Back
[00:13:30] <s1dev> any ideas on what would be cheaper? buying a 3 axis mill and turning it into a CNC or building everything myself?
[00:22:46] <L33TG33KG34R> tjb1: slisgrinder is me
[00:29:04] <ursa> cheaper to build your own, if you have the tools to do the job right.
[00:29:34] <ursa> Asking once again: Anybody know of an open source PID controller for routers?
[01:17:16] <Thetawaves> ursa, i have seen one
[01:17:21] <Thetawaves> ursa, can't remember the name
[01:17:24] <ursa> where?
[01:17:27] <Thetawaves> i'll think about it
[01:17:34] <archivist_> s1dev, or get a used machine and retrofit, I made my own machine from bits of machines
[01:17:45] <Thetawaves> i don't know if 'open source', but some guy was doing it out of his home pretty open like
[01:18:18] <archivist_> er Linucxcnc is opensource
[01:18:19] <ursa> the guy in Australia selling a proprietary version seems to be working out of home.
[01:19:02] <ursa> Linuxcnc is for CNC control of mills, routers, and lathes, but doesn't provide high-accuracy router speed control
[01:20:21] <Thetawaves> http://www.vhipe.com/product-private/SuperPID.htm
[01:20:33] <Thetawaves> is that what you are talking about?
[01:20:59] <ursa> that is the proprietary one. It's gotten good reviews, and isn't that expensive. However, there is no way to tinker with control parameters
[01:21:22] <Thetawaves> :(
[01:21:23] <ursa> and from past experience with PID controls, I know that the default parameters are almost always wrong
[01:21:42] <archivist_> who says you cannot add that loop to linuxcnc :)
[01:22:24] <ursa> You have any idea how nasty it is to do PID control? A few years ago, an airconditioner manufacturer tried, and dropped
[01:22:30] <ursa> a million bucks before giving up.
[01:22:50] <archivist_> PID is already in the software
[01:23:10] <ursa> on the other hand, there is a PID controller already there . . . . but that leaves the circuitry, and I'm no EE
[01:23:34] <Thetawaves> what is so hard about PID?
[01:24:00] <ursa> all kinds of problems turning physical measurements into digital values, for starters
[01:24:02] <Thetawaves> 1 mil on tuning tools orr??
[01:24:19] <Thetawaves> quantization errors?
[01:24:33] <ursa> no, a mil on trying to develop a controller for their equipment
[01:24:55] <ursa> quantization errors, hysteresis, nonlinearity, . . . lots of problems here.
[01:24:57] <Thetawaves> lolwut
[01:25:21] <archivist_> just buy an encoder and fit, the spot of paint method is liable to noise from dirt
[01:25:51] <ursa> As another example, there is a steam heating control on the market whose default parameters simply don't work. However,
[01:26:02] <ursa> they can be changed, and I know values that do work.
[01:26:46] <ursa> It is not only an encoder, but doing the proper power control, which is tricky, because it is AC not DC
[01:26:51] <archivist_> PID always depends on the whole system, a set of figures may not travel
[01:27:08] <archivist_> triac control
[01:27:32] <ursa> Yes, triac control. It's tricky--not difficult, actually, but tricky
[01:28:02] <ursa> I wouldn't be quite so adamant if I knew that the SuperPID controller was adaptive, but the mfctr doesn't say that it is.
[01:30:16] <ursa> btw, there is a PID library for Arduino that does include adaptive code. I'd use that if I had the rest of the circuitry.
[01:30:54] <archivist_> you dont need arduino!, its all in linuxcnc
[01:31:19] <ursa> I don't know if linuxcnc contains adaptive code. Does it?
[01:31:21] <archivist_> you just need the feedback and power control in hardware
[01:31:53] <archivist_> what do you mean by adaptive
[01:32:23] <ursa> it is possible to have code that monitors the PID response and modifies parameters to optimize performance.
[01:33:19] <archivist_> there are tools to help you set the pid values
[01:33:58] <ursa> That provides a set of PID values that are good for the conditions under which they are measured, not all field conditions.
[01:34:28] <ursa> the PID values given by those kinds of tools are generally pretty good, and one can live with them, but adaptive control still does better.
[01:35:08] <archivist_> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Tuning_EMC2/HAL_PID_Loops
[01:35:45] <ursa> TNX for the url, but :) I'm still looking for the hardware.
[01:36:04] <ursa> BTW, if you know what you are doing, the only tool you really need for tuning a PID loop is a stopwatch.
[01:37:10] <archivist_> not true we get loops much tighter with a scope and looking at the following error
[01:38:13] <ursa> you can use the scope as a stopwatch. To tune the loop, you need to little more than watch the number of overshoots and their timing.
[01:38:24] <ursa> the rest is calculation
[01:38:59] <ursa> Oh, and, yes, I've done it, numerous times, but never in a system that operates so fast I need a scope.
[01:42:37] <archivist_> linuxcnc includes halscope to watch the parameters
[01:43:25] <ursa> Means the tools are all there--measure the error, and get the timing. But that still doesn't give me the circuitry, which is what I need.
[01:44:17] <archivist_> well you just need a pwm to ac motor speed control and feedback electronics
[01:45:11] <ursa> and the circuitry for the encoder. Yeah, that's the list, and I don't have those components.
[01:45:32] <ursa> No problem finding them for DC motors, but for 5 amp AC motors, it's another story altogether
[01:50:50] <DJ9DJ> moin
[01:54:30] <archivist_> ursa, you may fine it easier to find a dc speed control input than pwm (it is easy to filter pwm to a dc voltage)
[01:54:35] <archivist_> find
[01:55:14] <archivist_> I have a 6 amp one on my desk here for ac motor
[01:58:52] <ursa> yes, pwm to dc is easy. Can you give me a brand and model for your speed controller?
[02:01:17] <archivist_> it is one of these
http://www.stockelectronics.org.uk/BMS-linked.html
[02:01:44] <ursa> You know, I wonder what is across the pot's in those cheapo open-loop controls that are all over the place . . .
[02:03:17] <archivist_> those are triac dimmers probably should be modifiable with the addition of some opto safety
[02:05:21] <ursa> if the pot is supplying 2-10v, wouldn't need opto safety.
[02:05:49] <ursa> just disconnect the voltage source and connect to the output of the pwm to dc filter
[02:06:04] <archivist_> no those cheap things have no isolation
[02:06:07] <archivist_> http://www.alldatasheet.com/datasheet-pdf/pdf/138518/COSMO/KMOC3061.html
[02:06:49] <ursa> wouldn't need isolation--that should be in the pwm to dc filter
[02:07:01] <archivist_> the pot on those will be at mains
[02:08:14] <ursa> if there is mains across the pot, then they aren't going to be controlled by 2-10v. If there is main across the pot, it is probably not usable for this purpose
[02:10:21] <archivist_> that is why I pointed at that chip documentation as the opto between the control and triac
[02:11:51] <ursa> getting too complicated. But you've given me a bunch of good ideas that I'm going to follow up before I drop $150 on SuperPID. I think I can do the job
[02:11:57] <ursa> a lot cheaper, and better.
[02:12:11] <ursa> thanks to your comments, that is.
[02:14:30] <archivist_> those cheap ones are basicly
http://electroschematics.com/444/motor-speed-regulator-with-triac/
[02:17:01] <ursa> that kind of thing would be difficult to make work here, if it can be done at all.
[02:18:17] <ursa> In any case, given your comments, I have enough to go on. A commercial encoder won't work, because there is no place to mount it.
[02:18:45] <ursa> but that is a problem that can be worked around. There are lots of circuits for speed detection kicking around.
[02:24:31] <archivist_> reflective optos are cheap enough
[02:24:58] <archivist_> but keeping the dust away is not trivial
[02:27:15] <archivist_> anyway the opto isolated MOC30xx series of triac driver chips are a starting point google for application notes for each chip you consider
[02:27:48] <ursa> yes, dust is going to be a PITA. However, because of the construction of my router, I am thinking of putting the sensor on
[02:28:14] <ursa> a swinging arm so that cleaning it off is no big deal.
[02:28:44] <ursa> As for the tirac chips, yeah, that is a good starting point. Will probably do a good bit of googling tomorrow.
[02:30:45] <ursa> btw, cncerning dust, the plastics and aluminum I am working with create very little. Healthy sized chips, but almost no dust.
[06:24:58] <Loetmichel> mornin'
[06:25:12] <micges> ahoy
[07:13:39] <s1dev> well, what's a good sized spindle that I could buy for <$300-400
[07:17:23] <Valen> depends what you want to do
[07:17:44] <s1dev> like what I want to cut?
[07:18:02] <Valen> yes
[07:18:51] <s1dev> metals mostly, steel + some soft stuff, and maybe stainless steel if I can
[07:19:34] <archivist_> what sort of machine tool
[07:20:07] <s1dev> honestly, I have no idea
[07:21:10] <Valen> for $500 your going to want to get an old low speed spindle i reckon
[07:21:23] <archivist_> my milling machine has a belt driven spindle taken from a lathe I scrapped, is that cheap enough
[07:21:28] <Valen> any of the newer ones (ebay) are all super high speed, and have no real bearings
[07:22:00] <s1dev> would craigslist be a good place to look?
[07:22:49] <archivist_> at that price maybe you could get a complete machine
[07:23:21] <archivist_> Valen, did you ever get your bevel gears?
[07:23:29] <s1dev> so where do the people selling completely machines get their spindle motors?
[07:24:08] <archivist_> complete machines have motors :)
[07:27:39] <archivist_> I just used a 3 phase motor with a pulley and added a vfd
[07:29:07] <jdh> motors are easy (vs. spindles)
[07:30:29] <archivist_> belt drive to lathe headstock mounted on slides vertically was my solution
[07:33:33] <archivist_> s1dev, re the earlier question build v buy, I built from scraps and scrounged parts, but bought steppers and drives and some metal
[07:36:01] <s1dev> I don't think I could find partially broken machines on craigslist
[07:36:21] <archivist_> local scrap yard and ebay
[07:36:24] <Valen> I reckon you will have to build
[07:36:37] <Valen> just because what you want doesnt exist for less than ~$5k
[07:36:48] <archivist_> I documented some of mine
http://www.archivist.info/cnc/
[07:38:46] <s1dev> thanks
[07:39:09] <archivist_> it does not have to be pretty :)
[07:42:08] <jdh> how did you align your Z rails?
[07:43:44] <archivist_> precision level for the left rail to be vertical, then used that and the carriage to get the other drilled and fitted
[07:45:00] <archivist_> and the level had an accident afterwards machine vibration made it slide off onto the floor
[07:45:22] <archivist_> new bubble cost me £80
[07:47:02] <Valen> :-<
[07:47:21] <archivist_> actually one of these
http://www.ebay.co.uk/itm/Hilger-Watts-Precision-Clinometer-TB9008-1-with-case-/150655237207?pt=LH_DefaultDomain_3&hash=item2313c07c57
[07:51:26] <archivist_> also used a square level
[07:52:11] <archivist_> but as the floor moved it was easier to measure actual angles and transfer
[08:18:33] <archivist_> s1dev, I dont like my current column, it will be replaced with one off a
http://www.collection.archivist.info/pollard.jpg
[08:19:23] <jdh> is your current one a lathe bed? or a bed from something?
[08:20:17] <archivist_> yes it is a lathe bed with added angle to help with torsional stiffness
[08:20:38] <archivist_> from the lathe the head came from
[08:21:21] <archivist_> top left
http://www.archivist.info/cnc/stagetwo/ a hobymat MD65
[08:21:46] <archivist_> revolting machine but headstock usable
[08:24:08] <jdh> your pictures always make me feel better about my own desk/workspace/tabletop/etc clutter :)
[08:24:21] <jdh> I am not the only one !
[08:30:17] <jthornton> mmmm buttery cheese sticks
[08:31:45] <archivist_> jdh, er wot...99% of the normal people have "some" clutter
[08:33:14] * jthornton is always impressed with the resourcefulness of archivist_
[08:33:30] <jthornton> and it nice to have everything within arms reach
[08:33:39] <archivist_> jdh I used to visit houses in my job as TV repair engineer, empty bare houses are not normal :)
[08:34:14] <archivist_> jthornton, you must have long arms :)
[08:35:06] <jthornton> well nothing is within arms reach for me
[08:36:43] <jdh> I feel I have an inordinate amount of clutter.
[08:37:28] <jdh> pretty much any flat space accumulates stuff. If the stuff is relatively flat, it will accumulate more stuff
[08:37:43] <jthornton> I try and put away 2 things for every 1 I take out but it never seems to work
[08:38:26] <archivist_> I have run out of "away" to put stuph
[08:39:13] <jdh> I've had two arduinos on breadboards on my desk talking to each other over a cheap RF link for 2 years or so. I see them blinking occasionally when I move stuff around.
[09:24:12] <skunkworks> random question - is there a way to see the currently copying file that rsync is copying?
[09:25:17] <Jymmm> verbose
[09:25:20] <Jymmm> -v
[09:26:26] <skunkworks> Jymmm, the rsync is already running as a cron
[09:26:42] <Jymmm> -P
[09:27:02] <Jymmm> then you need to have it setup to log that you could tail
[09:27:29] <skunkworks> ok. so no magic command to see it when already run. thanks
[09:27:31] <Jymmm> or add the -P to show progess during a transfer
[09:28:06] <Jymmm> not unless you enabled logging
[10:05:07] <Valen> skunkworks: perhaps a lsof | grep?
[10:30:05] <skunkworks> That works for what I need - lsof | grep media
[10:30:16] <skunkworks> I will have to thank him when valen gets back
[11:14:13] <Gabe_W> hey
[11:15:03] <Gabe_W> quick question, i ordered an encoder, and wasn't paying attention to the output type. I recieved an open collector output, how does one hook that up to the parallel port is it possible?
[11:15:55] <DJ9DJ> hey
[11:16:09] <DJ9DJ> open collector can just pull the signal to low
[11:16:40] <DJ9DJ> so you may need a pull-up resistor which pulls your signal to high the other times
[11:17:01] <DJ9DJ> resistor between your signal and +5v, i think
[11:17:31] <Gabe_W> so i have 5v to pin 1 on the parport
[11:17:47] <Gabe_W> 10-in and 11-in are A and B
[11:18:06] <Gabe_W> use a 470Ohm resistor on pin 1
[11:18:46] <DJ9DJ> so you may need two pullup resistors. one for output A, the other for B :)
[11:19:18] <Gabe_W> okay should of payed attention and bought a line driver like normal
[12:05:02] <IchGuckLive> hi all B)
[12:06:12] <Loetmichel> re @ home
[12:14:10] <IchGuckLive> Loetmichel: hast du den vfd auch selbst gebaut
[12:14:59] <Loetmichel> vfd?
[12:15:05] <Loetmichel> wo?
[12:15:14] <IchGuckLive> an der spindel
[12:15:28] <IchGuckLive> oh lets talk in english sorry
[12:15:43] <IchGuckLive> the spindel control to your chinese one
[12:25:47] <DJ9DJ> sorry, i cannot understand you! ;)
[12:27:12] <IchGuckLive> B)
[12:29:31] <Gabe_W> okay no luck with the open collector encoder, even with pullups
[12:30:51] <IchGuckLive> Gabe_W: lathe spindel
[12:31:00] <IchGuckLive> or servodrive
[12:31:37] <Gabe_W> servo, just bought it
[12:31:55] <Gabe_W> here is the literature
http://www.encoder.com/literature/datasheet-260.pdf
[12:32:55] <Gabe_W> 7200 cpr
[12:33:18] <IchGuckLive> full number please
[12:35:38] <Gabe_W> really, hold on
[12:36:06] <Gabe_W> 260-N-T-10-S-7200-Q-OC-1-SMJ-SF-1-N
[12:36:14] <IchGuckLive> or only the connector type
[12:36:19] <IchGuckLive> ok
[12:36:23] <Gabe_W> okay SMJ
[12:36:25] <Gabe_W> the 5 pin
[12:36:47] <pcw_home> Thats 28,800 counts per rev in quadrature so software counting is not likely to work above about 60 RPM
[12:36:59] <IchGuckLive> did you remove the connector ?
[12:37:20] <andypugh> "N" makes it less useful to me.
[12:37:48] <Gabe_W> i don't follow?
[12:37:53] <IchGuckLive> andypugh: maybe it is cheep
[12:38:13] <Gabe_W> $370.00
[12:38:21] <IchGuckLive> whow
[12:38:33] <andypugh> I might have a use for an incremental + commutation encoder.
[12:38:47] <jdh> you say they never had taxes?
[12:39:01] <andypugh> Gabe_W: It probably needs the correct supply voltage and pull-up resistance
[12:39:01] <Gabe_W> i have pin 1 giong to the +5v
[12:39:23] <Gabe_W> its says 5v-20v
[12:39:46] <archivist_> Gabe_W, that will be outputting pulses far to fast for a parallel port to see
[12:39:59] <Gabe_W> not at low rpms
[12:40:12] <Gabe_W> i had a 2500 cpr on there before with no problems
[12:40:31] <andypugh> The A' B' and Z' are to p-port gnd?
[12:40:48] <Gabe_W> i have the to 10 in and 11in and 12 in
[12:42:52] <Gabe_W> at a 48khz base-thread should be able to spin about 90rpms
[12:43:13] <Gabe_W> 1.3 rotations a second
[12:43:40] <pcw_home> assuming no quadrature distortion
[12:44:01] <Gabe_W> i only need 60rpms at most
[12:44:25] <Gabe_W> probably way lower than that
[12:44:37] <Gabe_W> so do i have it hooked up wrong?
[12:44:55] <Gabe_W> pin 1=5v
[12:45:03] <Gabe_W> gnd=gnd
[12:45:05] <andypugh> I don't know, how do youhave it hooked up?
[12:45:17] <Gabe_W> pin 10=A
[12:45:20] <Gabe_W> pin 11=B
[12:45:24] <pcw_home> check that both input toggle with halmeter
[12:45:25] <Gabe_W> pin 12=Z
[12:45:48] <Gabe_W> i know the port is working i just took the other encoder off, but i did check to see that
[12:45:59] <andypugh> Gabe A' B' and Z' ?
[12:46:08] <Gabe_W> yes
[12:46:29] <andypugh> That wasn't a yes-no question
[12:47:02] <Gabe_W> pin 10=A,pin 11=B,pin 12=Z
[12:47:30] <Gabe_W> with a resistor between the encoder and pin A,B,C
[12:47:30] <IchGuckLive> A' B' and Z' GND ?
[12:47:57] <andypugh> Ah, it's the 5-pin version without the complementary channels.
[12:48:06] <Gabe_W> yes
[12:48:09] <Gabe_W> single ended
[12:48:42] <Gabe_W> the previous encoder was a line-driver output
[12:49:09] <pcw_home> if it was differntial you _would not_ want to ground the 'a,'b 'z lines!
[12:49:25] <archivist_> should be three resistors +5 to a b c separately
[12:49:46] <pcw_home> (that would short 1/2 the line driver outputs to ground)
[12:50:02] <Gabe_W> i have a 470Ohm resistor
[12:50:06] <Gabe_W> maybe i'll try a higher one
[12:50:25] <archivist_> one? three off
[12:51:13] <Gabe_W> ?
[12:51:53] <archivist_> each line needs its own pullup
[12:52:41] <andypugh> Gabe_W: Draw us a picture of what you have?
[12:59:19] <Gabe_W> as soon as i touch the A pin to the pin 10 it drops to ground
[12:59:38] * Jymmm hugs his new ladder =)
[13:00:23] <Gabe_W> im getting frustrated, i know its something im not thinking about
[13:00:27] <pcw_home> thats possible (you have 50% chance of the output being ant a low level)
[13:00:38] <pcw_home> s/ant/at/
[13:00:50] <Gabe_W> but when i rotate the shaft it stays low
[13:01:34] <Gabe_W> maybe i can't use the input pins
[13:01:42] <pcw_home> sounds like a wiring error (you do have three pullup resistors?)
[13:01:52] <Gabe_W> yes
[13:01:59] <Gabe_W> i tried 470,560, and a 1k
[13:02:27] <Gabe_W> i might bust the arduino out, and see if i can't get it working with that first before i go any further
[13:02:38] <Gabe_W> it has built in pullups
[13:02:40] <pcw_home> well disconnect from Parallel port and check with DVM
[13:02:49] <Gabe_W> i left my scope at home
[13:02:53] <Gabe_W> i feel like im in the dark
[13:03:01] <Gabe_W> all i have is a regular voltage meter
[13:03:15] <pcw_home> you can use the ohms range to check without pullups
[13:03:34] <Gabe_W> so when the sensor picks up a mark what does it do?
[13:03:44] <Gabe_W> allows current to flow?
[13:03:48] <pcw_home> low Ohms when "on"
[13:04:07] <pcw_home> close to infinite when "off"
[13:04:19] <Gabe_W> probe between the A pin and GND?
[13:04:29] <pcw_home> Yes + on A
[13:04:38] <Gabe_W> okay
[13:04:41] <Gabe_W> i'll be back
[13:04:49] <pcw_home> (of course 5V to encoder needed)
[13:05:46] <pcw_home> also check your 5V to pullup resistor common (a problem here would cause your symptoms)
[13:07:17] <pcw_home> Also assuming you have a modern DVM (an ancient analog VOM might put to much current through the outputs RX1 range)
[13:07:32] <pcw_home> in Rx1 range
[13:08:09] <archivist_> ancient analog had the voltage reversed sometimes/always!
[13:08:49] <archivist_> as it measured the current on resistance ranges not voltage
[13:13:42] <pcw_home> Yes I remember that. Dont Know if the Simpson I have is like that or not (it has a reversing switch)
[13:14:06] <Gabe_W> okay
[13:14:40] <Gabe_W> i am having a brain fart day i feel, first i disconnected it to check the A to grnd, and realized it probably doesn't work without 5v
[13:15:35] <DJ9DJ> lol
[13:15:35] <DJ9DJ> brain fart day? :-D
[13:15:35] <Gabe_W> imagine that
[13:15:35] <Gabe_W> so i reconected the 5v and Gnd
[13:15:35] <archivist_> pcw did mention that too :)
[13:15:35] <Gabe_W> lol
[13:15:35] <Gabe_W> i had already walked away
[13:16:05] <Gabe_W> i checked the A to gnd and it should infinite
[13:16:15] <Gabe_W> no matter what i did, and i checked B and Z the same also
[13:16:23] <Gabe_W> i have 5v at the 5v and Gnd wire
[13:16:39] <Gabe_W> i have 1v at the A and B and Z wires(if that matter's)
[13:17:08] <Gabe_W> but it never changed resistance, i also did a diode check
[13:17:15] <Gabe_W> with it completely disconected
[13:18:23] <Gabe_W> frustrating, frustrating...i'll never leave home without my scope again
[13:19:25] <Gabe_W> i need to read, and get a better understanding on "open collector outputs", maybe this would help
[13:20:13] <pcw_home> They should be switches to GND (the collector of a NPN transistor)
[13:20:36] <DJ9DJ> https://en.wikipedia.org/wiki/Open_Collector
[13:21:23] <Gabe_W> the diode check showed continuity from a to grund but not vice versa
[13:21:28] <archivist_> er what <Gabe_W> i have 5v at the 5v and Gnd wire I hope you mean 0v at the gnd pin
[13:21:48] <DJ9DJ> i think he meant 5v between 5v and gnd :D
[13:21:51] <Gabe_W> i do, im ignorant
[13:22:00] <Gabe_W> bare with me
[13:22:13] <archivist_> just chekin :)
[13:22:52] <pcw_home> diode check makes sense (most integrated NPNs have a parasitic diode from collector to GND)
[13:22:52] <andypugh> I don't bare with no-one. (for which they are generally grateful)
[13:23:12] <pcw_home> Ha Ha
[13:23:53] <Gabe_W> so basically when the switch is open, current will flow from the A pin to gnd, and when its closed it will not
[13:24:09] <Gabe_W> its a transistor
[13:24:21] <Gabe_W> duh
[13:24:38] * FinboySlick is punished by his brain when it imagiens andypugh pole dancing for some reason.
[13:24:55] <FinboySlick> :s/imagiens/imagines/
[13:25:25] <Gabe_W> so shouldn't the pull up be from gnd too 5v?
[13:25:31] <Gabe_W> to*
[13:26:11] <archivist_> 5v to A
[13:26:48] <Gabe_W> if the resistor is to large what will i see?
[13:27:39] <pcw_home> slow rise times and more susceptibility to noise
[13:28:16] <Gabe_W> i'll go try to fiddle around some more
[13:33:19] <andypugh> FinboySlick: If it helps, here is a different Andy Pugh to imagine pole-dancing:
http://www.linkedin.com/profile/view?id=194984441&authType=name&authToken=-WQ8&goback=%2Econ
[13:34:30] <FinboySlick> andypugh: Thanks.
[13:35:44] <FinboySlick> andypugh: Why does it have to be a different one though? It'd be pretty cool if you were also a hot babe on top of being a genius.
[13:36:07] <FinboySlick> With a cool british accent on top of it, I might even fall in love.
[13:36:09] <andypugh> If I was a hot babe I wouldn't get anything done, though.
[13:36:29] <jdh> how hot would you have to be?
[13:38:31] <archivist_> andypugh gets hot by sliding on the tarmac
[13:38:49] <archivist_> healed yet?
[13:39:38] <andypugh> Mainly. Some parts are still sore.
[13:39:55] <andypugh> I think I am getting old, I used to heal a lot faster.
[13:40:43] <archivist_> fall off on grass it is softer...dont ask me how I know
[13:41:11] <pcw_home> You will know you're old when some parts stay sore...
[13:43:33] <jdh> forever?
[13:44:01] <archivist_> probably
[13:44:04] <Gabe_W> do you think the picture at the bottom of this data sheet showing the M12 connector is looking at the male or female?
http://www.encoder.com/literature/datasheet-260.pdf
[13:45:08] <archivist_> Gabe_W, get a magnifier and a bright light, may well be marked in the moulding
[13:46:02] <pcw_home> Looks like the male to me
[13:47:08] <Gabe_W> i have a 10x loop but i would have to take the encoder off to see the pin numbers
[13:47:19] <Gabe_W> i think im going to give up
[13:47:37] <Gabe_W> bring my scope in tomorrow
[13:48:28] <Gabe_W> i have literally made zero progress
[13:48:35] <archivist_> the table has the colour codes of the wires
[13:48:48] <Gabe_W> that would be great if i would be using the cable they sell
[13:49:00] <Gabe_W> but i got mine from mcmaster
[13:49:27] <Gabe_W> not the same colors
[13:49:35] <Gabe_W> i had to probe the pins
[13:50:15] <Gabe_W> screw this, im getting the arduino through a program in it, see what it can read
[13:50:31] <Gabe_W> i can't wait until tomorrow, it will be all i think about all night
[13:50:43] <archivist_> arduino is wasted time
[13:51:51] <jdh> I have a female M12 8pin cable that is opposite of the data sheet.
[13:55:00] <pcw_home> You might pull the cable and check with ohmmeter the the pinout to the male connector is as shown (stick a resistor lead or some such in the receptacle)
[13:55:10] <pcw_home> that the
[13:58:27] <andypugh> Gabe_W: It will be showing the part that they supply.
[13:59:14] <andypugh> And it seems to be pretty clearly the view looking in to the connector.
[13:59:19] <pcw_home> Right, the cable docs will show the female
[13:59:46] <Gabe_W> okay well there is one problem
[14:00:02] <andypugh> (The encoder connector, that is)
[14:00:20] <Gabe_W> i matched the wires to that picture, it should be the mirror image
[14:00:58] <pcw_home> A common enough error....
[14:01:27] <Gabe_W> hopefully i didn't mess anything up
[14:01:44] <pcw_home> probably not
[14:01:52] <jdh> the second set always works better.
[14:02:14] <Gabe_W> lol, it took 2 weeks for them to make this one
[14:02:33] <Gabe_W> i don't want a second one
[14:02:44] <jdh> you had an m12 cable made?
[14:03:29] <pcw_home> mirror image would just connect 5V to B (which would be open)
[14:03:44] <Gabe_W> the encoder
[14:04:02] <Gabe_W> it was a "sepcial order" item
[14:04:06] <Gabe_W> so they said
[14:04:41] <Gabe_W> which meant bend over and pucker up, because we are going to ream you for this product
[14:05:47] <jdh> for the funky CPR?
[14:06:55] <pcw_home> Its not funky if you're Babylonian
[14:09:05] <jdh> We have systems with 750 count encoders that do cw/ccw and are still stuck with them. (gets harder to source them)
[14:09:46] <pcw_home> 7200 lines is 1/80 of a degree per count
[14:10:18] <pcw_home> CW/CCW is pretty odd
[14:11:47] <jdh> the 750 is the more odd part evidently. It only goes one direction so the cw/ccw isn't very interesting.
[14:12:06] <pcw_home> and a little-lump-in-the-cable adapter would be easy
[14:15:55] <tjb1> Who here is running a plasma table with a water table?
[14:16:04] <jdh> you!
[14:18:00] <tjb1> That doesnt help :P
[14:19:32] <tjb1> Holy balls…$450 for 2 piece of 5x5' 11 gauge with bent ends
[14:19:44] <tjb1> I could buy the metal and the brake almost for that price
[14:23:25] <JT-Shop-2> me
[14:31:27] <tjb1> JT-Shop-2: What did you pay for you water pan
[14:31:32] <Gabe_W> its 28,800 counts total
[14:31:42] <Gabe_W> and it still doesn't work
[14:31:44] <Gabe_W> so its zero counts
[14:31:45] <tjb1> I called my metal supplier and she said $140 just for a 5x10 sheet of 11 gauge which is pretty reasonable
[14:31:54] <tjb1> Ill just weld the sides on
[14:32:25] <Gabe_W> actually when i hooked it up in mirror from before the voltage between 5v and gnd dropped to 2v
[14:32:56] <Gabe_W> so i just twisted all the wire's together and hooked to the ac main
[14:35:21] <Gabe_W> don't worry the breaker was off...
[14:35:30] <Gabe_W> but i felt better
[14:36:01] <JT-Shop-2> tjb1: I have no idea
[14:36:09] <JT-Shop-2> it was several years ago
[14:36:20] <tjb1> Hmm ok
[14:40:07] <Gabe_W> what does the pull up resistor do?
[14:42:46] <archivist_> pulls the open collector to 5v, it supplies the collector current
[14:43:14] <Gabe_W> it doesn't seem to work
[14:43:32] <pcw_home> The open collector output is a switch to GND. When the switch is on it will be at GND potential
[14:43:33] <pcw_home> When off the output is unconnected, so the pull-up resistor "pulls up" the output to 5V
[14:43:33] <Gabe_W> as soon as i touch the A or B channel wire to the in pin, it drops to gnd
[14:43:53] <Gabe_W> no matter what size resistor i use
[14:44:14] <archivist_> if it is currently low then that is correct
[14:44:15] <pcw_home> that sounds like a parallel port issue
[14:44:39] <pcw_home> Yes and theres that
[14:45:14] <pcw_home> is the parallel port pin high with just the pullup?
[14:45:18] <archivist_> you need to rotate the encoder, very very slowly to see it changing
[14:45:48] <andypugh> Gabe_W: Which "the pin" do you mean?
[14:51:13] <tjb1> Hey JT-Shop -
http://www.youtube.com/watch?v=pD3zaRatH5M
[14:52:07] <andypugh> That's pretty speedy
[14:52:23] <tjb1> What is?
[14:52:30] <andypugh> Seems like a lot of Z travel for a plasma?
[14:52:32] <JT-Shop> nice work
[14:52:41] <JT-Shop> I think you need more Z :)
[14:52:53] <L84Supper> Gabe_W:
http://denethor.wlu.ca/pc200/logic/ttl_open_pullup.png here's a simplified schematic of an open collector circuit, the transistor has been replaced with a switch as an equivalent
[14:53:13] <JT-Shop> I actually wish I had 4" of Z travel
[14:53:17] <tjb1> Well with the 8" carriage there I only have 7
[14:53:27] <tjb1> Thats if I run it to both extremes
[14:53:46] <L84Supper> when the collector of the transistor is Open the output is pulled up to the Vsupply
[14:53:58] <tjb1> And there is ~5" between the top of water table and bottom of gantry
[14:54:11] <tjb1> so only 2 inch extra which will allow about an inch each side for limits and stuff
[14:54:12] <JT-Shop> seems like you don't need any more Z than what fits under the gantry
[14:54:28] <L84Supper> when the collector is closed it pulls the output down near the Ground
[14:54:48] <JT-Shop> being able to raise the torch up to change tips will be a nice feature
[14:54:53] <andypugh> I would have thought you don't technically need any more than the thickness of material you expect to cut?
[14:55:05] <tjb1> Its about good JT, andypugh I will also have the DWP611 router on it
[14:55:21] <tjb1> and I wanted to use the 8" carriage with a router instead of the 4"
[14:55:50] <andypugh> It definitely doesn't hurt.
[14:56:18] <andypugh> If you need to drill, and so need a chuck, suddenly you run out of Z rather quickly
[14:58:04] <DaViruz> i usually use a collet chuck for less z waste
[14:58:37] <andypugh> Router collets seem to only come in a very few sizes though.
[14:59:21] <andypugh> And I am serioulsy thinking about making a BT30 collet with the taper cavity bored up insie the BT taper.
[14:59:55] <DaViruz> making collets is hard :(
[15:00:08] <tjb1> JT-Shop, care to help me figure out the numbers I need for the rack and pinion configuration
[15:00:14] <DaViruz> at least to me
[15:00:38] <andypugh> I wouldn't make the collet. Just the tapered cavity. Possibly in-situ bored by the magic of CNC.
[15:01:14] <JT-Shop> numbers?
[15:01:43] <tjb1> For the stepconf
[15:02:02] <JT-Shop> how many steps per inch?
[15:11:16] <tjb1> How to get the leadscrew pitch for the rack and pinion
[15:12:01] <L84Supper> what spot diameters are found with the CO2 lasers used by people here for cutting?
[15:12:11] <andypugh> tjb1: Stepconf tries to do all the sums for you, but it might be easier to just set all values to 1 and just enter microstep and use the steps-per-inch as motor steps per rev.
[15:12:15] <JT-Shop> how many teeth on the pinon?
[15:12:25] <tjb1> 22 JT-Shop
[15:12:37] <JT-Shop> what is the pitch of the rack?
[15:12:40] <tjb1> 20
[15:13:06] <tjb1> andypugh: if I did that it wouldnt be correct would it?
[15:13:10] <andypugh> tjb1: Alternatively, use the real 200 steps per rev, and enter the travel-per-rev as leadscrew pitch.
[15:13:26] <tjb1> CNCRP says it is 22 teeth on a 1" pitch circle
[15:13:39] <tjb1> so I got 3.14" as the Diametral pitch
[15:13:45] <andypugh> Pitch circle is irrelevant
[15:13:46] <tjb1> So one rev would be 3.14"?
[15:13:47] <JT-Shop> any other gearing
[15:13:57] <tjb1> Its a 2.73:1 reduction from the stepper
[15:14:41] <andypugh> Actually, ignore what I just said.
[15:14:56] <JT-Shop> so 200 steps per rev X microsteping X 2.73 = one rev at the output shaft
[15:15:11] <JT-Shop> the rack is 20 teeth per inch?
[15:15:15] <tjb1> Not sure
[15:15:19] <tjb1> I cant find that info
[15:15:22] <andypugh> If your pitch circle is 1" then the travel is 1" per rev.
[15:15:37] <JT-Shop> there you go
[15:15:41] <andypugh> (as long as pi =1)
[15:16:14] <andypugh> (You were right with the 3.14, but use more digits)
[15:16:29] <tjb1> So it would be pi is the distance travel per rev?
[15:16:38] <tjb1> 3.142.......
[15:17:08] <andypugh> OK, starting from the beginning. You have a 22DP gear. 22 teeth on a 1" dia pitch circle.
[15:17:09] <tjb1> and then for pulley teeth(motor:leadscrew): I put in 2.73:1
[15:17:22] <tjb1> andypugh: this is exactly what it says -
http://pastebin.com/raw.php?i=TN9dCv5L
[15:17:49] <andypugh> Pitch circle circumference is 3.1415…"
[15:17:51] <tjb1> 22 tooth gear with 1" pitch circle
[15:18:08] <tjb1> 20 pitch rack 20 degree pressure angle
[15:18:38] <andypugh> So you don't need any more numbers. You travel 3.14…" inch per rev, so the effective pitch is 3.14"
[15:18:55] <tjb1> Actuallay reading that now that I am not dead tired it appears I need to take 3.142/2.73
[15:18:59] <tjb1> *actually
[15:19:37] <andypugh> However, Stepconf in Imperial is expting TPI, not pitch. So your "leadscrew pitch" is 1/3.14
[15:19:45] <tjb1> So 1.15076653978 linear travel per rev?
[15:19:56] <andypugh> Sounds about right, yes.
[15:20:15] <tjb1> So the pitch is something under 1 lol
[15:21:55] <andypugh> Put it this way, the final SCALE value in the INI neesd to be either pi/2.73 or 2.73/pi and it will be pretty obvious if it is wrong.
[15:22:46] <andypugh> (all x microstep x steps/rev)
[15:23:19] <tjb1> Well I get .8689 rev per inch the way I did it
[15:23:34] <tjb1> 2.73/pi = .8689
[15:24:01] <andypugh> Use more digits
[15:24:25] <tjb1> 0.86898598928
[15:24:46] <andypugh> That should be enough. But they are free and the computer doesn't care.
[15:24:53] <skunkworks> does pitch circle matter? If the 22 tooth pinion - doesn't it go 22 teeth per rev on the rack? (never thought about racks before...)
[15:25:44] <andypugh> skunkworks: Just consider the gear and rack as a wheel and track. The wheel diameter is the PCD. That's all there is to it.
[15:25:47] <skunkworks> or 1.1 inches per rev?
[15:26:08] <tjb1> andy, do I need to put 2.73:1 in the boxes above?
[15:26:24] <skunkworks> huh
[15:26:25] <tjb1> Pulley teeth (Motor:Leadscrew):
[15:26:50] <andypugh> skunkworks: Yes, it does 22 teeth per rev, but how far is 22 teeth? That depends on the pitch, which is in terms of DP.
[15:27:36] <skunkworks> ah - ok
[15:28:00] <andypugh> And a 22 tooth 22DP gear and a 132-tooth 132DP gear both move pi inches per rev.
[15:29:16] <tjb1> andypugh: :P
[15:29:26] <tjb1> 2.73:1 in the Pulley Teeth?
[15:30:00] <andypugh> tjb1: Or use the actual tooth counts
[15:30:18] <tjb1> Which would boil down to 2.73:1
[15:31:01] <andypugh> Isn't 273 prime? That makes me think there is rounding there too.
[15:31:19] <tjb1> Have to wait until I get back home to count them
[15:31:22] <andypugh> (assuming it isn't a 273:100 ratio)
[15:32:18] <andypugh> Ah, no, Wikipedia says that it is 3 x 7 x 13
[15:32:27] <tjb1> Well thanks, I will let you know how it goes
[15:32:32] <tjb1> Probably wednesday night
[15:33:00] <tjb1> I also couldnt get Velocity over 1.1 in / s in stepconf
[15:38:44] <andypugh> Well, isn't this interesting?
http://en.wikipedia.org/wiki/Yan_Tan_Tethera
[15:39:29] <andypugh> tjb1: It was clearly moving a lot faster than that in the video
[15:39:36] <tjb1> ;)
[15:39:43] <tjb1> Couldnt get it higher in the step conf test axis
[15:40:44] <DJ9DJ> gn8
[15:51:45] <Gabe_W> Okay well end of the day, and though i stopped working on it because of other obligations. I will be bringing my scope in tomorrow and really look at whats going on with the encoder
[15:53:05] <Gabe_W> would cable length cause an issue?
[15:53:10] <Gabe_W> its a 12ft cable
[15:53:28] <Gabe_W> extremely long for what i need
[15:54:22] <JT-Shop> shielded cable?
[15:55:20] <Gabe_W> yes
[15:55:32] <JT-Shop> grounded on one end only?
[15:55:55] <Gabe_W> yes, i guess i will be able to tell tomorrow
[15:56:02] <Gabe_W> see the noise in the scope
[15:58:59] <Gabe_W> well im off building linux from scratch at home so thats why i haven't been on the forumn
[15:59:34] <Gabe_W> went to IMTS last weekend, amazing how much things change in 2 years
[15:59:58] <Gabe_W> i want a 3d printer
[16:00:35] <Gabe_W> it blow's my mind how it can make a working unit moving parts and all
[16:50:48] <andypugh> eBay is a strange place. I just bought
http://www.ebay.co.uk/itm/320843776796 for £140 from a bricks-and-mortar shoppe.
[16:51:49] <andypugh> They even say "RRP £220" then have the price at £289
[16:59:18] <JT-Shop> what does the 700c imply?
[16:59:27] <andypugh> Just a wheel size
[17:01:17] <JT-Shop> this is my new bike
http://www.giant-bicycles.com/en-us/bikes/model/revel.1.black.blue/9042/48918/
[17:02:17] <andypugh> Your bicycle taste differs from mine as much as your motorcycle taste :-)
[17:02:44] <andypugh> I like fast, light, uncomfortable and inpractical.
[17:02:54] <JT-Shop> somewhat dangerous on the highway in my opion
[17:03:15] <JT-Shop> and we have 26 miles of trails just a couple of miles from my house
[17:03:48] <andypugh> I cycle to work, as long as the weather is good. If the weather is bad I use the R1
[17:05:33] <JT-Shop> I walk to work no matter what the weather
[17:05:44] <JT-Shop> how far do you cycle to work?
[17:05:55] <andypugh> 5 miles
[17:06:14] <JT-Shop> that won't take long at all
[17:06:15] <andypugh> loadrt lcd
[17:06:18] <andypugh> Doh
[17:06:25] <JT-Shop> ok it's loaded
[17:06:56] <JT-Shop> do they have proper places to ride a bike or do you have to share with cars?
[17:08:23] <andypugh> Roads are the proper place to ride a bike.
[17:09:03] <andypugh> I strongly resist the notion that bicycles belong on the pavemant, or only on special tracks.
[17:12:26] <JT-Shop> the problem around here are the roads are high speed 65 mph and if they are 2 lane 55 mph so you might get run over if your not careful
[17:12:54] <JT-Shop> there are folks that ride on the shoulder of the road but not many
[17:13:27] <andypugh> Most of my journey is urban 30/40 limit.
[17:27:03] <JT-Shop> that sounds like a safe place to ride a bike
[17:31:39] <s1dev> what do most people use for moving the table on a CNC mill, a screw?
[17:32:17] <Jymmm> For skunkworks... a forklift.
[17:33:03] <tjb1> Jymm!
[17:33:07] <tjb1> +m
[17:33:39] <tjb1> http://www.youtube.com/watch?v=pD3zaRatH5M
[17:33:46] <andypugh> s1dev: On a milling machine, a ballscrew is customary
[17:34:22] <andypugh> Because the cutting forces tend to be quite high, and travel speeds low.
[17:34:28] <s1dev> thanks
[17:34:44] <s1dev> nema23s are typically used, correct?
[17:35:01] <andypugh> Only on the very smallest machines
[17:35:32] <s1dev> and on the larger machines?
[17:35:49] <andypugh> Ac servos, DC servos, hydraulics.
[17:36:39] <andypugh> I think
https://www.youtube.com/watch?v=35tHYaDUmZQ (LinuxCNC) is using large DC servos, as it is quite old.
https://www.youtube.com/watch?v=35tHYaDUmZQ
[17:37:13] <andypugh> NEMA 23s would struggle a bit
[17:37:14] <tjb1> Jymmm: isnt that exciting
[17:38:30] <andypugh> (I love the way that that machine looks to not even be working hard, but the pile of chips grows rather rapidly)
[17:38:40] <Jymmm> tjb1: Can't tell, tell the cameraman to quit fucking jumping up and down =)
[17:38:56] <tjb1> Hard to hold a keyboard and a phone Jymmm ;)
[17:39:12] <Jymmm> tjb1: duct tape the phone to something STILL
[17:39:21] <Jymmm> tjb1: congrats btw =)
[17:39:33] <tjb1> thanks :D
[17:40:11] <Jymmm> Damn FCC... H U R R Y U P ! ! !
[17:40:34] <tjb1> What did the FCC do?
[17:41:53] <Jymmm> tjb1: NOTHING, that's the problem =)
[17:42:01] <tjb1> What are you trying to get done
[17:42:37] <andypugh> s1dev: However, my little mill runs NEMA23 steppers, yes. But the one I am workikng on to replace it is 500W servo motots.
[17:42:37] <andypugh>
[17:42:37] <Jymmm> I took my General and passed exam on saturday and just waiting for the records to get updated
[17:42:55] <Jymmm> Missed Extra by 3 quesitons
[17:42:58] <s1dev> 500W.... that's a lot of power....
[17:43:07] <Jymmm> will study and take that this weekend
[17:43:53] <andypugh> s1dev: It's a bigger machine. But not huge.
[17:44:20] <s1dev> what's the size?
[17:44:23] <tjb1> Whats this for Jymmm
[17:44:32] <andypugh> Servos typically run at a fraction of their rated power.
[17:44:40] <Jymmm> tjb1: amateur radio / ham radio
[17:44:43] <s1dev> what voltage?
[17:44:52] <tjb1> Cool
[17:45:28] <andypugh> s1dev: No good video yet (it still doesn't run) but
http://www.youtube.com/watch?v=pxrzJ_KfcQ0&feature=share&list=UUexvgsGz_QFvOublovDYoTQ
[17:45:37] <s1dev> thanks
[17:45:47] <Jymmm> tjb1: I'm surprised that I didn't just get 3 questions right, one of them being my name!
[17:46:32] <tjb1> You missed a question that was your name?
[17:47:44] <andypugh> s1dev: Mine are running on simple rectified mains, so about 300V
[17:48:34] <s1dev> any idea on whether NEMA23s would work good for a 12x24" table for a mill?
[17:48:50] <andypugh> But I also have some running on 24V. it maily depends on what motors were cheap on eBay
[17:49:34] <andypugh> I think that a machine that size would be a bit better on 34s. Or preferably servos.
[17:50:02] <andypugh> But if you are prepared to cut and move slowly, 23s would work.
[17:51:05] <andypugh> What is the machine?
[17:51:44] <s1dev> 3 axis mill
[17:52:31] <andypugh> Make? Model? Mass?
[17:52:48] <s1dev> my thought is to build a machine
[17:53:21] <andypugh> What do you want to machine?
[17:53:55] <s1dev> soft metals and steel mostly
[17:54:44] <s1dev> MAYBE stainless steel, but that wouldn't happen very often
[17:57:47] <andypugh> A machine to do that probably wants to be made of fairly big lumps of cast iron.
[17:58:50] <andypugh> Which isn't a problem (iron castings are fairly cheap) but machinign the castings gets you into something of a bootrstrap problem
[17:59:37] <toastydeath> not to mention machining castings FOR machines is a specialized skill in of itself
[18:00:35] <andypugh> It's almost certainly going to be easier and better to buy a base machine and retrofit/convert
[18:00:47] <s1dev> alright
[18:01:34] <s1dev> any brands to avoid?
[18:02:16] <andypugh> How much space are you looking to occupy?
[18:02:42] <s1dev> not quite sure, but less than a room :P
[18:03:03] <andypugh> And what sort of things do you want to make?
[18:03:48] <s1dev> not anything specific in mind but I'll only need 3 axis
[18:05:25] <s1dev> possibly a 4th,
[18:05:34] <s1dev> but that can be added later
[18:07:00] <toastydeath> generally speaking, a knee mill is considered the bee's knees for most hobbyists - the bridgeport form factor
[18:07:15] <andypugh> It's perfectly OK to just want to make a CNC machine, but if you want to make jewellry then the advice will be different from if you want to make race car engine parts.
[18:08:11] <andypugh> Though it is worth noting that very tiny accurate stuff tends to need a very stiff machine.
[18:09:10] <s1dev> I'm not going to be making small stuff, so that's good, and the table size was just a thought, maybe once I get a mill I could use that to make parts to make it larger
[18:09:21] <s1dev> *small mill
[18:09:21] <andypugh> For example the guy making these:
http://www.f1-2000.co.uk/index.php?f=conrods Is using really quite a big (but very accurate) machine.
[18:09:36] <toastydeath> accuracy and size are not terribly related
[18:10:27] <toastydeath> also, it's worth noting that the machine he's using appears to be a Deckel, which has always been a very high accuracy machine meant for die work
[18:10:41] <s1dev> I don't think tolerances of an F1 would be needed too
[18:11:07] <JT-Shop> up next the button box... last part of the BP conversion YIPPIE!
[18:11:33] <andypugh> I have handled one of his conrods, they are _lovely_ (I am helping with a LinuxCNC conversion on his lathe)
[18:15:53] <andypugh> toastydeath: And yes, it's a Deckel FP2A:
http://www.f1-2000.co.uk/index.php?f=new_workshop
[18:16:10] <toastydeath> nice stuff
[18:16:20] <andypugh> Which is a rather huge machine considereing the 300mm cube work envelope.
[18:16:23] <toastydeath> criticism of deckel, it's way overpriced for what it is
[18:16:44] <toastydeath> it's a very nice machine but they go for way too much money, lots of older people with money to drive the price up want them
[18:17:13] <s1dev> do you think I could get this, create parts for a larger table out of steel, then use the spindle from that to create a CNC using the larger parts for the table
[18:17:14] <s1dev> http://www.grizzly.com/products/Mini-Milling-Machine/G8689
[18:17:38] <toastydeath> s1dev, i'd recommend not using steel, and the answer is probably no if you want any degree of accuracy
[18:17:51] <toastydeath> Steel and machine tools don't mix in terms of construction
[18:18:00] <s1dev> what would you use?
[18:18:05] <toastydeath> gray iron
[18:18:39] <toastydeath> also, you are limited to the accuracy of your machine
[18:19:04] <s1dev> what would be limiting there?
[18:19:05] <toastydeath> if you start with a $500 mill, that's the maximum accuracy you will have - and the error tolerances stack quite badly when a whole machine is put together
[18:19:15] <andypugh> Steel is OK, really. Stiffness is very close to iron (unlike Aluminium which is only about 30% of steel/iron)
[18:19:23] <toastydeath> steel is bad because of chatter
[18:19:49] <toastydeath> you lose a large amount of the static stiffness to dynamic effects
[18:20:12] <toastydeath> gray iron and epoxy/granite composites are much, much better in that regard
[18:20:38] <toastydeath> if you're building forging equipment, however, steel is the way to go - it makes for a more efficient hammer
[18:20:49] <toastydeath> press equipment doesn't much care
[18:20:57] <andypugh> I suspect that the problem with steel is not that it is really less damped (though it is) but that it's really easy to make iron castings thick and strong. Fabricated steel will always be less contuiguous.
[18:21:08] <toastydeath> it actually makes a huge difference
[18:21:27] <toastydeath> I don't have any of the papers on the subject anymore, but the damping of cast iron is the main reason for its continued use.
[18:22:03] <s1dev> so what parts of
http://www.grizzly.com/products/Mini-Milling-Machine/G8689 make it bad to use as a base for a CNC?
[18:22:06] <andypugh> s1dev: How long have you been planning this project?
[18:22:07] <toastydeath> so much so that modern machine castings are designed with FEA to be dynamically stiff at the expense of static stiffness
[18:22:19] <s1dev> andypugh very early stages
[18:22:19] <toastydeath> "bad?"
[18:22:38] <toastydeath> depends on what your goals are and what you plan to do
[18:22:44] <s1dev> detrimental to accuracy
[18:22:44] <toastydeath> for a lot of people that machine is fine
[18:22:51] <s1dev> and what is "any degree of accuracy"?
[18:23:00] <toastydeath> s1dev, it's all subjective
[18:23:02] <andypugh> s1dev: Can I suggest that you read, and discuss, and plan for at least a month before pressing any "buy" buttons?
[18:23:15] <Tom_itx> at least it has a solid column
[18:23:17] <Connor> I like me G0704...
[18:23:18] <toastydeath> I've run machines that locate accurately to .0001" within the envelope
[18:23:26] <toastydeath> whereas most knee mills aren't anywhere near that
[18:23:31] <toastydeath> and the machine you're looking at is worse still
[18:23:35] <jdh> s1dev:
http://www.hossmachine.info/cnc_conversion.html
[18:23:36] <s1dev> andypugh I'm definitely going to be reading, discussing, and planning for atleast several months
[18:23:37] <jdh> same mill
[18:23:37] <Tom_itx> toastydeath at what time of day? :)
[18:23:45] <Tom_itx> i've found temps will alter the accuracy
[18:23:46] <toastydeath> Tom_itx, any time of day, they've got coolers on them
[18:23:50] <Tom_itx> yeah
[18:24:05] <Tom_itx> they have to
[18:24:40] <Tom_itx> our little okuma kadet would vary maybe .0001 or so throuout the day
[18:24:50] <s1dev> toastydeath my calipers can't even read to .0001" so that's pretty useless to me
[18:24:52] <toastydeath> okuma is a great machine
[18:24:59] <Tom_itx> yeah i like them
[18:25:00] <toastydeath> s1dev, but you can understand how accuracy is subjective
[18:25:23] <toastydeath> everyone in this room has a different view of what accurate is, and you'll have to figure out what you need if you want feedback that is tailored to your goals
[18:25:33] <Tom_itx> we had one with a pallet changer but most didn't
[18:25:45] <toastydeath> i personally hate tiny desktop mills and would never buy one, but there are loads of people who absolutely love them and work on them every day
[18:25:55] <s1dev> yes, that was why I was confused by " and the answer is probably no if you want any degree of accuracy"
[18:25:59] <Tom_itx> get what fits the budget
[18:26:05] <jdh> and what fits the space
[18:26:14] <andypugh> s1dev: This is my machine 3 years ago:
http://www.youtube.com/watch?v=95B8PDlnajk&feature=share&list=UUexvgsGz_QFvOublovDYoTQ It's a chinese Mini Mill / Lathe and I was reasonably happy with it at the time (all NEMA 23, for referece). But with experience the deficiencies for the sort of thing I want to make (and I stress that last bit) became obvious.
[18:26:18] <Tom_itx> it kills me to play with my sherline but it's what i got
[18:26:23] <toastydeath> s1dev, using a desktop mill means making lots of adjustments to it constantly
[18:26:28] <s1dev> maybe to .005"?
[18:26:43] <s1dev> I'm just throwing out a number, but that seems to be accurate enough
[18:26:44] <toastydeath> whereas on a big machine that's by a good manufacturer, if I type in "5.000" that's where it goes
[18:27:17] <toastydeath> less accurate machines mean more shimming and bending the part to get it to cut straight, more cutting and measuring
[18:27:31] <andypugh> I like my lathe as long as I don't go too near to the Colchester Student that th Motor Club have :-)
[18:28:19] <toastydeath> smaller machines mean more fitting work and careful design so you can use loose tolerances
[18:28:32] <toastydeath> assemble in place and machine as an assembly, stuff like that
[18:28:56] <toastydeath> some people love that level of craftsmanship, others don't
[18:29:11] <Tom_itx> some have to tolerate it too :)
[18:30:17] <toastydeath> take a trip to some of the home shop websites and forums
[18:30:27] <toastydeath> stay out of the CNC stuff, just to see how people do a lot of machining in general
[18:30:33] <Tom_itx> i could show you one
[18:30:49] <toastydeath> and you'll get an idea, as you watch more project pictures and whatnot, the working differences and limitations of each
[18:31:34] <s1dev> suggestions on the small ~$600 mills?
[18:31:54] <toastydeath> the problem is this - if you're buying new, pretty much everything is in the same category around that price mark.
[18:32:10] <Tom_itx> http://www.users.qwest.net/~kmaxon/page/side/mill100_137.htm
[18:32:13] <toastydeath> different models have different horrifying downsides to them, and you really have to investigate on the forum each different one
[18:32:13] <Tom_itx> i envy him
[18:32:27] <toastydeath> to see which one is most popular, best, etc
[18:32:49] <toastydeath> used, you could probably find a local bridgeport or knee mill for about that if you hang out long enough
[18:33:04] <toastydeath> but at that point you're buying a relative unknown
[18:33:49] <andypugh> s1dev: Where are you, in a global sense?
[18:33:55] <s1dev> United States
[18:34:11] <s1dev> or to be more specific, Texas
[18:34:28] <toastydeath> lotsa oilfield and turret lathes around those parts
[18:34:34] <toastydeath> should have plenty of machine tools
[18:34:55] <s1dev> oilfields are out West unfortunately
[18:35:19] <toastydeath> it's okay, the industry permiates the coasts
[18:35:34] <toastydeath> sucks if you're stuck in the middle/panhandle, though
[18:35:59] <s1dev> craigslist maybe?
[18:36:02] <toastydeath> nah
[18:36:06] <toastydeath> i mean keep an eye on it
[18:36:09] <Tom_itx> toastydeath where are you?
[18:36:11] <toastydeath> but auctions, estate sales, etc
[18:36:14] <toastydeath> near DC
[18:36:16] <toastydeath> usa
[18:36:18] <Tom_itx> k
[18:36:23] <s1dev> ah
[18:36:30] <s1dev> alright, thanks
[18:36:31] <toastydeath> asset sales for closing shops
[18:36:46] <toastydeath> it's an Internet Person Mistake to think that people with machine tools use the internet to sell the majority of them
[18:36:58] <toastydeath> most machine tool deals are done without having any trace of the machine on the internet
[18:37:27] <Tom_itx> often hard to find
[18:37:34] <toastydeath> brb food
[18:38:02] <JT-Shop> be right over
[18:38:16] <Tom_itx> you didn't even hear the menu
[18:38:26] <toastydeath> yeah i'm going to a burrito joint
[18:38:27] <JT-Shop> LOL does it matter?
[18:38:29] <toastydeath> it's not even in my house
[18:39:20] * JT-Shop tries to remember where the sealtight is...
[18:40:16] <andypugh> s1dev: Big budget-stretch but:
http://www.ebay.com/itm/Bridgeport-CNC-milling-machine-Textron-Vertical-Mill-Series-1-/271058277814?pt=LH_DefaultDomain_0&hash=item3f1c5515b6
[18:41:27] <Tom_itx> newer than the boss 5 we had
[18:42:02] <Tom_itx> tooling looks like crap
[18:42:17] <andypugh> Might clean up, might not
[18:42:57] <Tom_itx> pretty rusty
[18:43:28] <andypugh> it's _possible_ that the tapers (inside the rack) are cleaner.
[18:43:54] <andypugh> Do they do rust in Texas?
[18:44:04] <Tom_itx> depends how far south
[18:44:11] <Tom_itx> near the coast they would
[18:44:26] <Tom_itx> likely worn out milling oil rig stuff
[18:46:00] <andypugh> I suspect that even a worn-out bridgeport beats my machine.
[18:46:10] <Tom_itx> yeah it's right on the coast
[18:46:19] <Tom_itx> real close to Houston/Galveston
[18:46:38] <andypugh> At least the BP slides fitted each other once. Mine were wrong from the factory.
[18:47:02] <Tom_itx> Galveston is on a peninsula
[18:47:23] <JT-Shop> I got my lathe from a dealer in Ft Worth
[18:48:46] <Tom_itx> i saw the water line on the buildings in Galveston after Katrina
[18:49:02] <Tom_itx> well above the doors
[18:49:26] <s1dev> maybe the mill is rusted on the inside then
[18:49:35] <JT-Shop> getting close
http://imagebin.org/228780
[18:50:28] <andypugh> They claim that the slides and screws are like new. That's plausible as they are protected from coolant. You would certainly want to look before biding.
[18:50:38] <Tom_itx> i'd give $1200 for that one :)
[18:51:08] <Tom_itx> andypugh yeah, it's rather humid down there all the time
[18:51:09] <s1dev> why is it that old Bridgeports are $1-2K yet new Bridgeports are $15K+
[18:51:20] <Tom_itx> get a Jet
[18:51:22] <andypugh> There is a bit of a gap between Sherline and Bridgeport with not a lot of good stuff in it.
[18:51:26] <JT-Shop> quick switch tooling
[18:52:07] <s1dev> quick switch tooling is worth several grand?
[18:52:11] <andypugh> My Harrison is nicely in the middlem but they are not common
[18:52:26] <JT-Shop> they are about $100 each when not rusty
[18:54:01] <s1dev> ah
[18:54:15] <andypugh> JT-Shop: I have a feeling that one of the components does gray conversion, but is named in a way to conceal the fact?
[18:54:20] <JT-Shop> my BP in the photo I just showed uses them
[18:54:36] <JT-Shop> uh yea I think so
[18:55:23] <JT-Shop> let me go look on the CHNC
[18:56:12] <JT-Shop> mux16 does
[18:56:28] <andypugh> Yeah, that would work
[18:56:45] <JT-Shop> I think it is the only one that has gray code option
[18:58:11] <JT-Shop> Tom_itx: see the BP?
[18:59:06] * JT-Shop sees why andy asked that question
[19:00:00] <andypugh> We ought to have a gray-code converter comp really.
[19:00:37] <JT-Shop> it would be easier to find
[19:01:27] <andypugh> arbitrary width, preferably
[19:01:49] <JT-Shop> makes more sense than tacking it onto mux
[19:07:04] <andypugh> Night all
[19:07:14] <JT-Shop> good night
[19:08:53] <Tom_itx> i saw yours
[19:10:07] <Tom_itx> that looks like a short table
[19:10:16] <JT-Shop> mine?
[19:10:22] <Tom_itx> yeah
[19:10:25] <Tom_itx> maybe it's just the pic
[19:11:02] <JT-Shop> 42x9 with 24" of X
[19:11:29] <JT-Shop> 6' long counting the servo lol
[19:13:25] <JT-Shop> I just mounted the button box
[19:13:35] <JT-Shop> starting to clean up the mess...
[19:13:35] <s1dev> looking at these small mills what's the difference between a #3 Morse Taper and an R8 taper?
[19:14:05] <Tom_itx> oh #3 is pretty weak
[19:14:08] <JT-Shop> morse is usually for drills like a tail stock on a lathe
[19:14:11] <Tom_itx> R8 has a drawbar
[19:14:29] <JT-Shop> R8 is the typical bridgeport and clone type of collet
[19:14:34] <Tom_itx> #3 won't hold much side force
[19:14:38] <Tom_itx> for milling
[19:14:40] <s1dev> so R* is ideal?
[19:14:43] <s1dev> *8
[19:14:50] <Tom_itx> common
[19:15:13] <Tom_itx> we used 40 and 50 taper
[19:15:42] <Tom_itx> so somewhere inbetween is good
[19:18:33] <s1dev> http://www.ebay.com/itm/TREE-10-1-2-x-42-VARIABLE-SPEED-VERTICAL-MILLING-MACHINE-2UVR-/160883320556?pt=LH_DefaultDomain_0&hash=item257564a6ec
[19:18:39] <s1dev> too bad it's in the wrong state
[19:19:51] <s1dev> and the wrong voltage as well
[19:22:10] <Tom_itx> does mesa make a small standalone vfd card?
[19:24:09] <Tom_itx> most of those are likely gonna be 3 phase
[19:27:39] <djdelorie> why is 3-phase a problem?
[19:28:17] <Tom_itx> generally not run to residential addresses
[19:28:28] <Tom_itx> some cnc don't like phase inverters
[19:28:48] <djdelorie> even digital ones? I mean, my bldc motors are all three-phase...
[19:29:08] <Tom_itx> it's all generated though
[19:29:13] <Tom_itx> not from the supply
[19:29:52] <djdelorie> well yeah, but it's functionally the same as provided three-phase (and maybe cleaner ;) if done right
[19:31:43] <toastydeath> digital phase converters are okay for most machines
[19:31:53] * djdelorie wonders if my bldc drivers could drive my bridgeport...
[19:31:59] <toastydeath> 2% error is generally considered good enough for a cnc
[19:32:11] <toastydeath> in voltage between phases
[19:32:16] <djdelorie> 10% error is considered good enough for home power delivery...
[19:32:29] <toastydeath> there's only one phase
[19:32:39] <djdelorie> (good thing too :)
[19:32:56] <toastydeath> machines are a little more tolerant of total voltage error but not much
[19:33:00] <toastydeath> 5-6%
[19:33:31] <Tom_itx> we had the power company put a monitor on the feed for a couple weeks when the new shop was built
[19:33:37] <toastydeath> ^^ this
[19:33:49] <toastydeath> machine installers will do the same thing if you buy from a dealer
[19:44:11] <JT-Shop> how can the voltage be wrong when you can pick between all three?
[19:45:13] <JT-Shop> s1dev: for a spindle motor 3hp and less you just get a VFD that is single phase in 3 phase out
[19:45:31] <JT-Shop> I use a GS2 as LinuxCNC has a driver for it
[19:45:44] <Tom_itx> jt does mesa have vfd's for small motors?
[19:45:58] <JT-Shop> I've not seen any
[19:45:58] <Tom_itx> i was thinking about my sherline spindle
[19:46:06] <JT-Shop> Automation Direct does
[19:46:10] <Tom_itx> someone said i could replace the pot with something
[19:46:14] <JT-Shop> is your sherline 3 phase?
[19:46:17] <Tom_itx> no
[19:46:23] <Tom_itx> dc motor
[19:46:27] <Tom_itx> 90v? maybe
[19:46:30] <Tom_itx> i dunno
[19:46:32] <JT-Shop> line voltage pot?
[19:46:38] <Tom_itx> no
[19:46:43] <Tom_itx> i don't think so
[19:46:49] <Tom_itx> i haven't torn into it yet
[19:47:46] * JT-Shop looks at the mess around the BP and ponders cleaning up
[19:48:03] * djdelorie must be donig something right, as he just got a GS2 also...
[19:48:39] <L33TG33KG34R> tjb1: you were asking if someone here was slisgrinder?
[19:48:44] <tjb1> yes
[19:48:50] <L33TG33KG34R> why?
[19:49:09] <tjb1> Thats you
[19:49:14] <L33TG33KG34R> yeah, just wondering
[19:49:38] <tjb1> Whats your problem with the 220v
[19:49:58] <L33TG33KG34R> well I need to use a VFD and it needs 220V
[19:50:35] <tjb1> yes
[19:50:48] <L33TG33KG34R> so I am pondering whether or not I should route a 12-3 wire from the panel or if I should use 2 existing plugs that are on either side of the 220V center tapped and gang my breaker
[19:51:08] <djdelorie> never use two separate plugs
[19:51:30] <L33TG33KG34R> why do you say that?
[19:51:36] <jdh> Tom_itx:
http://www.cnc4pc.com/Store/osc/product_info.php?cPath=25&products_id=303
[19:51:44] <djdelorie> unplugging one doesn't make the machine safe to fiddle with
[19:52:08] <tronwizard> and can make the plug out dangerous
[19:52:09] <L33TG33KG34R> yes, I would need to unplug both
[19:52:20] <djdelorie> and if you plug one of of the two plugs in the "wrong" way or place, bad things might happen
[19:52:26] <JT-Shop> djdelorie: got the GS2 up and going?
[19:52:44] <djdelorie> not yet. Just got the BP delivered last week, the GS2 is still on the counter
[19:53:03] <djdelorie> http://www.delorie.com/photos/bridgeport/
[19:53:14] <L33TG33KG34R> well, its really simple, take the hots of the 2 plugs and the ground and done. You have 220V with ground to connect to the VFD
[19:53:18] <JT-Shop> IIRC if you don't set the communication protocall they won't talk
[19:53:25] <Tom_itx> jdh, is the voltage divider in the manual speed control 0-10v?
[19:53:37] <djdelorie> first step is to just get a fixed speed working, the mill isn't CNC
[19:53:42] <jdh> Tom: dunno about yours, but most of them are.
[19:53:43] <Tom_itx> i haven't even taken it apart yet
[19:53:46] <tjb1> Im pretty sure you cant just take 2 hots and get 220
[19:53:53] <JT-Shop> was it that clean when you got it?
[19:53:59] <tjb1> You need to get the opposite hot wires off the transformer
[19:54:07] <tjb1> Otherwise youll just have 120v
[19:54:21] <djdelorie> yes, the only thing I've done with it so far is reassemble it
[19:54:37] <L33TG33KG34R> I know, I made sure the plugs are on the opposite side of the center tap from the breaker panel
[19:54:50] <JT-Shop> djdelorie: you can set it for 60 cycles and jsut use the manual speed changer
[19:54:55] <djdelorie> L33T: also, worry about ground loops, faults, etc. Please just run a new 12-3 to a dedicated outlet
[19:55:13] <L33TG33KG34R> that is what I will end up doing. Not my house to begin with
[19:55:30] <djdelorie> JT-Shop: that was the plan. The GS2 will eventually be shared with the lathe, once it has a 3-phase motor, to give it reversibility too.
[19:56:10] <ReadError> hey anyone here done a will-call order with mcmaster?
[19:56:12] <JT-Shop> ah it is a step pulley model
[19:56:19] <djdelorie> The BP is a potential CNC target, if I get to that point the GS2 will get to show its full functionality...
[19:56:21] <ReadError> ssi ?
[19:56:39] <jdh> Read: I have a friend in atlanta that does it, just goes and picks it up.
[19:56:53] <jdh> but, he also gets same-day UPS delivery if he orders early
[19:57:14] <ReadError> jdh
[19:57:18] <ReadError> i placed an order
[19:57:24] <ReadError> was just wondering if they send you an email
[19:57:29] <ReadError> like ohai its ready to be picked up
[19:57:32] <JT-Shop> djdelorie: nice clean BP
[19:57:32] <ReadError> or you just show up
[19:57:35] <jdh> oh, never asked.
[19:57:45] <djdelorie> I'd say "thanks" but it wasn't my doing ;-)
[19:57:47] <ReadError> i should throw in some endmills and stuff
[19:57:52] <ReadError> g10 kills them
[19:57:55] <ReadError> (fast)
[19:57:59] <JT-Shop> good score then
[19:58:09] <djdelorie> It was a one-owner machine, he took good care of it. Lots of oil throughout, too :-)
[19:58:23] <JT-Shop> lucky score then
[19:58:34] <JT-Shop> say goodnight Gracie
[19:58:34] <djdelorie> and it was nearby, although it still took four hours and heavy equipment to move it to my shop
[19:58:51] * JT-Shop heads inside for the night
[19:58:56] <JT-Shop> see you guys later
[19:59:34] <ReadError> how much was shipping djdelorie?
[20:02:02] <djdelorie> $425 to move it 10 miles or so, from his garage to my basement (the excavator guy again). Well worth it. it was a beast to lift and manage, and my basement is not a walk-in.
[20:03:12] <Bigun913> hello all
[20:05:44] <s1dev> djdelorie: what all was involved in moving it?
[20:05:52] <s1dev> how do you get it on the truck?
[20:06:09] <djdelorie> it had to be pulled out of they other guy's garage into the open, and he had a dirt driveway.
[20:06:27] <djdelorie> it had to be lifted onto a trailer, and it exceeded the strength of the excavator so it was done in two parts.
[20:06:27] <s1dev> how was that managed?
[20:06:53] <djdelorie> er, rollers, boards to slide on, and an expert touch with an excavator
[20:07:09] <Bigun913> Sorry to interrupt Can anyone here help me with tool touchoff
[20:07:09] <s1dev> and how was it moved into the basement?
[20:07:22] <djdelorie> then the two parts were lowered into my bulkhead, and shifted out of the way with an engine hoist
[20:07:39] <djdelorie> like this:
http://www.delorie.com/photos/southbend-lathe/img_2418.html
[20:08:01] <djdelorie> (that stairwell is removable as a module, the excavator set it aside on the lawn)
[20:08:07] <jdh> Bigun: ask the question, someone might answer
[20:10:06] <Bigun913> I dont understand how to set up tool touchoff. I am new to linux. I have the machine running and cutting but would love to be able to do auto z setting. Can anyone point me to any resources or help?
[20:11:02] <Tom_itx> http://www.gnipsel.com/linuxcnc/g-code/gen05.html
[20:11:05] <Tom_itx> yes
[20:12:10] <Bigun913> Tom_itx is that yes for me
[20:12:20] * Tom_itx nods
[20:14:09] <Bigun913> Would you chat
[20:15:01] <djdelorie> Bigun913: you *are* chatting...
[20:16:02] <jdh> do you have fixed tool holders?
[20:16:14] <Bigun913> Ok I guess I am I meant Tom_itx would you like to chat privately so I dont tie up the channel?
[20:17:09] <Tom_itx> no need
[20:17:12] <Tom_itx> just ask here
[20:17:14] <Bigun913> ok
[20:17:20] <Tom_itx> i'm in and out
[20:17:32] <Bigun913> ok
[20:19:31] <Bigun913> I have a 48 by 48 cnc mill that I built from scratch. I have it up and running. I would like to do auto z setting but cant figure out the tool touchoff. I am new to linux and dont understand the process to setup touchoff.
[20:20:08] <jdh> do you have a tool height setter, or do you have fixed tool holders?
[20:20:39] <Bigun913> I have a fixed tool holder
[20:20:58] <Bigun913> I am using a router
[20:21:12] <Tom_itx> so you need to reset the tool each tool change
[20:21:41] <Tom_itx> you should read jt's guide
[20:21:43] <Bigun913> That is correct I would like to use a plate on the surface of the material and reset the z dim.
[20:22:01] <Bigun913> where do i find it?
[20:22:10] <Tom_itx> in the link i gave you
[20:22:29] <Bigun913> the gnipsel.com?
[20:22:52] <Tom_itx> yup
[20:22:53] <jdh> that doesn't do auto setting.
[20:23:32] <jdh> http://softsolder.com/2010/04/14/emc2-ugliest-tool-length-probe-station-ever/
[20:23:44] <jdh> that uses a touch-off switch
[20:23:53] <Bigun913> I came from mach 3 and like linuxcnc much more but could never get the linits or toolsetting to work
[20:24:12] <djdelorie> limits are easy to do, but not as obvious to understand
[20:25:00] <Bigun913> I have changed to a c10 board and I think that was my major problem I had was the first board I used
[20:26:27] <Bigun913> This new board was much simpler and seems to work much better
[20:29:09] <Bigun913> Thanks for the pointers I am gonna go read those thanks!!
[20:29:43] <skunkworks__> logger[psha]:
[20:29:43] <logger[psha]> skunkworks__: Log stored at
http://psha.org.ru/irc/%23linuxcnc/2012-09-18.html
[20:47:09] <L33TG33KG34R> if I just want to control 4 steppers, a hitachi x200 or some vfd, coolant pump, mister along with inputs for e-stop, 6 limit switches
[20:47:17] <L33TG33KG34R> is the 7i76 enough for that?
[20:48:50] <jdh> should be.
[20:49:02] <jdh> do you really need individual limit switches?
[20:49:36] <jdh> plenty of inputs, but more wiring hassle
[20:49:59] <Valen> home switches perhaps
[20:50:22] <L33TG33KG34R> home switch + limit switch
[20:50:27] <L33TG33KG34R> for each axis
[20:50:42] <L33TG33KG34R> I am an ENT student so wiring isn't a problem
[20:50:51] <jdh> ear nose & throat?
[20:50:54] <Valen> you say that
[20:51:00] <Valen> but your wrong lol
[20:52:03] <L33TG33KG34R> you're* and no. You have no idea what they put us through
[20:52:11] <jp_> one limit signal per axis is more than enough
[20:52:42] <L33TG33KG34R> but in that extreme is the 7i76 enough? Also how would I interface the 7i76 with a VFD?
[20:52:49] <jdh> it has 32 inputs
[20:53:08] <jp_> analog perhaps or modbus if you have it
[20:53:26] <jdh> it has an analog spindle output also.
[20:53:37] <jp_> the &i76 has analog out for spindle purpose
[20:53:40] <jdh> it's like they designed the card just for this!
[20:53:43] <djdelorie> I wired all six limit switches to just one input
[20:54:16] <djdelorie> I think.... maybe one input per axis...
[20:54:24] <jdh> I did one per axis
[20:54:54] <jp_> looking at the machine can tell yo what one tripped anyways
[20:55:02] <skunkworks__> Valen: your idea of lsof | grep media worked great
[20:55:45] <jp_> anyone used kelling's stepper drives?
[20:56:17] <jdh> I have some
[20:56:25] <jp_> how do you like them
[20:56:44] <jdh> I got the cheap ones. Might have preferred the slightly better ones, but they work fine.
[20:56:51] <Tom_itx> i've used em
[20:57:01] <Tom_itx> the 282 in oz ones
[20:57:07] <Tom_itx> on my sherline work good
[20:57:13] <jdh> drives, or motors?
[20:57:15] <jp_> they are just re branded imports anyways aren't they
[20:57:19] <jp_> drives
[20:57:30] <jdh> they are rebranded leadshines?
[20:57:36] <Tom_itx> oh, i got gecko drivers
[20:57:40] <Tom_itx> probably
[20:57:47] <Tom_itx> they all look the same
[20:58:34] <Tom_itx> i like the geckos
[20:58:41] <jp_> jdh, what do you use for a PSU?
[20:58:58] <jdh> generic 48v switcher. Same as the one keling sells, but from ebay
[20:59:16] <jp_> geckos seem over priced
[20:59:26] <jdh> I'd rather have a gecko
[20:59:28] <jdh> but, I'm cheap.
[20:59:32] <jp_> probably last forever
[20:59:46] <Tom_itx> yeah but that was one thing i didn't want 'cheap' on
[21:00:31] <jp_> can say much the drives i have are not cheap by any means
[21:00:33] <Tom_itx> and i can use them on a bigger mill
[21:00:48] <jp_> just looking for low cost units for small stuff
[21:01:03] <jdh> I thought about replacing mine, the better ones are cheap enough but the ones I have seem to work fine.
[21:01:04] <Tom_itx> they're probably fine for that
[21:01:47] <jp_> im sure they are i bought one of those small side steps for my extruder and they work great
[21:02:37] <Tom_itx> probably a better choice than the pololu's
[21:02:50] <Tom_itx> i hear they blow alot
[21:03:03] <Tom_itx> probably from stupidity but still
[21:03:36] <jp_> all that reprap stuff seems lightweight hate to try a nema 23 on it
[21:06:51] <Tom_itx> that little gear drive stepper looks ok
[21:07:04] <jp_> Tom_itx, have you tested out your hotend?
[21:08:40] <Tom_itx> no
[21:09:12] <Tom_itx> i made those for the heck of it and slowly may collect parts to make one
[21:09:18] <jp_> my 3mm one works great. 1.75 however is a pain in the ass
[21:09:35] <Tom_itx> i won't use 1.75 from what i hear
[21:10:15] <jp_> dont its not worth it spent 2 days and no i plan to give up on it
[21:10:36] <jp_> 3mm just worked out of the gate
[21:27:28] <skorket> do you guys suggest geda over kicad? Or something else entirely?
[21:27:40] <djdelorie> I prefer geda, but I'm biased...
[21:27:58] <skorket> why is that?
[21:29:51] <archivist_> I started the #kicad channel next door that probably makes me biased too
[21:30:29] <djdelorie> skorket: I'm one of the maintainers
[21:30:30] <jdh> pcb-ware-wars!
[21:30:49] <djdelorie> nah, try both and use whatever you want.
[21:31:01] <L84Supper> skorket, the pro world is pretty much Cadence vs Altium
[21:32:14] <archivist_> the last pro I used was dos PCAD, I still prefer it fo rease of use
[21:32:18] <L84Supper> I've been using Capture and Allegro for years for the most complex designs, but the cost is >$10K
[21:32:48] <skorket> yeah, I'm an amateur/hobbyist/poor person, so nothing that expensive for me
[21:33:04] <L84Supper> use whatever you like
[21:33:13] <jdh> eagle?
[21:33:16] <skorket> I remember picking geda a while ago but now I'm not sure why and I wanted to get some opinions
[21:33:27] <archivist_> eagle is limited pcb size
[21:33:37] <L84Supper> I see more open hardware designs that use Kicad
[21:33:45] <jdh> yeah, but I'm pretty limited.
[21:33:53] <L84Supper> for schematic capture
[21:34:09] <archivist_> watching the irc channel and I see steady growth in use
[21:35:23] <L84Supper> i hear that some people actually suffer through using it for hand routing very dense 3 mil trace/space PCB's
[21:37:41] <Tom_itx> i use eagle
[21:37:44] <L84Supper> the pro tools have lots of utils for automating, simulating and design rule checking
[21:38:12] <Tom_itx> i looked at pads
[21:38:33] <L84Supper> I'm also spoiled by large libraries vs virtually none for Kicad
[21:38:53] <Tom_itx> i've got well over 300 eagle libs
[21:39:34] <Tom_itx> majority of them are in the package now though
[21:42:47] <Thetawaves> did you get eagle standard?
[21:43:21] <Thetawaves> i have eagle light
[21:43:40] <Tom_itx> i paid for it years back and still use that ver for most of my stuff unless someone is asking for help
[21:43:41] <Thetawaves> i probably won't spend 600$ for 6 signal layers
[21:43:46] <Tom_itx> i have 2 or 3 ver loaded
[21:44:30] <Tom_itx> i've probably used it since ver 2 or so
[21:44:31] <Thetawaves> i make the footprint for roughly half my parts anyway, so switching to kicad shouldn't be that difficult
[21:44:38] <L84Supper> Eagle Light is only 2 layers max ?!
[21:44:43] <Tom_itx> yeah
[21:45:00] <Tom_itx> how many do you need?
[21:45:01] <L84Supper> oh only $69
[21:45:05] <Thetawaves> 'only two'
[21:45:35] <Tom_itx> if you're making 4 layer boards you're likely above the hobby level
[21:46:07] <Tom_itx> it's rather reasonable compared to pads or altium
[21:46:16] <djdelorie> I do many 4-layer boards, and it's hobby...
[21:46:32] <Tom_itx> i can do 16
[21:46:38] <Tom_itx> but haven't
[21:46:43] <L84Supper> whats PADS cost these days? <$10K
[21:46:46] <djdelorie> I've tested pcb with 56 layers
[21:46:58] <Tom_itx> that's just nuts
[21:47:14] <djdelorie> yes it was. I did a test board that was 0.25 miles on a side too. My house was very small on that one...
[21:47:21] <Tom_itx> but you get into BGA and that's what you get
[21:48:15] <djdelorie> on a 64-bit machine, I think pcb can do a board the size of our solar system...
[21:50:10] <L84Supper> PCB 100,000:1 scale?
[21:50:22] <djdelorie> 2^63 nanometers
[21:52:15] <L84Supper> PCB's with 1 atom res
[21:52:52] <djdelorie> it gave us 1E-5 inch resolution without rounding issues
[21:53:43] <djdelorie> on a 32-bit build, you're limited to PCBs up to 6 feet across ;-)
[21:56:15] <Valen> the lack of libs for kicad are what sucks
[21:56:24] <Valen> apart from that it does the job well enough
[21:57:54] <L84Supper> does kicad have a util for flooding areas with copper based on some spacing rule from traces of holes?
[21:58:22] <L84Supper> of/or
[21:59:36] <L84Supper> last time i checked I didn't see a nice way to handle ground and power planes
[22:05:04] <L84Supper> well the PCB software is going to start looking a bit different soon when you can 3D print PCB's with no machining or plating required
[22:21:30] <Thetawaves> that sounds stupid
[22:21:45] <Thetawaves> when i can etch a thousand boards at a time
[22:22:25] <s1dev> Thetawaves: at what cost?
[22:22:38] <L84Supper> etch 1-2 layers
[22:23:00] <Thetawaves> huh?
[22:23:02] <L84Supper> then you stil have to plate the vias and holes
[22:24:52] <s1dev> I wonder if anyone has looked into electroplating boards
[22:25:10] <s1dev> instead of etching
[22:27:13] <L84Supper> thats how multilayer boards are done, they sandwich core that's been plated and prepreg, then also plate the through holes and vias
[22:29:31] <Tom_itx> ever been thru a board house?
[22:29:38] <Tom_itx> they got big vats to plate in
[22:31:55] <L84Supper> http://www.youtube.com/watch?v=naXcRXo0tdU&feature=related 30+ minute vid of all the steps
[22:32:52] <Valen> they don't electroplate the copper onto the boards, its glued on
[22:32:55] <Valen> well epoxy
[22:33:08] <Valen> very hard to electroplate to stuff thats non conductive
[22:33:21] <Valen> i spose you might be able to do some kind of metal spraying
[22:35:03] <L84Supper> for microwave we use PCB cores made of teflon and fibergalss plated with copper
[22:35:29] <Valen> sure its plated not glued foil?
[22:36:36] <Tom_itx> my understanding is it's plated
[22:39:53] <L84Supper> I'm currently working on printing the traces directly onto 3d printed core layers, a 3d PCB printer that only uses fluids
[22:41:47] <L84Supper> Valen:
http://www.polyflon.com/Plating%20Plastics.htm re your question about plating vs bonding
[22:42:28] <L84Supper> not that all teflon pcb's are made this way, some are plated and some are laminates
[22:56:42] <L84Supper> http://ideas.repec.org/a/wsi/srlxxx/v14y2007i02p241-253.html here's an abstract on the electroless copper plating of PTFE, I can't a free version yet
[22:58:03] <L84Supper> basically they exposing the PTFE films to a sodium naphthalenide (Na/naphtha) etchant, and esterification of 4,4'-azobis(4-cyanopentanoic acid) (ACP) with the hydroxyl groups covalently linked to the surface, followed by the surface-initiated free radical polymerization of 4-vinylpyridine (4 VP)
[22:58:41] <L84Supper> it's not really plated directly to the PTFE since they modify the surface of the PTFE quite a bit first
[23:55:56] <tjb1> sheetz run!!!
