* fenn nudges kanzure
[00:11:38] <kanzure> http://designfiles.org/~bryan/photos/2010-05-28_puma500.jpg
i bought that today
and need to figure out what i'm doing
some more pics (no he didnt pay that much) http://cgi.ebay.com/Unimate-PUMA500-Robot-Arm-JR3-67M25A-U560-hand-/290373583741?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item439b9d437d
i think it would make a fine coat rack
other pics: http://designfiles.org/~bryan/photos/puma560/
sorry wrong window
fail fail fail
fenn: Here's a good use for that... http://www.youtube.com/watch?v=92Ye2Q8U-rY
should this work or am i wrong. net estop-loop hm2_7i43.0.gpio.024.in_not iocontrol.0.emc-enable-in
Jymmm: splattering nerd brains all over my concrete floor?
how else are you going to get nerd brains out
jpm: Looks OK, what is it meant to do?
anyone have any magic tricks to get carbide drills out of steel? :)
torch and LN2
heat the steel, it'll expand
The LN2 is only for fun =)
andypugh: Wuss, that's too easy
I have drilled through from the other side and punched the stub out once. But I guess if that was easy you wouldn't have been using a carbide drill, or have broken it.
celeron55: didn't know that .. I have been google avoidant.. which is similar to being facebook avoidant
not quite sure why the drill failed, we drilled the rest of the part at higher but close speeds/feeds, was "fun" crushing the drill this morning, drilling from other side is not really an option, so i removed as much material around it as i could, eventually i was able to force it out sideways, luckily it was the middle hole of 3 holes predrilling a pocket :)
but crushing is such an annoying practice in steel, in stainless, they really break and you blow them out junk by junk, in steel, it embedds them :>
* kanzure is also in austin
oh, so you are
2012, Fomula One for you.
but that aside, just cool to see people in texas :)
* Endeavour is in Austin
Endeavour: we should meet up some time
feel free to call me.. 512 203 0507
i'm one of the guys doing the austin hackerspace
[01:06:53] <kanzure> http://austinhackerspace.com/
I'm working on CNC and 3D printing stuff.
Sorry, 'www.austinhackerspace.com' does not exist or is not available.
[01:07:36] <kanzure> http://austinhackerspace.org/
note the giant, unfinished mechmate in the corner of the pan pic :/
kanzure I note a complete lack of floor planning
there's a floor plan but the guy who has it doesn't want to post it up on the netz
Netz, german for net :P
my new keyboard claims to keep out germans
but it'll work in a sandstorm
while I was reading the instructions I thought I'd have paid what I paid for it just for the instructions I was laughing so hard
well let me finish up this buffer over here I'm dting to see if it still runs or not
Can anyone give me some advice on "* Building EMC2 realtime", as on the wiki toward the bottom of http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Rastering_With_A_Laser
What machines do you guys here have?
I haven't done much with building my own EMC2 using "make" and "--enable-run-in-place" and so forth. I was following the above link's build procedure, and I was OK until "make", when I got back:
Makefile:31: Makefile.inc: No such file or directory <newline> Makefile:42: *** Makefile.inc must specify RTPREFIX and other variables. Stop.
Any idea what I'm doing wrong? Or forgot to do?
Is there any positional feedback in EMC2?
Endeavour: There can be, if you wish
Has it been done?
Oh yes, many times
Is it built in, or is there some jury-rigging to be done?
In fact, in one case that I know of, feedback *twice*! See http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Combining_Two_Feedback_Devices_On_One_Axis
Of course, most people only use one feedback device per axis. Or none, if they're stepping. There's also the half-feedback Gecko step/servo drive thing, but that's going to the trouble and expense of a servo drive with fewer benefits from EMC2, in my opinion.
And some people keep an eye on their steppers with feedback.
On my earlier Q, I'm getting a little closer to the answer, but John needs to make parts so I'll go to the office. The problem may have to do with me not yet running the emc2 install script. I've downloaded it and made 2.3 and 2.4 versions, but I'm afraid to switch back to 2.4 apt-get sources without advice, since this machine is one of the 2.4.0 parallel port victims. BRB...
I'm hoping to build my own v2.4_branch ("2.4.1"?) to bypass 2.4.0 and the parallel port issue on this PC.
Running 2.3.5 standard version in the meantime.
I'm working on 3D printing. I've had some issue with the relative axis positioning
If a motor skips or something gets briefly stuck, the rest of the print is ruined.
Feedback would be nice to have.
You are stepping right now, then?
Yes, we plan to add feedback as soon as we can.
Just laying the foundations and trying to figure out how best to tackle the problem.
EMC2 seems far superior to RepRap.
Ha, EMC2 is far superior to just about any motion control software, IMO. And it's free/open!
Endeavour if you are printing why are you losing steps?
pfred1: I've had the extruder sometimes catch on stuff.
that'd be a problem with the file you are running then wouldn't it?
Yes and no.
well I'd think you would want the extruder above the work
The main thing is if a motor skips, somehow, it would be good to have a means of self-correcting.
most stepper people just throw bigger motors at it and call it a day
once you go closed loop steppers are sort of dumb to keep on dealing with
The extruder generally is, but you can end up with flecks of stuff sometimes the are a little bit further above the work layer. I've had a print jam up because of that and end up frame shifted.
how strong are the motors you are using? right now they sound a bit marginal to me
I mean if a fleck stalls it out
seriously the simplest solution is always the best
closed loop is a lot more involved than bolting on stronger motors
I just got a pair of 220 oz/in double stack nema 23s for $14 a piece
and they'll plow right through your run of the mill fleck
It's not quite as you're imagining it. The jamming is fairly rare, but is possible. My primary concern is if, for whatever reason, the stepper skips, that it doesn't ruin the print.
to the tune of about 300 pounds of force at the lead
just toss way more power on your lead than you'll ever need
all I'm saying is with steppers what I'm saying is the popular solution
once you go closed loop you might as well run servos too
the only advantage steppers have is they run open loop
other than that they suck quite frankly
Also of concern is accuracy from layer to layer. Without absolute positioning, how do you know you're actually where you "think" you are?
how accurate are your leadscrews? are they fully mapped?
Going to be using Acmes on the new design.
til you move up to ground ballscrews I wouldn't sweat super accuracy yet
though some acmes are pretty accurate
they typically cost what some ballscrews do
only reason to use them is to negate backdrive
but as long as you don't miss steps things are working out for you right?
Yeah, should be.
thats the beauty of CNC it "just works"
So you don't like steppers - I don't have much familiarity with servos.
well steppers are like motors made in the land of the backwards
they're the strongest when they're not moving and the faster you go the less torque they have
and that to me is just plain silly
Would you recommend servos over the steppers? If so, what does that entail?
lots of money
Steppers are rather cheaper?
that is the big drawback to servos
Guess steppers it is then. They're already expensive enough.
or everyone would run them
I can get a motor and a drive for $25
If you don't get them surplus they're already around $50ea.
well yeah surplus of course
Surplus decent ones are $15ish, which isn't terrible.
that same $25 won't get you an encoder for a servo
So what's the advantage of servos?
they have full torque at their top speed
and they have absolute positional accuracy
because they are closed loop
but you pay for it
now if you're a big deal commercial outfit then it is well worth it
they do seem harder to setup too people talk about tuning which soulds like black magic to me
also at a certian power point steppers are impractable
you can get servos in multi kilowatt power output
my steppers are .001 of a KW :)
on the bright side most printers do have steppers in them
We'll see how it goes.
Main reason I was looking at absolute was the one jam I experienced with a different design, and then one fellow's very vocal concern and insistence on putting in absolute positioning.
I can see with a subtractive technique having issues but not in an additive one
like pushing a tool through work material OK then sure you could stall but if you are just moving your extruder it should never happen
if it does its not the stepper its a problem with the file you're running
or just a problem with your machine's mechanics
its called cutting air
when you are above the work
so what sort of resolution are you looking for?
like are you making a plastic extruder?
Yeah. 0.35mm layer thickness or so.
machine I am working on is using 1/2-10 lead and at that I have 16,000 steps an inch
As I said, someone kept on virulently insisting that absolute positioning was necessary and that the hysteresis of stepper motors was going to be a major accuracy problem.
* Endeavour shrugs
I didn't really think it would be such a big deal, but I figured I'd look into it.
If it's not, then I won't worry about it.
show them this: http://www.youtube.com/watch?v=cgbeyNNBZ68
thats my motor reversing at 833 RPM
software takes care of acceleration
Endeavour your friend is an idiot
You have no idea.
But he's not my friend.
or greater minds than his have already worked all of this out
but yeah the software ramps the motor up and down
and emc knows all about all of this
you can set your acceleration to whatever you want
Well, I'm certainly glad that part of work has been taken care of then. ;)
in that video it was set a bit high
I think it accelerated to full speed in a quarter of a second or something
was pretty cool though huh?
thats what you can get out of stepper motors if you're a hack in a garage
this is 2010 after all
maybe in the 1970s your friends concerns would have been more valid than they are today but we're beyond all of that now
these ain't your daddy's stepper motor drivers we're playing with anymore!
what kind of motor drivers do you have?
I bet if you had like a G540 you could rip up the track
Using some drivers from Allegro, probably.
oh you're building?
I forgot the exact number, I think it was A997 or something like that.
you read a lot about them?
in microstepping mode they do seem to have a little bit of positional accuracy issues
though its not cumulative
one rev is still one rev
but I think they have something like a 40% variance in some of the microstepping modes
which is abysmal
yeah i know it
That's what it was.
I wasn't aware of that.
well I built my own and i avoided that chip like the plague
I used a TB6560AHQ
thats what was in that video
Don't like the stuff from Allegro?
way easier way higher performance
I've built a driver using their SLA7026M
but really allegro seems to be getting further from hobbyiest market
their stuff is getting harder ot get and their later designs are geared towards automated assembly
you need heat vias on the boards and whatnot using them
which isn't garage hack friendly
but really you can get TB6560AHQ based Chinese boards for next to nothing now a days
maybe $60 for a 4 axis with built in parallel port interface
Is there a prefab board with that chip on it that you're aware of or have experience with?
sorta makes building your own an exercise in idiocy
ok I'll tell you the big secret
with TB6560s you hear they handle 40 volts
they'll blow up like chinese foreworks somewhere around 32 volts input
so just stay under that and you're golden
but with their decay mode you don't need to run higher voltage and get pretty good performance out of them
in the video I was running at 24 volts
I'm going to get 72 IP/M rapids on my 1/2-10 leads
solid forget about stalling speed
like 300 pounds of force
but I'm going to twin drive two of my axis
so double that
KimK, You around?
people use the chinese tb6560 boards and as long as you're not a moron about it they work fine
I might try that out.
Endeavour , davidf: Yes, just got back, sorry for the delay Endeavour, are you getting your questions answered?
Endeavour if you want to build yourself the chips cost $4.73 a piece at digikey
I'd prefer not if something viable already exists.
Endeavour and i can give you the schematic I made for my driver
but yeah don't be stupid just buiy a board
the other huge drawback to the chinese boards is the documentation
but if you just use your head and a multimeter you can figure out what is what
and the chip datasheet of course
I'll tell you the other goofy deal with it all
the chinese boards come with sense resistors that set them up for I think 1.25 amps out though they claim they're setup for 3 amps
but the right sense resistors aren't the easiest things to find!
you can get them for $1.90 a pop from digikey
davidf: How's it going?
you *NEED* non-inductive sense resistors
I got one that came with jumpers for 1.5/3amp, in theory
atmega yeah what i read over and over is they're mostly locked at 1.25
personally I wouldn't put in any resistor that'd drive it past 2 amps
say around .22 ohms
mine doesn't do anything if you put the jumper on thehigh current setting
I've ben doing that kicad tutorial. I think it is pretty good.
Have you played with it yet?
when i tried it it crashed out on me at the third step
Not much of a library though.
I'd like nothing more than to be able to use FOSS designing boards
but so far I'm addicted to Eagle
KimK did you try kicad yet?
honestly I don't want to make a board bigger than 3x5 anyways
you have no idea how much that limitation has made me design better boards
pfred I thought the free V was limited to 2 X 3. It's 3 X 5??
davidf: No, not since we last talked, haven't had time.
is it only 2x3?
nah let me measure one
Just finished watching A Few Good Men.
It is actually pretty darn good. The PCB part is hard to use till you learn the few hot keys, then it's real nice.
its about 3x4
Keep in mind that the free version of Eagle can *open* any size Eagle board. (In case you're receiving a design from someone using a paid version.)
will it print it?
all i can say is its a great size to make boards
I think so, it's just writing/saving that's an issue, I think, not sure though.
well eagle won't let you place the part in the board layout editor
From the Eagle site:
The following limitations apply to the EAGLE Light Edition in general:
* The useable board area is limited to 100 x 80 mm (4 x 3.2 inches).
* Only two signal layers can be used (Top and Bottom).
* The schematic editor can only create one sheet.
yeah its a bit bigger on teh 3 side
That's pretty good for free.. :)
It's a download right?
once what i did was I exported subcircuits in PNG glued them all together in the gimp then printed that to a PDF
then i redesigned the circuit to fit anyways
Oh too smart.
davidf: Yes, pretty good for free, and there's a defined upgrade path when needed, so I liked that, as we discussed earlier.
pfred1: Are any of those eBay boards particularly good or bad?
so again the board size limitation forced me to design a better board
There seem to be a few options.
So what about the libraries? Plenty of parts, terminals etc?
Endeavour I'm not even sure if the ebay boards are the best option
davidf: I think it's in Synaptic, you just need to click.
Would you recommend something else?
Endeavour look for ones with big diodes by the motor connectors I think they're the better ones
What do you think of those HobbyCNC boards?
they might block the BEMF or something
they're SLA70XX based unipolar
Endeavor: First I would look for a Parker Zeta 4. They are awesome if you can find a used one.
depends what motors you get which would be better the hobby boards will run a bit higher voltage
Endeavour: I'm told a lot of steppers and drives come from Keiling (spelling?)
I have three. Often they go for $100 to $200.00 Worth every cent.
davidf you're lucky to get those at $50 a piece
What's special about them?
super high volts
Work directly from the power line (120 V)
They put out up to 90 V
Probably overkill for this application. Using NEMA 17 steppers.
high volts with steppers means you can run faster and still have torque
No transformers needed. Thats a savings right there and super convenient. Excellent Docs from the website
davidf he's building a printer not running a bridgeport
Can't hurt though, and you'd have the capability of 4 amps at 90 V for big motors as well
True but it is worth taking a look. Never know
If the price is right, why not?
yeah amaze your friends by whipping past your limit switch and twisting your machine into a pretzel!
Nah, don't be silly.
there's a video on youtube where a guy is rapiding his machine and knocks a limit switch right off
[03:25:17] <Endeavour> http://cgi.ebay.com/4-Axis-CNC-Stepper-Motor-Driver-Controller-router-mill-/230480175819?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item35a9b062cb
I have that board.
They are perfectly well behaved.
I'll probably just use it for now.
That's not the fault of the driver I assure you.
It's the A3977 though.
They only do what you tell 'em.
do some googling about that and the step inaccuracy
that and I don't know they're surface mount 2.5 amp max and you really need to design a board well to get them to perform right
davidf well the joker was running mach3 too
I use these on neme 23's on my Taig Mill and 7 X 12 lathe. 60 ipm is no prob, and I can get way more but it is limited by the puter
John's BP2 is running 4 Kelings, I believe. They seem OK. KL-8078's, 24-80V, 2.8-7.8A in 8 increments: http://www.kelinginc.net/KLDriver.html
yeah my computer for whatever reason can't stream past 25 KHz step pulses
davidf whats doin?
though if i lie about the microstepping mode it can
but I can't use the 4X1-2 phase sequencing because of the pulse limit
it limits me to .8 I/S
It might have been a little better if John got the 9082's instead of the 8078's since we've got them cranked all the way up, but they're OK.
hey ther Valen
How is everything tomorrow afternoon?
Down under :)
Valen its winter by you now isn't it?
KimK, what driver is Johnrunning?
yeah getting cooler
i'm cleaning the house atm so i'm only "around"
pfred ther you go then :)
[03:34:30] <Endeavour> http://hydraraptor.blogspot.com/2009/07/lessons-from-a3977.html
davidf: this is our machine btw http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?JakeAndRussells
if you dont have your heart set on steppers
Endeavour if you're building its a sucky package to work with
[03:36:31] <davidf> http://hydraraptor.blogspot.com/2009/07/lessons-from-a3977.html
pfred1: That fellow had mixed results.
OOPS. BAD MOUSE CLICK. SORRY
Endeavour yeah the oscillograms are giving me nightmares I had issues with dealing with noise on my drivers
Endeavour if stepper motors are one thing they're noise pigs
pfred1: I'll use the board I have for testing. If it doesn't work out I'll get something else. Don't doubt you, but I'd prefer to avoid spending extra money if I can avoid it.
Yes. Yes they are. :(
a pair of coils switching on and off tend to make a lot of noise in the process
How is it standard inkjet printers are as quiet as they are?
I mean electrical noise
they're isolation mounted to rubber belts and they're not pushed too hard and they're well engineered to boot
davidf: Driver? You mean where are the step & direction pulses coming from?
Endeavour they've had a few million since then to get better at it
plus small steppers make small noises typically
I think that's a chain printer, but I'm not sure.
Certainly. Would be nice to get quieter steppers than those available, though.
Going to have to see if it's possible to dampen their noise.
I got a little stepper motor out of a floppy drive thing is pretty silent
of course it has no torque
Kimk Yeah I wondered what kind of arrangement is moving the Bridgeport axes. I didn't understand what you were talking about. Unfamiliar with the part numbers you mentioned
it can also spin up to silly fast speeds
I think I've gotten it to do like 5,000 RPM
Adding a flywheel can make a huge improvement in RPM sometimes.
bigger motors just seem to run slower for me
best I've ever gotten a larger motor to spin is around 2,500 RPM
but that is at 24 volts too
That's prety fast...
breathe on it and it stalls
most high speed spindles tend to be low torque
We'll see how the project ends up working out.
davidf: John is using the Universal Stepper Controller from Pico-Systems: http://www.pico-systems.com/univstep.html
I am told that it can be converted to servo use, though I am not sure if that is a factory operation or just a different driver. Here's the very similar servo version: http://www.pico-systems.com/univpwm.html
Thanks everyone for the info tonight.
[03:51:13] <pfred1> http://www.youtube.com/watch?v=GU2GaSMPxNI
I'll keep you all posted.
Moving away from steppers I've got one other issue I was bouncing off folks in here last night.
If anyone wants a giid deal on a Zeta4, here's one:
[03:52:09] <davidf> http://cgi.ebay.com/Parker-Hannifin-Compumotor-Zeta-4-Drive-/110532907906?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item19bc465382
good not glid. geez.
There are three heaters I need to control; it would be nice to use the PID algorithms built in to EMC2, but it seems like it would be a pain - too many pins on a parallel port would be required to interface with 10 temp sensors.
I might snap that up myself. $13.00 S/H
Endeavour dual p-port cards are $15
Yeah. I'm just thinking it might be easier to put in a small microcontroller to take care of housekeeping tasks and leave EMC2 to control the axes alone.
and they're usually better than the built in ports as far as voltage swing and current sinking go
Having a bunch of parallel ports seems somewhat kludgy. (Though adding in an extra micro isn't exactly elegant, either)
Endeavour: What voltage and current are the heaters, and what is the temperature feedback device?
works well with EMC2
KimK: 12V, for one of them, don't recall the others
Was going to be driving them via a FET connected to the voltage rail.
after that you're looking at masa I/O anything boards
Endeavour, What wattage elements?
The measurement would be provided by a temp sensor, or possibly a thermistor.
The Mesa stuff is all fairly pricey, isn't it?
davidf: Let me go check. (Why?)
OK, so these are all pretty small heaters we're talking about here?
well if you can't do what you have to do with a parallel port it is the next thing
and i don't think they're too bad for what they are
* KimK is a big Mesa fan
Endeavour: since my boss will pay, i find them inexpensive
220W Max for one of them, davidf, but I won't run it that high most of the time.
he throws an amazing amount of money at this, and then sometimes i have to cut down my order of taps from 4 to 2.....
well they are a tad more than a PCI p-port card thats for sure
* KimK Oops: not the Mesa 7i43 though. No parallel ports for me, thanks.
Heh, indeed, pfred1
You could use a solid state relay. For a low power element like that it should be pretty cheap. The one I use for my kiln element (25 Amps) was $25.00
I hear bad things about solid state relays
The nice thing there is they are optically coupled, and a three volt 10 mA control signal is all you need to drive it.
SWPadnos: btw, on the yaskawa drives, you had it 100% right, the guy's answer was C is a index pulse and then included a pdf with only page 5-28, as you already knew, only new info added was their actual torque.
I've used an SSR before. Either way, triggering a transistor or otherwise isn't a big deal.
The problem with the temp readings is that there are about 10 sensors.
SSR's are OK as long as any switched loads are properly supressed, snubbed, etc.
So then those would have to run through an A/D
The type I have is very reliable. I've been using the same relay for ten years on my kiln and it is 3 times the wattage of Endeavour 's
Would require a lot of ports.
any switched loads --> any switched inductive loads
KimK SSRs are OK as long as the SSR doesn't blow out
Actually It is a lot bigger than I really need.
KimK when they work whats to worry about?
Just need proper heat sink. These are set up for easy mount to a sink with 2 screws. Totally self contained in a sealed box (potted)
I'll see if I can find the mfg if you want.
pfred1: I'm a little confused, are you speaking for or against SSRs?
KimK I'm all for them until they go then they're not too swift
Heaters are not inductive, so that is not an issue. And, the way to do it right is to use zero-crossing swiching.
pfred1: Haha, OK
KimK but seriously if my health depended on it I think I'd go with a mechanical relay
Zero-crossing switching for resistive loads (heaters?) only.
You just need to sense the line voltage and then AND your on/off control signal with a signal that is high when the AC crosses zero volts.
Endeavour so I guess your printer is going to be a bit more than a hot glue gun on a threaded rod?
That avaoids switching the ssr with power on it. Then it goes fully on and in that stare the internal resistance is near zero, so it makes it real easy on the ssr.
pfred1, I respectfully disagree. They just need to be used properly and driven properly.
davidf hey its my health
yes point well taken.
I'll take a properly specced and tested solid state device over a mechanical one any day
Yay for our side. :)
not that i haven't seen mechanical relays do bad things as well but at least i can see them doing it
had enough switches weld and break to have done with them
pfred1, did you grok what I was saying about zero-crossing switching?
davidf don't know if I grokked it but sure I can see how that would be a big help to the device
pfred1: A wee bit, yes.
Anyway, so taking that many sensor readings seems unreasonable.
most SSR's will be zero crossing inherently any way
mercury wetted contacts go a long ways torawds aleviating pitting in mechanical relays too
Sorry thats an obscure reference to a Ray Bradury (?) I think, sci fi classic.
stranger in a strange land
Probably will end up just using a micro for housekeeping tasks like heating.
Groking was what the martian intuitives did to fully understand something. :)
Ah yes, thank you. Too many years ago.
Endeavour, I'd suggest that the micro and have it talk serial to a userspace thingie on the PC is probably your best bet
you can PWM the heaters then with triacs and such and get much nicer control and faster heat up
valen, true. They are common but just slightly more expensive.
Endeavour, please remind me what is this thing your are building?
Valen: That could work, but the only issue is that I'd like to have one piece of software doing everything.
davidf, all the SSR's i've seen are based on triacs, they cant turn off while current is flowing
davidf: 3D Printer.
Endeavour: thats not the linux way ;->
Valen: The RepRap unfortunately has like three pieces of software doing everything. Would like to avoid doing that.
Valen one shot one kill
Eh, I really don't like having to deal with all that stuff, I like having one UI with all the relevant features presented.
Endeavour, EMC is built up of a number of small chunks
the axis front end that you see is just that a front end to the underlying doohickeys and thingamabobs
Certainly, but the part that you use to operate everything is fairly centralized.
For a heater the ability to turn off 120 times per second will give plenty of accuracy.
Its not a problem to integrate your micro based heater controller to HAL and thence on to something like pyvcp
davidf: The heater accuracy isn't too critical. :)
(or just skip hal, its not gaining you much, but would be nice as an interlock)
For most applications a sample rate of one or two seconds is more that adequate.
Valen: I'll have to read more into that. I read the getting started guide, but this is all pretty new to me.
couldn't you do heat like it was spindle speed ocntrol or something?
Endeavour, that's actually whay I asked about wattage. Here's my suggestion FWIW
just say instead of RPMs it is degrees?
pfred1: Problem is getting temp data. Most sensors are analog voltages or digital signals.
davidf: 220 for one, 72 for the other.
emc takes spindle encoders
The third will be a hair dryer that's run with a PID to heat the ambient air.
the 4th will be a phone line with 911 in case of smoke
what was that old kernel error message printer is on fire?
Why not just use a bang-bang plain old electric stove-top temperature control unit? $10.00 at the appliance store. Calibrate it and hand-mark the dial , set it and forget it. You can monitor it with an off the shelf thermometer.
Heh. Mostly metal construction, and the heaters are never intended to get particularly hot; everything well below 100C
davidf thats what I was thinking why even tie it in
Reasonably accurate temperature control is important.
but i guess if they're doing thin layers they want the control
You can get reasonably accurate control that way.
If you get too hot you'll burn the plastic and ruin your extruder.
but lets say you're running thick then thin sections
Too cold and it won't melt, and you'll ruin the extruder too.
you might need to vary the heat in a job
Think about when you turn a pan of water just to barely simmering. That is like, + / - 2 degrees.
On the ambient and bed heaters, it's not quite so critical, but it's good to be able to have quick and easy control of it.
I can see how dynamic control could be needed
yah so use a PIC to handle the controlls, serial it to a python/hal module and feed it up from there
OK, so if you want real PID control, and analog only, you do a PID controller with one lm324 op amp and a few caps and resistors.
Easier to do it in software. ;)
I've done it myself. It is not all that hard to design a circuit to do that.
But then you have digital to convert.
not for me it isn't
No, certainly not, but if I can just write an equation in code it's nicer than adding hardware.
I got a drawer full of LM324s over here
Endeavour: Mesa has 4-input & 6-input analog inputs (-10/+10, etc.) available at a pretty reasonable price. Take a look at their 7i33 & 7i38 daughter cards (there may be others too) and then pick a controller, maybe from the 5i20, 5i22, 5i23 group?
Endeavour: You still have to get your temp sensor into +10/-10V format though.
But you will have analog sensor circuitry, A/D conversion, interfacing, etc etc. That is way more electronic overhead than a discrete Analog PID.
The Micro can take care of it.
Just read the analog voltage at the pin, which gives me a value from 0-1023
No, it can't.
Yes, it can. I've done it before.
most micros have built in ADC's
It can only see digital. You've got to convert the analog signal first.
no, they come with built in Analog to digital converters
Oh sorry. I misread. I was thinking computer.
and hardware PWM outputs
So then in code you can just turn it back to a basic decimal value
Run it through a PID algorithm
Turn it to digital again
Then spit it back out through an analog output.
Could somebody help me understand a mesa issue but most likley a hal issue
if your keen, you put some phat resistors and diodes onto the 240V Ac line and you use a schmidt trigger input pin to pickup the zero crossing
KimK: I'm not sure why I want to use a Mesa board?
(the wussy way out is to use an opto but eh)
then you turn the triacs attached to the various legs on at a % through the wave form
you could still use EMC's PID to do it all
Endeavour: Oh, well, I thought you wanted to do everything through EMC2, but maybe I misunderstood. PID tuning live on the fly through EMC2 is nice though.
Just take the sensor data and send it over to the machine via serial, Valen?
KimK: Ah, I see.
You wouldn't need the analog output though. The PID control can be a full on, full off signal, that turns on and off like say if you want 25% power, you just output a 25% duty cycle pulse train. That's fine for heaters.
Endeavour: And your Mesa board would no doubt serve as your motion interface as well.
you could do it that way, then you push that up to HAL from python, run it through the PID there and push the results back out
it wont be "real time" but then who cares for a heater right?
jpm: Go ahead
Yeah, doesn't matter for the heater.
and you can still use EMC's pid loop
KimK: My axes are already being driven by EMC2 via a parallel port to a driver board, though?
hell even use G code to manage the temperatures
make sure "cuts" on the heater lines have a slow feed rate ;->
Endeavour: Oh, OK, you have an existing machine, I thought this was in the planning stages.
KimK: No, it is still in planning, I'm just having a little trouble understanding.
It is true, PID doesn't care or evaen know waht the H*&^ it is running. It just needs sensor feedback and an output signal.
you do not *have* to use a parallel port
we dont use one at all
I'm not sure why I'd need a Mesa board to drive the steppers
mesa have hardware that will run steppers and servos generally better than a parallel port
and you can easily add ~100I/O lines to it
Oh, I see.
Valen mesa board link pls?
Well, any external step generator board will free up the PC from pulse generation and will be able to pulse a lot faster anyway, so that's a good thing.
I need to see this small miracle.
[04:40:54] <Valen> http://www.mesanet.com/
5i23,7i43 and 7I40 is what we have
Kim: I was testing my boards today 7i43 7i33 and 7i37 had everyting working great untill the outputs on my 7i37 just stopped working battled with it for a bit and thought it might be a wiring issue. Turns out i eneded up having to setp hm2_7i34.0.gpio.is_output true in my hal file
yeah that one screwed me over
davidf: Look at their "Anything I/O" series, thats the one.
dunno why you suddenly needed to change that
Why would it work for a bit out of the box and then just stop
Dont get me wrong i love the mesa emc combo just weird and would like to understand
jpm: Yeah, I don't know why that would suddenly change either. So what else were you changing at the time?
i have NFI ask PCW ;->
Well, I'm done with my research for the day.
Lots of stuff to digest.
Endeavour: Hope we helped. Come back anytime.
Thanks for your help guys, I appreciate it.
KimK i was trying to link an output to the estop state and an input for machine on
I'll probably build a DIY Li'l CNC machine first to get my feet wet with EMC2, then start seriously integrating it with this other project.
Endeavour, Maybe try the etch-a-scetch thing?
Endeavour: Well, that should be fun. Or you can just use the motors, doing what I heard one person call "just spinning post-it notes".
davidf: Sarcasm, or are you serious?
You have small nema 17's around right?
Endeavour: I think there are some YouTube videos with small steppers mounted on an Etch-A-Sketch.
Just attach one to each knob on an etch-a-scetch and draw away. :)
Ah, use it as a plotter.
yup. :) dirt cheap way to play.
And good visual feedback to make sure it is doing what you tell i to.
jpm: So your setup worked at first, then not. Did you use any kind of wizard or autoconfig to get started?
* KimK hasn't used those, but has heard they have improved a lot lately
KimK: my basic config is based of the 7i43 servo config and then added mpg, estop and machine on functions
KimK: it all does work verry well just a quirk and probably someting i did just wondered if anyone has had it happen to them
jpm: Were there some pins which were pre-defined in some way as inputs or outputs and that was OK until you hand-edited? (I'm kind of grasping at straws here...)
Endeavour, KimK pfred1 Valen Well now, here's an interesting bit of infomation. Do look. :)
[04:57:17] <davidf> http://axis.unpy.net/etchcnc
It's cradek .
KimK: ther was one gpio in the original config then i added about 5 or six more. when the outputs were added the seemed to work fine for about an hours worth of playing around. I cant remember what i did when they stopped working but they seem fine now
The etch a sketch is cool my son wants one
jpm dumb question, do you have any proto-board wiring in this?
Could it just be a bad connection?
Davidf: no i went to great trouble to wire this like i would a normal panel so all wires are crimped and ferruled in terminal blocks
Davidf: That was my first reaction my second was DAM i blew the mosfets
jpm well theres your problem
too much neatness causes problems ;->
jpm: Another thing you might check is the external +5 input on the daughter card(s). Sometimes the daughter cards can seem to be powered due to leakage from other circuits, but they're really not. They need proper +5.
Proper +5 as in not an open circuit.
I will probably setup a blog with pictures showing my setup tommorow
jpm: I'll look forward to it
yeah, the +5v through the ribbon isn't enough
I almost thought maybe a different firmware got loaded on the 7i43
jpm I doubt this is your problem but I,ve meant to mention it and keep forgetting...
I did hit the wrong config one time and selected the 200k version of the 7i43
A few years ago when I was setting up emc for the first time I kept getting intermittent unexplainable anomalies.
I went through the whole gammut of possibilities with cradek SWPadnos and jepler etc. for hours on end over several days time checking the computer, attached hardware, etc couldn't figure it out...
I am using the ribbon cables right now but am planning on externally powering the boards
finally got a different computer and all was fine so I figured it was just incompatible for some still undiscovered reason.
A year later It happened again on the "good" computer.
jpm I'd be looking at that, with 3 boards the voltage was like 3V or something coming out of the ribbon
I do have to say for the cost it's still a sweet setup
I had a DOH! moment when I realized I had a usb flash drive plugged in.
I will i just have to make up the cables
Davidf: The flash drive was causing you problems?
Strangely, it only caused problems like once in a blue moon, maybe five seconds out of 30 minutes running, and no periodicity at all.
Sometimes it would run flawlessly for a 1/2 hour, and than like, say on a 45 deg cut, the x asix would freeze, and the y would keep right on doing what it should.
Very weird. Stopped after I removed it.
Yes, that was it. I checked thoroughly, the problem returns witn the drive in, and never shows ip when it's out.
guess its loading the 5V line too much or something
I mention it because it can be extremely intermittent.
Why cant all the bits and bytes just get along LOL
pfred1 there arent any atm
Could be, or grabbing the uP occasionally for housekeeping I don't know, and don't care, just as long as I remember to take it out.
pfred1: I'll remember that parallel port from earlier; thanks.
Im just glad i figured it out somewhat and the board is fine
Endeavour yeah it has the EMC2 recommended chip on it
I use one
Someday I want to build a BIG cnc router, at least 4 feet by 8 feet.
well now I'm the proud owner of a vintage B&D heavy duty buffer
why? because i got it cheap
I should take a pic of it what a tank!
guy up the road was closing up his TV repair shop and retiring and selling a bunch of junk
so I bought a bunch of junk off him it was one of the items
* Endeavour wishes he had a tank.
* Endeavour bids on a Soviet T-90
oh "what a tank." Thought you said WITH a tank. LOL
its like 60s vintage all cast body
I have a grinder just like it
this thing was loaded with polish gunk
I had to strip it all down and clean it out with gum cutter
uhuh. That was what they all were in my youth back in ancient Rome...
yeah you know what its like all solid metal
I love old bomb power tools like that
And the fact that it still works at age 50 says something.
oh its running like its brand new now
I had the field winding out the works cleaned it up
Anyone going to the 2010 EMC fest in Ann Arbor, Michigan June 22-25? http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EMC_Fest_2010
took off its winter coat I call it
I had a summer job in an electric motor rewinding shop back in college
i have one motor here I
i have one motor here I'd like to rewind
One day, we got a huge ass 5 HP motor circa 1900 AD. Had a dated patent plate on it...
tesla himself might have made that one
That thing was HUGE. You've seen 5 HP motors nowadays. Not Giant.
being as he did his AC in 1897
though his motors were 90% efficient
This thing was like 30 inches long, 20 inches dia., Solid cast iron case, i inch shaft...
they had bigger horses back in those days
later all thanks for the help
but if that motor was made in 1900 it was just 3 years after AC was released
Anyway, we were told it had been running get this, continuously 24-7 from 1900 until the day we got it in "for maintenance."
was probably a westinghouse too
there's a lightbulb in a firehouse thats been on for over 100 years
There was n-o-t-h-i-n-g wrong with it. We cleaned it up, put new felt in the bearing wells, oiled it and sent it out.
AC motors are awesome
when you take them apart i mean whats to go wrong?
That thing ran for 85 years straight and had nothing needed.
its all about main bearing surface area
and it couldn't overheat it was just too massive
But do you think you could buy one today that would run continuously till 2100?
you can blow serious cash on nice motors
and supposedly we can alloy metals better today
You're right. They were bronze sleeve bearings. Very efficient if made properly.
really metallurgy took a huge leap due to the space age
before that people just winged it
Well these guys winged pretty well imo.
when we decided to shoot for the Moon though they knuckled down and started taking things a lot more seriously
just think they didn't even have high speed steel back then
I would take sitting in a tin can and being shot 100,000 mile pretty seriously too. :)
more like a TV dinner container
most of the lander was the thickness of 3 sheets of aluminum foil
I saw the friedship 7 capsule in St. Louis at the Planetarium.
it was the lunar lander that you had to be a complete nut to fly
they had to put pans on the floor when they built it because if you dropped a screwdriver it
they had to put pans on the floor when they built it because if you dropped a screwdriver it'd go right through the floor
niel and buzz wished they were in a tin can I'm sure
They have it on disply where you can clibm a ladder and look inside the cockpit. It is awe-inspiring to see the hand-wired, point-poit electronis dash with a gazillion tobble switches and analog meters looking just like something you would make in your basement in your wildest weekend project dream.
I saw a 60 minutes program where they got into a warehouse of soviet stuff that never got launched
Pretty amazing the race to the moon in hindsight. The chances of those guys not making it back from the moon were pretty high, look at Apollo 11!
it looked like a steam fitter took acid
13 made it
11 was the one that roasted on the pad
I got to ride on the earth lunar rover
No, that was Apollo 1, wasn't it?
Apollo 11 was fine.
They barely made it back
I thought it was before they landed
one burnt up on the pad
Each of those was very risky .. the technology was not quite there yet
hey we haven't been out of low earth orbit since
pfred1: You're right, Apollo 13's troubles began before they landed, so they never did.
if instead of the shuttles they kept on just ubilding sat 5s lots say we'd be better off today
Apollo 11 could have been avoided.
shuttles have the worst safety record of any vehicle
They had some weird design decisions.
The doors open inwards, so when the capsule as pressurized you couldn't open the door.
I lost track did only 2 or did 3 blow up?
And it was a pure oxygen environment, so everything burned up incredibly fast.
Even stuff that was normally non-flammable burned up in 100% oxygen.
You guys are really into this stuff.
Yes, that was before the extensive redesigns due to the Apollo 1 fire.
davidf you ever been to the Cape?
I'm not super into it. I just know random stuff.
davidf oh its cool its like visiting the pyramids you just know its historical and it seems almost as old now
[05:44:45] <Dave911> http://en.wikipedia.org/wiki/Apollo_1
Apollo 1 burned on the pad ..
like all the pads they launched from are long defunct
I've been to the Cape several times ...
I'd like to go see one of the last launches.
Particularly of Endeavour.
But before our presence in space ends I'd like to see one of the launches.
Dave911 amazing how they use something then thats it they just let it rot away
I went and saw a shuttle launch once ... and then went to visit the cape and the President's motorcade totally screwed things up
Many of them are shut down but they still launch spy satellites from there ..
Stopped traffic for hours ... you would not believe it ..
I got goddard by me they launch stuff
down by my favorite named plase assawoman
The pres made an unannounced visit and we waited for hours in traffic and finally abandoned the cape for that day and went to the beach south of there .
its really strange place its all these mud flats
one thing they don't need at the cape is any security
that place is crawling with gators so bad they must feed the things!
U live in Fl?
We had tickets to go see the launch of the first shuttle that blew up... we couldn't go and I am glad we didn't
we used to be able to see the shuttle go up from my aunts house
Ever seen coral castle?
where is that?
was asking pfred1 ...
my favorite beach in fl got wiped out in a hurricane
was called the bathtub
it had 6 concentric rings of coral radiating out from it
but its gone now
coral castle is a very interesting place from what I've heard...
but when it was there it was paradise
directly across from grand bahama
pfred1, see q above.
never heard of it
fl is a big state though
and I've been to most of it at one time or another
It's a place built long ago by some guy supposedly as a tribute to his girlfriend or something like that
its on land?
It has these gigantic structures carved from coral limestone..
is it down by the keys?
nobody knows how he did it
coral is pretty soft
Its like a gigantic rock garden with HUGE ass sculptures, that are supposedly IMPOSSIBE in scale to move or erect from solid stone.
No one has ever figured out how he did it.
hehe, he tore it down and moved it once too :)
hey if you want a mystery in stone figure this one out: http://www.world-mysteries.com/mpl_6.htm
[05:56:03] <Dave911> http://www.sciforums.com/showthread.php?t=65028
more about it
there's stuff carved into those stones so accurately I don't think we could do it today
yesh I've seen a tv doc on the first one.
its like they had laser CNC or something
no there's no explaining some stonework
don't you think it's possible that humans were incredibly advanced long ago, and somehow threw themselves into the stone-age?
Ever read a canticle for Liebowitz?
what blows my mind is written records are so short
oy, booting 188.8.131.52-rtai on this laptop for fun, yikes
I read the sequel too
Tell him pfred1
but if we, or kim jun eel, or i'manutjob of iran nuke this place, there won't be much left but glass after a few thousand years
I didn't know there was a sequel.
what i remember most about that book is people would eat electronic components to gain wisdom
some you wouldn't live though
but yeah it was basically a post apocalyptic story
airbrush, I'll give you the short version
I think the sequal was something about an electronic pope
what if "we" left, and what was left behind evolved into us now
It's a post apol. story as pfred1 says. Liebowithtz is this monk who discovers a scrap of paper in the desert that says someting in a strange undecipherable language
in the grand scheme of things we evolved pretty fast
wasn't is a schematic?
that he illuminated
In actuality it say if I recall, Pound Pastrami, loaf of bread bring home for Emma.
I sort of remember something about an illuminated schematic in the book
An entire religion is immediately founded upon the "sacred document." and it goes on from there.
its good but i prefer cyberpunk today if I read that sort of thing
gibson stephenson that sort of stuff
So there ya go. Plausibe that we are only the great great great, ^ 3 grandkids of folks a dumb as we are.
its stunning how short what we'd consider "modern" really is
I wonder what the post apocalyptic machines will be like
I mean in WWII they had trouble with 2 way radio
trouble to the point where people died
Lets hope ther wont be an era like that.
and that was 1943
I saw an article in the magazine "Boy's Life" about a 14 year old kid that goes to MIT and has built his own 3d printer
I am interested in the reprap thing.
the project does seem to be evolving but I think they've made some poor choices too
It seems like it would be useful for the odd, one-off plastic part,
but so slow and inefficient...
but the progress they've made in terms of work envelope is impressive
at the sema show they had vendors showing 3d printers that could use 14 different types of plastics and 2 types of powdered metal
thats my theory do really bad the first time around because everyone loves improvement
I've been thinking about an off-shoot alternative, tell me what you think...
There is very low melting metal alloy available. Heard of this ?
I need to take a better picture of my sci-fi book collection
Some of it is like 180 deg F melting point and several steps above that depending on composition.
No not mercury. It is hard solid metal with an extremely low melting point.
seems most metal printers use scintering
the coolest printers use pairs of lasers
I don't know about those.
where the beams converge you get solidification
either in plastics or metals
heh, my cousing just got laid off, he was an engineer at an aerospace firm, he was showing me some 3d sealed bearings they had printed, they were plastic but they worked and seemed to turn smooooth
airbrush they probably had converging laser printers
But what I'm thinking is why not use the same setup, but substitute an alloy wire for the plastic fiament, and extrude molds that can cast the plastic.
he called it stereo-lityography <sic>
stereo being two lasers most likely
yes stero lith has been around for a long time
20 years or so
he said they've had it for over 12 years
20 years isn't that long
It is basically a UV epoxy resin in a vat, and lasers converge to give enough enargy at the intersection of the beam to solidify the resin/
goes by fast
who here saw the video of the screwball italian and his house printer?
I missed that one
It can be highly accurate rapid pototyping. Probably could be done well with EMC if you're a laser buff which I'm not
he laid down layers of sand then shot hardener where he wanted solid
What kind of hardener?
heck if i know
was probably an epoxy
A real house?
came out sort of syrupy
I saw where a guy built a house out of cardboard
he made these bizarre sculptures mostly
I'll find it
bales of hay
Actually hay bales is very old tech and quite efficient.
Can be stucko'd and very log - term viable highly insulated housing.
any of you guys do machining as a primary occupation?
long not log
I did once
I am self-under-employed.
I make tools for glass artists.
hehe, me too atm, I've had 4 jobs in the last 18 months
I made tools for machinists
3 of them in metalworking, one in construction
what kind of tools?
hold down clamps and wheel dressers
worked for an outfit called J&S Tool
i knew henry personally
who's most likely dead today
part of me wants to go work at progressive, the other side of me would dread those hours
along with machining in this country it seems
metalwoking sux atm
we're priced out of the market
its insane what we charge
beack 25 years ago when i did it we were insane
a lot of shops around here aren't much more than factories that happen to use lathes and mills for their machines
oh we did production
I'd do runs of 10,000
there are still a few nice job shops, but not many
sit there drilling the same hole over and over for a week
tell me about it
least i know how long it takes me to drill 10,000 holes huh?
I had a killer job at cci, it was a machine shop in the roush racing complex, I would set up and run 3 of one part, 6 of another, heh the day flew by
at one shop I worked at they had an OKK horizontal cnc, that thing is a beast
place i worked at was just depressing I mean you expected the place ot go under any day
heh, most of them around here have
there were 6 people working there when at its height there were 40
we had rooms of tools and no one in them
there are a few signs of life starting to show, but not much
like i say today there's a lot more money in tearing down machine shops than there is in putting htem up
airbrush: economy must have picked up months ago, we can't find machinists
darn I think you're right about that
morfic yeah who wants to do it?
it pays good money...
morfic: you mean production workers who'll beat their hands all to hell for 10.00 and hour?
then of course, my boss is a hardass, so it's that extra added fun :)
airbrush yeah friday nite was always splinter nite for me
i don't mean production shops, a job shop
davidf: Geronimo, TX
sit there for hours trying to dig all the splinters out
I was at one shop, and everybody on the line but me had scars down their palms from the operation
my hands *are* splinter magnets, i give you thaty
seriously, need qualified machinists and cant find them?
the good ol' 'no-see-em' good sharp calipers sometimes work *<B)
davidf: nope, when economy went south, we had people come from way up north, no car, nothing, would walk from motel to work, and other such crazyness
my man henry had the technique he'd grind them out
Trick is get them out immediately, don't wait 24 hrs
morfic, you should do like the shops do around here, fill the place up with mexicans!!!
problem is, not everyone running around calling themselves a machinist is one
far as I'm concerned you sit there and run a machine day after day you're a machinist
that's an operator
a lot of orders coming through are Qty. 1
oh i see if you're not smacking fixtures around with a hammer now and again you're not a machinist?
Drill holes all day doesnt equate with designing and building drill presses.
not what I'm saying, to me a machinist is someone you can hand a print to, and he goes off, sets the fixtures up, programs the machine (if cnc) and produces a part in tolerance
machinists don't design things they read prints
engineers design things
i expect a cnc machinist to be able to program g-code by hand, set, and fixture machines
they should have a working knowledge of different types of fixture design and be able to pick and build one appropriate for the quantity and type of job being done
he must also be able to "tweak" the code, especially that from cam programs
all the best cnc machinists i've worked with have also been accomplished manual machinists, but i don't think that's a strict requirement anymore.
no, but it helps, a lot
I'm thinking' of going to school for mobile web design, graphic arts, or somethin, anything but manufacturing
I cant imagine coding a machine without the ability to use it manually
an operator knows how to adjust tool offsets to compensate for wear, how to change parts and check them; a set up guy can pick a program out, ideally do some minor modification to it, and set the machine
airbrush whatever you do make sure it can't be outsourced
airbrush think service
yeh, you're right about that
because you can't compete with the chinese or the indians
well its not you its this country
If you ask him to decide what surface speed to program for a particular metal should he be able to do that? How about if you ask him to make a die for a punch press, should he know how to do that, and figure proper clearance between male / female parts, or just write code to punch it and / or pull a handle?
toastydeath: been long since i produced a part on a manual machine, but it is how i started out, and i bet i can teach a manual guy how to run a CNC a lot faster than you can teach a cnc guy how to run a manual, even if it's still the same thing, just that the computer is one time made from silicon and the other time from organic cells, considering the way stuff is layed out is pretty much the same, it should work the same both ways
morfic, i agree, i'm 24 and i started on manual machines
good ol' sloppy ways in the manual world :)
even though the majority of my limited experience is in cnc programming rather than manual parts
davidf, you're kidding yourself about the chinese not being able to do the work
I remember when "made in japan" was a sign of inferior quality
davidf, obviously a machinist should know what surface speed to program the machine to
and if you're looking for a die for a punch press, you need to be talking to a tool and diemaker
airbrush I still have a set of punches stamped japan that are like hard rubber
and all the tool and diemakers i know can, and do, exactly that
it's scary how little the chinese bid on some parts we also bid on, we have some good customers, and i hope we keep them for a while, and it is astonishing they still order "new" parts (some we need to program) with date of 1970 or earlier
just like a machinists makes a part from a print without external input, a diemaker will make a die for a part with nothing but the part print
and if you're ever asked to polish an injection mold, you need a buzz, and a set of headphones with pink floyd, you still might loose your sanity though
morfic here's something scary when i was in that shop they'd buy chinese lathe centers and resurface them and sell them like we made them!
haha one of the diemakers i used to work with refused to work on moulds for exactly that reason
he was pretty much die work only
pfred1: o_O what company?
morfic J&S Tool
morfic they'd buy the crap right out of harbor freight
just for the litler centers they could get
airbrush, for the record I don't recall saying aything about chinese...
company still exists but I doubt they make anything at all anymore
even back then it mainly survived on diamond sales
If you ask him to decide what surface speed to program for a particular metal should he be able to do that? How about if you ask him to make a die for a punch press, should he know how to do that, and figure proper clearance between male / female parts, or just write code to punch it and / or pull a handle?
I was assuming you meant chinese
Oh. That was not in reference to chinese mfg. at all.
oh, my bad
well I'm stoked my buffer works and as an odd aside when i had it all apart I noticed it has japanese bearings in it
and old B&D polisher
I used to hear about the incompetence of various "foreign" shops a lot
airbrush, np it was in response to this and a couple others"
far as I'm concerned you sit there and run a machine day after day you're a machinist
no offense pfred1
I bought an adapter for one of my welders genuine Miller part paid top dollar for it thing wasn't even close to being able to fit so I took it back second one same deal anyone can be a screw up
hey, I'm GOOD at screwin' up, just ask my ex
scary fact is, chinese are getting damn good.
well whether they are or they aren't I try not to send them too much money
not at screwing up. didnt read that first. LOL
yep, and they work for 100 a month, 7 days a week, 12 hours a day
I bought a keyboard today says made in china on it but I only paid $1 for it at a yard sale
me too but sometimes its a must just to keep my biz going if I need a new capability.
and pregnant with no shoes. :)
I buy top quality tools at flea markets for prices even the chinese can't compete with
there are some fine old tools
the buffer i rebuilt tonite cost me $5
beat that chinaman!
hehe, we used to have a joke: what do you call 40 indians with files?
a pakistanii mill
I've seen those people make guns in the dirt between their feet with all hand tools
davidf yes really
In person? yipes.
davidf no it was in a documentary
oh good. :)
I waqs gettin worried. LOL
davidf just because i wasn't there in person doesn't mean it didn't happen though
davidf we did it too 300 years ago you know?
or maybe you don't
only difference is they're making AKs today ;)
japan has all these hole in the wall workshops big companies farm out processes to
with dirt floors and tin roofs
wonder which one they farmed the toyota throttle to
its true they probably did
there'd be like this one super old blind guy in there with a grinding wheel doing amazing things
probably the same outfit that made the shut-off safty for that oil rig :P
unsafty u mean?
I saw some ancient guy in another ducumentary doing some process on the zirconia knives
big factory would blank them out he'd finish them
linday books has some old machining manuals I'm gonna order one day
you can download a lot of them
anything over 100 years old is public domain
let me get you some links sec
I gotta get the filenames
they have some on things like deep hole drilling of rifle barrels, turning a lathe feed screw by hand, ..
oh there's a whole website where a guy shows how to make a kentucky longrifle the old fashioned way
you should see his rifler
big old wooden twisty thing
did he do it with a broach?
he dows it on a hand lathe with a piece of broken file and splinters of wood he explains the whole process
did he make the barrel by beating the hot metal around a round rod?
but yeah its basically broached
forge the lock and spring?
yeah he starts out with rod bundle
that sounds so cool
his guns are like 60 calibre
worth a fortune
think he gets like 10 grand a pop for them
I always said I'd like to make a doomsday machine, and see what would happen if I ran it backwards
for a smokepole?
There's a Co. in Illinois that makes long rifles. Family owned since colonial days.
Interesting way they get the barrel straight..
hrm there should be a PDF file
They bend it after drilling till they can put a rod through.
says "full text format: not available" :(
davidf, is that how the ww1 barrels were done?
airbrush yeah they've changed it but if you browse around yo ushould find simething
ok, will do
[07:18:25] <pfred1> http://magic.lib.msu.edu/search~S39?/dMachine-shop+practice./dmachine+shop+practice/-3%2C-1%2C0%2CB/exact&FF=dmachine+theory&1%2C170%2C
looks like it leads to files
I have a friend that has been to the shop. He told me about it. They make everything by hand exactly the way they did it 200 years ago.
Even the screws.
there used to be a guy like that at old salem
He asked them what size the little screws were.
heh professor roy has set up a school/shop less than 100 miles from here
dude, that last link is awesome
They said, no idea. We have to send extra ones with each rifle so they will fit, LOL.
airbrush welcome to the 21st century have a nice stay
but yeah anything over 100 years old is no copyright
have you guys noticed how much Roy Underhill has improved from when he first started?
or should I say duh?
you've never heard of roy underhill?
the woodwright's shop, on public tv
uses nothing but hand tools
the stuff he's done lately has had wicked joinery, and so thight you can't see the seams
I remember when he first came on tv, he would cut and beat his hands all to pieces, and the stuff he made looked rough
airbrush here's the toy i picked up last weekend: http://img46.imageshack.us/img46/8965/hornplane.jpg
its austrian and the iron is made by two cherries
unfortunate that that kind of care is only commercially viable as an oddity or colletor's item.
davidf you've never heard of a tool manufacturer called two cheries?
they make the finest german tool steel
I'm not a machinist btw
they're more a woodworking thing
their chisels typically go for around $20 a piece
dude, that thing is wicked
[07:28:43] <pfred1> http://www.diefenbacher.com/german.htm
yeah it was a nice snag for $15
thing doesn't evne have any cracks in it
its from 1900
I have a complete set of the stanley universal plane with all the blades
you mean a 45?
is it in the box?
my uncle left it to me, he was an antique tool dealer
I have a couple oldies like that too
best 45 i ever saw was in the box missing 2 blades
it's in an old chest with the sliding tills, also has some other hand tools
well they came in a cardboard box
I think if it was blue it was probably a later one from the 40s
you guys are into this I can see
the really old ones are black I believe
thatsn ot to say that yours is worthless
I have a right handed and a left handed hatchet
not by a long shot
you mean broad axe
a nice plow plane, and an old witches tooth is in the chest too
I need a router plane
i'm not sure, but on ax is flat on the left side, the other is flat on the right side
I had one, not complete but I think I still have it
broad axes came handed like that
here's one drawer of my planes: http://img72.imageshack.us/img72/6417/pict0822h.jpg
router plane that is...
a fella at my woodcarving club made a nice plow plane, he took first place in the carving contest a couple years ago with it
seems I remember there's a right hand and a left hend part and I only have one of them does that make sense?
I made the handles for a couple of those planes
Speaking of old stuff, pfred1 while you're at garage sales...
either of you ever go to that big flea market in hillsville va on labor day?
I never have
I saw on the news the other day and old Atari game cartridge called Air Force sold for $31,000 on ebay.
I've seen boxes of those things
garage sales. You never know.
I made this saw handle: http://img541.imageshack.us/img541/6763/sawh.jpg
I took a piece of clay and squeezed it then carved it into wood
worked pretty well
well I traced the clay onto the stock i cut it out of
Wow you are really into this stuff. I think you might qualify as an official geek. :)
davidf oh this is nothing
that was just an afternoon
Just curious mind to say your age?
Ah. Now you start to comput
OK now you both begin to make sense to me. 58 here
For some reason I got the impression early on you were both like 24
I was once
but i did some pretty stupid stuff then
I don't remember much of my youth
let me find a picture
I thought geez these guys are into wierd stuff for youngsters. LOL
I still feel young mentally
but physical ....... things hurt at times in strange places, ya know?
Well from what I've seen you are, Just kidding ;)
here's me when I was 24: http://img249.imageshack.us/img249/781/tapepick.jpg
it took me 2 years to build that car
Yes tell me about it. Starting at the prostate. No don't bother
Yes, got emc working on 10.04!
.. Had to tell someone. lol
that picture was taken the day after I painted it and I'm picking some masking tape off it
davidf now that is how i wasted my youth
Better than how I did LOL.
but back then I rather fancied myself quite the hotrodder
cliff tell us bout it!!!
I was quite the hot rod. Not as productive.
O, where do I start... Recompile the kernel about 10 times... recompile rtai couple... downgrade grub.. and viola...
[07:51:30] <pfred1> http://img185.imageshack.us/img185/5063/engineg.jpg
Patience, patience, and sticktuitivness
actually its lists long lists
cliff, you mean you've got emc to work on a computer? not hooked up to a machine?
pfred1, what car is that?
heh you can even see the notebook in that picture
why did you have to recompile the kernel so many times?
I just got emc running realtime on the new ubuntu.. I just have to copy the configs over to make sure it will still run the machine..
its a 1966 Volvo P1800S
Not familiar with that Diesel?
ah, ok, congradulations
I don't think my dual weber DCOE sidedrafts would appreciate diesel very much ;)
oh yes 66 volvo. classic, had a friend who's dad had a brand new one, red, in 66.
though how bad i jetted them out they'd probably pass the stuff
was all the rage. I remember everyone in the subdivision was agog
poor mans corvette
the corvette would have been a better investment
though my fave vette is the 63 split window of course
I do remember that car.
my volvo is still pretty fast
still have it?
let me find a newer picture of it
wow, gtg it's too late, later all
[07:59:14] <pfred1> http://img109.imageshack.us/img109/6161/p1010022f.jpg
thats what everyone says it looks like a Hotwheels car
Once I ran out of gas in my Chrysler new Yorker circa 1969 and happened to have a gal of #2 diesel in the trunk ( for solvent) poured it in and drove home smokin like a factory. LOL
least your car got a good lube out of it
Surprised it ran all the way home.
Not after though for a while. ;)
but yeah i like to make or fix stuff
I'm always making or fixing something
ir fixing on making something
Cool! just got a $200 sale.
like this CNC project
did you see my motor drivers I made?
Nice when that happens
show & tell time
[08:05:08] <pfred1> http://www.youtube.com/watch?v=cgbeyNNBZ68
yes I did see that the other day.
[08:06:32] <pfred1> http://img404.imageshack.us/img404/8343/tb6560board.jpg
yeah the video is memorable
noisy cuss ain't it?
best i can figure it its going about 833 RPM in that video
Making your own stuff is a real satisfaction. Even if you can throw money at it instead, it is rewarding and contributes to the quality of one's life, no?
one of my personal quotes goes like this, "Any asshole can spend money."
But it takes a special asshole to waste countless hours to save a couple bucks!
this asshole wishes he could spend more. :)
Also wishes he had countless more hours in which to do it.
well I read a book many years ago and what was in it stuck with me it said that a hobby should pass time enjoyably for a minimum of financial outlay or something like that
and really its not like I'm living or dying over this stuff i do it just for something to do
I don't like to play golf or watch TV
Ben Franklin would have endorsed that Philosophy I'm sure
he'd have drank and whored i know him!
and spent his last dime doing it too
Ben didn't like TV either
but I've always had a thing for electronics
So did a lot of English ladies I think.
they have things for electronics?
typed while you were loom up there and fir it out
look not loom
I think ben liked the french ladies
its because of ben that electronics is backwards today
he was a bad guesser
he had a 50/50 shot and he blew it!
u mean about the dir of electron flow
conventional current flow
see I'm not a complete dummy
we can thank ben for that
that drunken whore
actually I'm EE and Co E BS
I just read datasheets and sometimes blow stuff up
but more often than not it works like I want it to
Engr's do that too, right after they write them. :)
We don't say blow up though
Scream and go to Heaven is more common.
you would if you saw how i blow stuff up
1 min back
my latest bench power supply thinks its a mini welder or something
its only supposed to put out about 300 watts but sometimes i don't know i htink it tries harder than that
this is what the insides of it looks like: http://img713.imageshack.us/img713/2244/pict0579s.jpg
Don't see much there but that huge-ass cap 31,000 microfreds?
there's 3 TO-3 NPN power transistors on the back of the heatsink
In || ?
yeah let me find the schematic
Matbe the || thing is a problem
[08:26:41] <pfred1> http://hmin.tripod.com/als/andysm/drawings/supplies/ps2-30v.gif
thing puts out some juice let me tell you
I built it to mess around with all this stepper motor stuff
Does it perform as u want?
here's me load testing it: http://img196.imageshack.us/img196/6414/pict0576h.jpg
once it did that i was like it should do it
each bulb eats about 2 amps
200 watt bulbs?
nah they're car bulbs
1157s I think
Oh ok It's low volts
13.7 in that picture but yes they're 12V rated
I figured if I cranked the supply up any more I'd blow them out
they were looking to be on the verge to me
For hi amp low V I'd probly go with a Marine battery and a charger. Huge wattage any time You want
SAimple and gigantic current capabiliy
yes but I was looking for a bit more voltage 12 is easy
so two in series
500 amps :)
thats actually how I got the input voltage
in the tan box is a transformer i got out of an old computer and I seriesed two of the secondaries
they were for the 5 volt rail I guess
they were like 13V a piece
trans was out of a PDP 11/34 minicomputer
so you know it wanted some 5V
probably about 40 or so amps worth
I learned on that one Im old
you learned on the el cheapo model did you?
heh remember my plane drawer?
8 inch floppies. Hi tech stuff in those days
its a gutted HDD from the same computer
12" glass platters
I still got ths spindles laying around someplace
always wanted to make a go kart out of them
I didn't come across those
Oh, what the disks layed on?
I always get the coolest parts out of dinosaur minis
yeah what the platters spun on
was inside the HDDs
We didn't have the opportunity to rip apart the latest technology.
one thing i splurged on was a solder pot for stripping boards
I use a solder sucker
with this sucker i pull a 64 pin DIP in 1 second flat: http://img408.imageshack.us/img408/1200/solderpot.jpg
when you have dozens of boards to strip solder suckers suck
I imagine that's right
it set me back a couple bucks but is well worth it
its nice for tinning irons too
I call it my own personal volcano
Where did you say you are?
I live in Delaware now
I used to live in New jersey
but they're crazy!
someone ought to round up everyone involved in government there and just shoot them
because they're all organized criminals as far as I'm concerned
yeah they tax you for breathing in that state
its insane least it was where I was
15,000 a year in property tax on a 1/4 acre?
Well that's no surprise. Whole country's nuts imo
well we're not nuts yet in DE
heck I think my water tax in jersey was higher than all I pay here
BUT 15 K is off the charts
yes it is and thats why i had to go
and why I could shoot them all and sleep like a babe afterwards too
Crazy. It's entitlements
well they sure felt entitled
but I wasn't getting squat for it
40% not paying taxes means 60% paying !00%
hey we all have to live within our means right?
That's why the natl debt is at record heights
where I was at it kept going downhill and the prices kept on going up so I got sick and tired of it and got out
maybe i live in a bug ridden swamp now but its cheap
its MY bug ridden swamp!
I know exactly what you mean.
and hey when the water level got a bit too high for my comfort last year I called the state and they were here lickety split to take care of it for me
if I'd still been in jersey I'd be doing the backstroke still
I couldn't believe it I called they were here the next day
Hey pfred1 I gotta get some shuteye
So its 3:45 AM here and 4:45 AM There, I suggest we both need a life. LOL
its memorial weekend
my swamp: http://img171.imageshack.us/img171/9926/flood2i.jpg
OK I'll stay up all weekend then,
yeah I've done it but not this weekend
since they reditched me out I've been high and dry
whole county flooded out this past year
but i believe it was unusual
That looks like your Volvo in the pic.
nah my volvo is in my backyard
which is past that treeline
I got a whole other clearing a few hundred feet beyond there
thats where i have my meth lab and stuff you know?
OK I'll talk w you later on pfred1 got to get horizontal. :)
don't want to see a pic of my meth lab?
Give me a break
well here's the back of it the front is funnier but http://img46.imageshack.us/img46/4763/45332352.jpg
You're quite silly
nope I'm serious http://img263.imageshack.us/img263/3218/p1010018sx.jpg
shouldn't everyone have their very own meth lab in the backyard?
I don't care i can't see it from my house
all right on that note I'll let you go nite
damm Gateway :/
i ran a file with G10 L2 P4 x5.7 y0 z0, then closed it and open a completely different file. the offset were still active
how do i cancel g54
offsets are stored
i know, but i thought g80 canceled that
g80 cancels modal motion
so all my file should start with G10 L2 P4 x0 y0 z0?
[12:11:31] <JT-Dev> http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G92,-G92.1,-G92.2,
take a look at G92.1 and ,2
and if your using G10 L2 you might put G10 L2 P4 x0 y0 z0 at the end of the file
thanks, would putting g92.1 at the start of all file do that/
or .2(which ever needed)
I always practice that if I put an offset somewhere in a file I also remove it before the file is done
unless I want it for the next file...
it only leads to disaster to leave an offset in the var file that was set by a gcode file
cool, thanks and what does the L2,L20(don't remember the others) do ?
i read that part and don't understand any of it :)
L20 makes the current location be equal to the axis numbers
axis numbers? the movements? x1y1
so if you move to X1 and say G10 L20 P1 X2 it makes the G54 system at that point your at now equal to 2 for the X axis
doesn't l2 do that?
G10 L20 is similar to G10 L2 except that instead of setting the offset/entry to the given value, it is set to a calculated value that makes the current coordinates become the given value.
calculated from what?
I think LOL
x-chat on 10.04
xchat on old street
hows that archivist?
JT-Dev: to me it sounds like measuring diameter on lathe, then telling control X1.435 [MEASURE] (G10 L20 that is)
JT-Dev: done anything with your pokeys55 yet?
jthornton, Im on 8.04
morfic I have only tested it out so far
JT-Dev: n/m, should have looked just one comment further up
Is it possible to use EMC2 to control the device over modern serial-to-uart chips like FT232? No realtime is needed for my task and serial protocol is known
JT-Dev: i was thinking of G10 L10
_TwisT_: yes, you just write a userland nonrealtime program that has hal pins
cradek: cool, and how to launch EMC2 on non-realtime kernel without emulation mode? Is it possible?
you can build emc2 in sim mode, which does not require realtime and cannot control realtime hardware
but it still can be used to control non-realtime, right?
what is the hardware? you won't be able to do motion that way
_TwisT_, what are you doing?
jthornton: just playing around with a custom hardware (microcontroller+stepper drivers) :)
that's the wrong approach for getting motion from emc
motion from emc is done by controlling the machine's position or velocity directly in realtime
you do want realtime for coordinated motion
is it a machine or do you just want to see the steppers spin?
this is an 2 axis plotter, actually
so there is no way to control it by means of serial connection, right?
However, the top speed is 115200baud
this is actual info?
cradek, what IRC client did you say you were using that allowed you to search for text?
but other clients might have search too
try /lastlog text
seems like on x-chat Ctrl-f works but I don't like the interface on x-chat
hm, interp_arc.cc might be useful, if it makes a difference to me to have endpoint and center point of arc, still hoping for a quiet day where i can find the math for arc/line or line/line intersections though
jthornton: irssi should do /lastlog and is a text/console client
alex_joni: like a scan line algorithm, finding where scanline at X intersects a line or arc segment at Z
alex_joni: maybe i should ignore arc/line intersections for now and see if i can get somewhere with only line/line intersections
cradek: and what about FT232 serial to USB converters (virtual COM ports)? Data rate is up to 3 mbps, as specified in datasheet... I think, info on wiki is about real serial ports and not about virtual ports, right?
but there is no realtime drivers for that chips, as far as I can see...
realtime and USB does'nt sound so friendly :/
alex_joni: this would be used in a G71, for every [currentX-DOC] it would iterate through all defined line segments, and return the Z at which the scanline intersects a segment, then at that X, the machine can move that far -stockallowance in Z
morfic: try #cam
and talking to awallin
I know he did some line/arc, arc/arc, whatnot/whatnot intersections lately
bricofoy: i always wonder that when i see people wanting to run emc2 off usb drives
alex_joni: ah, thanks
_TwisT_: did you actually read the page cradek linked?
And I've found only speed limit problem
3.3. RS232 Serial
Many PCs have a common, easy-to-program chip for RS232 serial interfaces, the 16550 FIFO. However, the top speed is 115200baud. In 200us, only 16 usable bits can be transmitted in each direction (10 bit times per character, 2 characters = 173uS), which is not enough to send DAC commands and retrieve position feedback for multiple axes.
USB doesn't help there
it does have a higher baudrate, but it works in bursts
once very couple of msec's
which is even worse than 16550 FIFO
Some years ago I worked with FT devices
ah, now I see.., :(
was on windows, not inux
but we experienced some problems with the drivers
sometimes they look like frozen for some time
never found out why
but this was not critical for us
and I believe, that will not be critical for me :)
I have only 2 axes and no servos, just steppers
alex_joni:having a small problem with homing to index with stepgen in position mode,the first time the axis homes it gets ferror,hit f2 and rehome the axis and it works every time?
servo control sounds a bit realtime-needing for me...
but I have no servos :)
just two steppers, 200 steps per turn
so is stepper control
_TwisT_: you want to send steps via rs232 or to send position to some ucontroller to generate steps?
200 setps/rev, with microstepping driver
whats on the table?
_TwisT_: right what?
_TwisT_: that was a multiple choice question
you still need something quite fast
custom UC with custom firmware and with known serial protocol
the problem with USB is you can have something fast, but experience some delay from times to times
_TwisT_, deleay on steppers is far worse
[13:33:45] <jlmjvm> http://pastebin.com/n6BTGTnP
can work only if your custom UC generates movement for itself, but if it only converts step pulses created by EMC
jlmjvm: some halscope traces of position fb vs. cmd and encoder values would help
you will have problem with the USB latency
_TwisT_: as far I know via standard rs232 (no usb use) you can reliably send position to one axis 500 times/sec (in realtime)
bricofoy: my UC generates step pulses for both steppers by itself, I need only to control direction and speed via serial
_TwisT_: you can do what you want, but you'll have problems with latency, easiest to do is write python userspace hal component using pyserial
micges: and my two steppers is 200 turns/sec each
micges: ok, thanks :)
200 turns/sec ?
how did you come up with that?
steps/sec? or steps/turn ?
200 steps per turn
alex_joni:will try,only used the scope with the hal manual exercise
jlmjvm: it's not that hard to use
try to trigger on ferror
and have fb, cmd, and encoder (position and index) as the channels
cradek, should someone answer cogoman on the mailing list that something like that is already there as we use the interpolated position for screw cutting with low count encoders
_TwisT_: if you send speed commands to your steppers in nonrealtime, you will have no idea exactly how many steps were produced, so you will have unknown position that will get worse and worse
archivist: I didn't read it...
cradek, he points at http://machinedesign.com/article/new-algorithm-thwarts-velocity-disturbances-in-motion-control-schemes-0721
Dog beach party http://www.youtube.com/watch_popup?v=pkPNa4DBFHI
frallzor: Don't make me short out your controller punk!
good news ! frozen-bubble over internet works good while machining :)
SWPadnos: Interesting... Never seen MPPT in a solar controller before... http://www.golandcentury.com/faqs/58-solar-controller/118-mppt-vs-pwm
I work with solar panels
MPPT are becoming something frequent
"OutBack" produce good MPPT solar battery chargers
So do "Trace Engineering"
well, "Xantrex", now
their changed their name
and MPPT are present in all grid-connexion solar inverter
cradek: you are right, but I have an plotter with a small spindle loads, so I think steppers shouldn't miss a steps...
you totally missed my point
imagine you are driving your car and every so often I'd call you on the cell phone and tell you which way to turn and how fast to go until I called again next time. Do you see why with this system I couldn't guide you to a destination?
it would work, except once in a while the next time would happen to be in a tunnel, where your cell won't have any reception
alex_joni: well, it wouldn't work anyway :-)
it would be fun to watch and bet on how long it takes to crash :P
Eric_K is now known as EricKeller
you would have passed the junction ehen the call comes
with careful timing and very slow speeds we might make a turn or two, but you can see that the method is doomed to failure
cradek: oh, and you forgot to say he needs to drive blindfolded
yes of course - he has only my phonecalls for guidance
and his speedometer
_TwisT_, your method is going to take longer to implement and not work, you could just follow the manual
like landing a plane in the fog with blinking instruments :P
bricofoy: unfortunately that doesn't always work out ok
bricofoy: at least 9 here: http://en.wikipedia.org/wiki/List_of_accidents_and_incidents_involving_commercial_aircraft
alex_joni, I know you've to be really lucky to succed in this
was just to enforce your point :)
[17:19:38] <frallzor> http://pici.se/p/large/rQXgnAIMx/
first attempt of milling a picture =)
Yup, looks like a picture.
I have a 3 axis router I am bringing to life with my first installation of EMC. I am jogging it around and can execute basic g code
Still have several issues
One issue is this:
If I interpolate a 5" circle at 75 inches per minute and the software generates G code of just a line or two of circular moves then the tool moves 75 inches per minute just as planned
But if I do an ellipse or if for any reason the software breaks it up into a million line segments
EMC won't obey the feed speed.
It will vary from 40 to 60 as the x and y axis move around the circle
It seems consistent. Any time there are a bunch of lines that are nearly tangential from segment to segment on a contour, EMC is (in my configuration) unable to follow programmed speed.
Maybe it can't process the lines fast enough?
Maybe it is somehow trying to stop at the end of each line segment?
not unless you have blending turned off
Where would I find the blending variable and see if it is off or on?
you can improve that in a few ways: most obvious is to improve the gcode, next most obvious is to use the G64 P(tolerance) setting, next you should make sure your machine's accelerations are set as high as possible
[18:43:27] <EricKeller> http://www.linuxcnc.org/docview/html//common_User_Concepts.html
Cleaning up the code is not an option unless you can direct me to more intelligent software. Most of my work is compound curves and the software I have tried all breaks it up into segments
EricKeller, I will try that link Thanks
You guys are great. According to the docs the default mode is G64 "blend without tolerance" and that means the machine will never move at a velocity such that it can't stop at the end of the next consecutive move.
No wonder it refuses to go very fast along the contures.
So my next question is "how do I get it to look ahead more than one move" and make it's decisions further ahead.
I could easily apply a tolerance of .005 or so for 99% of my work. As long as the path repeats when I am cutting multiple passes, it would never be an issue
But why is this little caveat of just one line look ahead for planning to stop.
setting a tolerance will help a lot - test it and see what you get
also, like I said, make sure your machine's acceleration is set as high as possible. it's very important for short segments.
The machine is out in my shop, and I don't have internet out there. I will go try it and report back but could be 30 minutes
changing to newest tip of 2.4 branch got rid of weird moves after stop ( was running 3Dchips under 2.4.0~pre )
I have the kernel 184.108.40.206 patched wit rtai-3.8.1 I can enable acpi in minimal form, for example active some option of the subsystem acpi
with emc2-2.4.0 version
Reporting back from the shop on G64 P.005
Well, I couldn't duplicate the problem.
I fiddled with it, but even without adding the G64 with the P tolerance, it ran at full speed.
At least I learned a little basic G code.
Thanks for the help cradek and EricKeller
I'm out for now
need help with pyvcp anyone around?
lots around... just ask your question ;)
It might help if you look at www.linuxemc2.pmiautomation.com first
I don't see any pyvcp panel on the left side
for some reason i cannot get my spindle spped to connect to the bar widget
that's the right side ;)
how did you try to connect it?
net spindle-vel-cmd => abs.0.in net absolute-spindle-vel abs.0.out => pyvcp.spindle-speed
and in my hal file net spindle-vel-cmd => hm2_7i43.0.pwmgen.03.value
jpm: can you do a halcmd show sig spindle-vel-cmd in a terminal?
Spindle-vel-cmd is valid and its value = my pwmgen
but the pyvcp signal = 0
i am asking the spindle to be at 1200 right now
jpm: please open a terminal
then type 'halcmd show sig spindle-vel-cmd'
and pastebin the output if it's longer than 2-3 lines
float 1200 spindle-vel-cmd ==> hm2_7i43.0.pwmgen.03.value <== motion.spindle-speed-out float 20 spindle-vel-cmd-rps <== motion.spindle-speed-out-rps
I don't see the connection to pyvcp
nor the connection to abs.0.
can you pastebin your hal file(s) ?
I appoigize Have not used paste bin. I know it's like the postgui hal is just ignored
i will try
jpm: it's really easy
[21:02:02] <jpm> http://pastebin.ca/1874023
that is the post gui i will do the other now
[21:03:18] <jpm> http://pastebin.ca/1874024
please do the INI file too
[21:07:25] <jpm> http://pastebin.ca/1874028
#POSTGUI_HALFILE = custom_postgui.hal
that's commented out
and it's in the wrong section
needs to be under [HAL] not [HALUI]
* alex_joni is off to bed
good night all
Have a good one
KimK: posted som picts today not much but visit www.linuxemc2.pmiautomation.com
jpm: Nice, thanks for posting.
No problem i plan on adding to it
KimK: any idea on how to add simulated encoders to my system for troubleshooting until i get the servos hooked up
In hal you can just write new values to numeric variables, I think. What, 32-bit signed maybe? Whatever you're using. I think that would be easier than trying to fake a quadrature input, but you could probably do that too.
I think there's a lookup table function of some kind in hal, maybe that could be your quad-faker? You'd have to walk up and down each table entry, though, no skipping. Unless you want to simulate a defective encoder.
Kimk: do you know how the simulated encoder works
Either way should allow you to creep the encoder a little bit, and you should get a little bit of analog output. Up to whatever following error limits you have set (probably set pretty loose at this point?)
So maybe you could go up to "quite a bit" of analog output, LOL.
Oh, ha! There's a simulated encoder in hal already? Man, these guys are great! They've been there, done that, and got the T-shirt already!
No, I don't know how the simulated encoder works, but if it's there already, I'll take a quick look.
Yes, there is a simulated encoder function. You can set it up to match the real encoders and wire the cmd-pos into it, and the feedback out, and you get a system that you can check out, then when you get the real encoders, just do a swap.
Of course, it is no use at all for tuning servos or anything. (fine for a stepper system though, which you intend to add encoders to later).
Yes, it's all right there at http://www.linuxcnc.org/docview/html//hal_rtcomps.html#sec:Simulated-Encoder
thanks for pointing that out. Speed controlled by one pin, so you could write to it using halshow or connect the control pin to a PyVCP knob (I'd recommend having zero in the middle and going plus/minus).
I am currently making a new Z-axis drive for my mill. I am afraid my heart isn't really in it though, as the job is making it all too clear that the column is as stiff as a wet noodle, and so the final result will never be satisfactory.
That Deckel on eBay looks more and more tempting, but it might be a shame to CNC-retrofit such a nice machine.
andypugh: Maybe you need to make a stiffer column too?
hmm anyone every seen ball end mills for woodworking?
Yes, I think I've seen those. For routing, etc.?
I seem to lack an iron foundry, so making a better column is tricky.
And yes, I have seen many ball-ended router cutters.
Sometimes just a flat (half-shaft) at those kinds of RPMs.
andypugh care to show some?
I cant find any regular ones for 3d finishing, just crappy router ones
I am looking now. I know where there is a box with some in...
Ah, hang on, I only know of router cutters too.
I want "regular" you know =)
andypugh: What would be the (approximate) finished dimensions of an improved column, and what would be the proposed improvements over the existing one?
No good to you?
thats the kind I can find everywhere
andypugh:do you have a stepper config with sim encoders?
KimK: I guess about twice as wide, with 4x the wall thickness would be a start.
andypugh: And what size are we talking here? And what wall thickness?
jlmjvm: No, I have never felt the urge to use one, though I do recall discussing a setup on here. I can't find it in the archives though
KimK: You can see my mill in this video. I am concluding that it is a toy. http://www.youtube.com/watch?v=ZhICrb0Tbn4
andypugh: Yes, I saw that before (glad to see it again!), and toy or not, I'm very glad you posted it, that is just excellent. I learned a lot about gear hobbing just by watching your video. (Reminds me of "You can observe a lot just by watching." --Yogi Berra)
andypugh: John is here and asks how far are you from Grantham? (near Lincolnshire?) (about 100 mi north of London?) John has some in-laws there.
I am about 100 miles from Grantham.
Oops, we have to get to the hardware store before they close to get 2 feet of 1 5/8 in. (40mm?) clear vinyl hose for a surface grinder. Back in 45min maybe?
Gramtham is a place I have been through or past many times, but I have never stopped.
hmm make a t-rex skeleton with a 500mm long head maybe? :P
any suggestions on a project for a 6"x10" piece of 1/8" thick Copper?
elmo40, use it for a cutting board?
maybe form it into a frying pan? ;)
make a CNC roll former.
just pond it into a frying pan shape
though, like aluminum, I bet it would not be too healthy
nothing is healthy when consuming it.
you don't consume the pan, silly
heat it up = release of stuff
ya, to a thousand degrees or so
people have the same argument about Teflon®
Copper pans are fairly normal, just expensive
you need to BURN the teflon and have it turn into a gas.
they release small small amounts of toxins
andypugh, its copper with a stainless ligning thats common
I have one of those
No, you can buy solid copper cookware
But it's expensive and a pig to clean.
so... arc metal deposit Stainless onto my new copper base? :)
teflon is safe enough for normal cooking purposes
but w/e :P
my dream cookware is something like this: http://www.titaniumexclusive.com/1.html
teflon is safe
the temp gap between not so safe and safe cooking is not as big as I would like
you can consume it, when it scratches off.
it goes right through you... your stomach can not digest it
WalterN: what do you mean? a gas burning stove still won't get hot enough to cause Teflon to turn into a gas.
elmo40, ideal for me would be copper with a 316 stainless coating
not 309ss ?
even with polyester thats found in much cookware heat causes relase of some toxins
316 is basically the best stainless to use
though it would not work with an induction stove
That "Titanium" cookware isn't actually Titanium, right? That would be silly.
"titanium surface finish"
they have an Al base
titanium is good stuff
arc spray deposited Ti coating
just picked up a set of Henckels. THE BEST knives out there.
WalterN: no, electroplating is a dipping process
I am very fond of titanium. I bought my camera mainly because it was titanium. But it has very poor heat conduction, so is a daft choice for cooking.
arc spray deposit (or whatever it is called) is molten Ti sprayed onto the Al base
And the _best_ knives are made by little men in Sheffield.
[23:09:39] <WalterN> http://www.cookingforengineers.com/article/275/Saber-Kitchen-Knives
[23:09:43] <WalterN> http://www.saberknives.com/shop/store
these knives have also been criogenically treated.
not too many websites have the 'font +/-' buttons on them.
What are they made of? Unless it is a metastable stainless (lean 316 etc) then cryogenic treatment is a gimmick. Anyway, you don't want your kitchen knives too hard, you want to be able to tune the edge for the job in hand.
andypugh, if it was 316 the knife would be too soft to hold a good edge.. heh
Depends on the treatment. Cold-worked 316 can hit 3GPa
oh yeah? well my GPA is 3.5!
andypugh: what is the max Rockwell hardness?
3GPa is the Tensile or Shear ?
anyways. toll steel is never stainless. will not keep an edge
it is a soft metal
tool steel is ..what?
oh, maybe hardened vs not hardened
you will find it very difficult to find a stainless bolt higher then grade 5 ! (imperial scale) while high carbon bolts can reach way past that.
Martensitic stainless will, and if you google "stainless tool steel" you will find some very nice materials
eah, but just add a lil slat water, and you'll have shit in the morning =)
316 is one of the few metals/alloys that hydrogen peroxide will not readily react with
I invented a new steel formulation a few years ago, ye know. A plain-carbon cold-rolling steel with traces of Erbium.
cryogenics is not only for Austenetic steel.
to turn it more Martensitic
F1 teams cryo their entire engines! block and all
remember, just because it is 'room temperature' doesn't mean the heat treatment cycle has to stop. there is still ~ 290 degrees lower to go.
Yes, well. I am unconvinced it achieves very much. Not many things get _faster_ at lower temperatures.
they heat treat the block. rough machine it. heat treat to stress relieve. fine machine it to +/- 0.005 or so, then cryo it. then machine it to finished size.
long process. but worth it!
why do you have to get faster? and what is it you wanted to get faster?
cryo is an extension to heat treatment. not a replacement
Heat treatment is all about chemical and metallurgical processes in the material. You can end up with some metastable phases still there are room temperature, and going colder can help to push them over the edge (which is why it works with austenitic stainless, to push it to martensite if it is still metastable at RT). I am not aware of any such phase transformations in aluminium, that is all chemical (AFAIK, I am a ferrous
metallurgist, Ali is a bit of a mystery). All that cooling Ali down does is make the chemical processes run more slowly.
I think the theory behind cryo non-ferrous metals (I have also heard plastics) is that thermal expansion causes minor but permanent deformation of the material.
temporarily shrink the random gaps between atoms, making them more aligned
theory, as with Gravity, has been tested and measured
ries_ is now known as ries
i can see it as a stress-relief thing. (almost)
some companies claim better machining of Al after treatment.
better surface finish
never tried it myself.
Some people claim that pyramids sharpen razor blades :-)
liquid nitrogen isn't something to mess with ;)
it is stone...
Vectrics software is awesome, I just discovred custom textures option :P
Liquid Nitrogen is great fun. Safer than boiling water. It amuses me that people put on safety gear to handle it, then take it all off to make a cup of tea,
[23:53:26] <frallzor> http://pici.se/p/HPGBskIYb/?size=fullsize
In Autodesk Inventor, if you make a large balck-chrome sphere the texture is a reflection of a black Dodge Viper.
Im having a blast locating kick ass photos off wood :P
Fun thing to do (in an open area). Half-fill a 3-litre coke bottle (the PET ones) with LiqN2. Put the lid on and walk away. Eventually there will be a collossal bang, and absolutely no sign of the bottle at all. I think they atomise.