nte doesn't make chips, what is the real part number for that chip
maybe people could help you easier
is that a unipolar driver? if it is, maximum sink current is the max current you want do pull through your motor.
It's a bipolar, part number is 054c?
Heres the docs for it. http://www.nteinc.com/specs/1800to1899/pdf/nte1857.pdf
I have everything hooked up except for pin 6 I'm not sure what to do with it.
I was hoping that setting that pin correctly would keep my stepper from pulling to much current
Sorry, pin 1 is also in question. Whats a clamp diode?
I have a diode to connect between pin 1 and pin 16 (V+) but which direction should it face? From 1 towards V+ or vice versa?
humm, it makes sense that it would go into V+, isn't that where extra electricity should go. I just dont want to cook it right off the bat. I had to order this thing and wait a week, plus it was $12.
Can anyone direct me to a better channel for this question?
IRC isn't the best way to learn electronics
I'm not 'learning' electronics persey, I just have a few terms I dont understand.
As long as I can read I can learn. The problem is most docs I read assume I already know certain terms.
right, the assume that you have "learned electronics"
I can see it in my mind even if I don;t know what to call it.
It was worth asking
At least I wont be wishing I had asked.
Google is absolutly fruitless for this IC
one of the big problems in talking about electronics on IRC is that a picture (schematic) is worth at least several thousand words
Well I have no schematic
Or I'd be golden
I got two pins that I dont know what they do.
The other 14 are already hooked up
have you drawn a schematic showing what you are hooking up?
just pencil and paper
circuits are visual things - topology and all that
Yeah I agree
you were asking about clamp diodes, but the datasheet says the part already has them
But I if say dirt road and youve always lived in NY city then you might say 'what?'
Right jmk, I just want to understand it.
you know what an H bridge is?
Why, if the diodes are provided can I also jump one from pins 1 & 16?
I know what it is but not how it is lol.
I'm still reading the datasheet
I went ahead and put a diode going from pin 1 into pin 16 thinking that excess electricity should probably go there lol.
It's now pin 8 and 6.
I do not understand the pupose of these two.
6 sets the coil current
Ok thats what I thought
now youve confirned that.
sets maximum current you mean?
it doesnt matter Ill figure that out, how bout pin 8?
8 determines whether you are doing half or full stepping
oops, thats 9
* jmkasunich can't read
New e-mail free UPS enco over $200 with code HFSP27
btw, feel free to 'quote the text'. I poured over that doc at least a hundred times.
half stepping involves going from 2 coils energized to only one, then two again, etc
when only one is energized, you have two choices for the other coil - leave it open circuited, or short it
shortinbg it would be low right
low resitance; impedance?
pin 8 low = high impedance = open
pin 8 hi = low impedance = shorted
right on man thanks!
okay now what are the implications of either?
to be honest, I dunno
Not to be a pain in the ass or anything.
probably depends on the motor
Do you think I can hurt anything by making the wrong choice?
this is not a chopper drive, it would be used with high resistance low current motors
okay then I wont sweat it.
you should be able to run the motor either way, and see which has better torque, etc
Man youve helped tremendously
the difference would probably only show up at higher speeds
Dont care about torque. Im only hooking this to emc (an old printer motor) so I can see the basics at work.
also, you can connect pin 1 directly to 16, the diode isn't needed
okay, I'll remove the diode since it may be backwards lol
using a diode, zener, or resistor there comes under the heading of advanced tweaking, and you can't really do it correctly unless you are examining the waveforms with an oscilliscope anyway
Well they provide a formula for figuring the value of your resistor that goes from pin 6 to ground (the maximum sink setting). So now I just have to read the details of my motor and get my calculator.
hmm, I don't think that bias input (pin 6) is iontended to be used to set the actual motor current
more of a current limit
Which should be 700ma
the actual motor current nornally would be determined by the supply voltage and winding resistance (ohms law)
since thats all the IC can take
not that simple
Dude! Dont say that
I was almost home free
if you hooked that to a 1 ohm stepper, and a 12 volt supply, you'd toast the chip no matter what you do with the bias pin
because at 350mA, the stepper would drop only 0.35V
right that would be 12 amps, no?
all the remaining supply voltage would be dropped across the chip
if you hooked the stepper directly across the 12V supply (no chip) you'd get 12A
if you hook the stepper to the chip, the chip will try to limit current to 350 mA (or whatever you set with the bias input)
okay, that makes sense.
so you'd get 350mA thru the motor, 0.35V across the motor, and the remaining 1.65 V dropped across the chip
11.65V * 350 mA = 4 watts (and thats just for one winding)
8 W would toast that chip in short order
it's really intended for motors that have high resistance, such that power supply voltage divided by resistance = less than 350 mA
I think I've overlooked something
(trying to answer before I read the whole datasheet)
Hey your allowed
never mind, I had it right the first time
lol you asswipe
Get my hopes up
I saw "clock" in the datasheet, and started to wonder if it was actually a chopper chip
but thats just the step input
no such luck, I knew that
dunno why they call it clock
step pin out of computer (I'm assuming)
chopper drives have a real clock (to me clock is a constant frequency thing, not the step command)
Yeah that is coming. But until after xmas I'm on a tight budget and love the idea of building my own circuit and acually understanding it.
based on thermal resistance and rated temps, the absolute max power for that chip is something like 2 watts
So my voltage divided by resistance should be less than .35 right or .035?
liek 12V and 36 ohms
Great man, thanks for taking time to explain that to me.
you should read the "power dissipation" section
Now I have enough to at least dig deeper. Before I was at a complete stall.
breaking those rules is the most likely way to let the smoke out
I did but didn't totally get it. I guess that's what you've been talking about with excess voltage being dropped across the chip.
How's it hangin toasty?
You know your the one in school, you could jump in at any time and help out jmk.
* jmkasunich is out of school, getting dumber by the day
jmk, one thing occurs to me as I sit here with my calculator.
Wait nm, my own mind answered me lol
oh wait no it did'nt.
If I have two coils at 350ma each then my voltage/resitance can be .7 or under right?
each coil is independent
damn gotta be 350
do you have a motor?
yeah and it says 10ohm which is 1.2A
that is a much bigger motor than this chip is designed to drive
Well I got floppy motors too.
lemme look at those
floppy motors probably aren't labeled very well, you'll have to measure them
(got an ohmmeter?)
also, floppy motors might be 5 wire, which can only be used unipolar - depends on the motor
I have three four wire floppy motors and a meter :D
I have a measurement now to google its meaning.
okay, best I can tell I got 17.3 ohm's across my floppy motor. Does that sound right?
according to those numbers I can't even run a tiny floppy motor with that chip. Even if I can supply the minum voltage of 7.5V.
What kind of tiny motors is this thing made for. Something has to be wrong.
how big it the motor? an inch cube? inch and a half?
The floppy motor is .585 round. The printer motor I have rated at 10 ohms is 42mm.
the floppy one is certainly small enough, but also designed for a 5V supply (or less)
I guess the floppy one is 15mm
[01:36:54] <jmk-st> http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&storeId=10001&catalogId=10001&productId=237884
Which means I cant use the chip I have.
12 V 0.3A
40 ohms per phase
Make a voltage divider to parallel the motor to one of the divider resistors
if you're gonna burn power in resistors, just add series resistors to the motor
I didnt think I could do that
sure you can
you need a minimum of 35 or so ohms
you got 10
add 25 ohms in series with each winding
the resistors need to be 5W or larger - the actual power at 350 mA is 0.35 * 0.35 * 25 = 3.06
note that this is NOT the most efficent way to run the motor, just _a_ way to run the motor with the chip you have
that chip was designed with motors like the Jameco one I posted in mind
By in series you mean 25 ohms from one pole of a coil to it's other pole. Or do you mean just along the V+ going to the coil?
Nice link btw, I might be ordering one of those $8 motors.
I know what series and parallel are lol
no motor terminal goes to V+
I'm just wanting to be absolutly clear.
I know lol
I meant from one of the power pins
as in power not pin 16
then why did you write "Or do you mean just along the V+ going to the coil?"
the resistor doesn't go in the power supply
I was trying to save time and not be absolutly clear.
it (they, you need two) goes in series with the windings
normally you connect one end of winding 1 to L1, and the other end to L2
if you are adding a series resistor, you connect one end of the winding to L1, and the other end to one end of the resistor
then you connect the other end of the resistor to L2
Thats what I meant by V+ believe it or not lmao
ditto for the other winding, other resistor, and L3/L4
I cant speak it man
I see it and I'm not sure if Im right but I dont even know how to speak it.
gack - 1mS latency
But okay. If I can do that then I got it.
You scared me for a minute.
I thought I had bought a chip for nano motors
btw have you guys ever saw the nano motors made from bacteria flagellum(sp?)
Like 30 rpms instantly and forward reverse with virtualy no lag.
jmk-st: One more question. Now that I know I can wire the resistors in and make the motors .35A should I still figure and set the maximum (pin 6) at .35? I dont understand why I should have to tell it that .35 is all it can take, why did they integrate that into the chip?
to be honest, not entirely sure
oh, I take that back, I do know (I think)
okay so it;s not just me
the output transistors are bipolars, and need base current proportional to the load current.
if they biased the chip to always provide enough base current to let the outputs handle 350mA, they'd be wasting a lot of current if the load is only 50mA
so they let you choose the bias level
So I should still figure that maximum to be @ .35A I guess what confuses me is that if I had made that chip I would say in the docs 'use a whatever ohm resitor here to set the maximum at .35A, the chips max capacity'
what if the user has a 50mA motor
then they would set the bias so the chip can deliver just a hair more than 50mA, instead of 350mA
which would save power and chip heating
Well that way you would know that if you were gonna be way under that you could, as you said, save power but if you were gonna max it out like me then you could just set it to max.
they assume that you know how to calculate the ohms
(datasheets are targeted at engineers who are designing chips into products, NOT at hobbyists
Well you've given me more than enough info to fry my chip with peace.
If I cook it now then it won'y be because I did'nt risk sounding like a retard and asking for help.
Or spelling like one.
cook her up
Now to tackle hal :O
I can't believe I just did that
crap is NEVER good
I have a PC power splitter cable, and I wanted to use one end of it to power something else
so I cut the wires close to the other connector to leave max length
except I cut at the wrong end
* dogfishguzzler doesnt think it's funny
* dogfishguzzler laughs away from the screen
damn, all my resistprs are too big.
Well atleast I know.
I can't thank you enough jmk
Sorry, I got another question. Last one tonight. I also have a 30V power suplly from the printer I got the motors from. I'm assuming from what Ive read here tonight that adding a 100 ohm resistor to each windings lead I will end up with 142ma of draw (using my 10 ohm motor)
Is that right?
motors aren't resistors
the resistance is only dominant at low-to-no speed
when the motor starts turning, it'll be a generator, so the voltage across the motor coil will be reduced, so the current will go down as well
I need an ebay feedback blurb generator program
wow I hit a lull in ebay servo drive purchasing, similar items to what I bought are going for 2-3 times as much
"limit switches for beginners" http://cgi.ebay.com/CNC-Stepper-motor-BOOKLET-LIMIT-SWITCHES-FOR-BEGINNERS_W0QQitemZ330196780970QQihZ014QQcategoryZ71394QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
this is a great deal, but I don't have $900 http://cgi.ebay.com/Allen-Bradley-Servo-Drive-1398-DDM-009X-DN-Lot-5-units_W0QQitemZ200182348170QQihZ010QQcategoryZ78191QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
went to lowes to buy crimp terminals, of course my 12 year old son notices the butt splices
ask him for me whether butt splices or bastard files are funnier
he doesn't know what a bastard is
well some help he is
butt splices make sense though
dunno how to explain a bastard file
yeah no idea here either
although if you want to hurt a bastard, and all you have is files, a bastard file is the one to choose
does it just mean the same as 'single cut'?
a hand cut rasp is the tool of choice though
[02:44:14] <jmk-st> http://www.answerbag.com/a_view/32467
some answers are worse than not knowing
the intarweb to the rescue
I'm not buying what they are selling. "I dunno" is a lot better answer
wonder if my wife would notice that I bought $325 worth of servomotors
only if you share the bank account/credit card
"you bastard, where's my file?"
"you bastard, why did you spend $325 on servo motors?"
more like "you bastard, why did you spend $325"
"...on something besides me?"
she' mostly concerned about the rate of spending
that's what makes it so tricky :)
need a lowpass filter
that's what paypal is supposed to be for, but I'm not selling enough junk
jmk: what do you need a lowpass filter for?
everyone needs a lowpass filter
wholepair: it filters out spikes in cash flow
oh - I def don't need one of those then - never have spikes in my cash flow
we're talking about outgo, not income
might be able to use a shielded and grounded wallet then -
no need for that, just coat the thing in 20' of duct tape
or put all your money in a ferrite core
it would fi
I saw a thread on cnczone where someone proposed running mach3 in wine on a linux box - anyone ever tried it? anythoughts?
really bad idea
really really really bad
might sorta work
uh, work for what?
work at all
as in spit out pulses
I would never do it - just wondered what your thoughts were - in my endeavors to setup a laptop with emc for replaceing my desktop I compiled a kernel and emc2.2.2 using instructions from here: http://www.britishideas.com/2007/11/07/how-to-install-emc2-on-unbuntu-710-gutsy-from-scratch/
laptops often have bad realtime performance so you can't actually run any machinery with them
generalization, but probably true /\
thats what I keep hearing and reading but I thought I would play around with it for a while - as a learning experience -
I get an unexpected realtime error when emc2 starts, but it never seems to recur while running. This is ok, correct?
the error is suppressed for the rest of the run
it might be happening 50 (or 500) times a second
so latency tests?
I think it even says that in the long version of the message (in dmesg)
Hello, I am looking for a python script that generates gear profiles in gcode. Is there one out there?
Hugomatic: I don't know - maybe ask google - what do you need the profiles for?
(do you remember seeing one and just can't find it now?)
[03:05:35] <cradek> http://www.bookcase.com/library/software/msdos.desktop.cad.html
I've used "geargen" which appears to be on this page
I'd like to make gears on my mill, for an eventual robotic arm...
it was a long time ago
this may or may not be the one I used.
it may have generated dxf or something else that I converted to gcode afterward
and of course it's not python
thanks.... I have come accross it. Its not open source either. This could be a nice EMC project for me
use a gear cutter - I don't understand how one can make real gears on a mill by interpolating - unless they are uber big
wholepair: you are probably right. Except for toys, I'd have to get such a cutter.
if you have wire edm I bet you can make all sorts of gears
ah - yes - then a gcode program would be benificial
this was the end result of my laptop tests - older desktop on left - newer laptop on right -> http://www.uoregon.edu/~jgarman/latency2.png
yay for old desktops
overruns are bad I hear - I wont test that knowledge though, and just stick to the desktop
what happens when the supply of old desktops with parrallel ports is exhausted - what other hardware is EMC2 been used with -
my bridgeport is run by a computer made in '83. I don't see any reason why, once it's converted to EMC, I can't use my P3 to run it for 25 more years
so other people are worrying about this more than I am
(most critically, people selling parport based hardware)
wholepair: there are also parport boards that plug int pci and pci-e slots
yeah - im just thinking outloud - I was looking at this the other day and thought it would be an imrovement to the parralel port: http://www.ethernut.de/en/hardware/index.html
and its opensource -
bios issues in new computers may kill desktop realtime before the parport shortage is really a problem
power management and other things have probably killed it already
can't predict the future
microsoft+intel+RIAA+MPAA may try to kill linux
steves_logging is now known as steve_stallings
hi chris, and everyone...
been a while since I caught up enough to interact with current threads
how is your shop coming?
well, get with the program(s)
"caught up" being a relative term
(though I'm in the same boat - been traveling lately)
steve_stallings: I had a blast from the past this week - rememeber that linear motor I gave you?
shop for me is still a project waiting to get started, the new office is almost done, hope to move in this January
I found the little electronic boxes that go with the linear scales
jmk - yes, that motor is actually on a shelf that I pass by frequently
I'll try to remember to bring those boxes to the next workshop
they're lonely here
anyone here used a Compaq EVO DS510S (Intel 8456 chipset) with EMC? Is the on board video going to be an issue? Local used dealer has several cheap and the form factor seems favorable for demo machine.
you should get one and test it for us :)
a local used dealer here let me try my CD on some hardware
probably will get at least one, it can be a bench machine for testing boards with DOS utilities if nothing else
you can get a pretty good idea quickly
something I could try with a "live" CD easily?
boot it, run latency_test, run glxgears, jiggle some windows around. if it passes that you're really pretty sure it's fine
or, just run emc (stepper, parport) and jog around for a bit
they seem like a neat package, two real PCI slots laying sideways, real parallel, serial, network on MB
if you don't get the realtime warning, you're probably ok
wonder why enco has 3/16 but not 1/4 30-taper end mill holders
1/4 inch is an oddball size for endmills
3/8 and 3/16 are the most common shanks
my spot drills are 1/4, but I guess you're right I don't have much else that size
what I really need is 3 or 4 albrecht drill chucks on 30-taper adapters...
not if I see them first!
3 or 4?
oh, so you can set tool lengths
sure, he has quick change tooling
why not go for er collets then
fine for mills, but drills need to be all sorts of weird sizes, think drill for tapping
er collets are better than most for collapse range, but still not as quick as a drill chuck
if he wants a bunch of chucks, its because he doesn't intend to change drills often, and is going to be setting up a tool table with the lengths
even my mom knows my lathe is going to kick ass
50 amp plasma on the way and some 48" 1/2 inch ACME
I work in a small shop - have three axis mill run by emc and a tig welder - when ever I have emc running and the machine on and then go to weld the machine starts jumping around - (scary)
if the plasma wont cut my turbine discs properly im going to shit then order a laser
wholpair is that even normal ?
yes it is normal - there is noise getting through somewhere - so I never do both at the same time -
why is it scary ?
and why are you tig welding with a cnc ?
Im not tig wleding with a cnc - these are two seperate tools - sometimes I still have the machine on when I weld - but - get this - when I was testing a new laptop to run my machine with the problem went away -
it must have more copper in it
forget how that goes
brb check my cookies
it must have better shielding and filtering on the power supply -
or worse grounding on the case
enco's website is nice. I keep a shopping list in my 'cart' and when it gets over $50 (free shipping threshold) I order. Their system remembers the cart for you.
ie, a ground loop went away
the laptop probably eliminates a ground loop
what he said
oh good - then how do I elimanate the ground loop on the desktop computer?
find it, then get rid of it ;)
how about if I just run an extension cord from the neigbors...
biiiiiig loop ;)
you want the other extreme - desktop power coming from the exact same place as the rest of the power and grounding
we had the craziest problem with a ground loop using an a/d converter
we fixed it by plugging everything into the same power strip
I do have the desktop power coming from the same service box in the shop - its on a 20 amp breaker - the tig welder is on a 200 amp
I cant plug the welder into the same power strip - not an option
you don't want to do that
and make sure that shield cables are grounded only at the PC / controller end
all power for the mill needs to come from the same place
you want the milling machine and PC plugged into the same source
not the welder and the PC
power should come from the sun
Thom Hartmann "The Last Hours of Ancient Sunlight"
welder probably dumping too much current into ground/neutral
does the tig have HF start?
his thesis is that all power does come from the sun (even nuclear, I asked him)
that's a little silly
I think he was giving me the brush-off because some cute 20-something was gushing all over his book
yes, I thought he was a bit stupid at that point
yes it does have an HF start - which pin is the shield ussually on?
the sun has rays of energy for growing wonderfull memories
SWPadnos: which sun?
shield of what?
shield from my death laser
his main thesis is pretty reasonable, that oil/gas/coal are basically sequestered sunlight, and we're using it all up
* Unit41 puts it to a sharks head
shield of the cable is ussually connected to one of the pins right?
thats a pretty generic statement
depends on the cable, the connector, and the intended use
the parrallel port cable -
better connectors have shields themselves, and the cable shield goes to the connector shield
the shield should be connected at the control end only, and should not connect to the case/grounding at the motor/machine end
don't you want to break that shield connection at the machine end?
so you aren't connecting to ground via the machine chassis
but worst case is no connection at all
I think connection at both ends may be worse at times
but I certainly could be wrong about that
you are correct, sir, it depends
SWPadnos: certainly can be worse sometimes
wholepair - do you have optoisolated inputs on your motor drivers, or on your limits switch inputs?
shielding and noise is a combination of art and science
but it's mostly voodoo
the system uses a cambel break out board - with opto isolation
connected on one end is best because it ignores the voodoo
OK, that isolates the limit inputs, now what motor drivers are you using
OK, you can acheive a reasonable isolation and shielding setup with care
I recommend that the limit switch wiring be shielded with the shield connected to the ground of the Campbell board
fruitcakes will have to wait for Christmas, best effort at noise rejection can be attempted now
SWPadnos: you know of any simple windows utility to do snmp walks by chance?
I was just listening to http://www.19.5degs.com/lyrics/Fruitcakes/86485
the step and direction signals from the Campbell board are optoisolated at the inputs of the Gecko drivers
I just cut the drain wires on my limit switch cables thinking that shielding wasn't really needed for that
I think 3M or HP has some, but I don't think it's necessarily simple, and I'm pretty sure it's not free (by a long shot)
SWPadnos: this is just for testing, nothing like openview
ther emay be downloadable demos, but I don't expect them to fall into the "simple" category
the negative side of the filter capacitor for the motor power feeding the Gecko drivers is a good point to choose for a single point ground
SWPadnos: yeah, everything I've foudn so far is just overkill.
this is complicated by the fact that the computer and the machine driver both need a safety ground
steve: single point ground for what? the limit switches
you should try to power the computer, machine controller, computer monitor, and any other computer connected devices from the same AC mains circuit and same safety ground
Jymmm, google found this page http://pcwin.com/software/Snmp_Walk/index-1.htm
the single point ground should connect to the ground signals of the Campbell board, the shields of the limit switch cables, and the shield of the encoder cables
SWPadnos: I tried it, great if you happen to KNOW the OID's =) Just ofund this... http://software.ccschmidt.de/itmpro.html
the shields should NOT connect to the machine frame ground
I was just looking at the campbell breakout board and there is a stpgnd and a pcgnd that is connected to nothing...
moment while I pull down the Campbell docs....
TP4 and TP2
SWPadnos: Found this too, but I just can't figure out how to display only remote ip, instead of local machine it's running on http://snmprg.sourceforge.net/
Jymm: go to one of those what is my ip websites
wholepair - I do not see the signal labels you mentioned in the manual and I do not have an actual board, can you look at the PDF manual and tell me where they are?
steve: hld on
pcgnd makes sense, stpgnd sounds like step ground, but that does not make sense
for rev 1.7 - looking at it now - http://www.campbelldesigns.com/files/breakout-user-guide.pdf
grounding and shielding are sort of black magic, sometimes more art than science
there are two main approaches when using PC's and breakout boards
first is total isolation, this avoids ground loops, but total isolation is a myth at the high frequencies involved in a arc welder or HF start TIG torch
the second approach is a good star ground, my favored approach - this may control noise better, but powering the PC and the machine controller from the same mains circuit and safety ground is essential
so - those two pins seem to be for testing - page eight of the above pdf
it is often impossible to get the actual machine frame on the same safety ground as the PC because the machine may be powered from a 3 phase circuit, so I recommend not connecting shields to the machine frame as part of avoiding ground loops
and as to my earlier question it says on page 9 that pin 18 and 26 are pcground
OK, my comment about connecting the Campbell board ground would apply to just the limit switch grounds
the machine control box is on 110 and grounded to box - the pc is pluged in right under that - the motor is pluges into a single phase 220 - andso - I need a parralel port cable that doesnt have pin 18 and 26
the step and direction outputs will probably be +5 common and it does not matter since these signals are optoisolated the the Gecko drives anyway
wholepair - you parallel prot cable MUST have ground on pins 18 thru 25 (or 26 if considering a ribbon with pin 26 to shield)
which motor is 220 volts? spindle?
spindle - yes
is the box containing the breakout board and Geckos bolted (grounded) to the machine frame?
OK, collision of philosophi
OK, collision of ideas
my discussion assumed that the star ground, including the box, would not be connected to the machine frame
the real world presents situations where all the proposed rules cannot always be followed
I assume the spindle motor has a ground wire in its power cable?
spindle motor, metal or plastic
OK, for safety you need metal to metal contact from the spindle motor housing to the machine frame
that means that the machine frame must have a safety ground on the 220 circuit
and IMO the 220V power cable should provide the safety ground for the entire machine
ok - so what I am hearing is that the 110 the provides power and ground to the control system should not be conected to the enclosure?
the best situation is expensive, that is a 220 to 120 transformer in the machine to power the PC, monitor, and control box
first - how many pieces do you have?
machine frame is one
PC case is two
drive box is three
unless the PC is inside the drive box
is the drive box bolted to the machine frame? or separate?
is the PC mounted to or in the drive box, or separate?>
the pc is seperate - darned - I was so excited when the laptop didnt produce the same jitter while welding - but then it turned out to be a bad realtime candidate
so you have three pieces?
wholepair - do you have a multimeter and know how to use it?
two pieces - but the encloser that is connected to the machine has two cables - one 110 for the control circuitry and one 220 for the motor - the motor is turned on forward and reverse by relays on the breakout board
I have one crapy multimeter - and I can do basic stuf with it - voltage/current/resistance/continuity
so there is a box bolted to the machine, with two line cords, one for 220 and one for 120, plus a cable going to the computer which is separate
and the computer has its own power cord
is that correct?
you wind up with three plugs, one 220 and two 120?
when I run glxgears and move windows around I still only get 35 microseconds jitter, too much?
thats a little more than ideal for software stepping, but fine for servos
jmkasunich: yes, three plugs... I found an old picture of my machine, http://www.cnczone.com/forums/showpost.php?p=167248&postcount=9
so why does starting up emc give me the warning
it will work for software stepping, but your max speed won't be super high
I'm using m5i20
eric_Unter: I don't know - the actual numbers that caused that message are in the system log, reading that might shed some light
wholepair - two issues here, safety and noise, they often conflict
you can break the ground loop between the computer and the machine frame, but it take work and may pose risks if you have grounding problems
you can use the multimeter to test for continuity between the PC' safety ground to the 220 ground while everything is unplugged from the power mains
personally I'd try to tie everything to the machine
I do agree with JMK here, just explaining the alternate approach that some people take
another observation - looking at the photo you posted - all the wires going to the geckos are bundled together
encoders, step/dir, power, and motor all together isn't good
you did say gecko 320s, right? servo drives?
we seem to be far down the path, and as I look back I do not see answers to some other questions
yes - 320's
are the encoder cables shielded?
don't look that way
encoder cable shields should not connect to machine frame, only to Gecko negative power rail
every single cable run should be looked at, and grounded as appropriate for _that_ run - there are no one-size-fits-all solutions
OT - here are some pictures of my wakeup call this morning at 7 AM, right in front of my house - http://www.metalworking.com/personal/
no one really hurt, but it is a major problem, rush hour traffic back up and the speeders coming up the hill don't see the stopped cars until too late
they seem to never learn
powering down for a second
see ya in a bit maybe
steve_stallings is now known as steves_logging
cradek, jepler, how did the storm go?
Guest191 is now known as skunkworks_
skunkworks_: not too severe here
jepler: is there a reason you named it latenc-test instead of latency_test?
you know what I mean ;)
I though I was the only one that first typed it in latency_test - now I see cradek did the same for steve stallings.
skunkworks_: apparently I was just trying to trick everyone..
cradek_ is now known as cradek
lewing is now known as lewing_
lewin1 is now known as lewing
skunkworks__ is now known as skunkworks_
Dallu1 is now known as Dallur
I need some help..
Does anyone have any config files for sherline lathe :P ??
I`m trying to config this and i have a lot of problems ;p
does sherline sell a cnc lathe now?
or is it homebrew?
I thought the nist lathe was sherline? wasn't it?
my company bought a CNC
SHERLINE 4410 ;p
Mestric couase i`m from europe ;]
skunkworks_: true, but I didn't think it was a stock item
They are giving BDi 4.51 and EMC1
i instaled EMC2 and i`m tring to start it
GNoMeK: interesting. there is no lathe specific support in EMC1
does the lathe come with some kind of sensor to sense the spindle position for cutting threads?
only x and z axis stearing
that's very disappointing
seems like they leave a lot of work for you to do
i think only monualmove can cutt theards ;]
it doesn't have gears, you can't cut threads
my boss bought it
and i need to start it ;p
what do you need it to do? If you need threading, you will have to buy additional hardware to retrofit the things it needs
but this lathe win not need to cutt theards ;]
i need to congigure EMC to metric, and all maximum axceletarion
i See you have Polish surename
there is a new tool called "stepconf" with the latest release of emc2
er... i guess it's on the cnc menu
yes but i tried to configure it, i used sherline putputs
and when i test axis it doesn`t move
So why i asked if anyone have a complete config file ;]
if not i will keep trying
tommorow at work ;]
But i`m kind of tired doing it all time ;]
maybe you do not have the right pinout. none of us know how it is wired, so you will have to find that out
cradek: sherline is still shipping EMC1 so its not surprising it doesn't do threading
probably X is on pins 2,3 but Z may be 4,5 or 6,7
fenn: yeah I know. I wondered if they at least had the necessary hardware on it.
There is an option Sherline outputs i tried this
[19:57:21] <skunkworks_> http://www.electronicsam.com/images/KandT/cncworkshop/nistlathe.JPG
i will need to chand it and try
GNoMeK: yes but it may not work for the lathe. We don't know how it's wired.
i think that encoder's kinda overkill :)
fenn: not to mention the six home/limit switches
Skunworks: mylathe loks loke this
cradek: what did that have for drives and such?
a commercial driver box called 'mighty drive'
but you don't think that came with it originally?
don't think so
* archivist likes the swarf holes on the limit switches
archivist: actually they are potted I think
reminds me of here at work , 3 phase 440 at floor level
03cradek 07TRUNK * 10emc2/src/emc/rs274ngc/ (interp_read.cc interp_o_word.cc interp_internal.hh): repeat/endrepeat O words to make simple loops simpler than using while
What is it connected to the motor on thephoto ?
neer motor ??
next to motor(engine) i dont know how to name it
on the spindle? that's an encoder which senses the spindle position. this is needed for cutting threads
and its connected to computer ??
Sorry for the stupid questions ;]
the computer reads the spindle position and follows it with the Z axis to cut threads
but i`m kind of knew with mini lathes
is it limited by stepper motor speed or the parport read frequency?
you can think of it as "electronic gearing"
AND EMC can handel it ?
fenn: depends on the thread pitch
GNoMeK: emc2 can do this, yes
Tell me one think i gotcontrolecontrolig 4 axes so how i connecnt this to computer ??
fenn: I think I was reading that encoder in software up to about 1krpm, nearing top speed in low gearing
GNoMeK: I don't understand your question
I have controler with my lathe where i can connect 4 axis, x,y,z,a. And i connect it whit LPT to computer, and how i connect encoder to computer?? to conotrl spindle ??
the controller was designed to run stepper motors so it probably doesn't have any inputs
there are several ways to connect an encoder but it's a complicated topic. if you do not need this, I recommend you don't worry about it at first
fenn is right, it will require different hardware or wiring
ok now i need to configure my lathe then i will think about sth else
you could call sherline and ask them for the pinout of the parallel port driver, to be sure how to connect it in stepconf
I think they should help you with this if the machine is new.
My boss bought it from reseler so they told my to call resler and he is not responding for my emails
"Because of software support considerations, the complete CNC systems with computer are available only by direct order from Sherline." <- wonder what that means
When i use example nistlathe i can control it manually
oh? that's very promising
now i need to change it into metric
maybe i will try it in this way
does your lathe have metric lead screws?
OK then the scale will be slightly wrong, you are right, you will have to change it
I`m from Europe Poland
with cnc it doesn't really matter whether the screws are metric or english
(but you do have to set the scale properly)
but i have some errors with nistlathe so i will try to change it
i will make a program and try to do it
good luck, come back with more specific questions when you get farther along
tel me what time at your country ??
if you put G21 in your program, you can program in mm, even though the machine is configured as inch
it is noon to 4pm in the US
2:18 pm here
9 pm here ;p
lewin1 is now known as lewing
Ok i wille try tommorow
I hope you get it working
is this it ? http://www.sherline.com/8760pg.htm
Thank you very much
they give pinouts
wonder what C1 and C2 are
Shunkwors : thanks i will try it
in emc2 there is sherline output to choose and i tried it
i will se if its like in sherline website
yikes - 600 for 4 L297-298 drives..
I think they are not L298 because they say unipolar
ah - even worse ;)
the complete package is $2225
* fenn thinks he could buy a nice car with that
well a car anyway
i wonder how much i've spent on lathes in the last five years..
Ok thank you guys ;p
i think i will show up tommorow with more questions ;]
bye bye guys ;]
and thanks one more time
lurk and see the various questions
[20:29:11] <fenn> http://fennetic.net/machines/?lathe_construction_costs
yikes - fenn has a spreadsheet
i do.. somewhere
you and my wife would get along well..
it's supposed to be attached to that page, but it isnt
wheelium and pistonium ;)
it is much better than canium
here is the real list of costs: http://fennetic.net/pub/irc/lathe-bom-sort.html
or as it is called down south - sodacanium
most expensive single item was shipping, followed closely by a box of screws
sorry - I think it is cokecanium
in the book he doesn't use a micrometer or anything, just plain calipers
not even a drill press
metal casting has got the be the most satisfying hobby per dollar
er, per dollar*dolor
I bet.. dad now had a bunch of grandpa's casting equipment..
or whatever utilitarianist unit of effort you choose
I don't know if he has all the paterns for the mill and shaper still
hi againg :P i have one more question
you're better off doing lost foam anyway
now - yes.. It would be easy with even a crappy 3 axis mill.
then you get square sides, castings to size, no icky sand-pounding and schlepping
GNoMeK: jump in
Anyone have ant parameters like max acceleration and such thinks for sherline lathe ???
i will need it to configure stepconf
look in the nist-lathe.ini
I would start with the nist lathe settings...
what fenn said ;)
ok i will try it
ok now i realy gone ;p lol
those values are inches/s^2
you want mm/s^2
so multiply by 25.4
i will try and tommorow i will write if its work :]
Ok bye ;]
i used to think the "hanson" brand taps at the hardware store were "the good stuff" that old farts always said you were supposed to get, but i think they really suck now that i've used real good tap
heh - it is the best you can usually buy locally
the local farm and barn sells them here
fenn: I'm surprised a machine shop would touch anything for $20. did you know someone or did you just walk in off the street?
apparently they do a lot of odd jobs for the university, geologists and stuff
i called around and usually the answer was "well we can do it, but it'll be $400 to stop the current run"
[21:06:15] <skunkworks_> http://www.electronicsam.com/images/KandT/cncworkshop/fennalum.JPG
I understand they're busy, but "we'll call you when we have time to do it for a reasonable cost" would be nice instead.
surely everyone has busy times and idle times
yeah they just said 'put it on the shelf and we'll call you when its ready'
real simple though, i just needed a casting fly-cut
maybe or maybe not... depends on how hard it is to mount to the mill table
i was doing it on a drill press and a piece of plate glass, and the glass was bending enough to make it dig in
could be 5 minutes or a half hour
so i decided to play it safe
we do the same thing for heavy machine moving.. Call them up and say - whenever you have some down time - this is what we need. (usually get a good deal)
that was when i learned that you can cut glass with a carbide bit
I don't understand how you would use glass to mill on a drill press
just to make the table bigger so you can slide the casting along?
* skunkworks_ wasn't getting the picture either
yeah pretty much
the regular table has those slots in it, and high spots fall into them
doesn't sound very easy.
so you want a totally flat surface
oh, I see
but the bed was 24" long, and so the weight of the casting (and a piece of metal i had bolted to it to reduce vibration/keep it from snagging) would bend the glass
24"+ of travel is a fairly large table to do it in one pass
(my not-tiny mill is only 18)
in hind-sight i should have just put down some wax paper and epoxy and smoosh the casting into the epoxy
actually, just use a steel tube instead :P
i jack and pack and clean a side first
archivist: this is an order of magnitude more primitive
think whitworth plates
jack and pack means use of bits of scrap
apparently plate glass isn't perfectly flat - it follows the curvature of earth's gravitational field
if made by the float process
never thought about that but it's obvious if you imagine a piece of float glass on the ocean
I like my $2x surface plate
got mine from army surplus
I think the amount of curvature on a 3' piece of glass is less significant than the surface finish
ie, if you polish it with 100000 grit polish, the variation in taht 1/100000 inch is greater than the curvature
03seb 07TRUNK * 10emc2/src/hal/drivers/mesa7i43-firmware/ (7 files): initial commit of upstream files from Mesa Electronics
03seb 07TRUNK * 10emc2/src/hal/drivers/mesa7i43-firmware/SOURCE/ (6 files): initial checkin of upstream source for the eppio fpga firmware, from Mesa Electronics
03seb 07TRUNK * 10emc2/src/hal/drivers/ (mesa7i43-gpio.h mesa7i43_gpio.comp Submakefile): new driver for the Mesa 7i43 FPGA board with EPP interface
just for reference is 100000 grit finer then house hold dust?
dust varies a lot
isn't there a "RMS" figure for typical/average dust size?
.01 to 100 micron == 0.4 to 4000 microinch (the usual RMS unit)
but what is the typical/average size of dust?
you should ask google/yahoo answers
according to this, the number of particles goes up at the low end of the range (0.3-0.5um) http://www.chps.net/info/iaq_papers/PaperIV.8.pdf
but it might go higher at even smaller sizes
hmmmm final polishing with a dusty rag ;)
this may interest anyone looking for the size of dust particles: http://hypertextbook.com/facts/2003/MarinaBolotovsky.shtml
but unfortunately it doesn't actually answer the question
the best answer probally is that it varies but down in a certain range, I just wonder if you are better off polishing with a dusty cloth
no, if you try to polish with mixed grit you just get scratches
and dust doesn't cut most things
dust isn't necessarily sand (or silica/quartz)
so you can't polish hard-ish things with it
seem to remember somewhere that dust contain a lot of hard stuff (garnet bits, etc)
you can make fumed silica with some gaseous decomposition reaction
* ds2 does not like to think about stuff that small :/
turns out that you shouldn't try to spin your brushless motor by hand if it's powered up
or what, it doesn't move?
turns out that there is finite probability your middle finger will get smashed flattr than a pancake
did you get it running yet?
and that will cause problems typing the letter e, the most frequently used letter in the English language
I did get it running
good, then you can cnc a prosthetic finger :)
middle finger hurt so much I figured I might as well figure out why it wasn't working
i discovered a couple days ago that carbide lathe tools really can cut through skin with enough force
inserts i mean
was it on a lathe?
i was tapping in the chuck, and had the tool post parked in front of the chuck so i could get the tailstock close enough
and somehow my finger got wedged between the chuck and the tool
hate that :)
was just spinning the chuck by hand, fortunately
03seb 07TRUNK * 10emc2/src/hal/drivers/rbf2h.py:
Make LGPL work, and a minor cosmetic fix for GPL.
(I discovered these when shamelessly stealing rbf2h.py for mesa7i43's
split my thumbnail about halfway up
I was thinking wrong about my home switch locations
they should probably be near the center of travel if possible
except on the spindle
what if it's on the wrong side of the switch?
switch gotta stay on
that's the only limitation
so the switch is on for half of the travel?
you can't use it as a limit then ;)
limits are going to throw a contactor and depower machine, so no sharing
the mazak has a home switch in the center, (for x at least) but it seems to be "momentary" or whatever you call it
it's not odd.. older robots all did it like that
home switch on for half the travel
nowadays they don't have a limit switch at all ;)
no more homing is actually lots nicer
03seb 07TRUNK * 10emc2/debian/changelog: Added a note to the changelog about the new 7i43 driver, added myself to AUTHORS
03seb 07TRUNK * 10emc2/docs/AUTHORS: Added a note to the changelog about the new 7i43 driver, added myself to AUTHORS
they remember the position inside a revolution, when you power up and the position still matches, it assumes nothing happend, and the old position is good to go
if it doesn't match, you need to move it manually close to a predefined position, and searches for the zero pos nearby
I was thinking resolvers had been dropped
eric__u: why would you think that?
just an observation from servo motor market
I hardly saw AC motors without resolvers
servo's I mean..
most have encoders now
resolvers are better in crappy enviroment
eric__u: we used to have encoders on older motors
don't see the difference really, neither see the environment
eric__u: oh.. slap it with a hammer, and you'll notice the difference :D
try running an encoder submerced in oil ;)
drop it, maybe
i wonder if there's anything to gene's story about hall sensors dying when exposed to wd-40
cant see it myself
I have some that aren't that well sealed though
i mean for stuff that isn't some weird ceramic chip
eric__u: the main advantage of a resolver against an encoder imo is that you know the current position
you only get that with abs encoders..
i'm thinking about using these magnetic absolute position sensors (based on multiple hall effect sensors)
(and usually they use the resolver in the drive to be able to drive the motor..)
they are more expensive though
eric__u: might be.. but so are AC servos vs. steppers :P
the chips are $10 and you get 12 bits per rev at 10kHz sample rate
if you need them.. gotta have them
fenn: managed to rig something up with the samples you got?
but that's just me
are your robots direct drive?
JMK work for Baldor now?
alex_joni: just drive your stepper with an AC servo drive :)
eric__u: rack & pinion for linear axes
and helical gears for rotaries usually
* fenn is still trying to figure out how harmonic drives work
[22:46:02] <alex_joni> http://www.roymech.co.uk/Useful_Tables/Drive/Harmonic_Gears.html
the rigid ones are quite similar :)
* alex_joni heads to bed
good night all
Newbie problem. Have Ubuntu dapper and emc 2-2.0.5 uptodate. Compiles ok. Wont run stepper mm due to shmget failure. ??
whoah that's an old version...
did you install the package, or from the linuxcnc cd, or from source?
don't remember. This was an interrupted project about 6 months old.
well that makes it pretty hard for me to guess what's wrong
[23:06:51] <cradek> http://www.linuxcnc.org/content/view/21/4/lang,en/
you could get the latest cd here, and you'll have emc2.2, and it'll work
Sorry, I noticed that you and JMK had found a fixed a lot of these before, so I was hoping. If I keep dapper, should I just load a newer version?
this cd is still based on dapper since it's the current LTS release
you should be able to go to linuxcnc.org and run the install script
probably quicker than downloading the entire cd
but it's hard to say what state the machine is in currently
(if you don't remember anything about how it got there)
I have to run, thanks for helping fenn
Thanks. I'll try that. All I know is that SPM is happy.
fenn? down load emc2 from linuxorg. says it is up to date. rebooted. same action. Have both emc2 and emc2-2.0.5 directories, as intention to modify code. Will they confuse each other?
did you update the repositories to point to emc2.2?
will, but how
[23:41:02] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.2