Re: [Emc-users] about to give up on the pi's

[Steve Traugott]

Has anyone tried an (Intel) Edison yet?  It's an x86 machine, and comes in
at $100 if you add the arduino-pinout carrier board to give you all the I/O
pins:  https://www.sparkfun.com/products/13097

On Thu, May 4, 2017 at 2:38 PM, Gene Heskett  wrote:

> On Thursday 04 May 2017 13:14:07 Jon Elson wrote:
>
> > On 05/03/2017 11:36 PM, Gene Heskett wrote:
> > > On Wednesday 03 May 2017 23:44:02 Jon Elson wrote:
> > >> On 05/03/2017 09:28 PM, Gene Heskett wrote:
> > >>> The killer in the beaglebone soup is the cost of the "capes".
> > >>> You've over $200 plus psu's etc before it can turn 2 motors at the
> > >>> same time.
> > >>
> > >> Umm, the CRAMPS board is $79.95, and I sell the 8825 drivers
> > >> for $5 each, but I get them from China for about $1.30
> > >> each.
> > >
> > > The 8825 does what?  I need to take another look at your site I
> > > guess.
> >
> > 2 A 35 V stepper driver.
>
> Whereas I'm at 43 volts and 5 amps on x(8 wire motor in parallel), 65V
> and around 3.5A on z.
>
> And fixing to add another 20 volts to x, its tapped out at about 40 ipm.
> Fairly small fine threaded ball screw.
>
> > >> The BeagleBone is $55,
> > >
> > > And how many cores & how fast is its arm?
> >
> > I think it is only one core, 1 GHz.  Works fine for
> > Machinekit/LinuxCNC control of Cartesian machine, but ssh -X
> > connection from a machine with a screen is noticeably slower
> > than an X86 PC with directly-connected screen.
>
> Where this pi has 4 cores, 64 bit, and is running at 1.2GHz
> And twice the memory.  Needs more yet.  So I think I'll stick with it
> unless this last experiment fails. dd is still munching away at creating
> a clone of the only u-sd that will still boot it.
>
> Then 2 boxes of deck stuff just walked up and plopped themselves down on
> the front deck, which I am suspecting I'll need to modify the gazebo as
> its too tall to fit under the roof overhang. Might wind up ripping out
> some greenery and extending the deck, outward by a foot or so and east 5
> feet as it will block the front door otherwise.  Fun and games with
> 2x8's for framing and 2x4's for flooring.
>
> If it ever happens. Dee said no.  So I've spent the afternoon assembling
> the deck furniture, which came in flatish pieces. Got the corner chair
> piece and 2 chairs, but still have 2 in the box when it started a light
> rain, and my back said enough already. So dinner is warming in the
> microwave.
> >
> > The Beagle Bone has 2 PRU processors (200 MHz real time
> > 32-bit CPUs) that Charles Steinkuehler set up the framework
> > for so that LinuxCNC can do fast stepping, PWM and encoder
> > counting.  Not quite FPGA performance, but WAY faster than
> > software stepping.  So, that relieves the ARM CPU from the
> > base thread tasks.
> >
> > Jon
> >
> Thanks Jon.
>
> 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] Need recommendation on encoder cable --> Ethernet cable?

[Steve Traugott]

What I'm trying to say is that when you say "ethernet cable", most people
mean UTP, unshielded twisted pair, generally rated CAT-5 or CAT-6 these
days.  Most ethernet patch cords are UTP.  You can get the same thing in a
shielded version as well, which you may find you want:

https://www.google.com/search?q=shielded+cat-5=shop


On Mon, Oct 17, 2016 at 12:33 PM, Nicklas Karlsson <
nicklas.karlsso...@gmail.com> wrote:

> > ...and for encoder cables, you may want shielded twisted pair (has a foil
> > layer under the jacket).  I've had issues with unshielded twisted pair
> > picking up motor EMF and adding extra encoder counts.  ;-)
>
> That's why ethernet cables and differential signals are a good choice.
>
> The PWM signals in the servos are really horrible then it come to
> electrical disturbance.
>
> 
> --
> 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] Need recommendation on encoder cable --> Ethernet cable?

[Steve Traugott]

...and for encoder cables, you may want shielded twisted pair (has a foil
layer under the jacket).  I've had issues with unshielded twisted pair
picking up motor EMF and adding extra encoder counts.  ;-)

On Mon, Oct 17, 2016 at 10:45 AM, Chris Albertson  wrote:

> But today in real life, no one makes their own patch cables.  But every
> installer has to make their own terminators for the on-wall cable because
> you can't pull wire that has a bulky connector on it.  So the only tools
> and parts they commonly sell are for the in-wall kind of solid wire.
>
> As for oil resistance, Buy the "plenum" type.  They use teflon insolation
> and don't cost that much more.These are sold because of fire codes in
> some kinds of buildings, they don't create toxic smoke in a fire.  Yes they
> do make Plenum patch cables with flexible wire and no PVC.
>
> On Mon, Oct 17, 2016 at 10:10 AM, Przemek Klosowski <
> przemek.klosow...@gmail.com> wrote:
>
> > On Mon, Oct 17, 2016 at 12:40 PM, Chris Albertson
> >  wrote:
> > > Ethernet patch cable might work. .  Don't try and make you own.
> > The pre-made ones come in any
> > > length from about 12" to 25 feet and they have those over molded train
> > You're absolutely right about buying cables---the cost is so low that
> > it doesn't make sense. The patch cords actually get down to few inches
> > (3-4"), which is great for local cabling in tight spaces.
> >
> >
> > >The common kind of RF45 termination tools work only with the
> > > solid in-wall kind of wire so don't use them.
> >
> > Actually, it's not the termination tools, but the terminals and plugs
> > themselves---they are still insulation-displacing crimps, but the
> > shape of the spades is different. In the beginning of the computer era
> > we used to make our own cables and used the same tools with different
> > plugs---it was a PITA because they looked the same (translucent
> > plastic). Unless you opened a new marked box, we used to inspect the
> > connector and it wasn't easy---you had to squint at the shape of the
> > tiny gold-plated terminals through wavy plastic.
> >
> > If you used the wrong type for your wire, you'd get a broken wire
> > touching the terminal by a tension in the plastic insulation: it might
> > work for a little bit, but eventually the lack  of gas-tight
> > metal-metal connection would cause it to fail.
> >
> > I'm saying this because someone might actually make a patch cable in a
> > pinch, and that's what you have to pay attention to.
> >
> > 
> > --
> > 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
>
--
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] Clone a HD

[Steve Traugott]

John, you're right about booting from the live CD.  You usually don't want
to use dd to copy from a mounted (live) filesystem; since it operates at a
block level, any write activity on the source drive during the copy is
likely to produce a corrupt copy. When doing what you want to do, I usually
boot the machine off of a CD and use dd.  If the new drive is larger, or if
I must run in the live machine, I'll partition the new drive with cfdisk
and use rsync to copy the files, then run grub-install to write the boot
track.  In either case, you don't need tar.  Tar doesn't do the right thing
with special files anyway, such as device inodes and sockets, so don't use
it.  (I know much of what I just now said may not make a lot of sense yet.)

Whatever you do, don't get the source and destination drives mixed up.  ;-)


If you want something like a recipe, this google search (on the various
stack exchange web sites) will help:


https://www.google.com/search?q=stack+exchange+clone+drive+linux+dd+rsync

If you want more background and/or have an unusual situation, you'll want
to google for:

- dd
- rsync
- grub
- dmesg  # shows recent hardware activity
- cat /proc/partitions  # shows currently available partitions
- mount # makes an available partition active and attaches it somewhere in
the live filesystem tree
- umount # the Linux equivalent of what you need to do before ejecting
removable media
- chroot

For a boot CD, I usually use systemrescuecd, because it pretty much gives
me a vanilla Linux machine -- I've been working with *nix since the early
80's, so anything more than that just gets in my way for a simple copy.
I've never used clonezilla, but as far as I understand it's got more drive
cloning tools added for what you want to do.  It's also got a lot of
network-based bits in there, the kind of thing needed for multi-machine
deployments, which adds a lot of complexity, but you can just ignore all
that for your use case.

Finally, here's my own completely untested example showing what I usually
would do if using rsync.  I'm probably missing some steps here, and I've
simplified stuff, so use this only as a guide, not a recipe:

Assuming sdc is the source drive and sdd is the destination:

cfdisk /dev/sdd   # make a big Linux partition called sdd1, and a swap
partition roughly 2x RAM size called sdd2
mkfs.ext4 /dev/sdd1  # format sdd1  (caution -- make dead sure sdd1 is the
destination)
mkswap /dev/sdd2  # format swap partition (caution)
mkdir /mnt/sdc1
mkdir /mnt/sdd1
mount /dev/sdc1 /mnt/src  # make source accessible under /mnt/src
mount /dev/sdd1 /mnt/dst  # make destination accessible under /mnt/dst
ls -la /mnt/src  # confirm you're looking at the old drive
ls -la /mnt/dst # confirm this is the new drive -- only thing showing
should be an empty lost+found directory
rsync -HaSx  /mnt/src/ /mnt/dst/  # do the copy -- all of these flags are
important, as well as the trailing slashes
rsync -PHaSvx /mnt/src/ mnt/dst/  # alternative, verbose version of the
above command
mount -t proc none /mnt/dst/proc   # get things set up for the chroot
mount -o bind /dev /mnt/dst/dev
mount -t sysfs sys /mnt/dst/sys
chroot /mnt/dst/ /bin/bash  # start a new shell using /mnt/dst as /
grub-install /dev/sdd  # install boot loader (caution)
exit   # leave the chroot
umount /mnt/dst/proc  # tear down the chroot
umount /mnt/dst/dev
umount /mnt/dst/sys
umount /dev/sdc1  # detach the source and target drives from the live
filesystem tree
umount /dev/sdd1

All of that mount and chroot stuff towards the end is to enable you to use
the new drive's version of grub to install the boot loader in the boot
track of the new drive.  Google for 'chroot grub-install' for more details
about this.

Steve

On Sat, Feb 20, 2016 at 9:28 AM,  wrote:

>
> >> On Feb 20, 2016, at 11:56 AM, Chris Albertson <
> albertson.ch...@gmail.com> wrote:
> >> Note that to close a device you need identical hardware.
> >
> > To clarify, you don’t need “identical” hardware.  You need a disk that
> is the same size or larger to clone a disk.
> > -Tom
>
> And, I should add, that if the disk is larger the extra space is not used
> (due to the partition map).  There are tricks to add that space (or expand
> into it).
> -Tom
>
>
>
> --
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=272487151=/4140
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Site24x7 APM Insight: Get Deep Visibility into Application 

Re: [Emc-users] Clone a HD

[Steve Traugott]

Argh.  Got the mkdir's wrong.  Here's a corrected version, no guarantees
etc.:

Assuming sdc is the source drive and sdd is the destination:
>
> cfdisk /dev/sdd   # make a big Linux partition called sdd1, and a swap
> partition roughly 2x RAM size called sdd2
> mkfs.ext4 /dev/sdd1  # format sdd1  (caution -- make dead sure sdd1 is the
> destination)
> mkswap /dev/sdd2  # format swap partition (caution)
> mkdir /mnt/src
> mkdir /mnt/dst
> mount /dev/sdc1 /mnt/src  # make source accessible under /mnt/src
> mount /dev/sdd1 /mnt/dst  # make destination accessible under /mnt/dst
> ls -la /mnt/src  # confirm you're looking at the old drive
> ls -la /mnt/dst # confirm this is the new drive -- only thing showing
> should be an empty lost+found directory
> rsync -HaSx  /mnt/src/ /mnt/dst/  # do the copy -- all of these flags are
> important, as well as the trailing slashes
> rsync -PHaSvx /mnt/src/ mnt/dst/  # alternative, verbose version of the
> above command
> mount -t proc none /mnt/dst/proc   # get things set up for the chroot
> mount -o bind /dev /mnt/dst/dev
> mount -t sysfs sys /mnt/dst/sys
> chroot /mnt/dst/ /bin/bash  # start a new shell using /mnt/dst as /
> grub-install /dev/sdd  # install boot loader (caution)
> exit   # leave the chroot
> umount /mnt/dst/proc  # tear down the chroot
> umount /mnt/dst/dev
> umount /mnt/dst/sys
> umount /dev/sdc1  # detach the source and target drives from the live
> filesystem tree
> umount /dev/sdd1
>
> All of that mount and chroot stuff towards the end is to enable you to use
> the new drive's version of grub to install the boot loader in the boot
> track of the new drive.  Google for 'chroot grub-install' for more details
> about this.
>

Steve
--
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151=/4140
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Clone a HD

[Steve Traugott]

Also, you may need to edit /etc/fstab after booting the new machine for it
to find your new swap partition.  Roughly:

free  # if it shows a swap size of 0, then...
sudo nano /etc/fstab  #  find the swap line and change /dev/whatever to
/dev/sda2
sudo swapon -a  # activate the new swap
free # check it

...assuming that after you've booted the new drive in the new machine it
came up as /dev/sda rather than /dev/sdd, of course.

I'm probably giving you just enough info here to be dangerous.  ;-)

Steve


On Sat, Feb 20, 2016 at 11:05 AM, Steve Traugott <stev...@t7a.org> wrote:

> Argh.  Got the mkdir's wrong.  Here's a corrected version, no guarantees
> etc.:
>
> Assuming sdc is the source drive and sdd is the destination:
>>
>> cfdisk /dev/sdd   # make a big Linux partition called sdd1, and a swap
>> partition roughly 2x RAM size called sdd2
>> mkfs.ext4 /dev/sdd1  # format sdd1  (caution -- make dead sure sdd1 is
>> the destination)
>> mkswap /dev/sdd2  # format swap partition (caution)
>> mkdir /mnt/src
>> mkdir /mnt/dst
>> mount /dev/sdc1 /mnt/src  # make source accessible under /mnt/src
>> mount /dev/sdd1 /mnt/dst  # make destination accessible under /mnt/dst
>> ls -la /mnt/src  # confirm you're looking at the old drive
>> ls -la /mnt/dst # confirm this is the new drive -- only thing showing
>> should be an empty lost+found directory
>> rsync -HaSx  /mnt/src/ /mnt/dst/  # do the copy -- all of these flags are
>> important, as well as the trailing slashes
>> rsync -PHaSvx /mnt/src/ mnt/dst/  # alternative, verbose version of the
>> above command
>> mount -t proc none /mnt/dst/proc   # get things set up for the chroot
>> mount -o bind /dev /mnt/dst/dev
>> mount -t sysfs sys /mnt/dst/sys
>> chroot /mnt/dst/ /bin/bash  # start a new shell using /mnt/dst as /
>> grub-install /dev/sdd  # install boot loader (caution)
>> exit   # leave the chroot
>> umount /mnt/dst/proc  # tear down the chroot
>> umount /mnt/dst/dev
>> umount /mnt/dst/sys
>> umount /dev/sdc1  # detach the source and target drives from the live
>> filesystem tree
>> umount /dev/sdd1
>>
>> All of that mount and chroot stuff towards the end is to enable you to
>> use the new drive's version of grub to install the boot loader in the boot
>> track of the new drive.  Google for 'chroot grub-install' for more details
>> about this.
>>
>
> Steve
>
--
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151=/4140
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Micro-switch repeatability

[Steve Traugott]

I also remember someone, could have been at reprap.org or cncrouterparts.com,
doing repeatability tests, and getting similar surprisingly consistent
results with cheap switches.

On Mon, Feb 15, 2016 at 5:04 PM, Len Shelton  wrote:

> We don't use levers, we activate the plungers directly. The key is to
> design the switch mount such that a physical stop would prevent wrecking
> the plastic switch on a rapid jogs ooops.
>
>  >Len
>
>
>
> On 2/15/2016 6:37 PM, Bertho Stultiens wrote:
> > On 02/16/2016 01:16 AM, Len Shelton wrote:
> >> We in fact see about 2-3 tenths of a thou repeatibility with a slow
> >> approach on cheap chinese snap switches.
> > Just like what others seem to report.
> >
> > Is that on direct activation or using an positional offset on a lever?
> >
> > Using a lever seems like introducing a variable due to spring forces.
> >
>
>
>
> --
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=272487151=/4140
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151=/4140
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Well, thats the end of aliexpress.com for me

[Steve Traugott]

Interesting.  My wife and I run an electronics manufacturing business.  We
tend to prefer mouser *because* their parametric search works so well --
I've always thought it was better than Digikey's.

I think it probably comes down to getting the hang of a particular web
site's parametric search.  For instance, in Mouser's case, you need to make
sure you select a category after entering anything in the search box,
otherwise you don't get the parametric search UI.  This makes sense; they
can't really show the parameter selections for a random assortment of parts
and equipment in an assortment of categories, because the set of parameter
selection boxes would explode sideways across the page.

Adapting John's Digikey example to Mouser:  Type "MOSFET" in the search
field, check the "stocked" box, hit submit.  That gives you a list of
categories.  Click on the MOSFET category; that opens up the parametric
search.  Select TO-92, select sub-1-ohm.  That gives 8 matches, all of
which are available in qty 1.

>From http://www.mouser.com/helppage/#nomin: "No minimum order dollar amount
on products normally stocked in our warehouse."

Steve

On Sun, Jan 31, 2016 at 3:51 PM, John Kasunich 
wrote:

>
>
> On Sun, Jan 31, 2016, at 06:18 PM, Gene Heskett wrote:
> > On Sunday 31 January 2016 13:43:50 Jon Elson wrote:
> >
> > > I can't imagine why you'd buy commodity semis from a Chinese
> > > source when there are great distributors like Digi-Key and
> > > Mouser, that are quite happy to serve small orders.
> > >
> > Both of the above have minimum orders for me that are quite costly for a
> > 10 pak of anything in the transistor 10 cent to 5 dollar price range.
> > That and their sites search engines return what looks to me to be random
> > from dice throws results.  If I go searching for small hexfets I get
> > 20,000 hits, none of which give me the data to make a choice
> > semi-intelligently.  Digikey is slightly better, but not enough to spend
> > hours  wading thru the selections while still hiding the important data.
> > More often than not, clicking on a fine tune option, such as any case
> > smaller than a to-220 returns zero results and I know that there are
> > smaller versions such as could drive an ice cube relay at 12 or 24
> > volts.  That can be easily put in a to-92, or even in a SMD package, so
> > why can't it be found?
>
> I find this rather confusing.  Mouser is meh, but IMHO Digikey's search
> engine is simply outstanding.  For example:
>
> I type MOSFET into the search box.  It returns a list of categories.
> I pick "FETS - Single".  It returns a list of 39000+ parts.
> I click the "In stock" box and "apply filters".  List is now 17,999 parts.
> Scroll over to "Mounting Type", select "Thru hole" and apply filters.
> 3293 parts.
> Scroll over to "Supplier Device Package" and select the four TO-92 variants
> (ctrl-click for multiple selections).  Apply filters, 123 parts.
> Under packaging, select "Bulk" and "Cut tape", (the other variants are for
> large quantities).  Apply filters, down to 106.
> Select 40 to 100V in the drain-to-source voltage box, down to 59.
> Select all the sub-1-ohm ones in the Rds-on box, down to 7.
> Put "1" in the desired quantity box.  Hit the up arrow in the price column
> to sort cheapest first (giving a quantity moves anything with a minimum
> order larger than that quantity items to the bottom of the list).
> Cheapest part is at the top of the list, MicroChip TN0604N3-G.  Click
> on the PDF icon in the 2nd column and I'm looking at a data sheet.
> Click on the links in the 4th or 5th column and I'm looking at the
> details page for that part.  Add 1 to cart.  Place order.  There is no
> minimum total order (although it sucks to spend $6 on shipping when
> you are ordering a $1 part).
>
> It took me less time to find that part than it took to type out what I did.
>
> Digikey is absolutely my first stop for just about any part, both for
> hobby stuff and for my day job.  Like McMaster, their prices may
> be a bit higher than some other sources, but usually not enough to
> matter, and the search engine more than makes up for it.
>
> John Kasunich
>
>   John Kasunich
>   jmkasun...@fastmail.fm
>
>
> --
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=267308311=/4140
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: 

Re: [Emc-users] VFD Tweaking

[Steve Traugott]

Here's a better VFD/motor matching overview, with a pretty picture of
what's going on (scroll down until you get to the schematic showing the
three output PWM waves):

http://www.franklin-controls.com/pump/blog/motors/

On Thu, Jan 7, 2016 at 5:29 PM, Steve Traugott <stev...@t7a.org> wrote:

> On Thu, Jan 7, 2016 at 5:14 PM, Brian Morel <brianmore...@gmail.com>
> wrote:
>
>> As far as the current issue, you need to be careful using old motors with
>> a vfd.  The insulation ratings of the older motors were not designed to
>> handle the voltage spikes that can occur with running motors fairly
>> deviated from their original hz designs.
>>
>
> This sounds to me like a likely culprit.  The high-frequency,
> high-amperage switching on the output side of a VFD causes high voltage
> transients which, if the motor isn't rated for VFD duty, can pierce the
> winding insulation and cause internal arcing and carbon tracks.  You may be
> seeing higher current simply because the windings are now shorting together
> in a few spots.
>
> I remember spending some time searching for a good 5 HP motor with the
> right insulation rating to run on a VFD -- they cost a bit more, and are
> harder to find.  I don't now remember what the magic designation is, but a
> quick google search just now tells me it may have been "class H"
> insulation.  In several subsequent years of running said motor and VFD,
> I've never had any issues.
>
> This looks like a pretty good overview:
> http://energy.gov/sites/prod/files/2014/04/f15/motor_tip_sheet14.pdf
>
> Steve
>
--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] VFD Tweaking

[Steve Traugott]

On Thu, Jan 7, 2016 at 5:14 PM, Brian Morel  wrote:

> As far as the current issue, you need to be careful using old motors with
> a vfd.  The insulation ratings of the older motors were not designed to
> handle the voltage spikes that can occur with running motors fairly
> deviated from their original hz designs.
>

This sounds to me like a likely culprit.  The high-frequency, high-amperage
switching on the output side of a VFD causes high voltage transients which,
if the motor isn't rated for VFD duty, can pierce the winding insulation
and cause internal arcing and carbon tracks.  You may be seeing higher
current simply because the windings are now shorting together in a few
spots.

I remember spending some time searching for a good 5 HP motor with the
right insulation rating to run on a VFD -- they cost a bit more, and are
harder to find.  I don't now remember what the magic designation is, but a
quick google search just now tells me it may have been "class H"
insulation.  In several subsequent years of running said motor and VFD,
I've never had any issues.

This looks like a pretty good overview:
http://energy.gov/sites/prod/files/2014/04/f15/motor_tip_sheet14.pdf

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

Re: [Emc-users] debris collection

[Steve Traugott]

For a muffler, loosely stuff some random open-cell foam rubber scraps into
a 5-gallon bucket.  Attach an outlet hose to the vacuum and stick the other
end into your bucket full of foam rubber.  Works surprisingly well relative
to the time invested.

In the case of the bucket max, it also might work to invert the muffler
bucket upside down on top of it, cutting a hole in the bottom for outlet
air and in the side for inlet hose and cord.  That would kill the noise
from the motor itself as well as from the exhaust air.

On Mon, Nov 30, 2015 at 4:52 PM, Gene Heskett  wrote:

> On Monday 30 November 2015 11:46:28 Dave Cole wrote:
>
> > Have you considered putting the vacuum outside?
> >
> > Dave
>
> >From where it is, that would be a good 25' of 1.5" pvc, Dave.
>
> And a huge heat leak if I pump that much air outside, that garage is
> quite well insulated, 6" walls are full of cocoon, and about a foot of
> it on the ceiling with an extra layer of 2" styro on the inside of the
> garage door, and even in the subzero weather, one cheap electric heater
> keeps it above 60F.  Plug in a 2nd one, and it will be up past 80 in 1
> hour.  Besides, where its sitting, its only a few seconds to open the
> catch bucket and retrieve the buttons it sucks up.  Longer to shake the
> dust off them.  But I'll give some thought to moving only the vacuum &
> leaving the cyclone where it is.  The obvious place would be inside a
> 30" square opening in the front wall of the house where the electric
> meter is, but its not going to be a straight shot.  Its only about 6
> feet to that wall, but that wall is busier than all get out because
> thats where all the electrical is for the standby generator.
>
> That might mean piping its exhaust back in for the heat recovery this
> time of the year & is why I mentioned muffler designs I could cobble up
> out of pvc.
>
> > On 11/30/2015 11:31 AM, Gene Heskett wrote:
> > > Greetings all;
> > >
> > > Todays story from West (by God) Virginia.
> > >
> > > I have an Oneida Dust Deputy on the table saw, bolted to the side of
> > > a 16 gallon shop vac.  It works very well, with filters in the
> > > shop-vac lasting a year or so because 99.9% of the dust goes into
> > > the 5 gallon bucket under the Deputy, with the main vac not even
> > > getting enough to cover the bottom in a long time.
> > >
> > > But its both ex$pen$ive, and bulky, always in the way.  Way too big
> > > to try and move it to my new mill.
> > >
> > > So, since I already had a bucket-max, (Shop-Vac product) which is a
> > > vac motor mounted in a 5 gallon buckets lid, and had been using it
> > > for a sucker on an R.O. hand sander with limited success, limited by
> > > the life of a $15 to $30 filter.
> > >
> > > So I thought I'd see just how efficient a cylindrical cyclone might
> > > work.
> > >
> > > At Lowes, a section of 4" sewer pipe, a floor flange, a 4" to 2"
> > > adaptor, some 2 to 1.5 adaptors were bought.  Sawed off about a foot
> > > of 4", clamped in on the mills table and milled the OD of a 1.5"
> > > white plastic pipe into it, about 4" from one end so that its outer
> > > wall matched the side of the 4", a tangent entry, glued the floor
> > > flange to it, screwed that to the middle of a bucket lid, then cut
> > > out the lid to the inner diameter of the flange.  Then solvent glued
> > > a section of the 1.5" into the side of the 4", leaving enough length
> > > to reach the wall and a 1.5" riser pipe up to the level of the top
> > > of the machine where an elbow and short 1.5 nipple connected to a
> > > radiator hose for flexability, which jumpers to some more plastic
> > > pipe bolted to the side of the machine head. Suitable pieces get it
> > > to the vicinity of the tool.
> > >
> > > On top of the 4", the 4 to 2 adaptor was glued on and a 2 to 1.5 was
> > > bored out on the lathe so the 1.5" could be driven on in and its
> > > bottom end placed about 3" below the side pipes entry all for a
> > > tight friction fit.  This was then connected to the bucket-max,
> > > machineing stuff on the lathe as required, with a couple elbows and
> > > a horizontal run so the buckets could sit side by side on the floor.
> > > Very little of it is glued so it can be popped apart for bucket
> > > dumping etc.  And it all sits under the left end of the machines
> > > table and out of traffic.
> > >
> > > Its now been collecting at least 99% of the machining dust, and the
> > > buttons I'm making, plus the majority of the dust and finger cutouts
> > > from the last few box parts machined.  The filter in the bucket-max
> > > is still pretty clean after 3 weeks use, with almost nothing in its
> > > bucket. The bucket under this contraption now has about 2.5" of
> > > debris in it.
> > >
> > > The floor around the mill is staying pretty clean too, so I'd have
> > > to say its a resounding success at "dust control".  Biggest problem
> > > is the noise, I can't hear the phone when its running. The
> > > bucket-max