If not it seems like the perfect application for a
> Kalman filter.
> 
> But maybe you don't combine them but use the linear DRO for position loop
> and the motor shaft encoders for velocity.
> 
> This is a common problem I think with robot arms.  The joint has an angle
> sensor but the motor has a shaft encoder.   So the control loops might be
> nested.
> 
> It reminds my the old saying the "A man with a watch knows what time it is,
> a man with two watches is never sure of the time."
> 

We did something like this several years ago at Stuart's shop in Wichita, on a 
big Giddings and Lewis boring mill.

We used two PID loops, with their outputs summed.  The position command went to 
both loops.  The feedback for one loop came from the motor encoder, and the 
feedback for the other loop came from the linear scale.

The motor loop was tuned as normal, except that the I-gain was kept at zero.
The linear scale loop was tuned using ONLY I-gain.  So the linear scale loop 
corrected the fairly small steady-state errors due to things like the lead 
screw heating up (10 foot long screw, it adds up).  It also compensates for 
backlash in the screw, and if there is much of that it leads to disturbances on 
direction reversal.  Isn't going to fix a clapped out machine, but can improve 
the accuracy of a tight machine.  


-- 
  John Kasunich
  jmkasun...@fastmail.fm

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Gene Heskett]

On Tuesday 22 May 2018 08:12:51 andy pugh wrote:

> On 22 May 2018 at 09:55, Gene Heskett  wrote:
> > I don't think I've got one of those seesaws,
>
> Steel rule.
>
I still don't have one. Micrometer, dial indicator, or digital caliper do 
it here.

> https://www.youtube.com/watch?v=m3WWoQDGO2c

Some error, the body of that chuck he has clamped the gage to, is NOT the 
bore holding the tap. Its motion will be the centerline of the spindle. 
Jaw wear may not be holding the tap dead on.  Not to mention that the 
tip of the inverted tap he used is forged, not precision ground, and 
could easily be off center by more than that. I have some in my 
collection that are at least 20 thou off.  Flea market stuff I have to 
sharpen everytime I use them.

Theres right, good enough for the girls we go with, and wrong.

His chances of being wrong are not quite zero but is good enough.  My use 
of the g38.2 to find locations has worked well for over 15 years as long 
as the screws are good. So I'd have to call it good enough for the girls 
I've gone with. :) Its allowed me to drill halfway thru a pcb, turn the 
pcb over and finish the hole from the other side with no visible offsets 
where they meet in the middle of the boards thickness.

-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Andy Pugh]

On 22 May 2018, at 13:39, Gene Heskett  wrote:

>> Steel rule.
>> 
> 
> Gotcha, aka feeler blade,

> https://www.youtube.com/watchv=m3WWoQDGO2c

No. Watch the video. 
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Gene Heskett]

On Tuesday 22 May 2018 08:12:51 andy pugh wrote:

> On 22 May 2018 at 09:55, Gene Heskett  wrote:
> > I don't think I've got one of those seesaws,
>
> Steel rule.
>
Gotcha, aka feeler blade, or find the sides with g38.2's and find the 
center with a calculation. I like the last better as that can split a mm 
into fine pieces. And I'd only have to do it once if recorded.

> https://www.youtube.com/watch?v=m3WWoQDGO2c



-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Gene Heskett]

On Tuesday 22 May 2018 07:06:16 Erik Christiansen wrote:

> On 22.05.18 04:55, Gene Heskett wrote:
> > On Tuesday 22 May 2018 00:51:08 Erik Christiansen wrote:
> > > On 21.05.18 22:19, Gene Heskett wrote:
> >
> > Thanks Erik. Writing about this makes me think of better ways to do
> > it.
>
> Yep, thoughts can stay in a groove if not nudged. All that I was
> referring to with the "seesaw" is: Grab a little 6" steel rule, pinch
> the middle of it between a turned pointy bit of rod in the spindle's
> chuck and the top of the horizontal tap hat. When the table is wound
> either side of centre, the rule will tilt like a seesaw, being
> horizontal only when the point is centered. It's most sensitive with a
> reasonably sharp point. The method won't split a thou, but I have no
> way of lining the DTI up with the spindle centre, and its reading
> won't split a thou on the top of the cylinder, either.
>
> Erik
A variation on that is a contact in an insulator in the mill spindle. Go 
low enough to hit it to one side, run the spindle at 500-1k revs so the 
wobble in the conact wire is equalized, and do a g38.2, first fast, then 
back away a couple thou, run the g38.2 again and record the contact. 
Pick it up enough to clear the top and run it to the other side and back 
down to the same z, the repeat the g38.2 sequence running y in the 
opposite direction, recording the 2nd slow contact. Subtract one from 
the other, and /2. That I can do right in the code, but its yet another 
tool change that ought to be done as the hole is being drilled. but I 
think it would be accurate enough for each 90 degrees the table is 
driven. Do that and record it, then sweep that spot with a 1/4" mill to 
make a flat spot so the drill will settle quickly on first contact, 
drill the hole. The flat would be a consistent height, so the drill can 
be brought to -0.002" or so, pause while it centers, then peck to depth.
Turn table to 90, set new y from previous table, repeat the drill cycle. 
Turn table to 180, wash rinse and repeat, ditto for 270. Run the table 
back to -5, then to 0 to tske up the tables horrible backlash, change 
the tool to the tap and tap all 4 holes. Done. Mount the tap you sized 
the center hole for, add and tighten the grub screws, and stamp the size 
and tpmm on the face of the brass so it can be easily read while still 
sitting in the storage box. Drop in a pocket of the storage box I've 
already made. The tap is protected from contact with other taps, further 
dulling it, and it makes a good handle to retrieve it out of the pocket 
its in for storage. I made a box with 100 pockets, the pockets made from 
the square plastic extrusions used for deck rail balusters. All glued 
into one block with plastic solvent glue, and wrapped with 5/16" x 4" 
wide solid oak panel frame. 30 minutes or less per hat made, depending 
on how long I spend searching for the correct sized drill for the center 
hole for the shank of the tap.  Or for the grub screws drill tap...

With 3 common drill sizes standards about, that can be fun, not.

Time to skinny dip, make starter fluid, aka coffee, and take care of the 
missus.

Later & thanks Erik. 

-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[andy pugh]

On 22 May 2018 at 09:55, Gene Heskett  wrote:

> I don't think I've got one of those seesaws,

Steel rule.

https://www.youtube.com/watch?v=m3WWoQDGO2c


-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] No Charge Pump signal

[andy pugh]

On 22 May 2018 at 07:29, John Dammeyer  wrote:

> If I understand the hal file correctly the charge pump code is loaded, the
> estop_out pin and charge-pump.enable are dependent on the user-enable-out
> which is controlled by the ENABLE button on the top left part of the screen.

If you look at charge-pump.enable with Halmeter does it change state?
You could also Halscope parport pin 17, that should indicate if HAL
_thinks_ it is twiddling the pin.

It is possible that the scope might not see anything without pull-up
or pull-down on the pin. P-port pins are a bit funny that way I seem
to recall (I haven't used one for ages)

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Erik Christiansen]

On 22.05.18 04:55, Gene Heskett wrote:
> On Tuesday 22 May 2018 00:51:08 Erik Christiansen wrote:
> 
> > On 21.05.18 22:19, Gene Heskett wrote:
> Thanks Erik. Writing about this makes me think of better ways to do it.

Yep, thoughts can stay in a groove if not nudged. All that I was
referring to with the "seesaw" is: Grab a little 6" steel rule, pinch
the middle of it between a turned pointy bit of rod in the spindle's
chuck and the top of the horizontal tap hat. When the table is wound
either side of centre, the rule will tilt like a seesaw, being
horizontal only when the point is centered. It's most sensitive with a
reasonably sharp point. The method won't split a thou, but I have no way
of lining the DTI up with the spindle centre, and its reading won't
split a thou on the top of the cylinder, either.

Erik

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] No Charge Pump signal

[Gene Heskett]

On Tuesday 22 May 2018 02:29:33 John Dammeyer wrote:

> I've used the stepconf wizard to assign to DB25-17.
>
> The hal file ends up with this:
>
> loadrt charge_pump
> net estop-out charge-pump.enable iocontrol.0.user-enable-out
> net charge-pump <= charge-pump.out
>
> addf charge-pump base-thread
>
> net charge-pump => parport.0.pin-17-out
>
> net estop-out <= iocontrol.0.user-enable-out
> net estop-ext => iocontrol.0.emc-enable-in
>
> A scope on the DB-25 pin 17 does not show any sort of pulse stream.  I
> know the signal is making it from the PC out to the Breakout board
> because if I assign pin 17 to say coolant instead of charge pump I can
> click on the flood on check box and toggle the flood output signal.
>
> If I understand the hal file correctly the charge pump code is loaded,
> the estop_out pin and charge-pump.enable are dependent on the
> user-enable-out which is controlled by the ENABLE button on the top
> left part of the screen. The charge-pump variable is set with
> charge-pump.out
> And the charge-pump variable  is put out to the pin-17.
>
> I can delete all those lines and then run stepconf wizard which
> recreates what I have above.
>
> I think I'm doing everything mentioned here.
> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?About_Charge_Pumps
>
> What's missing to enable charge pump?
>
> Thanks
> John
>
The charge pump has only one output, and I needed two in order to control 
the jig's location bar.  So I fed the charge-pump output to a couple 
and2's and gated them to control the outputs to drive the bucket thingy 
and relays that ran the motors, re-purposing the mist and flood 
functions.

I have since parked that jig, and am now switching a vacuum to collect 
swarf with it.

Use the halscope to check why the signal isn't getting out.

>
> --
> Check out the vibrant tech community on one of the world's
> most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> 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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] tap hats again

[Gene Heskett]

On Tuesday 22 May 2018 00:51:08 Erik Christiansen wrote:

> On 21.05.18 22:19, Gene Heskett wrote:
> > Greetings all;
> >
> > The first "tap hat" is done, and has a 5mmx.8 tap mounted in it.
>
> When the box of 10 different grubscrews arrived, it seemed like a good
> idea to write the size and _pitch_ over each compartment, on the
> transparent lid. That saves time later.
>
> > But I've got to profile how far off center that 3" chuck mounted to
> > my rotary table wobbles, so I can write a short thing to drill the 4
> > holes in the side that hold the grub screws, I must have been an
> > hour and a half piddling with that and actually get the grub screws
> > centered on the tap butt flats.
>
> For the original tap hats, all I used to center on the cylindrical top
> was the old 6" steel rule seesaw, with a pointy tip in the drill
> chuck. When the seesaw's level, it's centered. (Certainly well enough
> to hold the bum end of a tap already aligned in a close-fitting bore,
> I figure.)

I don't think I've got one of those seesaws, but I do have a dial so I 
can watch the movement as it turns, record the offsets and use them in 
the code. That would have to be done in a single touchoff as that table 
has no home switch. And it might be a good idea to sweep the top with an 
1/8" flat nosed mill to make a small flat to drill into. 3 tool changes. 
The drilling is throwing up a ridge around the hole just tall enough to 
interfere with inserting it smoothly into the R8.
>
> > Only about 50 more to go. :(
>
> ISTR that you've made them all out of a bar of ¾" brass, so the one
> centering ought to do for the lot? With an A axis to spin the 90°, you
> can drill 4, then tap 4, for ¼ of the tool changes.

Yes, moving the y a few thou to compensate for the chucks wobble.
>
> Erik
>
7/8" brass rod. More room for longer, or doubled, grub screws. But IIRC, 
I'm going to have to make a tube to hold them far enough away from the 
jaws else the spindle will hit them. Or maybe chuck the rear end of one 
of the modified 7/8" R8 collets, put the brass in the R8 backwards, 
using the side screw as a depth stop, and put the tables tailstock 
against the central hole. That will leave the body of the rod projecting 
and far enough from the chuck jaws that everything clears. It would 
improve the grip if I had an R8 socket and a drawbar to close the R8, 
but with the tailstock supporting the end, it ought to work. First thing 
I'll try today when I get to the garage.

Thanks Erik. Writing about this makes me think of better ways to do it.

-- 
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)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Very low backlash ballscrews [Was: Re: How does this work?

[Marcus Bowman]

I agree:

> It is not a quarter as hard to repack a ball nut as a lot of people make it
> out to be.
> If you can repack the bearings on your bicycle, you will have no problem.
> You can probably find good general instructions for repacking a ball nut in
> the steering section of your autos service manual.
> 
> If you are not careful the balls can get pushed out of place when putting
> the nut back on the screw.
> If this is a problem, you can turn a plastic mandrel with an OD just shy of
> the Minor Diameter of the screw.
> Insert this into the nut and use it to transfer the nut to the screw.

If you have to cut the screw to machine to length, the additional section of 
screw can be used to hold the balls in place. If both ends of screw and 
cut-off-screw-end are reasonably flat and square, rotating the cut off part 
lets you line up the tracks, and you can wind the screw out as the cut-off part 
is held against the end as though it was one continuous length, so you end up 
with the screw free, and the cut-off part inside the nut, acting as a mandrel. 
I can do that on two of my axes. The third requires a plain mandrel as Kurt 
suggests. 
> 
> Those cheap ball nuts are nothing to be afraid of. They are dead simple and
> not much you can do wrong.
> You will almost certainly find that you nuts run much more smoothly after
> repacking them.
> And for me at least, the peace of mind that comes from knowing the nuts are
> good is well worth the 10 minutes it takes to check them.
> 

> Also, if you have a screw/nut that is oversized, it is fairly easy to get
> oversized balls to take up the slack.
> 
When a ball nut has been in use for so long that the backlash has increased, 
oversize balls are one way of returning it to a more reasonable fit. My 
supplier holds large stocks of balls of different sizes fro that purpose.
One other point:
I use THK nuts which are like sections of rectangular bar. They have a flat 
bottom with 4 tapped holes for securing the nut to a flat surface (so the 
flange is effectively at right angles to the ones shown in most catalogues. 
Sadly, that means they are in short supply, and therefore not cheap. They are, 
however, very easy to fit on the saddle of a mill. I use a little platform with 
a turned spigot to fit the usual hole provided for a standard leadscrew nut, 
allowing some rotation for self-alignment.
See:
https://tech.thk.com/en/products/pdf/en_a15_282.pdf
square nuts type BNT.

Marcus
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] No Charge Pump signal

[John Dammeyer]

I've used the stepconf wizard to assign to DB25-17.

The hal file ends up with this:

loadrt charge_pump
net estop-out charge-pump.enable iocontrol.0.user-enable-out
net charge-pump <= charge-pump.out

addf charge-pump base-thread

net charge-pump => parport.0.pin-17-out

net estop-out <= iocontrol.0.user-enable-out
net estop-ext => iocontrol.0.emc-enable-in

A scope on the DB-25 pin 17 does not show any sort of pulse stream.  I know
the signal is making it from the PC out to the Breakout board because if I
assign pin 17 to say coolant instead of charge pump I can click on the flood
on check box and toggle the flood output signal.

If I understand the hal file correctly the charge pump code is loaded, the
estop_out pin and charge-pump.enable are dependent on the user-enable-out
which is controlled by the ENABLE button on the top left part of the screen.
The charge-pump variable is set with charge-pump.out
And the charge-pump variable  is put out to the pin-17.

I can delete all those lines and then run stepconf wizard which recreates
what I have above.

I think I'm doing everything mentioned here.
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?About_Charge_Pumps

What's missing to enable charge pump?

Thanks
John



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users