n. It's like the
difference between being a programmer/hacker as opposed to someone who
uses a PC to update their Facebook status.
Back on topic If I was Rick, I'd find SOME way to put an encoder on
the spindle, even a low resolution encoder, and avoid patching LinuxCNC
on the sof
The metalcutting ones are not center cutting, they are roundover only. You
could probably get away with a shallow helical ramp in on wood, otherwise order
a center cutting end mill 1/16 or closer if you can find one, cut the pockets
to clearance height for the roundover, do the roundover and the
The VFD I'm using uses three wire control - FWD, REV, RUN/STOP, where the MESA
card is designed around FWD/REV and RUN/STOP two wire control, however the VFD
can set inputs for active high or low operation so you can set FWD for active
high, REV for active low and tie the two together to a sing
That's what I do, in fact I let the air run through for an hour or so to really
drive out moisture. Remember to put some RV antifreeze in the tanks and toilet
bowl as it helps to protect the seals on the valves from becoming brittle and
failing.
--Original Mail--
From: "Gregg Eshelman"
>> The FDA never approved use of Thalidomide in the US.
>> Practically all
>> US cases of birth defects from it were from mothers who were in
>> Europe when it was in widespread use.
> It's been re-approved in some places but only for use after the first
> trimester. I've also read some theory th
Be careful with filters, especially on return lines. If a filter on the
pressure side starts to clog you just loose coolant flow which doesn't hurt the
pump, but if it's in a return line you may find coolant overflowing from your
machine.
--Original Mail--
From: "Viesturs LÄcis"
To:
The typical $100-200 coolant units have a screen in the return inlet to the
tank to catch big chips, and a baffle in the tank so the smaller chips in the
coolant that pass through the return screen will settle to the bottom of the
tank in that chamber while the coolant will flow over the top of
Not when running multiple motors from a single VFD. I've read whitepapers from
VFD manufacturers showing sample configurations with something like eight
motors on a single large VFD, each with it's own OCP. I think the days of loose
motor connections blowing up a VFD are in the past and current
Since nearly all the features are just firmware in the VFD, eliminating them
wouldn't save any money. Every VFD I've worked with comes out of the box
pre-set as simply as you indicate, just connect it to a normal three phase
motor, power and press the run button and you'll get something like a 2
The VFD manufacturers have white papers specifically covering driving multiple
motors from one larger VFD so it's absolutely a supported configuration. You
need separate overcurrent protection for each connected motor though since the
VFD's OCP can't work to protect an individual motor in such a
Perhaps you could tweak the program with a clearance move and optional stop at
the expected insert change points?
--Original Mail--
From: "Jeff Johnson"
To:
Sent: Fri, 21 Nov 2014 09:40:04 -0500
Subject: [Emc-users] Question for larger machine users of LCNC
This is a question for peop
"or some reference plane that matches what you expect your table to be"
--Original Mail--
From: "Mark Wendt"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 19 Nov 2014 09:37:03 -0500
Subject: Re: [Emc-users] Screw Mapping
On Wed, Nov 19, 2014 at 9:29 AM, wrote:
> Usually you slid
Usually you slide the block with the diamond dresser across the table against a
guide, or some reference plane that matches what you expect your table to be.
On a surface grinder you would hold the dresser on your mag chuck and move the
cross feed axis, but there are other setups and some dresse
Perhaps a reducer bushing would be needed to adapt the center hole size.
Precise centering wouldn't be a big deal since you'd be dressing the wheel to
profile in-place inherently centering it. I suspect it would be cheap enough to
be worthwhile even if you still used a correction table to get th
Smaller dia grinding wheels have higher max RPMs, and it looks from the machine
pics like the saw blades are pretty small diameter so a grinding wheel in their
place would be similarly small. The necessary wheel is probably $20 or so. I
also think the offset correction table is probably workable
I'm not a grinding expert, but I believe it would work fine with the correct
selection of grinding wheel type. A quick look on www.nortonindustrial.com
shows many wheel compositions recommended for grinding aluminum. I bet a call
to one of their application engineer folks would get you a good re
How about temporarily replacing one of your saw blades with a grinding wheel,
dressing it to the correct angle to be flat across the width of the vacuum
table and then light grinding passes down the table? Your saw spindles should
be plenty rigid and high enough RPM, and dressing an angle on the
Is there a way to clamp a sander or similar to the head for use in flattening?
A solid clamping and a series of light cuts ought to flatten to match the axis.
--Original Mail--
From: "Mark Wendt"
To: "Enhanced Machine Controller (EMC)"
Sent: Tue, 18 Nov 2014 10:56:10 -0500
Subject: Re:
> I've installed homing switches on mine, but have yet to figure out a good
> way to calibrate the actual cutting tip position when swapping QC holders
> for this and that operations. What I need is another 4 or so tool holders
> so I never have to take a given tool out of its holder, just re-ch
Get the "freezer" zip-locs, they're heavier.
--Original Mail--
From: "Gene Heskett"
To:
Sent: Wed, 12 Nov 2014 11:16:23 -0500
Subject: Re: [Emc-users] Question re using magnetics for work holding
On Wednesday 12 November 2014 10:45:56 p...@wpnet.us did opine
And Gene did reply:
> Put t
Mag chucks are certainly common on things like surface grinders. Not quite the
same as using discrete magnets to stick something non-ferrous to a ferrous mill
table though.
--Original Mail--
From: "Todd Zuercher"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed,
Put them in a zip loc bag, and if they pickup some ferous chips turn the bag
inside out and pull away to remove the chips. Vacuum clamps work with just ~14
PSI of clamping force so I don't see why magnets wouldn't work for light cuts
as well.
--Original Mail--
From: "Gene Heskett"
To:
not under the Estop circuit control.
Alex
Il giorno 08/nov/2014 17:41, ha scritto:
> This is how I have it setup on the CNC lathe refit I've been working on.
> Using 120V on estop only because the relays that came with the machine were
> 120v coils, so use what I've got. PC/monit
This is how I have it setup on the CNC lathe refit I've been working on. Using
120V on estop only because the relays that came with the machine were 120v
coils, so use what I've got. PC/monitor power is the only thing not on the
estop system, just the main machine power switch
Excellent point, the good folks developing and enhancing LinuxCNC hear all our
troubles and complaints here, but probably far too little thanks for their
efforts.
Thank you!
--Original Mail--
From: "Gene Heskett"
To:
Sent: Fri, 7 Nov 2014 12:00:53 -0500
Subject: Re: [Emc-users] $64 q
I tend to think something like this is going to have to be the future of open
source. Volunteers to get a concept to something somewhat useable, generate
critical mass of users and then some sort of funding contribution so the
project can be advanced in a more "professional" manner with people p
Death to evil google. I don't use them for anything. Yes, I have an android
phone, but I have exterminated damned near everything tied to google on it, it
can't even update anything and I like it that way.
--Original Mail--
From: "Dave Cole"
To:
Sent: Thu, 30 Oct 2014 11:24:48 -0500
S
ely.
That's a nice sized lathe, about 11x30. I'd like to be able to get
something like a Denford Orac or a Magnaturn 612 or even a Compact 5 CNC
or PC that needs a refit.
---
This email is free
Movement at last! Now I can get on with axis tuning, then getting the coolant
pump going, etc.
Thanks
Pete C.
--Original Mail--
From: "andy pugh"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 29 Oct 2014 20:53:39 +
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25
is 24V. Is it safe to tie the grounds together?
Yes, its safe but possible undesirable as it makes the PC 5V and field power
ground common so you lose the 24V IO isolation
>
>
> --Original Mail--
> From: "Peter C. Wallace"
> To: "Enhanced Machine Controlle
Never mind, it must not be since it's live when the field power is off.
--Original Mail--
From:
To:
Sent: Wed, 29 Oct 2014 14:55:12 -0500
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
Actually, isn't that 5V output derived from the 24V field power input anyway?
Actually, isn't that 5V output derived from the 24V field power input anyway?
--Original Mail--
From:
To:
Sent: Wed, 29 Oct 2014 14:50:02 -0500
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
The encoder is powered from TB3 pin 21 and 23 since it's 5V and field power is
The encoder is powered from TB3 pin 21 and 23 since it's 5V and field power is
24V. Is it safe to tie the grounds together?
--Original Mail--
From: "Peter C. Wallace"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 29 Oct 2014 12:45:32 -0700 PDT
Subject: Re: [Emc-users] Lathe conver
Both show good in the Tek 'scope with the probes clipped onto the little bit of
wire accessible from the screw terminals on the 7i76. Hard to imagine the wire
clamped in the screw terminal but not making contact. With LinuxCNC running
doesn't it reset that count with each read so it only will sh
Ok, I verified the pin and changed the line in the HAL to:
net mpg_count hm2_5i25.0.7i76.0.0.enc0.count => axis.0.jog-counts
axis.2.jog-counts axis.1.jog-counts
With halcmd I can show hm2_5i25.0.7i76.0.0.enc0.count and see it go to 1
briefly when spinning the MPG, so it looks like the signal is
hm2_5i25.0.encoder.00. is the hardware high speed encoder?
hm2_5i25.0.encoder.01. is not the low speed encoder?
I've been looking through all the HostMot2 docs I can find including the
section on encoders and it's not very clear to me.
http://www.linuxcnc.org/docs/2.6/html/drivers/hostmot2.ht
I think the term brushless was inadvertently use initially. I can't debate the
merits of resolvers vs. encoders, but is anyone building new machines using
resolvers these days? In my mind at least if it has resolvers it's older and
unlikely to have brushless motors.
--Original Mail--
F
Even more confusing for the thread is we are both Petes :)
--Original Mail--
From: "Pete Matos"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 29 Oct 2014 13:18:49 -0400
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
Pete I will take your word on it you are the gur
http://wpnet.us/Harrison2.ini
http://wpnet.us/Harrison2.hal
--Original Mail--
From: "Peter C. Wallace"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 29 Oct 2014 10:04:36 -0700 PDT
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
On Wed, 29 Oct 2014, p...@wpnet.us w
ard set waiting for a conversion on my
much larger CNC mill after I'm done with the lathe. The mill has a Dynapath
control currently that works fine, but has small memory so I just drip feed it
serial from a PC currently.
--Original Mail--
From: "Pete Matos"
To: &qu
It's a Harrison Trainer manual/CNC 280. It previously had a conversion from
it's original dead controls to Microkinetics stuff and now I'm going to
LinuxCNC with the MESA 5i25/7i76 set, retaining the MK stepper drives since
they're fine. I'm doing it as a strictly CNC conversion since I have oth
I've messed with this a bit and don't have it working yet. Attached is my
current hal file hacked together from PNCConf generated bits, bits from docs
and manual bits (don't laugh, I'll clean it up once I get everything working).
The steps are all set to 0.1 just to ensure that even if the selec
All insured I'm sure, just an astronomical premium increase in the works...
--Original Mail--
From: "Pete Matos"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 29 Oct 2014 11:52:16 -0400
Subject: Re: [Emc-users] The Warm Glow of Mach 3
Jeez man that rocket explosion was EPIC!! A s
--Original Mail--
From: "Peter C. Wallace"
To: "Enhanced Machine Controller (EMC)"
Sent: Mon, 27 Oct 2014 14:03:16 -0700 PDT
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
...
> There's a brief mention on page 18 in the mode options
That is indeed brief. Well, I've got
--Original Mail--
From: "Peter C. Wallace"
To: "Enhanced Machine Controller (EMC)"
Sent: Mon, 27 Oct 2014 13:39:46 -0700 PDT
Subject: Re: [Emc-users] Lathe conversion with Mesa 5i25 / 7i76
...
> The 7I76 provides one high speed encoder counter for spindle use and 2 medium
> speed encode
The 7i76 only provides one encoder input since it's a stepper card not a servo
card. Can you point me in the direction of some docs on loading a software
encoder counter and configuring it for MPG use (one MPG, with axis and step
select switches). I'm new to LinuxCNC so it takes me a while to fi
I'm back at work on my lathe conversion, reviewing what I have so far, cleaning
up documentation, etc. I've run into an issue / question, hopefully I'm just
missing something.
I have my spindle encoder connected to the Mesa 7i76 TB3 encoder inputs, which
seems to be ok (still have to test opera
You could also go the way of the VAXbar... Perhaps modernize it ala RaspberryPi
and an automated drink mixing system?
--Original Mail--
From: "Dave Cole"
To:
Sent: Thu, 23 Oct 2014 10:06:17 -0400
Subject: Re: [Emc-users] part 2 - Mach3 to LinuxCNC
Years ago I used to tie control syste
Except that is totally false. See my other post for the reasons Mach3 is more
popular with new users.
--Original Mail--
From: "Len Shelton"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 22 Oct 2014 09:24:39 -0500
Subject: Re: [Emc-users] part 2 - Mach3 to LinuxCNC
Except that the
I think there is plenty of reason to care about why another control may be more
popular, including commerical/industrial controls. Looking at how other
controls do things and understanding why they may be more popular provides
valuable information on what might be improved in LinuxCNC. With larg
It goes back to "fit and finish" and ease of configuration.
In Mach3 the graphical configuration utility is integral, rather than a
separate program or programs (pncconf, stepconf,etc) which vary depending on
what setup you are using. In Mach3 the "I/O debug" is integral rather than
several se
Way back when I started in on home CNC, I did a trial between Mach3 and EMC
(pre-EMC2/LinuxCNC days), and Mach3 won resoundingly for it's vastly better
"fit and finish and useabiliy". Today I'm starting in with LinuxCNC again on a
lathe conversion and eventually a conversion on a CNC mill with a
o: "Enhanced Machine Controller (EMC)"
Sent: Fri, 10 Oct 2014 00:07:41 -0600
Subject: Re: [Emc-users] Oscilloscope + logic analyzer (PC based)
On 10/9/2014 12:36 PM, Gene Heskett wrote:
> Dunno, I suppose if one wanted to test the quality of a new and unknown
> quality cable it coul
I'd rather see Fry's and/or Microcenter expand.
--Original Mail--
From: "Mark Wendt"
To: "Enhanced Machine Controller (EMC)"
Sent: Sat, 20 Sep 2014 10:08:57 -0400
Subject: Re: [Emc-users] Hey Gene! Still looking for a pocket O'scope?
Yeah, I've heard the rumors of bankruptcy too. It
I made a new motor pulley for the lathe to correct the basic 50Hz/60Hz issue.
Now at 60.00Hz on the VFD spindle speed actually comes in a hair low, 965rpm
which according to my calculation means if I raise the frequency limit on the
VFD to 62.18Hz (it is a 60Hz rated motor) that should put it pr
Picture here: http://gympiemodelflyers.wordpress.com/articles/brushless-motors/
--Original Mail--
From: "Dave Caroline"
To: "Enhanced Machine Controller (EMC)"
Sent: Fri, 19 Sep 2014 17:53:53 +0100
Subject: Re: [Emc-users] brushless motor coil winder
I do have one relevant picture alre
The current RC brushless motors have a lot of poles and a lot of turns of thin
wire, like 120T.
--Original Mail--
From: "Kirk Wallace"
To:
Sent: Fri, 19 Sep 2014 09:43:11 -0700
Subject: Re: [Emc-users] brushless motor coil winder
On 09/19/2014 09:11 AM, Dave Caroline wrote:
> I dont th
I have a crew cab long bed dually diesel F350 (King Ranch too), hauling stuff
isn't an issue :)
--Original Mail--
From: "Jon Elson"
To: "Enhanced Machine Controller (EMC)"
Sent: Fri, 19 Sep 2014 10:37:58 -0500
Subject: Re: [Emc-users] LinuxCNC Fest - Houston, 10/18 - 10/24
On 09/18/2
I'm interested in attending as well, particularly if I can get a hotel that
isn't too expensive. I'd be coming from north of Dallas if you are interested
in carpooling Jon, I'm the same Pete C. from RCM.
Pete C.
--Original Mail--
From: "Jon Elson"
To: "Enhanced Machine Controller (EMC
I'm leaning towards making a new pulley. Analog scaling could probably
compensate for it, but ultimately when I do S1000 it does take the VFD up to
60Hz output as shown on the VFD display, which is giving a spindle speed of
closer to 1200. Making a new motor pulley .8x the current diameter shoul
Even with the encoder scale set at 2000 and the near scale set to 1.3 I can't
get spindle-at-speed to go true properly. Not sure what else is wrong, I'll
have to double check the encoder. Closed loop would be nice as it would
compensate for the probable gearing issue, though I could probably fix
More fiddling and it appears the encoder is 2000 CPR. At any given speed
selection the spindle is turning a bit faster than specified, with the
difference becoming greater the higher the speed. I believe this is due to it
being a UK lathe and running at 60Hz instead of 50Hz, at least the differe
I fiddled some more with scale and max and it's actually working now. Scale
1000 and max 1000 to match the current gear the lathe is in. I stepped it
through the speeds from s100 to s1000 by 100s and my optical tachometer
confirmed it was about right. Presumably if the encoder can be setup corre
I added the missing rpm in the hal config,
>From this: net spindle-vel-cmd-abs => hm2_5i25.0.7i76.0.0.spinout
To this: net spindle-vel-cmd-rpm-abs => hm2_5i25.0.7i76.0.0.spinout
And now the spindle runs in linuxCNC. There is still an issue since it seems
doing an S setting in MDI isn'
That gets back to my earlier post where there seems to be some mismatch in
naming. If others are using 2.6.3 though I would expect if it was an issue it
would have been found some time ago? This is what PnCConf created, not anything
I modified.
Searching the halcmd show I find:
float OUT
Ok, that red herring has been fried. I went through the troubleshooting steps
and disabled essentially everything disableable in the BIOS. Now the
"unexpected realtime delay" is gone and the latency test also is producing
better numbers, now ~14000.
The original spindle issue remains, no operat
No evidence, that's what I was told by someone who was helping me troubleshoot
the original spindle inop in LinuxCNC but working in PnCConf test issue.
Again I have no explanation as to why I can control the spindle just fine -
forward, reverse and any speed I set - in PnCConf but get no spindle
Ok, new information. It seems spindle operation is being disabled due to
complaints about "unexpected real time delay", without any visible notation in
LinuxCNC that anything has been disabled due to an error (poor form).
To summarize the setup:
The PC is one I built a year and
ittle memory so I drip feed
it from a PC, and I also need to add 4th axis to it.
--Original Mail--
From: "Gene Heskett"
To:
Sent: Thu, 18 Sep 2014 13:53:32 -0400
Subject: Re: [Emc-users] Question on "gearchange" use?
On Thursday 18 September 2014 09:48:12 p...@wpnet.
Mousing is also a lot slower than physical knobs and you have to look where you
are mousing and thus have to take your focus off the screen.
--Original Mail--
From: "Gene Heskett"
To:
Sent: Thu, 18 Sep 2014 13:47:18 -0400
Subject: Re: [Emc-users] Hey Gene! Still looking for a pocket O'
Ok, I'm starting to figure out the halcmd thing...
halcmd shows that the count mode is not being set correctly when testing in
PnCConf. When in linuxCNC halcmd shows that is is being set to TRUE (I edited
it in hal to use TRUE instead of 1). However in linuxCNC halcmd seems to be
showing that b
The jumpers are for single ended or differential connections to the encoder,
they are all left for single ended since that matches the encoder. The encoder
is only about 36" of wire away from the 7i76.
I don't have skype, I suppose I could drag a laptop out to the shop and see
about installing
>From the hal:
# ---Encoder feedback signals/setup---
setphm2_5i25.0.encoder.00.counter-mode 1
setphm2_5i25.0.encoder.00.filter 1
setphm2_5i25.0.encoder.00.index-invert 0
setphm2_5i25.0.encoder.00.index-mask 0
setphm2_5i25.0.encoder.00.index-mask-invert 0
setphm2_5i25.0.en
This seems to be pointing to one issue, when I do this in the open loop test,
the box that displays "hm2_5i25.0.encoder.00.position" just jumps between 0 and
0.00025 regardless of whether I have "single input encoder" checked.
Still not being able to get any spindle operation at all from LinuxCN
I'll try that today
--Original Mail--
From: "andy pugh"
To: "Enhanced Machine Controller (EMC)"
Sent: Thu, 18 Sep 2014 15:07:21 +0100
Subject: Re: [Emc-users] New to LinuxCNC and PnCConf and need some help...
On 18 September 2014 14:56, wrote:
> I figured I'd just get the spindle run
That makes sense, you know what voltage you send to the VFD and can measure the
spindle RPM so you can certainly figure which of the known gear ratios has been
selected. Of course my current issue is to get the spindle to do anything under
LinuxCNC :( It runs fine in the PnCConf test...
--
The VFD takes 0-10V, so the output min is set to 0 and max to 10. When testing
in the PnCConf open loop test the scope confirms that the output voltage is
correct for whatever I enter in the test box for DAC output i.e. 5 gives about
5V, 2 about 2V, etc. that all works fine. The current enable a
I'd also be interested in some samples. The lathe I am working on has 9 gear
ranges and it might be nice to utilize perhaps three of them to optimize torque
in the speed range desired. The spindle encoder is on the spindle so it will
give actual spindle speed no matter what range the gears are i
http://wpnet.us/Harrison4.hal
I did check the "single channel" box in the spindle setup in PnCConf, and I see
it did set the counter mode to 1. Even if it did try to do closed loop, I would
expect it to at least start the spindle and then fail, rather than not moving
at all. Even in the open lo
Sorry, I missed some mail, though I see it on Sourceforge. I ran into a max
mail per IP setting on my mail server so it stopped accepting the list mails.
I've adjusted the limit so hopefully that will fix it.
The update on the current state of things:
The HOME_OFFSET setting has solved the homi
I'll try changing the counter mode and see if that helps. Why can't I seem to
find documentation that clearly lists all variables / keywords and how they
operate and accepted values?
Thanks,
Pete C.
--Original Mail--
From: "andy pugh"
To: "Enhanced Machine Controller (EMC)"
Sent:
The Mesa 7i76 spindle I/O only operates as enable and dir as far as I know. The
notes say the enable output is hard linked with the enable for the analog
output. It all seems to work properly (outside if the instant reverse issue)
when testing in PnCConf's open loop test. Nothing in LinuxCNC whi
It works 100% correct if I set it up to home in the negative direction as it
seems to be designed to do. If I set it up to home in the positive direction I
find no way to tell it that the positive home/limit location is 18.0, not 0.0
and that's the basic problem. I don't want to change the direc
I guess I'm in the stone ages then, I'm still using a Tektronix 400MHz scope
with a *gasp* CRT.
--Original Mail--
From: "andy pugh"
To: "Enhanced Machine Controller (EMC)"
Sent: Wed, 17 Sep 2014 14:42:53 +0100
Subject: Re: [Emc-users] Hey Gene! Still looking for a pocket O'scope?
On 1
For conventional axis travels Z- is to the headstock and X- is infeed towrds
the material. When I set the homing direction the other way, it will find the +
home/limit switch ok, but it want's to set that home location as zero and if I
have home location as 18 it will continue off into overtrave
The spindle runs properly (except for the sudden reverse issue) in the PnCConf
open loop test. Enable, direction and control voltage all work peoperly, I can
change the control voltage output and get the expected speeds from the VFD.
There seems to be some disconnect / misconfig in getting it to
I have a Harrison T280 lathe I am retrofitting to LinuxCNC. This is a stepper
driven machine, the original controls were previously retrofitted with
MicroKinetics stuff and I'm re-retrofitting with LinuxCNC and Mesa 5i25 / 7i76
while keeping the MK stepper drives.
I did a clean install of 2.6.3
87 matches
Mail list logo
