#linuxcnc | Logs for 2012-03-19

Back
[00:00:28] * mikebronner cut angle or radius for shaped bits. perhaps what I'm thinking of is specified in the CAM software?
[00:03:21] <jdhnc> yes
[00:03:59] <jdhnc> the CAM should also compensate for radius
[00:05:33] <Tom_itx> you select that with the tool
[00:05:47] <Tom_itx> ball, bull flat etc and give the tool tip radius
[00:06:23] <Tom_itx> i've had a couple odd ones i fudged a bit on like a tapered cutter
[00:07:46] <mikebronner> ah :) so if I specify all my tools in the CAM software, I don't need to worry about the tool table in EMC2/AXIS, is that correct?
[00:07:51] <Tom_itx> then the cam software calculates the path accordingly
[00:08:15] <jdhnc> if your CAM requests tool changes, EMC will stop and prompt you to change tools
[00:08:17] <Tom_itx> ie if you're cutting a 3d surface with a bull nose cutter, it will know what the edge radius of the tool is
[00:08:53] <Tom_itx> you can enter descriptions in the tool table that match your gcode file
[00:09:52] <mikebronner> tom: without having to worry about the offsets, right? Just tool number and description?
[00:10:23] <Tom_itx> the length offset is important
[00:10:30] <Tom_itx> and i'd enter the diameter
[00:10:48] <Tom_itx> that's where your H word gets it's value1
[00:10:52] <Tom_itx> -1
[00:10:56] <mikebronner> won't that then conflict with the compensations calculated by the CAM software
[00:11:14] <Tom_itx> it depends how you set up your tooling and offsets
[00:11:29] <Tom_itx> i have a certain way i'm used to doing it and it works
[00:11:30] <Tom_itx> for me
[00:11:57] <Tom_itx> others use tool touch probes in linuxcnc and likely set the gcode file up differently
[00:12:22] <Tom_itx> you will find a way that works best for you
[00:12:49] <Tom_itx> for example, my z zero is usually set above the work surface
[00:12:50] <mikebronner> hmmm, ya, I guess I just need to get my hands dirty and discover a workflow :)
[00:13:15] <Tom_itx> but not everybody does it that way
[00:16:46] <mikebronner> I'll give this a shot, thanks for the advice Tom, jdhnc, and Jymmm
[00:41:20] -!- rob__H has quit [Ping timeout: 272 seconds]
[01:16:56] -!- AitalMAC [[email protected]] has joined #linuxcnc
[01:35:42] <elmo40> mikebronner: you *need* to set the tool length. that is very important. if you don't then how will the CNC know where to put the spindle to make the tool cut the part?
[01:36:12] <elmo40> as Tom_itx said... everyone has a different way of doing things ;)
[01:36:29] <mikebronner> elmo: if I set all my tool parameters in my CAM software, won't that be accounted for in the g-code?
[01:37:07] -!- mhaberler has quit [Quit: mhaberler]
[01:37:11] <elmo40> some do everything from the bed. make the bed 0 and measure everything from it... tool length, fixture height, material height... everything. Over the years it has been the most widely used method that I have seen
[01:37:27] <elmo40> tool height is not set in the CAM. it is set on the Machine.
[01:38:16] <mikebronner> i'm probably very confused at this stage, since all I have achieved so far is to cut out the sample provided with EMC2/AXIS
[01:38:33] <mikebronner> will EMC2 alert me if there is something I didn't do right?
[01:38:57] <mikebronner> and what is the difference between tool length and tool height?
[01:39:11] <elmo40> tool configuration (ball, flat, chamfer...) the CNC does 'know' what that it... all it does it interpret G-code into a movement. D and H is to tell the machine how far to offset(compensate).
[01:39:30] <elmo40> length and height _should_ mean the same thing ;)
[01:40:19] <elmo40> but if you are talking about CAM to Machine then length is usually from the bottom of the spindle to the tip of the cutter. On the Machine Height is what you set it as (off the table for 0 or off the part for 0)
[01:42:06] <mikebronner> hmmm, i'm probably not seeing what you mean. If I home my Z-Axis to the height of the table, then the machine knows where the bit is, same if I home it to the top of the part.
[01:42:19] <elmo40> how will it know?
[01:42:35] <elmo40> if you change the drill to a ball nose, how will it know the difference in height?
[01:42:38] <elmo40> you have to tell it.
[01:42:52] <elmo40> measure it, probe it, however you do it but do it all the same way with all the tools
[01:43:04] <mikebronner> ah, now I see what you mean
[01:43:21] <elmo40> you can home the Z and then measure from the table to the tip of the tool (I've seen that method. not an accurate way but works)
[01:43:56] <mikebronner> I believe CamBam accounts for this, as it is asking me for all my tool lengths. I will have to experiment and see how it reacts in EMC2 with different bit lengths.
[01:44:12] <mikebronner> thanks for the heads up and helping me understand :)
[01:51:06] <Tom_itx> there are several zeros to keep in mind. machine zero, the fixture offset zeros and the tool length zero point or better known as the Tool length offset
[01:51:23] <Tom_itx> fixture offset zero would be like a G54
[01:51:45] <Tom_itx> also shown in your gcode but set at the machine
[01:52:10] <Tom_itx> your gcode would just say to use that offset and the machine would know where that was
[01:52:36] <mikebronner> boy, i can see it will take some time before one masters this :)
[01:53:02] <Tom_itx> it seems confusing until you've done it once or twice
[01:54:08] <mikebronner> yea, i think that's what it will take … i'll need to make some trial items and learn from that
[01:54:18] <Tom_itx> the gcode doesn't know how long the tool is that's why you specify a H word for each tool
[01:54:33] <Tom_itx> then the gcode knows where the tool tip is
[01:56:35] <Tom_itx> when i do a tool change i specify a 'T' number and a corresponding 'H' number and based on that the gcode will look up the value stored at the H word address and apply that to the gcode for that tool
[01:57:24] <mikebronner> So why is it that the CAM software can't add these values if it has them in its tool table?
[01:57:50] <Tom_itx> ok back up a sec
[01:58:01] <Tom_itx> for example i specify T1 M6
[01:58:16] <Tom_itx> then G01 z.1 H1
[01:58:37] <Tom_itx> at that point the program will look up the number stored at H1 for that tool
[01:58:49] <Tom_itx> all the gcode knows is that is the offset for that tool
[01:59:09] <Tom_itx> the machine stores the actual offset number based on how long the actual tool is
[02:00:02] <mikebronner> I see, so this works like a look-up value and references the table in the machine each time it is used
[02:00:20] <Tom_itx> yes
[02:00:45] <Tom_itx> the same applies to the x y and z values of any work offset like G54 55 56 etc
[02:02:17] <elmo40> it may sound overwhelming right now, so many offsets and all, but it will sync in ;)
[02:03:37] <mikebronner> so this means that i have to maintain two sets of tool libraries: one in CAM for shapes, sizes, etc, and another in EMC2 for length
[02:03:57] <Tom_itx> in the gcode you just tell it what tool to use and what offset your workpiece is at then the program uses the stored lookup values at those locations
[02:04:45] <Tom_itx> my cam has a tool setup page for diameter, tip shape, feeds and speeds etc to use
[02:05:09] <Tom_itx> that information is output into the gcode
[02:05:52] <elmo40> in the g-code it will, for example, plunge to Z-0.500". If you change a tool that value will still be -0.500" but since the tool is of a different length it may cut too deep or not touch at all. This is why the machine has to offset for each tool. Then it will make every tool hit the (exact) same z-0.500"
[02:06:47] <mikebronner> ah, i see … theoretically the CAM software could do it, if it would change all occurrences of Z-0.500" according to the currently installed bit
[02:07:06] <Tom_itx> yes
[02:07:34] <elmo40> but that is just insane to keep track of
[02:07:37] <Tom_itx> in fact since i don't have a tool changer on my little mill, sometimes i break the program up into separate tools
[02:07:46] <elmo40> and a b!tch to maintain!
[02:07:49] <Tom_itx> yes
[02:08:07] <elmo40> now, to add to this you also have a 'wear' offset ;)
[02:08:13] <mikebronner> hehe, ya, especially if you have to maintain the geode manually
[02:08:28] <elmo40> once you set the tool length (which is machine values) you can fine tune the tooling with wear values.
[02:11:56] -!- Nick001 has quit [Read error: Connection reset by peer]
[02:12:32] -!- FinboySlick has quit [Quit: Leaving.]
[02:24:06] -!- phantoxe has quit [Read error: Connection reset by peer]
[02:24:27] -!- |n0b0dy| [|n0b0dy|[email protected]] has parted #linuxcnc
[02:24:53] -!- phantoxe [[email protected]] has joined #linuxcnc
[02:40:17] -!- iwoj has quit [Ping timeout: 244 seconds]
[02:42:09] <mikebronner> night guys
[02:42:29] -!- mikebronner [[email protected]] has parted #linuxcnc
[02:43:56] -!- elmo40 has quit [Quit: Leaving.]
[02:47:14] -!- elmo40 [elmo40!~Elmo40@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has joined #linuxcnc
[02:51:19] -!- elmo40 has quit [Ping timeout: 246 seconds]
[03:03:01] <ssi> sweet, these SSRs are 3-12VDC control
[03:03:30] <ssi> I think I'm gonna feed 5V back down to the machine from the PC power supply so I can just sink through the relays to an FPGA pin to switch everything
[03:03:45] <ssi> they're 3-32VDC actually
[03:03:58] <ssi> the way the machine is configured originally has them all running on 12VDC
[03:04:54] <ssi> and there's one SSR that has its 120V input hot whenever the machine is hot, and the control can switch that which activates a mechanical relay which makes the 120v inputs hot on all the other SSRs
[03:08:20] <pcw> If you use common 5V on SSR+ leads you should set the FPGA I/O to open drain mode
[03:09:17] <ssi> only 2.6mA through the SSR control in the on state, fwiw
[03:09:46] <pcw> is_output, invert_output, and is_opendrain attributes on all output bits
[03:09:52] <ssi> ok
[03:10:22] <pcw> (assuming you want all to be true=on)
[03:10:29] <ssi> yea, seems reasonable :)
[03:11:05] <ssi> open drain is basically Z state?
[03:11:21] <ssi> stiff low, weak high?
[03:11:23] <ssi> something like that?
[03:13:09] <pcw> open drain means sink only
[03:13:32] <ssi> gotcha
[03:13:43] <ssi> is there still the 24mA sink limit in open drain mode?
[03:14:04] <pcw> it does drive weakly high because of the 3.3K pullups
[03:14:16] <pcw> Yes 24ma max
[03:16:46] <pcw> That reminds me of a customer using one of our 3X20s (144 I/O)
[03:16:48] <pcw> They complained when their test file shorted out the 3.3V power supply on the 3X20
[03:16:50] <pcw> I asked they what their test file did and they said, just outputs the 50 MHz clock on every pin!
[03:17:46] <pcw> so dont try to source/sink 24 mA on every pin...
[03:17:56] <ssi> heheh
[03:18:28] <ssi> well I have worst case like eight SSRs to run
[03:18:34] <ssi> at 2.6mA apiece on 5v
[03:18:42] <ssi> they flow like 7.5mA on 12V
[03:18:59] <ssi> and actually some of those are mutex
[03:19:04] <ssi> like high clutch/low clutch
[03:19:30] <ssi> and I don't think I'll be using SSRs for the coolant and spindle like the original machine did
[03:19:38] <ssi> probably talk directly to the VFDs
[03:21:11] <pcw> all the outputs at 10mA or so is probably OK also they are not all likely to change at onece
[03:21:54] <ssi> also I'm using a '42 for IO breakout
[03:22:01] <ssi> which is what, 50R serially terminated?
[03:27:26] <pcw> Yes the 7I42 has a 50 Ohm series resistor and diode clamps to 0 and 4.4V
[03:30:21] -!- phillc54 [[email protected]] has joined #linuxcnc
[03:33:53] -!- phantoxe has quit []
[03:39:28] -!- Nick001 [[email protected]] has joined #linuxcnc
[03:44:40] -!- Gast200 [[email protected]] has joined #linuxcnc
[03:48:52] -!- koax has quit [Ping timeout: 265 seconds]
[03:53:40] -!- automata__ [[email protected]] has joined #linuxcnc
[04:03:28] -!- automata__ has quit [Ping timeout: 244 seconds]
[04:37:14] -!- iwoj [[email protected]] has joined #linuxcnc
[04:45:24] -!- A2Sheds has quit [Ping timeout: 265 seconds]
[04:50:33] -!- Gast200 has quit [Quit: Bye Bye]
[04:59:37] -!- AitalMAC has quit [Ping timeout: 245 seconds]
[04:59:58] -!- AitalMAC [[email protected]] has joined #linuxcnc
[05:00:37] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[05:01:08] -!- A2Sheds has quit [Client Quit]
[05:01:32] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[05:20:41] -!- factor has quit [Ping timeout: 265 seconds]
[05:24:05] -!- pcw has quit [Ping timeout: 246 seconds]
[05:24:44] -!- pcw_ has quit [Ping timeout: 245 seconds]
[05:25:14] <ssi> woooo
[05:25:19] <ssi> got my first valid signal wired up :)
[05:46:34] -!- psha[work] [psha[work][email protected]] has joined #linuxcnc
[05:59:25] -!- capricorn_one has quit [Remote host closed the connection]
[06:05:04] -!- ve7it has quit [Remote host closed the connection]
[06:13:05] -!- factor [factor!~factor@r74-195-218-139.msk1cmtc02.mskgok.ok.dh.suddenlink.net] has joined #linuxcnc
[06:23:08] <alex4nder> hey
[06:23:32] -!- ewidance [[email protected]] has parted #linuxcnc
[06:41:12] -!- vladimirek [[email protected]] has joined #linuxcnc
[06:46:56] -!- Tecan has quit [Quit: live long and phosphor]
[07:07:58] -!- vladimirek has quit [Remote host closed the connection]
[07:30:18] -!- mhaberler [[email protected]] has joined #linuxcnc
[07:37:44] -!- A2Sheds has quit [Ping timeout: 260 seconds]
[07:49:44] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[07:54:49] -!- DJ9DJ [DJ9DJ!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[07:55:01] <DJ9DJ> moin
[07:57:23] <alex4nder> yoh
[08:17:45] -!- ewidance [[email protected]] has joined #linuxcnc
[08:31:26] -!- vladimirek [[email protected]] has joined #linuxcnc
[08:34:04] -!- Valen [Valen!~Valen@2001:44b8:3199:9100:21e:8cff:fe63:55df] has joined #linuxcnc
[08:37:17] -!- rob_h [[email protected]] has joined #linuxcnc
[08:46:07] -!- vladimirek has quit [Remote host closed the connection]
[08:48:21] -!- superkuh has quit [Read error: Connection reset by peer]
[08:52:56] -!- Valen has quit [Quit: Leaving.]
[08:53:12] -!- Valen [Valen!~Valen@2001:44b8:3199:9100:21e:8cff:fe63:55df] has joined #linuxcnc
[08:56:53] -!- A2Sheds has quit [Ping timeout: 246 seconds]
[09:04:01] <Loetmichel> mornin'
[09:04:06] -!- mhaberler has quit [Quit: mhaberler]
[09:04:49] <Loetmichel> hrhr, my co-worker called me a "blacksmith" for using M8 * 65mm screws. normally our screw storage ends at M5...
[09:05:59] <Loetmichel> i told her that i had worked with 10kg hammers and 32mm wrenches. THAT is heavy metalworking, nbot the 8mm screwa ;-)
[09:06:07] -!- ewidance [[email protected]] has parted #linuxcnc
[09:14:41] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[09:22:42] <Valen> we use 32mm wrench all the time
[09:22:48] <Valen> for our teensy spindle ;->
[09:23:18] <Valen> now these http://en.wikipedia.org/wiki/File:SydneyHarbourBridgeNutMilsonsPoint.JPG are some nuts ;->
[09:23:38] <Valen> they stop this from floating away http://en.wikipedia.org/wiki/File:Sydney_Harbour_Bridge_from_Circular_Quay.jpg
[09:46:22] -!- ewidance [[email protected]] has joined #linuxcnc
[09:57:19] <archivist> a slight problem with nuts up in scotland http://news.bbc.co.uk/1/hi/scotland/edinburgh_and_east/7648230.stm
[09:57:36] -!- ewidance [[email protected]] has parted #linuxcnc
[09:57:40] -!- mk0 [[email protected]] has joined #linuxcnc
[10:20:37] -!- bedah [[email protected]] has joined #linuxcnc
[10:25:08] -!- phantoxe [[email protected]] has joined #linuxcnc
[10:29:50] <Loetmichel> Valen: our company produces Computers for military/services. we have screwa down to 1.2mm but only up to 5mm ;-)
[10:29:55] <Loetmichel> screws
[10:54:28] -!- Poincare has quit [Remote host closed the connection]
[10:58:50] -!- jthornton [[email protected]] has joined #linuxcnc
[11:03:19] -!- Poincare [Poincare!~jefffnode@2001:470:cb24:4::2] has joined #linuxcnc
[11:13:58] -!- Thetawaves has quit [Quit: This computer has gone to sleep]
[11:21:49] -!- servos4ever [servos4ever!~chatzilla@74-45-201-222.dr01.hnvr.mi.frontiernet.net] has joined #linuxcnc
[11:35:36] -!- erasmo has quit [Quit: Leaving]
[11:36:33] -!- alex_joni has quit [Quit: ayeee]
[12:02:30] -!- mhaberler [[email protected]] has joined #linuxcnc
[12:05:57] -!- alex_joni [alex_joni!~alex_joni@emc/board-of-directors/alexjoni] has joined #linuxcnc
[12:05:57] -!- mode/#linuxcnc [+v alex_joni] by ChanServ
[12:14:13] -!- phantoxe has quit [Ping timeout: 248 seconds]
[12:18:00] -!- stfwi [[email protected]] has joined #linuxcnc
[12:23:05] -!- phantoxe [[email protected]] has joined #linuxcnc
[12:27:15] -!- ries has quit [Read error: Connection reset by peer]
[12:32:56] -!- ries [[email protected]] has joined #linuxcnc
[12:33:11] -!- stfwi has quit [Ping timeout: 246 seconds]
[12:35:41] -!- ewidance [[email protected]] has joined #linuxcnc
[12:37:30] -!- ewidance [[email protected]] has parted #linuxcnc
[12:37:37] -!- ewidance has quit [Remote host closed the connection]
[12:38:16] -!- ewidance [[email protected]] has joined #linuxcnc
[12:39:35] -!- ewidance has quit [Client Quit]
[12:50:59] -!- sumpfralle [[email protected]] has joined #linuxcnc
[12:58:05] -!- FinboySlick [[email protected]] has joined #linuxcnc
[12:58:07] -!- mhaberler has quit [Read error: Connection reset by peer]
[12:58:21] -!- stfwi [[email protected]] has joined #linuxcnc
[12:59:24] -!- vin321 [[email protected]] has joined #linuxcnc
[13:07:23] -!- sumpfralle has quit [Remote host closed the connection]
[13:08:13] -!- sumpfralle [[email protected]] has joined #linuxcnc
[13:08:55] -!- skunkworks [[email protected]] has joined #linuxcnc
[13:45:17] -!- maximilian_h has quit [Ping timeout: 246 seconds]
[13:45:24] -!- mk0 has quit [Quit: http://neo-technical.wikispaces.com/emc2-arch]
[13:59:34] -!- stfwi_ [[email protected]] has joined #linuxcnc
[13:59:34] -!- stfwi has quit [Read error: Connection reset by peer]
[13:59:35] stfwi_ is now known as stfwi
[14:01:21] -!- psha[work] has quit [Quit: Lost terminal]
[14:03:17] -!- maximilian_h [[email protected]] has joined #linuxcnc
[14:04:07] <skunkworks> http://bbs.homeshopmachinist.net/showthread.php?t=53295&highlight=linuxcnc
[14:04:11] -!- Valen has quit [Quit: Leaving.]
[14:05:49] -!- Cylly [[email protected]] has joined #linuxcnc
[14:06:38] -!- Loetmichel has quit [Ping timeout: 246 seconds]
[14:20:13] -!- pjm [[email protected]] has joined #linuxcnc
[14:23:31] -!- pjm__ has quit [Ping timeout: 264 seconds]
[14:26:18] Cylly is now known as Loetmichel
[14:29:02] <mozmck> skunkworks: the quote in the last post is laughable.
[14:29:58] <mozmck> how many distributions does windows have? How many kernel versions? The guy's bias is pretty obvious.
[14:35:19] <Jymmm> Windows BOB FTW!!!
[14:36:05] <cradek> There is a [unnamed] guy who makes [product that doesn't work with linuxcnc]. He says linuxcnc sucks!!1!
[14:36:20] <cradek> I can't believe I scanned that whole thread for that punchline
[14:36:41] <mozmck> :)
[14:37:30] <skunkworks> heh
[14:37:41] <skunkworks> sorry - just random morning reading
[14:37:48] <Jymmm> blame skunkworks who coulda just linked to that posting directly http://bbs.homeshopmachinist.net/showpost.php?p=752148&postcount=37
[14:38:06] * Jymmm gives skunkworks a wedgy!
[14:38:08] <skunkworks> Then you would have missed my posts...
[14:38:18] <Jymmm> skunkworks: EXACTLY
[14:38:28] -!- pcw [[email protected]] has joined #linuxcnc
[14:38:31] <skunkworks> you guyzzzz
[14:39:02] <cradek> 10 years ago linux WAS by geeks for geeks. today, people who what "everyone" to be able to use it are crapping it all up
[14:39:15] <cradek> who want
[14:39:23] <cradek> gack, can't type
[14:39:45] <mozmck> Yeah, some folks think that unless it looks and acts like a "smart"phone nobody can use it.
[14:39:56] <Jymmm> I blame ubuntu with all their dependancies for that.
[14:41:49] -!- stfwi has quit [Read error: Connection reset by peer]
[14:42:05] -!- stfwi [[email protected]] has joined #linuxcnc
[14:42:35] <mozmck> I do like the search feature in the "dash" in unity, but not the lack of a real menu. Mint put a menu back and added a search box to it, which is nice.
[14:43:37] -!- pcw_ [[email protected]] has joined #linuxcnc
[14:44:02] <alex4nder> morning
[14:44:47] <Jymmm> cradek: PS doesn't have "block style" commenting does it? I couldn't find anything in the manuals.
[14:45:08] <cradek> not that I know of
[14:45:21] <Jymmm> cradek: Ok, thanks.
[14:45:22] -!- jthornton has quit [Quit: ChatZilla 0.9.88.1 [Firefox 10.0.2/20120215223356]]
[14:46:56] -!- jthornton [[email protected]] has joined #linuxcnc
[14:51:59] <Jymmm> cradek: Is '/rx { cos mul } def' == 'cos(r) * x' ?
[14:52:24] <Jymmm> This RPN confuses me sometimes =)
[14:52:43] <cradek> yes it's calculating r cos theta
[14:53:11] <cradek> put r theta on the stack and call rx to get the x rectangular coordinate
[14:53:55] <Jymmm> cradek: Realize that I have no clue on the maths, just breaking down parts of it and then googling =)
[14:54:31] <cradek> it would have been nice to write r theta -> x y, but I was too tired
[14:54:58] <Jymmm> cradek: hey, this works for me =)
[14:55:56] <Jymmm> I'm still dombfounded that something so simple is so complex. It's a fricken lampshade?! lol
[14:56:40] <cradek> yeah, the math isn't straightforward
[14:58:38] <alex4nder> deconstructed cones?
[14:58:56] <cradek> yes
[14:59:03] -!- pjm has quit [Read error: Connection reset by peer]
[14:59:23] <cradek> you can sure use right triangles to figure them out, but it doesn't come out very cleanly
[14:59:24] -!- pjm [[email protected]] has joined #linuxcnc
[14:59:29] <alex4nder> haha
[14:59:30] <Jymmm> cradek: And of course I have to make it more complex by attempting to make 'D' not at one end of 'h', but part way in the middle (still trapizodial in shape) http://img3.etsystatic.com/il_fullxfull.302444839.jpg
[15:00:05] <Jymmm> Yet still have 'h' the full height of the frame or a bit longer.
[15:00:42] <Jymmm> ignore the "hourglass'ish" figure of that frame.
[15:00:44] <cradek> you're trying to make a shape that bends to fit over that? That's not possible without folding or cutting.
[15:00:57] <Jymmm> cradek: no no no no.
[15:01:25] <pcw> Who needs trig? Pythagoras FTW
[15:01:40] <Jymmm> cradek: Still striaght lines, but the WIDEST part being half way up the height of the frame, instead of 'D' being at the bottom of it.
[15:01:51] <alex4nder> that's easy though
[15:02:01] <alex4nder> create the cone, and deconstruct it
[15:02:10] <alex4nder> you have constant slope
[15:02:39] <Jymmm> alex4nder: Great, now make a vector drawing of that theory of yours =)
[15:03:01] -!- tensaiteki has quit [Quit: Bye Bye]
[15:04:06] <alex4nder> Jymmm: haha,.. IRC consulting gets expensive. ;)
[15:04:42] <Jymmm> alex4nder: Oh, well in that case, let me insert m credit card here --> [ ]
[15:04:56] <alex4nder> haha
[15:05:33] <Jymmm> alex4nder: cradek was very kind enough to write a PS that does it already.
[15:06:02] <Jymmm> alex4nder: But, I'm clueless, and tyring to learn PS and the math all at the same time now.
[15:06:16] <Jymmm> 1 + 1 == 3, right?
[15:06:45] <cradek> close... 1 + 1 < 3
[15:07:13] <djdelorie> 1+3=3 for very large values of "1"
[15:07:14] <cradek> or 1 + 1 ~= 3
[15:07:36] <Jymmm> Ah, I guess I'll hae to get new batteries for my solar powered abacus then.
[15:17:53] <pcw> naive question: do any CAM programs generate NURBs and
[15:17:54] <pcw> would NURBS be away around the one block lookahead limitation for high speed profiling?
[15:19:07] <pcw> s/away/a way/
[15:24:56] <ssi> pcw: rhino3d claims to be nurbs modeling, and they have rhinocam plugin for it, but I don't really know if that answers your question
[15:28:04] <pcw> That may just be for input
[15:28:22] <ssi> yes, that may be true
[15:28:48] <ssi> does linuxcnc even support NURBS blocks?
[15:29:12] <ssi> well look at that, it does
[15:29:15] <ssi> neat-o
[15:29:19] <pcw> its has NURBS gcode (G5.2 G5.3)
[15:29:33] -!- psha [[email protected]] has joined #linuxcnc
[15:30:09] <pcw> not sure what limitations it has (number of control points block etc)
[15:30:20] <ssi> so I got the very first working signal hooked up on my HNC last night :D
[15:30:37] <jthornton> Sweet!
[15:30:56] <ssi> GPIO pin attached to the machine.is-on signal, connected to CR2, the SSR that drives the mechanical relay that puts line power on all the other SSRs
[15:31:10] <ssi> it's not much, but it's a start :)
[15:31:28] <pcw> Well thats a start! I will look into the resolver issues today on my test machine
[15:31:39] <ssi> thanks :)
[15:32:03] <ssi> I wish I had a way to identify for sure what the transformation ratio of the resolvers is
[15:32:43] <pcw> well maxoutput/input
[15:33:02] <ssi> it seems to be around 1:3
[15:33:31] <pcw> or (sin^2+coss^2)/input
[15:33:39] <ssi> and btw I did manage to scope each drive pin referenced to local ground and sum them
[15:33:42] <ssi> and the driver is fine :)
[15:33:45] <pcw> sqrt
[15:35:00] <pcw> OK ive never broken one despite being careless with our scratch monkey (plugging and unplugging resolver cables hot)
[15:36:35] <ssi> pcw: https://p.twimg.com/AoXC1HMCEAAtiut.jpg:large
[15:38:17] -!- stfwi has quit [Read error: No route to host]
[15:38:43] <pcw> The fact that you cannot change the excitation frequency means
[15:38:45] <pcw> either a bitfile error or a version error in LinuxCNC resolver.c
[15:38:47] <pcw> I'll get to the bottom of this today
[15:39:41] <ssi> I may try to hand-match some dividers
[15:39:57] <ssi> how stiff do they need to be?
[15:41:21] <pcw> you can just reduce the drive to make it work (once the other errors are sorted out)
[15:41:23] <pcw> but its better for noise immunity to have as much signal as you can get
[15:41:25] <pcw> so desensitizing the sine/cos inputs is the best way I will see about
[15:41:27] <pcw> making a H version of the 7I49 for high output resolvers
[15:41:52] -!- GoSebGo [[email protected]] has joined #linuxcnc
[15:43:27] <ssi> oje
[15:43:31] <ssi> oke even :)
[15:44:10] <pcw> the sine cos inputs are 100K input impedance so a low part of the divider being 2-10K is probably OK
[15:44:12] <cradek> pcw: [other product] has three output amplitude jumpers that give you 7 settings in the obvious way
[15:44:40] <cradek> takes a lot of real estate I bet, but it's a good scheme
[15:45:24] <pcw> I can set the output amplitude in software but its better noise wise to have as much signall as you can get
[15:45:46] <cradek> ah I see what you mean
[15:46:28] <pcw> that may be why the resolvers have such a high step up ratio (pulse noise from drives)
[15:48:30] <pcw> I will add AGC to the resolver firmware eventually but I'd like to get in the ballpark first
[15:50:19] <pcw> I already calculate the signal magnitude (radius) but just for a resolver OK flag
[15:50:26] <frysteev_> wooo resolver fun
[15:57:56] -!- GeorgeH has quit [Ping timeout: 246 seconds]
[16:07:23] -!- jthornton has quit [Quit: ChatZilla 0.9.88.1 [Firefox 10.0.2/20120215223356]]
[16:08:17] -!- mhaberler [[email protected]] has joined #linuxcnc
[16:09:38] -!- sumpfralle has quit [Ping timeout: 245 seconds]
[16:13:20] -!- JT-Shop [[email protected]] has joined #linuxcnc
[16:19:04] <mazafaka> ssi: Because we're grown-ups now, and it's our turn to decide what that means (http://xkcd.com/150/)
[16:32:30] <mazafaka> and romantic: http://xkcd.com/162/
[16:32:49] -!- ve7it [[email protected]] has joined #linuxcnc
[16:37:03] -!- automata__ [[email protected]] has joined #linuxcnc
[16:38:35] -!- JT-Shop has quit [Quit: ChatZilla 0.9.88.1 [Firefox 10.0.2/20120215223356]]
[16:40:59] -!- maximilian_h has quit [Ping timeout: 245 seconds]
[16:48:26] -!- mhaberler has quit [Ping timeout: 272 seconds]
[16:56:49] -!- maximilian_h [[email protected]] has joined #linuxcnc
[16:59:10] <skunkworks> http://www.youtube.com/watch?v=ynsxPLl9OPU
[16:59:52] -!- mhaberler [[email protected]] has joined #linuxcnc
[17:00:23] <skunkworks> http://www.youtube.com/watch?v=wW-v0k08m-g
[17:00:26] -!- m42 has quit [Quit: m42]
[17:01:21] <cradek> assuming those are supposed to be straight lines, that doesn't look right at all...
[17:01:40] <skunkworks> I was hoping he would show the preview...
[17:02:02] <skunkworks> But yes - I bet the constants are not right
[17:07:19] -!- mhaberler has quit [Ping timeout: 264 seconds]
[17:08:45] -!- syyl [[email protected]] has joined #linuxcnc
[17:10:59] -!- mhaberler [[email protected]] has joined #linuxcnc
[17:15:27] -!- mhaberler has quit [Ping timeout: 245 seconds]
[17:15:46] -!- mhaberler [[email protected]] has joined #linuxcnc
[17:19:20] -!- mhaberler_ [[email protected]] has joined #linuxcnc
[17:20:03] -!- mhaberler has quit [Ping timeout: 245 seconds]
[17:20:06] mhaberler_ is now known as mhaberler
[17:22:34] -!- mhaberler has quit [Client Quit]
[17:23:12] <ssi> mazafaka: as much of an xkcd fan as I am, I'm confused as to why you're quoting it at me ;)
[17:25:01] <mazafaka> just funny, these words are nice to explain something
[17:43:11] -!- tom3p [[email protected]] has joined #linuxcnc
[17:44:32] -!- syyl_ [[email protected]] has joined #linuxcnc
[17:45:29] <tom3p> i made a VM out of a dapper emc2 with APT360/vapt/postp, it runs, but the generic xorg.conf has no OpenGL.
[17:45:49] <tom3p> i get emc to start but stops w/o GL. how to add OGL ?
[17:46:40] -!- syyl has quit [Ping timeout: 252 seconds]
[17:47:21] -!- jthornton [[email protected]] has joined #linuxcnc
[17:50:50] -!- Thetawaves [[email protected]] has joined #linuxcnc
[17:57:38] -!- erasmo has quit [Read error: Connection reset by peer]
[18:00:03] -!- vladimirek [[email protected]] has joined #linuxcnc
[18:01:09] <ssi> So I guess I want to get into customizing the UI
[18:01:19] <ssi> I would like to add a collet open/close button and indicator
[18:01:25] <ssi> anyone have any tips on how to get started on that?
[18:01:47] <cpresser> ssi: take a look at the pyvcp-examples
[18:02:12] <jthornton> ssi: single button for open/close?
[18:02:23] <ssi> jthornton: not sure!
[18:02:56] <jthornton> I have two physical buttons on mine, one for open and one for close
[18:03:46] <cpresser> a checkbox might work. but this depends on you hardware and hal-wiring
[18:03:52] <jthornton> IIRC the valve is a detent one with two coils
[18:03:53] -!- IchGuckLive [[email protected]] has joined #linuxcnc
[18:04:03] <IchGuckLive> Hi all around the globe B)
[18:04:03] <jthornton> you could have two pyvcp push buttons
[18:04:53] <jthornton> ssi: there are some examples on the forum too in the hal examples section for single button toggle
[18:05:05] * jthornton hangs up the phone now :/
[18:05:09] -!- jthornton has quit [Quit: ChatZilla 0.9.88.1 [Firefox 10.0.2/20120215223356]]
[18:10:03] -!- mhaberler [[email protected]] has joined #linuxcnc
[18:14:36] <ssi> onoes
[18:15:18] <frysteev_> ssi: what did you do now?
[18:15:45] <ssi> I got boss'd and then jt runnoft
[18:17:05] <IchGuckLive> ssi pyvcp push buttons
[18:17:22] <ssi> yea I need to look into it
[18:17:26] <ssi> I haven't messed with pyvcp at all
[18:17:52] <IchGuckLive> what is the goal to reatch
[18:18:08] <tom3p> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?One_Button_Toggle
[18:18:45] <IchGuckLive> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?PyVCP
[18:21:07] -!- tom3p has quit [Read error: Connection reset by peer]
[18:21:30] -!- tom3p [[email protected]] has joined #linuxcnc
[18:22:41] <tom3p> never ever ever ever try to get a channel list with xchat, always hangs, window goes white. dangit forgot AGAIN
[18:23:42] -!- isssy [[email protected]] has joined #linuxcnc
[18:23:56] <IchGuckLive> tom3p: O.O
[18:32:32] <cpresser> if you want to have a 'more complex' logical function you can try to write your own python userspace component (at least for non-realtime-stuff like coolant, collets, ...)
[18:36:08] <ssi> looks simple enough
[18:36:10] -!- mhaberler has quit [Read error: Connection reset by peer]
[18:36:26] <IchGuckLive> ssi: it is as you are more into it
[18:37:25] <djdelorie> tom3p: my xchat has no problem downloading a channel list
[18:37:26] -!- erasmo has quit [Remote host closed the connection]
[18:37:43] -!- isssy has quit [Quit: Bye Bye]
[18:37:59] <tom3p> djdelorie, thx i been trying to debug this
[18:38:50] * djdelorie is running xchat 2.8.8 on Fedora 14
[18:45:19] * frysteev_ is running on very little sleep
[18:47:47] <ssi> me too!
[18:48:14] <Loetmichel> frysteev_: to less time or to much children?
[18:48:18] <Loetmichel> ;-)
[18:48:38] <frysteev_> my only childer run on gcode
[18:50:58] <Loetmichel> so to less time it is
[18:51:05] <Loetmichel> work related?
[18:51:22] <frysteev_> to many side jobs to pay the bills
[18:53:06] <Loetmichel> ouch
[18:53:20] <Loetmichel> AM i happy that my wife pays 90% of the bills
[18:53:40] <jdhnc> nice
[18:53:41] <Loetmichel> so my salary is nearly complete free for my hobby
[18:53:46] <frysteev_> SUGGA MOMMA
[18:54:04] <jdhnc> I need one of those... My wife is a public school teacher.
[18:54:59] -!- isssy [[email protected]] has joined #linuxcnc
[18:55:32] <Loetmichel> my wife too.... but worked in her universitiy time as a telesales. and got a call from a headhunter one day after graduation... now she is a "upper management" at SAP ;-)
[18:55:52] <IchGuckLive> jdhnc: so you are the star on dadys day with a little fast running cnc router in the classroom
[18:55:56] <frysteev_> "upper management" meaning shes on top?
[18:56:16] <Loetmichel> correction: she says: only middle mamagement, 5 reporting steps up to the CEO ;-)
[18:57:03] <jdhnc> she has two more degrees than I do ban makes less than half as much
[18:57:16] <jdhnc> s/ban/and/
[18:58:22] <ssi> jdhnc: nobody ever said public school would earn them big bucks :P
[18:58:27] <IchGuckLive> is there in the usa also childrens working days over the year where they meet real stuff of manufacturiing
[18:59:06] <Loetmichel> jdhnc: opposite here.. i earn 2800 eur a month... 1452eur after taxes... she has about triple ;-)
[18:59:10] <IchGuckLive> 7,8,9 grades
[18:59:33] <ssi> 50% taxes
[18:59:37] <ssi> and you still get to pay VAT
[18:59:40] <ssi> spectacular
[18:59:45] <Loetmichel> ssi; nearly
[18:59:54] <jdhnc> geez
[19:00:17] <IchGuckLive> ssi 19% in Germany VAT and 7 on fod
[19:00:23] <Loetmichel> taxes, unemplyment assurance ant all the other state fees
[19:00:24] <jdhnc> hmm... my wife makes more than 2800eu, maybe it's not so bad :)
[19:01:21] <Loetmichel> jdhnc: i am just head of production in a small military computer company
[19:01:27] * ssi hugs the US
[19:01:59] <Loetmichel> ssi: but its so much because in germany, if you are married, you can split the tax class
[19:02:26] <ssi> not sure what that means exactly, but ok :)
[19:02:28] <Loetmichel> both can have class4, so both gets roughly the same taxing as non-married
[19:03:20] <Loetmichel> or one can have five, the other three. then the one with five gets a hight tax rate and the other with three a lower than unmarried
[19:03:34] <IchGuckLive> TAX is good this runs the country
[19:03:36] <ssi> I see... maybe
[19:03:43] -!- phantoxe has quit []
[19:03:56] <Loetmichel> so i have class 5
[19:03:58] <Loetmichel> ;-)
[19:04:03] <ssi> Loetmichel: OH I see
[19:04:10] <ssi> so your wife gets taxed at the low rate, but she makes the money
[19:04:15] <ssi> that's bizarre, but interesting
[19:04:16] <Loetmichel> exactly
[19:05:04] <ssi> tax loopholes are an international language ;)
[19:05:58] <Loetmichel> its not THAT much oif a difference, but about 1k to 5k a jear may be in. (in our case, compared to 4/4)
[19:06:31] <IchGuckLive> ssi there are also lots of returns on the years taxreturn
[19:06:58] <ssi> yeah, good old interest free loans to the government
[19:07:01] <ssi> everyone's favorite pastime
[19:07:41] <Loetmichel> IchGuckLive: only if you have enough bills to "write off" to the IRS
[19:07:56] <IchGuckLive> agree
[19:09:02] <Loetmichel> most years we have to pay a little tax rather than getting returns, cause of the class3/5
[19:09:30] <Loetmichel> but so we are not loaning the government money interest free ;-)
[19:11:56] <ssi> k taxes are depressing
[19:12:02] <ssi> let's talk about something else :D
[19:12:18] <djdelorie> you could talk about cnc.. ;-)
[19:12:19] -!- mhaberler [[email protected]] has joined #linuxcnc
[19:12:23] <ssi> yeah, that :D
[19:12:26] <IchGuckLive> Loetmichel: homeoffice400/M,Grandmaschildrenscare 200/M,Grandpas sozialcare100/M and all money handreturns O.O
[19:12:33] <ssi> I'm about to start messing with pyvcp
[19:13:12] <IchGuckLive> bills as you selfproduce with a simple payment devels circle ;-)
[19:13:50] <IchGuckLive> actualy cost §eurs bank fee per month
[19:14:05] <IchGuckLive> and a big xmas present B)
[19:14:26] <Loetmichel> djdelorie: at the moment i am contemplating if i should mill the lower (copper) power rail again, with the pads rotatet by 10 degrees... on this two rings will flow about 150A peak... -> http://www.cyrom.org/palbum/main.php?g2_itemId=12940
[19:14:33] <IchGuckLive> im off by resttime GN8
[19:14:50] -!- IchGuckLive has quit [Quit: ChatZilla 0.9.87 [Firefox 10.0.2/20120216080748]]
[19:15:35] <ssi> Loetmichel: what are you working on?
[19:15:44] <ssi> looks like a multirotor copter
[19:15:59] <Loetmichel> it is
[19:16:05] <ssi> hehe nice
[19:16:08] <ssi> I built a couple before
[19:16:14] <ssi> but I had problems with them
[19:16:16] <djdelorie> I read an article about making electric motors for in-hub car motors, that used pie-shaped copper plates as the "electromagnets". One "turn", lots of amps.
[19:16:22] <Loetmichel> friend earthed his Octocopter. from above 100 meters
[19:16:23] <ssi> motors and props were too big; too much rotor inertia
[19:17:03] <Loetmichel> had send me the crashed parts, and i managed to salvage 6 motors and ESCs and the Flightcontrolboard.
[19:17:16] <Loetmichel> so i build him a new hexa with the salvaged parts
[19:17:29] * djdelorie is still working towards "first run" on his cnc table.
[19:17:30] <ssi> I want to use the IMU that I was using for my quadcopter and build an autopilot for a collective pitch gas heli
[19:17:45] <ssi> I was thinking about mounting a camera on it and doing auto aerial photography
[19:17:52] <djdelorie> a bad solder joint on a current sense resistor fried part of the current feedback sensing circuitry
[19:18:13] <Loetmichel> IMU?
[19:18:18] <ssi> inertial measurement unit
[19:18:25] <ssi> I have a sparkfun 9 axis IMU
[19:18:25] <Loetmichel> ah
[19:18:30] <ssi> gyro/accel/magnetometer
[19:18:32] <Loetmichel> the "flight control"
[19:18:34] <ssi> yea
[19:18:43] <ssi> well it's not a complete flight control in itself
[19:18:47] <ssi> it's just the measurement
[19:18:53] <ssi> the other thing I want to do is build a complete AHARS
[19:19:09] <ssi> or ADAHRS rather
[19:19:15] * Loetmichel has only parts from www.mikrokopter.de
[19:19:37] <ssi> I'd like to get into designing my own avionics
[19:19:41] <ssi> for fullscale planes
[19:19:55] <mhaberler> djdelorie: are you the one of djgpp fame?
[19:20:06] <Loetmichel> btw: momantary work done on hte big hexa: http://www.cyrom.org/palbum/main.php?g2_itemId=11789
[19:20:09] <djdelorie> that's me
[19:20:18] <mhaberler> whoa.. hats off!
[19:20:25] <djdelorie> :-)
[19:20:29] <Loetmichel> harhar
[19:20:32] <ssi> looks good :)
[19:20:32] <Loetmichel> ;-)
[19:20:44] <ssi> what does that read, 762g?
[19:21:05] <ssi> what's the motor power you're using?+
[19:21:06] <Loetmichel> the world is a village like we say in germany ;-)
[19:21:09] <mhaberler> just curious - what you'd be using linuxcnc for?
[19:21:17] <Loetmichel> ssi: yes, it does
[19:21:37] <Loetmichel> will be aabout 1200 grams TOW
[19:21:42] <Loetmichel> without payload
[19:22:10] -!- pcw has quit [Ping timeout: 260 seconds]
[19:22:22] -!- pcw_ has quit [Ping timeout: 252 seconds]
[19:23:20] <Loetmichel> and about 2.0 kW motor power... cvombined thust somewhrere in the range of 5600 grams ;-)
[19:23:56] <ssi> ~330W each motor?
[19:24:06] <mhaberler> djdelorie: you're entitled to a free bugfix or two by an old user (especially if I created the bugs in the first place, which is a great opportunity to excel ;)
[19:24:06] <ssi> the first one I built had 4x 600W motors, and those were WAY too big
[19:24:17] <ssi> the second one had 4x 180W, and still didn't work well
[19:24:25] <ssi> although I was using heavy rigid airplane propellors for rotors
[19:24:31] <ssi> and I think they just had too much inertia
[19:24:35] <ssi> never could tune the PID loops stable
[19:24:42] <Loetmichel> ssi correct
[19:24:49] <Loetmichel> 350W
[19:24:50] <Fox_Muldr> when i want to change from linuxcnc 2.4.x to 2.5 development i just have to change the repository in ubuuntu right?
[19:25:13] <ssi> Loetmichel: here's the first one:
[19:25:14] <ssi> Loetmichel: http://www.youtube.com/watch?v=zV5HZIDCn58
[19:25:44] <ssi> damn thing was terrifying
[19:25:51] <Loetmichel> harhar
[19:26:11] <ssi> it's probably on the order of 600mm across
[19:26:36] <ssi> the second: http://www.youtube.com/watch?v=4eROXtMq4Q4
[19:26:40] <ssi> obviously not finished
[19:27:04] <Loetmichel> my own actual quad: (onboard:) http://www.cyrom.org/MC/liwi230510.avi
[19:28:22] <Loetmichel> outside: http://www.cyrom.org/MC/050709/testfllug_vado2.avi
[19:28:39] <Loetmichel> (same onboard)
[19:28:42] -!- psha has quit [Quit: Lost terminal]
[19:28:44] <ssi> I'm not getting any video on those; just sound :?
[19:29:19] <Loetmichel> hmmm, wrong codec?
[19:29:38] <Loetmichel> sorry, havent them on youtube or something
[19:31:43] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=12171 <- last meeting of Kopterfreaks in germany. mine is the one in front wint the lilac front "girder" style outrigger
[19:32:04] -!- vladimirek has quit [Remote host closed the connection]
[19:32:17] <ssi> got it
[19:32:33] <ssi> haha I like the one next to it with the huge tall gear
[19:32:35] -!- vladimirek [[email protected]] has joined #linuxcnc
[19:33:17] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=12691 <. had milled more than i have flown anyway on that meeting
[19:33:30] <Loetmichel> (much crashes and broken parts to fix ;-)
[19:34:43] <Loetmichel> i have a gear in a similar height for mine. too, its necessary if you have a camera on a servo underneath ;-)
[19:36:35] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8259
[19:37:02] -!- JT-Shop [[email protected]] has joined #linuxcnc
[19:37:20] <Loetmichel> so, now i have to go to bed, wife is calling ;-)
[19:37:47] -!- JT-Shop has quit [Read error: Connection reset by peer]
[19:38:10] <jdhnc> guess we know who's in charge there!
[19:38:14] -!- JT-Shop [[email protected]] has joined #linuxcnc
[19:38:28] <Loetmichel> jdhnc: not narried?
[19:38:44] <Loetmichel> ' cause you sate the obvious ;-)
[19:38:46] <Loetmichel> stat
[19:38:47] <Loetmichel> e
[19:39:23] <jdhnc> 21 years
[19:39:30] <jdhnc> yes, I am aware :)
[19:44:17] * JT-Shop is a bit puzzled by the volts display on my plasma... mechanical setting up for a cut and the volts read out is almost double now after swapping from the 5i20 to the 5i25
[19:44:31] <djdelorie> free volts!
[19:44:32] * JT-Shop wonders what I might have done wrong?
[19:51:00] <mrsun> hmm, where to find the power factor for different materials for milling? :/
[19:51:09] <mrsun> pererably in some metric language
[19:51:36] <ssi> JT-Shop: what are you using to get tip voltage into your 5i?
[19:53:39] <JT-Shop> THC-AD, the only thing I changed AFAIK is the 5i20 for a 5i25/7i76
[19:54:09] <JT-Shop> getting ready to set up to measure the tip voltage at the plasma and work my way out from there
[19:54:49] <ssi> my table is running mach, sadly
[19:54:59] <ssi> I have an LC-THC from candcnc
[19:55:02] <ssi> and a G540
[19:55:02] <JT-Shop> now metric has its own language :?
[19:55:19] <ssi> built that machine before I got onboard the emc train :)
[19:55:37] <JT-Shop> it worked fine for years on the old setup so I can only assume I've don't something stupid
[19:55:54] <JT-Shop> you could always upgrade it
[19:55:58] <ssi> I plan to
[19:56:01] <ssi> but it's not high on the list
[19:56:02] <JT-Shop> for free :)
[19:56:15] <JT-Shop> I know that, any progress on the HNC?
[19:56:34] <ssi> well pcw is hopefully working on fixing the resolver firmware
[19:56:35] -!- syyl_ws [[email protected]] has joined #linuxcnc
[19:56:44] <ssi> plus I think we discovered that the hnc resolvers are high-output
[19:56:47] <ssi> maybe 1:3 or 1:4
[19:57:15] <ssi> I wired my first control pin last night, machine.is-on is wired to the main relay that switches power to all the other relays
[19:57:42] <ssi> I want to discuss how the turret works before I get to crazy in rewiring the interface box
[19:58:00] <ssi> the way it's set up now, there's five logic boards in the interface box on the lathe, and I'm pretty sure they handle all the decoding to run the turret
[19:58:05] -!- PCW [[email protected]] has joined #linuxcnc
[19:58:10] <ssi> I think I might like to run the turret more directly, but I don't fully understand how it works
[19:58:23] -!- zlog has quit [Remote host closed the connection]
[19:58:29] -!- Tom_itx has quit []
[19:59:45] <cradek> it has absolute position feedback on 4? signals, in something like binary if I remember right
[19:59:59] <ssi> yea BCD
[20:00:07] <ssi> but how do you actuate the up and rotate?
[20:00:14] <ssi> not sure what signals are involved
[20:00:21] <cradek> it has two bits to drive it: lift-and-spin, and pop-up-the-stop-that-stops-it-in-position
[20:00:34] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #linuxcnc
[20:00:35] <ssi> and I guess it only spins one direction
[20:00:38] <ssi> pneumatically spun?
[20:00:39] -!- zlog [[email protected]] has joined #linuxcnc
[20:00:52] <cradek> so you lift and spin, then when you see the feedback is right, you pop up the stop. then you wait a little bit and lower the turret
[20:00:54] <ssi> I noticed the original control would only ever rotate it CCW
[20:00:56] <cradek> yes, airmotor
[20:01:00] <PCW> ssi: try refetching the rmsvss6_8 bitfile, the old one is brokenated (now I have to find all the pinout files I copy pasted bad into with off by one register addresses)
[20:01:07] <cradek> yes it can only turn one way
[20:01:33] <ssi> cradek: what sort of signals does it take to actuate the pneumatics?
[20:02:03] <cradek> I think the solenoids are low voltage
[20:02:26] <cradek> 12 maybe? you need to check.
[20:02:52] <cradek> oh I forgot. there is a "turret down successfully" prox
[20:02:59] <ssi> PCW: got it, thanks
[20:03:12] <cradek> you must make sure it's fully down before you continue your program for obvious reasons
[20:03:16] <ssi> sure
[20:03:18] * JT-Shop goes to see if I have a pop up stop
[20:03:33] <JT-Shop> ssi: I do the tool change in ladder
[20:03:43] <cradek> JT-Shop: I don't know if it actually pops up, I haven't seen the mechanism
[20:03:44] <ssi> cradek: did you tell me that you're running your proxes off 12v?
[20:03:49] <ssi> JT-Shop: I don't know anything about ladder yet
[20:03:54] <ssi> JT-Shop: but I figured that's how it'd be done
[20:03:59] <cradek> ssi: I dunno for sure
[20:04:06] <cradek> why not try and see if it works
[20:04:12] <ssi> cradek: do you know what sort of signal to expect out of the prox?
[20:04:20] <ssi> sink I'd imagine?
[20:04:24] <jdhnc> most prox switches we use run off 10-30VDC
[20:05:20] <PCW> ssi: good bitfile has md5 csum that starts 550cd
[20:05:22] <PCW> bad bitfile will read some constant velocity but nothing else
[20:05:32] <ssi> PCW: yeah sounds familiar :)
[20:05:37] <ssi> new bitfile is 550cd
[20:05:43] <JT-Shop> cradek: but you fire a solenoid for something when you see the correct tool?
[20:05:46] <ssi> old one started d6eb8
[20:06:05] <PCW> good yep that ones verified bad
[20:06:11] <ssi> cool
[20:06:17] <ssi> maybe I'll have half-ass resolvers by tonight :D
[20:06:33] <ssi> I may try voltage-dividing the drive signals too
[20:07:12] <PCW> we will make a HV in version some time this week
[20:07:25] <ssi> PCW: should I just swap mine out for one of those in that case?
[20:07:30] <ssi> when they're ready
[20:07:46] <JT-Shop> yep there it is CB66 Red Turret Stop
[20:07:59] <PCW> Probably we would patch yours (since its used)
[20:08:03] <ssi> gotcha
[20:08:21] <PCW> but we can send a test one ahead
[20:08:35] <ssi> that'd be great :D
[20:09:45] <PCW> now to grep for baddness in the source files
[20:09:47] <PCW> (and figure out why my EtherHM2 crashes when I ping it with more than 256 -hdrsize bytes)
[20:10:54] <PCW> its an 8 bit processor so its got to be some multi-byte math I botched but I sure cant see it
[20:14:04] <JT-Shop> ssi: I used wsum to decode the turret encoder lines which are 1, 2, 4, and 8
[20:17:22] -!- automata_ [[email protected]] has joined #linuxcnc
[20:17:33] <ssi> I don't really understand why it's four lines
[20:17:40] <ssi> it's an eight position turret!
[20:17:49] <ssi> three bits should be plenty :)
[20:19:02] <JT-Shop> I think they want to see something other than 0 for a position
[20:19:45] <cradek> it goes to 0 (or something) when it's between positions
[20:20:08] <cradek> just spin it and watch halscope and you'll see
[20:20:34] <JT-Shop> anyway you could cut and paste from my config the things you need
[20:20:44] -!- automata [[email protected]] has joined #linuxcnc
[20:20:59] <cradek> yes - and then it stops in position ready to be lowered onto whatever locks it
[20:21:14] -!- automata__ has quit [Ping timeout: 272 seconds]
[20:21:23] <cradek> JT-Shop: ^
[20:21:29] <JT-Shop> I just checked and I"m using the lock too, I just forgot I did
[20:22:57] -!- automata_ has quit [Ping timeout: 244 seconds]
[20:23:04] <ssi> cradek: you know the logic boards in the interface box on the lathe itself?
[20:23:09] <ssi> cradek: are you 100% not using them for anything?
[20:23:47] <cradek> those ran everything that was operated by M, S, T codes. I definitely removed them.
[20:23:55] <ssi> ok good
[20:23:56] <ssi> in that case
[20:24:01] <ssi> I'm going to completely rearrange that section of the box
[20:24:18] <ssi> the way it's set up now, those five terminal strips are wired to the boards via that PCB backplane
[20:24:34] <ssi> I'm going to remove the boards and then use those terminal strips just for termination
[20:24:46] <ssi> it'll mean rearranging all the wires coming into those strips
[20:24:50] <cradek> I was tempted to do that too, but didn't
[20:25:08] <ssi> cradek: if you get a chance, I'd love to see a photo of what you did in that box
[20:25:10] <cradek> probably because the pcb itself connected a bunch of stuff I didn't want
[20:25:27] <ssi> well if you remove the logic boards, I sorta doubt anything's connected to anything else
[20:25:36] <ssi> you might have a stray trace that's going up to the cardedge connectors
[20:25:37] <ssi> but that's about it
[20:25:57] <cradek> heh I don't believe that for a second :-)
[20:26:08] <ssi> heheh
[20:26:49] <ssi> I spose I could rip that whole deal out and replace it with some two-sided terminal strips
[20:26:52] <ssi> probably a saner idea
[20:27:03] -!- vladimirek has quit [Remote host closed the connection]
[20:27:29] <cradek> iirc, I wanted that area for mounting the pc, pc power supply, and mesa cards
[20:27:37] <ssi> ah
[20:27:41] <cradek> I tapped lots of 4-40 holes, I know that
[20:27:46] <ssi> I mounted that stuff inside the old control
[20:27:49] <cradek> I hate mechanical stuff
[20:28:08] <ssi> what are you using for servo amps?
[20:28:10] <cradek> heh you're going to have PLENTY of room...
[20:28:14] <ssi> yeah I have plenty :D
[20:28:23] <cradek> the originals of course!
[20:28:33] <ssi> where did you put them?
[20:28:39] -!- A2Sheds has quit [Quit: puff of smoke]
[20:28:42] <cradek> in a smaller box
[20:28:45] <ssi> :D
[20:28:48] <ssi> original power supply too?
[20:28:51] <ssi> 90V?
[20:29:04] <cradek> the box was a "little" phone system
[20:29:11] <cradek> yeah it's in the external box
[20:31:17] <ssi> I'm using the original box and the original block plugs
[20:31:26] <ssi> I made the mistake of taking one of those plugs apart last night
[20:31:31] <ssi> they fall apart like legos
[20:31:33] <ssi> it was a nightmare
[20:32:14] <JT-Shop> http://i47.photobucket.com/albums/f163/johnplctech/Hardinge/interface02.jpg
[20:32:23] <JT-Shop> http://i47.photobucket.com/albums/f163/johnplctech/Hardinge/interface03.jpg
[20:32:26] <cradek> heh yeah I remember those plugs!
[20:32:59] <ssi> JT-Shop: rotary phase converter in the bottom?
[20:33:14] <ssi> ooh maybe that's what I need to do.... DIN rail and some din mountable terminal blocks
[20:33:27] <JT-Shop> yea
[20:33:36] <ssi> know of a good source for those?
[20:33:55] <JT-Shop> I use AutomationDirect the push in kind
[20:34:22] <ssi> do they make ones for tiny wire
[20:34:27] <ssi> like 20-26 ga?
[20:34:49] <JT-Shop> the ones I used work up to 12 ga
[20:34:57] <JT-Shop> small works fine
[20:35:41] <ssi> http://www.automationdirect.com/adc/Overview/Catalog/Terminal_Blocks/Screwless_DIN-Rail_Terminal_Blocks
[20:35:44] <ssi> those?
[20:36:06] <JT-Shop> loading the page ...............
[20:36:13] <ssi> oh right, dialupboy :D
[20:36:48] <jdhnc> I hate screwless terminal blocks
[20:37:20] <djdelorie> really? I've been swapping boards so much, I'm starting to hate the screwmore terminal blocks...
[20:37:43] <ssi> heheheh
[20:37:52] <mrsun> http://www.youtube.com/watch?v=EIikH-h5LFI me like, i wonder if it does metric :P
[20:38:45] <JT-Shop> ssi: yea, that's the ones I used
[20:38:46] * djdelorie wonders if it can convert RPM to metric...
[20:38:58] -!- automata_ [[email protected]] has joined #linuxcnc
[20:40:26] <ssi> single level ones are $24 for 50
[20:40:27] <ssi> that's not bad
[20:42:08] -!- cevad has quit [Ping timeout: 260 seconds]
[20:42:11] -!- automata has quit [Ping timeout: 244 seconds]
[20:42:28] -!- automata__ [[email protected]] has joined #linuxcnc
[20:43:49] -!- cevad [[email protected]] has joined #linuxcnc
[20:44:21] <ssi> whoa automation direct is local to me
[20:45:09] <jdhnc> that's either good, or really bad.
[20:45:16] <ssi> haha yes
[20:45:36] <ssi> well it means that I could order the shit for pickup tonight if I wanted
[20:45:48] -!- automata_ has quit [Ping timeout: 244 seconds]
[20:45:50] <ssi> but I dont really want to drive to cumming tonight, so I guess I'll have it shipped :P
[20:46:47] <ssi> it also means I have to pay sales tax
[20:46:47] <ssi> heh
[20:48:28] -!- automata [[email protected]] has joined #linuxcnc
[20:49:43] <JT-Shop> well that test proved that the THC-AD and my config is correct it showed 158 volts at the tip voltage terminals
[20:50:34] -!- automata__ has quit [Ping timeout: 245 seconds]
[20:51:14] -!- automata_ [[email protected]] has joined #linuxcnc
[20:52:01] <djdelorie> JT-Shop - just right for driving my motors :-)
[20:52:18] <ssi> you wouldn't want to drive your motors with what comes out of the tip of that torch ;)
[20:52:44] -!- automata__ [[email protected]] has joined #linuxcnc
[20:53:19] <djdelorie> nope!
[20:54:35] -!- automata has quit [Ping timeout: 244 seconds]
[20:55:49] -!- automata_ has quit [Ping timeout: 248 seconds]
[20:56:08] -!- automata_ [[email protected]] has joined #linuxcnc
[20:59:07] -!- automata__ has quit [Ping timeout: 246 seconds]
[20:59:31] -!- syyl_ws has quit [Quit: Verlassend]
[21:02:06] -!- automata_ has quit [Ping timeout: 252 seconds]
[21:05:30] -!- FinboySlick has quit [Quit: Leaving.]
[21:07:47] -!- isssy has quit [Quit: Bye Bye]
[21:14:44] -!- tom3p has quit [Quit: Ex-Chat]
[21:16:02] -!- mhaberler has quit [Quit: mhaberler]
[21:16:48] <JT-Shop> seems the older you get the more bread crumb trails you have to leave about
[21:18:27] <PCW> tell me about it... ... ... ...
[21:20:12] -!- GeorgeH [[email protected]] has joined #linuxcnc
[21:29:38] <JT-Shop> well it helps if I put the proper tip in the plasma torch... on with the testing
[21:30:11] * JT-Shop wonders how/what to limit with the thc comp...
[21:31:23] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[21:33:37] -!- mhaberler [[email protected]] has joined #linuxcnc
[21:35:26] -!- GeorgeH has quit [Read error: Connection reset by peer]
[21:37:03] <JT-Shop> PCW: the p gain seems to work :)
[21:37:19] * JT-Shop works on some kind of limits
[21:38:17] <PCW> well its always better to slow down when you get to where your going rather than always running at the same speed...
[21:40:34] <JT-Shop> thanks for the help on that
[21:40:52] -!- A2Sheds has quit [Quit: puff of smoke]
[21:41:15] <PCW> since its a pure velocity mode loop, a D term is not of any use
[21:41:16] <PCW> but yes next steps are to bound the error (or max velocity)
[21:41:18] <PCW> then a limit on acceleration (like limit2?)would be nice
[21:42:44] <PCW> (or maybe the stepgen can do that part)
[21:43:00] -!- phantoxe [[email protected]] has joined #linuxcnc
[21:43:11] <JT-Shop> it might do that
[21:44:28] <PCW> It likely you can set the gain (P term) higher if you bound the error (limit the proportional l band)
[21:46:56] <PCW> Having a proportional feedback is the advantage of an actual A-D measurement rather than just a hi/lo bit from a comparator
[21:47:21] -!- micges [[email protected]] has joined #linuxcnc
[21:47:34] <JT-Shop> stepgen has a maxaccel pin
[21:47:46] -!- micges [[email protected]] has parted #linuxcnc
[21:48:19] <PCW> yeah so that should work
[21:58:11] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #linuxcnc
[22:12:19] <DJ9DJ> gn8
[22:13:01] -!- DJ9DJ has quit [Quit: bye]
[22:22:04] -!- Valen [Valen!~Valen@2001:44b8:3199:9100:21e:8cff:fe63:55df] has joined #linuxcnc
[22:34:42] -!- raynerd [raynerd!~craynerd@host86-148-223-114.range86-148.btcentralplus.com] has joined #linuxcnc
[22:35:36] <raynerd> guys, what cutting tool should I be using for an mdf door sign. The cutters I have are too think or ripping the edge up rather than giving me a clean cut
[22:35:49] <raynerd> think = thick
[22:42:14] <cpresser> raynerd: you need other MDF :)
[22:42:31] <cpresser> there is MDF with 'deep milling capabilities' (really expensive)
[22:43:14] <cpresser> it would be easier to use another wood (a solid one, slow growing), they are much easier to mill
[22:51:08] -!- erasmo has quit [Remote host closed the connection]
[22:52:37] <kb8wmc> http://www.ebay.com/itm/ARM-Dso201-Portable-Pocket-sized-Mini-Nano-Handheld-Digital-Storage-Oscilloscope-/220976820974?pt=LH_DefaultDomain_0&hash=item33733eb2ee
[22:53:24] -!- Fox_Muldr has quit [Ping timeout: 265 seconds]
[22:54:03] <jdhnc> kb8: there is a 4 channel version also
[22:54:21] -!- skunkworks has quit []
[22:55:07] <kb8wmc> jdhnc: very good....my how technology has changed my perspective
[22:55:34] -!- Fox_Muldr [[email protected]] has joined #linuxcnc
[22:56:36] <jdhnc> 2 of the quad one are digital only, they would be nice for watching timing
[22:56:49] <jdhnc> but, $200ish is a little much for that.
[22:57:03] <kb8wmc> ouch!.....yes sir
[23:02:05] <kb8wmc> jdhnc: just think back to not actually that many years ago...some laboratory with a real oscilloscope...
[23:03:01] <jdhnc> we still have a bunch of crate based scopes with dinky little screens, even some round ones
[23:03:32] * djdelorie discovered that the mcu *on the controller* makes a better mixed-signal scope than the one on his desk...
[23:03:36] <Jymmm> jdhnc: and cloth wiring too!!!
[23:03:40] <kb8wmc> wow, brings back some memories huh?
[23:03:49] <djdelorie> "while you're gathering all that data, could you send me a copy?"
[23:03:52] <jdhnc> I had a Tek 465b that was pretty nice, but too big to keep out.
[23:04:28] <jdhnc> I ebayed it and bought a cheap chinese color scope that does everything I need (which isn't much)
[23:04:40] <kb8wmc> and today?....shirt pocket size
[23:05:01] -!- syyl_ has quit [Quit: Leaving]
[23:05:07] <Jymmm> and which will survive an EMP? LOL
[23:05:23] <kb8wmc> ah, excellent point
[23:06:02] <kb8wmc> well, if you listen to msm, that can never happen
[23:06:16] <Jymmm> MSM?
[23:06:35] <jdhnc> Jymm: media, it's what tea partiers call popular news
[23:07:17] <kb8wmc> yep, main stream media....lol
[23:07:38] -!- joe9 has quit [Quit: leaving]
[23:08:41] <kb8wmc> msm, honest, concise, a real credit to their profession....
[23:08:43] <Jymmm> I don't listen to MSM, I listen to the Feds
[23:09:07] <Jymmm> When the guberment warns of such things, I tend to listen a bit =)
[23:10:17] <Jymmm> http://cryptome.org/emp.htm
[23:10:33] <jdhnc> the feds are always gaming something, depending on who's controlling the message
[23:11:02] <kb8wmc> anyone and everyone even remotely associated with giver-ment is on my pay-no-mind list...they will all lie at the drop of a hat....factually, they would lie even if the truth was more believable
[23:11:56] <kb8wmc> none of them were hired because of their charm, wit, intelligence, or honesty
[23:14:14] <CareBear\> kb8wmc : The nano DSOs are fun, but you obviously get the analog frontend that you're paying for. :)
[23:15:15] <kb8wmc> okay....
[23:20:38] -!- bedah has quit [Quit: Ex-Chat]
[23:22:17] -!- elmo40 [elmo40!~Elmo40@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has joined #linuxcnc
[23:29:16] <cpresser> CareBear\: did you ever use such a nano-DSO?
[23:29:22] <alex4nder> kb8wmc: there's irony in a ham radio operator saying anything about the feds
[23:31:08] <kb8wmc> yes, your point is well taken, most of the HAM's I have every met, are either borderline communists, outright socialists, definitely facist, statist minded and cowards to the core...
[23:33:03] <alex4nder> and they spend their days talking about their hemorrhoids on teh 2 meter repeaters
[23:33:11] <kb8wmc> lol
[23:33:19] <kb8wmc> yeah, that may be so
[23:34:13] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[23:36:18] <kb8wmc> alex4nder: here is a link to kind of a cute cartoon for you....et al, of course.... http://www.youtube.com/watch?v=EUS1m5MSt9k
[23:43:12] -!- micges [[email protected]] has joined #linuxcnc
[23:43:22] -!- micges [[email protected]] has parted #linuxcnc
[23:48:26] -!- raynerd has quit [Ping timeout: 250 seconds]