#emc | Logs for 2011-01-13

[00:00:04] <andypugh> Sorry, was calibrating a touchscreen
[00:00:22] <andypugh> Was that a question or a suggestion?
[00:00:38] <JT-Shop> question
[00:01:23] <JT-Shop> I would assume for a knee type mill you would want to press as near the CG as possible
[00:01:26] <andypugh> Leadscrew, then. I figure that the leadscrew is connected to the bottom of the Knee, so pushing on that is equivalent to pushing on the bottom of the knee
[00:02:06] -!- theorbtwo has quit [Read error: Operation timed out]
[00:02:06] <andypugh> I started off thinking I could make a hole for a push-rod, then it occurred to me that that was silly when there already was one.
[00:03:28] -!- theorbtwo [theorbtwo!~theorb@] has joined #emc
[00:04:04] <andypugh> The bottom of the leadscrew rotates in manual mode (I want to keep that option) but then I probably wouldn't be using the counterbalance air in manual mode.
[00:04:26] <JT-Shop> so you are putting the pneumatic cylinder next to the jacking screw for the knee?
[00:04:40] <andypugh> Under it, was the idea.
[00:05:14] <andypugh> And not so much a pneumatic cylinder, as a tube with a piston in it.
[00:05:15] <JT-Shop> so you have some other means to move it with a ball screw or something?
[00:05:49] <andypugh> It's to counterbalance the weight, so that the motor isn't always taking enough current to hold a 200kg weight.
[00:06:49] <andypugh> It's a bit "lateral" I admit.
[00:07:29] <JT-Shop> I'm just trying to get a mental picture using my BridgePort Series 1 as the base
[00:07:41] <Valen> we found it important to grab the head on ours around its CG
[00:07:47] <Valen> which was well out from the back
[00:08:01] <Valen> stopped it from binding on the slides
[00:08:07] <andypugh> I imagine that is where the existing jackscrew is mounted.
[00:09:37] <andypugh> http://www.lathes.co.uk/harrisonmiller/
[00:09:42] <andypugh> Is the device.
[00:09:54] <andypugh> The Z screw winds down into the base.
[00:10:14] <Valen> looks like it'd be close but the CG will move depending on where the table is
[00:10:28] <Valen> my gut feeling would be a little inboard of the screw
[00:10:38] <andypugh> True, but the Z slides are nice and long
[00:11:09] <JT-Shop> I would put it inboard of the Z screw as well
[00:11:21] <andypugh> Put what inboard of the Z-screw?
[00:11:35] <JT-Shop> the tube to assist in lifting
[00:11:58] <andypugh> I was going to put that under the Z screw, in the base, pushing the Z screw up.
[00:12:09] <andypugh> :-)
[00:12:32] <JT-Shop> the z nut rotates and the screw is stationary!
[00:12:33] <Valen> might work
[00:12:38] <Valen> keep in mind the rotation
[00:12:40] <JT-Shop> I understand now :)
[00:12:45] <Valen> ahh duh
[00:12:46] <Valen> nvm
[00:13:00] <Valen> actually that *might* not be a good idea
[00:13:11] <Valen> I can see it possibly adding some bend to the screw
[00:13:20] <andypugh> Yeah. The "manual" mode is rotating screw, fixed nut. I am adding a CNC-rotated nut.
[00:13:37] <Valen> if you put some mojo on your piston to keep it straight perhaps
[00:13:44] <Valen> where you getting the spinning ballnut from?
[00:14:19] <andypugh> http://picasaweb.google.com/bodgesoc/Gibbs#5559905744520719218
[00:14:40] <andypugh> Though currently that is a spinning leadscrew nut.
[00:15:18] <andypugh> (All the iron parts are static, the only moving part you can see there is the pulley and belt)
[00:16:02] <andypugh> Hmm, though that motor is a bit in the way of my proposed cylinder.
[00:16:55] * JT-Shop heads in to cut the veggies for some Sausage Surprise for dinner
[00:18:35] <andypugh> I need to get the motor spinning properly (still not happy with the Resolver convertor) then I can see if the motor can drive the leadscrew, and how happy it is doing so. If that looks bad then I will convert to ballscrew, but then I think I will need a counterbalance, as otherwise the motor will always be working. (I have a feeling that the stiffness of the Acme will actually help the motor steady-state)
[00:19:42] <Valen> andypugh: a 400W scooter motor moved our 100kg Z axis without too much of an issue, direct drive
[00:20:29] <mendelbuild> hi
[00:22:29] <andypugh> AHoy
[00:22:48] <andypugh> Is your machine running with EMC2 now then?
[00:25:23] -!- Connor [Connor!~Connor@] has parted #emc
[00:28:27] -!- Connor [Connor!~Connor@] has joined #emc
[00:29:13] -!- davidf [davidf!~dave@173-26-244-213.client.mchsi.com] has joined #emc
[00:30:45] <davidf> Hi
[00:31:45] <davidf> Anybody have a suggestion for adding linear position feedback to my 7 X 12 mini lathe x-axis?
[00:32:18] <andypugh> Is it CNC?
[00:32:27] <skunkworks> davidf: first question is ' what are you trying to do'
[00:32:33] <skunkworks> ok 2nd question
[00:32:54] <davidf> yes, all set up but I'm doing something that needs like .0005 accuracy and repeatability.
[00:33:10] <andypugh> inch or mm?
[00:33:13] <andypugh> :-)
[00:33:27] <davidf> I/m making pin vises esentially.
[00:33:38] <davidf> inch
[00:33:56] <davidf> .0005 mm would be nice though :)
[00:34:06] <andypugh> Stepper?
[00:34:09] <davidf> yes
[00:34:12] <davidf> open loop
[00:34:37] <andypugh> The accuracy and repeatability _should_ be the same open-loop as closed-loop
[00:35:00] <davidf> I can make about two parts then I have to spend forever tweaking the tool offsets etc
[00:35:07] <andypugh> (And there is not a lot you can really do with your system if you do have feedback)
[00:35:34] <davidf> you mean if I do not have feedback?
[00:35:56] <andypugh> No, I mean, what are you going to do with the feedback?
[00:36:37] <andypugh> If something is moving between parts so that you lose size control, then it will do that regardless.
[00:36:41] <davidf> I had hoped that the emc would do whatever was necessary to put my cutter exactly where it is supposed to be.
[00:37:18] <davidf> I thought I'd put the sensor directly on the tool.
[00:37:46] <davidf> so if the carriage moved, it wouldn't matter
[00:37:49] <andypugh> People have tried steppers with encoders without huge success.
[00:38:09] <andypugh> Though what you are trying to do is a little different.
[00:38:17] <elmo40> other then a bulb burning out, what else could happen to an encoder such as this? http://www.atechauthority.com/products.asp?id=174
[00:38:31] <davidf> my machine just isnn't massive and tight enough to sit still for long.
[00:38:38] <andypugh> The fundamental problem is that the 7x12 lathes are rubbish.
[00:39:03] <davidf> yes but cheap rubbish.
[00:39:12] <davidf> all I can afford atm
[00:39:15] <andypugh> Are you still using the leadscrew, or a ballscrew?
[00:39:22] <elmo40> reason why I ask is the machine at work has this on its spindle and it is acting up (not performing spindle orientation and crashing the tool changer) and I asked for the old one. you think I could repair it? ;)
[00:39:38] <davidf> original screws and the nuts are tempermental.
[00:40:45] <andypugh> elmo40: Nothing to lose by trying
[00:40:53] <davidf> is it possible to read a serial port during program execution?
[00:40:56] <andypugh> davidf: I have a similar lathe to yours.
[00:41:07] <andypugh> I wasn't being rude.
[00:41:16] -!- robh__ has quit [Ping timeout: 250 seconds]
[00:41:21] <davidf> I know. No offense taken. :)
[00:42:32] <andypugh> I put a ballscrew on the X, but I am not sure that it is a huge improvement. But then the original leadscrew is tiny, and in the case of mine the nut was held in with a cross-head screw, with another to tilt the nut to takeup backlash (horrible)
[00:42:52] <davidf> reason I ask is if nothingf else I could maybe take a reading of position from a digital caliper once in awhile before a critica cut
[00:43:19] <davidf> that's how mine is done.
[00:43:43] <andypugh> If you can face it, you could pause the program every time before the finish cut, touch-off to the diameter, and then take the finish cut.
[00:43:47] <davidf> I thing I have as much of a problem from the carriage play anyway.
[00:43:56] <andypugh> Yes, mine rocks.
[00:44:07] <andypugh> Less than it did, though.
[00:44:45] <andypugh> I scraped it in again with blue, as the original builder didn't grasp that you were meant to fit both ways at the same time
[00:45:02] <andypugh> So both sides had a good contact, but not at the same time.
[00:45:07] <davidf> I adjusted it as tight as I can, even scraped the ways a bit. still maybe a few thou play though depending on how hard the metal I'm cutting is.
[00:46:10] <andypugh> I modified the Gibbs too, the original iron plates were nasty.
[00:46:12] <andypugh> http://picasaweb.google.com/bodgesoc/Gibbs#5403345449585651218
[00:46:32] <andypugh> (click back to the beginning for the manufacturing sequence)
[00:46:35] <davidf> andypugh what did you mean if I can face it? You mean if I can stand to do it or if I can face off a part?
[00:46:59] <andypugh> If you can stand to do it with every single part.
[00:47:25] <davidf> yes. but this is production work and it is slooooow
[00:47:35] <Valen> off to work I go
[00:47:38] <Valen> catchyas
[00:47:50] -!- Valen has quit [Quit: Leaving.]
[00:47:54] <andypugh> Production work on a 7x12? Oh My!
[00:48:13] <davidf> that is what I've been doing but even that doesn't produce perfect results every time.
[00:48:23] <davidf> yeah tell me about it.
[00:48:39] <andypugh> What sort of qantity?
[00:48:51] <andypugh> And what are you making?
[00:49:05] <davidf> it is ok for some stuff but this is pushing it past its capability im afraid.
[00:49:15] <davidf> hundreds hopefully
[00:49:49] <davidf> cant say exactly yet. going to apply for a provisional patent in a few days then I can say
[00:49:59] <andypugh> http://en.wikipedia.org/wiki/Tool_bit#Box_tool
[00:53:10] <davidf> box tool huh?
[00:53:36] <andypugh> Might be more fun make that than fiddle with the lathe.
[00:53:56] <andypugh> You could look at this page: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Combining_Two_Feedback_Devices_On_One_Axis
[00:53:56] <davidf> How does it keep pressure on the back of the stock as the diameter decreases?
[00:54:27] <andypugh> It's a single-pass tool. But you could use it for final sizing
[00:54:39] <davidf> I see.
[00:55:42] <davidf> the most critical part of this is threaded. So I need an accurate sizing pass followed by an accurate threading cycle.
[00:56:24] <davidf> And to make things a little bit more painful, it's stainless steel.
[00:56:44] <elmo40> >_<
[00:58:03] <andypugh> You ought to be able to rig up one of these: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Combining_Two_Feedback_Devices_On_One_Axis
[00:58:28] <andypugh> One with a serial link might be better, though I am not sure how easy the serial data is to read.
[01:00:19] <davidf> I dont know either. I just know there is a 4-lead serial port on most digital calipers.
[01:00:31] <andypugh> For production threading, you could use http://cgi.ebay.co.uk/3-4-CHS-Coventry-Die-Head-Diehead-Thread-Capstan-Lathe-/320641779987
[01:01:21] <andypugh> Though CNC single-point it probably better, if you can sort out your lathe.
[01:03:25] <davidf> well you know, I had forgotten all about using a die. duh. been so focused on tweaking the g76 cycle etc so long I forgot to even think about that. Could be I could get a die centered and straigt enough to work.
[01:03:50] <andypugh> http://www.shumatech.com/support/chinese_scales.htm
[01:04:41] <andypugh> The Coventry heads use 4 cutters, and you can tweak them to the right tolerance. The also release when they hit the end stop.
[01:04:57] <andypugh> (ie the blades pop out of mesh)
[01:05:46] -!- Dannyboy has quit [Remote host closed the connection]
[01:08:01] <davidf> so it cuts from all directions like a die but the cutters move like a single point tool?
[01:09:16] -!- Connor has quit [Quit: Leaving.]
[01:09:22] <andypugh> It is exactly like a die, but with the ability to change diameter by enough to disengage
[01:14:40] <davidf> I need to thread about 0.4 inches and then there is a shoulder .05 inch from that so a regular die wouldn't work but maybe I could cut cose first and finish with a chasing die?
[01:14:56] <andypugh> Anyway, in theory you could use a secondary scale and a slow-moving PID to put an offset on the tool position. However I think you would be better trying to work out what is making the lathe lose position.
[01:15:18] <davidf> also it is a non-standard size. 7 mm by 40 tpi
[01:15:42] <andypugh> You probably can't get a die head that small
[01:15:56] <andypugh> And that is a horribly non-standard size.
[01:16:19] <davidf> It's a Dremel shaft.
[01:16:29] <davidf> Cute of them huh?
[01:16:46] <andypugh> Hmm.
[01:17:56] <davidf> it is very close to .65 mm pitch or something like that that comes out to 40.64 tpi or so but it didn't seem to fit right until I used 40 exactly
[01:18:33] <andypugh> 9/32" x 40 tpi
[01:20:07] <davidf> well, the dremel shaft major diameter measures .277 inches exactly whatever that is.
[01:20:16] <andypugh> 9/32" is far more likely that M7. I think onlt Bugatti ever used M7.
[01:20:32] -!- emcrules_ has quit [Read error: Connection reset by peer]
[01:20:39] <davidf> whatever it is it is exactly .277 inch.
[01:20:56] <andypugh> .277 is a tad over both M7 and 9/32
[01:21:22] <davidf> I wouldnt be surprised if they used something ridiculous to help prevent copying
[01:21:47] <andypugh> Entirely possible. The interweb says it is 2.77
[01:22:17] <davidf> what's the interweb?
[01:22:47] <andypugh> I have an M14.3 x 0.45 thread holding my drawbar together, it is what worked best. You are not governed by engineering convention with a CNC lathe.
[01:22:58] <andypugh> http://www.rcgroups.com/forums/showthread.php?t=515450
[01:23:21] <andypugh> Sorry, interweb is a joke name for the internet.
[01:23:46] <davidf> right unless you want a die. :) I saw one for dremel shafts on ebay once but didn't buy it. Should have.
[01:23:59] <davidf> 'k. :)
[01:25:50] <davidf> anyway .277 and 40 tpi works pretty well - good enough for me. One part I made had zero visible runout at 7 inches out from the chuck.
[01:26:07] <davidf> I was amazed at that.
[01:26:43] <andypugh> Anyway, time to sleep.
[01:26:51] <andypugh> Night all.
[01:26:56] <davidf> But I've spent hours just to make 6 good parts. Not practical this way.
[01:27:03] -!- andypugh has quit [Quit: andypugh]
[01:27:11] <davidf> OK THanks for the help goodnight.
[01:35:06] -!- adb has quit [Ping timeout: 240 seconds]
[01:37:45] -!- emcrules_ [emcrules_!~jp1@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[01:40:09] <kilian-afk> I need a hint in the right direction. I have a dental-mill, and it's set to 125,000 steps-per-inch (which seems like a LOT)
[01:40:25] <kilian-afk> and I'm driving it with an old PC that can't generate lots and lots of steps-per-second,
[01:40:42] <kilian-afk> so I'm looking to build an external stepper generator of some sort.
[01:41:00] <kilian-afk> Is there an example of what direction to go with this?
[01:41:02] <skunkworks> buy one...
[01:41:14] <skunkworks> you can get one for $89ish
[01:41:21] <kilian-afk> Which one is your favorite?
[01:41:26] <skunkworks> mesa
[01:42:26] -!- davidf [davidf!~dave@173-26-244-213.client.mchsi.com] has parted #emc
[01:42:33] <skunkworks> http://mesanet.com/
[01:42:42] <skunkworks> look at the 7i43
[01:43:06] <skunkworks> or pico systems has a external step generator.
[01:46:43] <kilian-afk> I'm seeing a lot of mesa stepper motor drivers, but I have good drivers right now. Are you talking about softdmc on an FPGA card to generate ateps, or am I just not seeing it.
[01:48:13] -!- Poincare has quit [Ping timeout: 276 seconds]
[01:50:45] <skunkworks> yes the fpga
[01:51:13] <skunkworks> emc doesn't use softdmc - it has its own firmware that it installs from mesa
[01:51:50] <skunkworks> emc still does the motion control - the mesa hardware is just step generating and or pwm and encoder counting.
[01:52:17] <skunkworks> and i/o
[01:54:07] -!- Poincare [Poincare!~jefffnode@ip-188-118-11-57.reverse.destiny.be] has joined #emc
[01:56:13] <skunkworks> and or
[01:56:16] <skunkworks> :)
[02:00:08] <kilian-afk> OK, a 7i43 plugged in to a USB port, hostmot2 driver, load up some magic firmware to have 3 stepper generators and some I/O for limit switches, spindle enable and dust collector enable.
[02:00:18] <kilian-afk> Is that all supported?
[02:07:55] -!- theorbtwo has quit [Remote host closed the connection]
[02:08:00] -!- theorbtwo [theorbtwo!~theorb@] has joined #emc
[02:11:06] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[02:13:02] <elmo40> anyone know what kind of card/adaptor I would need to get a rotary encoder such as this working on my spindle? http://www.atechauthority.com/products.asp?id=174
[02:25:26] -!- celeron55 has quit [Ping timeout: 240 seconds]
[02:36:45] -!- emcrules_ has quit [Read error: Connection reset by peer]
[02:52:11] -!- emcrules_ [emcrules_!~jp1@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[02:52:18] -!- nullie [nullie!~nullie@nullie.telenet.ru] has joined #emc
[03:12:56] -!- emcrules_laptop [emcrules_laptop!~emcrules_@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[03:27:40] -!- mendelbuild has quit [Ping timeout: 250 seconds]
[03:36:06] -!- emcrules_laptop has quit [Read error: Connection reset by peer]
[03:36:06] -!- emcrules_ has quit [Read error: Connection reset by peer]
[03:37:17] -!- nullie has quit [Quit: Ex-Chat]
[03:38:03] <pcw_home> If you use low enough line count relative to you base thread rate, you can count it with a parallel port
[03:39:11] <elmo40> that count is measured with the latency program?
[03:39:21] <elmo40> max frequency, or something?
[03:51:39] -!- emcrules_ [emcrules_!~jp1@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[03:53:06] <pcw_home> Dont know exactly but I think the maximum count rate would be one count per sample so the base period gives the maximum rate
[03:53:08] <pcw_home> but: because of quadrature error and jitter you need some margin so maybe 1/2 the base rate
[04:13:35] -!- Connor [Connor!~Connor@] has joined #emc
[04:43:01] -!- nullie [nullie!~nullie@dhcp249-250.yandex.net] has joined #emc
[05:52:03] -!- dgarr has quit [Ping timeout: 260 seconds]
[06:39:12] -!- LawrenceG has quit [Remote host closed the connection]
[06:49:52] -!- psha [psha!~psha@] has joined #emc
[07:19:43] -!- emcrules_ has quit [Read error: Operation timed out]
[07:22:37] -!- emcrules [emcrules!~jp1@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[07:24:37] -!- mhaberler [mhaberler!~chatzilla@host88-93-static.38-79-b.business.telecomitalia.it] has joined #emc
[07:27:25] -!- celeron55 [celeron55!~perttu@a91-155-42-12.elisa-laajakaista.fi] has joined #emc
[07:35:59] -!- JT-Shop has quit [Read error: Connection reset by peer]
[07:36:33] -!- JT-Shop [JT-Shop!~chatzilla@216-41-143-50.semo.net] has joined #emc
[07:37:50] -!- Jymmm has quit [Read error: Operation timed out]
[07:38:41] -!- packrat has quit [Ping timeout: 260 seconds]
[07:39:06] -!- packrat [packrat!~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net] has joined #emc
[07:40:38] -!- Jymmm [Jymmm!~jymmm@unaffiliated/jymmm] has joined #emc
[07:40:47] -!- mhaberler has quit [Ping timeout: 240 seconds]
[07:42:11] -!- encryptio has quit [Ping timeout: 260 seconds]
[07:42:18] -!- encryptio [encryptio!~ckastorff@adsl-074-170-091-222.sip.mem.bellsouth.net] has joined #emc
[08:35:19] -!- e-jones [e-jones!~jkastner@nat/redhat/x-ctekhlzswytqwlwx] has joined #emc
[08:55:25] -!- e-jones has quit [Quit: Ex-Chat]
[09:06:08] -!- elmo40 has quit [Read error: Connection reset by peer]
[09:08:40] -!- elmo40 [elmo40!~Administr@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has joined #emc
[09:09:03] -!- funkyman75 [funkyman75!~chatzilla@adsl-dyn48.78-98-134.t-com.sk] has joined #emc
[09:10:28] <funkyman75> Hi all.Does anyone have an emc2 postprocessor for Edgecam ?
[09:25:45] -!- robh__ [robh__!~robert@5ace70af.bb.sky.com] has joined #emc
[09:26:38] -!- mhaberler [mhaberler!~chatzilla@host88-93-static.38-79-b.business.telecomitalia.it] has joined #emc
[09:46:53] -!- mhaberler has quit [Ping timeout: 255 seconds]
[09:47:41] -!- e-jones [e-jones!~jkastner@nat/redhat/x-ijtchxrfjyjdhdgc] has joined #emc
[09:48:12] -!- e-jones has quit [Read error: Connection reset by peer]
[10:40:12] -!- maximilian_h [maximilian_h!~bonsai@p549DDDE5.dip.t-dialin.net] has joined #emc
[10:40:53] <maximilian_h> Hello Guys
[10:41:20] <maximilian_h> I cleaned up my desk and put my notes about programming a 7i43 with a jtag cable in the wiki
[10:42:07] <maximilian_h> could somebody have a look at it and see if it is more or less coherent, it seems to me, but I might have missed something
[10:42:09] <maximilian_h> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?JTAG_Boundary_Scan_Example_For_A_Mesanet_7i43
[10:52:23] -!- awallin [awallin!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[10:54:16] <elmo40> funkyman75: what kind of Posts do you have for edgecam? I am sure pretty sure the Fanuc post will suffice.
[10:58:29] -!- psha has quit [Quit: Lost terminal]
[10:59:50] -!- maximilian_h has quit [Ping timeout: 240 seconds]
[11:02:21] <funkyman75> ok i'll try them
[11:04:33] -!- maximilian_h [maximilian_h!~bonsai@p549DDDE5.dip.t-dialin.net] has joined #emc
[11:14:36] -!- Valen [Valen!~Valen@] has joined #emc
[11:16:26] -!- funkyman75 has quit [Remote host closed the connection]
[11:29:02] -!- maximilian_h has quit [Ping timeout: 240 seconds]
[11:35:07] -!- maximilian_h [maximilian_h!~bonsai@p549DDDE5.dip.t-dialin.net] has joined #emc
[11:47:09] -!- psha[work] [psha[work]!~psha@qwe.s2s.msu.ru] has joined #emc
[11:56:53] -!- bootnecklad_ [bootnecklad_!~bootneckl@] has joined #emc
[12:00:01] -!- bootnecklad has quit [Ping timeout: 260 seconds]
[12:05:56] -!- logger[psha] has quit [Ping timeout: 250 seconds]
[12:07:12] -!- logger[psha] [logger[psha]!~loggerpsh@] has joined #emc
[12:45:59] -!- kent_cnc [kent_cnc!~cnc@87-194-141-154.bethere.co.uk] has joined #emc
[12:51:35] -!- mk0 [mk0!~x-ray@fiztech.basnet.by] has joined #emc
[12:57:29] -!- Valen has quit [Quit: Leaving.]
[12:58:22] -!- skunkworks has quit [Ping timeout: 240 seconds]
[13:02:40] -!- izua_ [izua_!~izua@unaffiliated/izua] has joined #emc
[13:06:10] -!- izua has quit [Ping timeout: 276 seconds]
[13:27:18] -!- awallin has quit [Remote host closed the connection]
[13:28:42] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[13:33:02] -!- maximilian_h has quit [Ping timeout: 250 seconds]
[13:43:44] -!- skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.] has joined #emc
[13:49:28] -!- dgarr [dgarr!~dgarrett@adsl-75-61-65-2.dsl.pltn13.sbcglobal.net] has joined #emc
[14:09:53] -!- WonTu [WonTu!~WonTu@p57B5411A.dip.t-dialin.net] has joined #emc
[14:10:07] -!- WonTu [WonTu!~WonTu@p57B5411A.dip.t-dialin.net] has parted #emc
[14:13:05] -!- wobblybootie [wobblybootie!5c11d183@gateway/web/freenode/ip.] has joined #emc
[14:33:46] -!- micges_work [micges_work!~x@aejk77.neoplus.adsl.tpnet.pl] has joined #emc
[14:35:53] -!- psha[work] has quit [Quit: Lost terminal]
[14:41:34] -!- micges_work has quit [Ping timeout: 255 seconds]
[14:44:53] -!- micges_work [micges_work!~x@aeji103.neoplus.adsl.tpnet.pl] has joined #emc
[14:56:42] -!- wobblybootie has quit [Quit: Page closed]
[15:16:28] -!- micges_work [micges_work!~x@aeji103.neoplus.adsl.tpnet.pl] has parted #emc
[15:22:31] -!- nullie has quit [Quit: Ex-Chat]
[15:31:38] -!- psha [psha!~psha@] has joined #emc
[15:51:46] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has joined #emc
[15:55:13] -!- mk0 has quit [Read error: Connection reset by peer]
[15:57:26] -!- Dannyboy [Dannyboy!~Dan@] has joined #emc
[15:58:46] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has parted #emc
[16:00:11] -!- grommit [grommit!~chatzilla@] has joined #emc
[16:01:09] <psha> grommit: hi
[16:01:30] <grommit> Hi psha!
[16:01:35] <grommit> How are things going?
[16:02:54] -!- bootnecklad [bootnecklad!~bootneckl@] has joined #emc
[16:03:20] -!- kent_cnc has quit [Quit: Ex-Chat]
[16:03:22] <psha> fine
[16:05:29] -!- bootnecklad_ has quit [Ping timeout: 264 seconds]
[16:06:48] <grommit> what are you working on these days?
[16:08:19] -!- jepler has quit [Quit: .]
[16:08:36] -!- jepler [jepler!~jepler@emc/developer/pdpc.professional.jepler] has joined #emc
[16:09:15] -!- adb has quit [Ping timeout: 246 seconds]
[16:09:44] <grommit> I am trying to set up a gantry machine config and have an error I can parse. Anyone know what this might be? It is likely a basic error as I borrowed from another config. http://pastebin.com/HqEQtRAp
[16:09:54] <grommit> can==can't
[16:12:41] <grommit> nevermind, figured it out, I had to remove NML_FILE=emc.nml from my ini file....
[16:19:10] -!- nullie [nullie!~nullie@nullie.telenet.ru] has joined #emc
[16:19:20] <psha> same
[16:19:23] <psha> gladevcp
[16:19:40] <psha> fighting to get working modular Gtk interface
[16:24:39] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[16:43:04] -!- nullie has quit [Quit: Ex-Chat]
[16:44:46] -!- brujo [brujo!~brujo@] has joined #emc
[16:47:49] <grommit> I have a charge_pump set up but it isn't working. I think the frequency of my base_thread is too fast for it. The board wants 3000-15000Hz. my base thread is 21600. So I am getting a 23kHz charge pump signal. Is there a way to just halve the signal?
[16:48:17] -!- mhaberler [mhaberler!~chatzilla@host88-93-static.38-79-b.business.telecomitalia.it] has joined #emc
[16:50:46] <psha> pwm?
[17:03:18] <cradek> or a very simple comp that has 4 states instead of 2
[17:03:38] -!- Connor has quit [Remote host closed the connection]
[17:04:17] <cradek> or just make another thread that's half as fast as your base thread
[17:04:43] <cradek> or put the charge pump func in the servo thread
[17:05:01] <cradek> er no, that's less than 3000
[17:06:57] -!- grommit has quit [Remote host closed the connection]
[17:21:12] -!- Connor [Connor!~Connor@] has joined #emc
[17:22:16] -!- mhaberler has quit [Ping timeout: 276 seconds]
[17:23:13] -!- odiug [odiug!~guido@p57BD0A5B.dip0.t-ipconnect.de] has joined #emc
[17:31:47] -!- odiug has quit [Ping timeout: 240 seconds]
[17:34:34] -!- Dallur [Dallur!~root@] has joined #emc
[17:50:55] -!- kbarry [kbarry!~chatzilla@173-13-173-249-sfba.hfc.comcastbusiness.net] has joined #emc
[18:17:29] -!- micges [micges!~micges@acti11.neoplus.adsl.tpnet.pl] has joined #emc
[18:43:55] -!- e3m [e3m!~IceChat7@] has joined #emc
[19:10:52] -!- motioncontrol [motioncontrol!~i@host56-90-dynamic.252-95-r.retail.telecomitalia.it] has joined #emc
[19:14:46] -!- Dallur has quit [Quit: Leaving.]
[19:17:02] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc
[19:23:22] <andypugh> Looking at the pluto_servo.comp driver I found an undocumented feature of "comp" called "modparam". Is this an omission or a direction?
[19:26:14] <andypugh> Sorry, wrong channel, I suspect.
[19:26:33] <psha> forget what he've just said!
[19:28:26] -!- grommit [grommit!~chatzilla@adsl-99-103-190-125.dsl.sfldmi.sbcglobal.net] has joined #emc
[19:29:06] <grommit> sorry, machine crashed earlier. Not sure what is going on, but I am here again, until I'm not ;-)
[19:32:21] <psha> grommit: have you read cradek's posts?
[19:32:39] <grommit> So, to pick up where I left off.... How do I create a new thread? I tried entering this in the kinematics.hal file: cp_period_nsec=[EMCMOT]CP_PERIOD but I get an error. This was in dmesg: [15720.581522] motmod: Unknown parameter `cp_period_nsec' Is there a special name I have to use?
[19:32:58] <grommit> Yes, he said I need to create a new thread, right?
[19:33:11] <psha> yes
[19:33:45] <grommit> Do you know the mechanics of creating a thread (that is not base or servo)?
[19:34:09] <psha> not yet ;)
[19:34:22] <grommit> me either ;-)
[19:35:06] <psha> loadrt threads period1=1000000 name1=thread
[19:35:22] <psha> found somewhere in documentation in hal section
[19:35:25] <psha> in hal-examples
[19:35:35] <grommit> cool, i'll take a look....
[19:35:51] <atmega> http://www.linuxcnc.org/docview/html//man/man9/threads.9.html
[19:36:13] <grommit> thx!
[19:47:47] <grommit> no love. I have this in my .hal file: loadrt threads name1=cp-thread period1=125000 but am getting an error: [16448.616028] HAL_LIB: ERROR: new thread period 125000 is less than existing thread period 993600
[19:48:06] <atmega> did you read the man page?
[19:48:07] <grommit> Not sure what thread I have that is 993600, other then the servo thread of 1000000...
[19:48:12] <grommit> yes
[19:48:43] <atmega> Threads must be created in order, from fastest to slowest.
[19:48:50] <atmega> that seems like a likely cause
[19:50:03] <grommit> I am only creating one thread. The base_thread and servo_thread are created by EMCMOT, not this command
[19:56:14] <atmega> out of curiosity, can you try making it reallly slow and see if it works?
[19:57:52] <grommit> I made it 1250000 (extra 0) and it does work so it is order. So the question becomes: given that EMCMOT starts base and servo threads, how do I use it to start another thread?
[19:58:17] <psha> don't start servo thread? :)
[20:00:35] <grommit> hmm, don't i need the servo thread for something?
[20:00:59] <psha> you may run it with loadrt threads later?
[20:01:03] <psha> after you thread
[20:04:33] <atmega> do you use it?
[20:05:26] <grommit> I tried removing the servo thread, and then adding it in with: loadrt threads name1=cp-thread period1=1250000 name2=servo-thread period2=1000000
[20:05:39] <grommit> but it fails and says there is a duplicate servo-thread
[20:05:58] -!- klick0 has quit [Read error: Connection reset by peer]
[20:06:10] <grommit> I removed it from the EMCMOT config... so I don't know where it gets created...
[20:06:45] <grommit> In kinematics.hal I only have: loadrt [EMCMOT]EMCMOT base_period_nsec=[EMCMOT]BASE_PERIOD
[20:07:12] <psha> maybe it's creating it even if you don't give period?
[20:07:13] <grommit> Are you asking if I use the servo thread?
[20:07:25] <grommit> must be...
[20:14:29] <atmega> I was
[20:14:32] -!- acemi [acemi!~acemi@unaffiliated/acemi] has joined #emc
[20:14:54] <atmega> but, I am basically clueless on teh subject so it was just more curiosity.
[20:15:34] <atmega> once again, I tried to order a mill from Grizzly, once again, they refuse to sell me one.
[20:16:00] <grommit> I tried taking the base period out of the EMCMOT part and then using: loadrt threads name1=base-thread period1=21600 name2=cp-thread period2=1250000 name2=servo-thread period2=1000000
[20:16:15] <grommit> but doesn't work, says it can't create base-thread
[20:17:57] <grommit> in fact it complains that base-thread period is less then 999830. So the servo thread is apparently getting created first regardless....
[20:22:33] -!- odiug [odiug!~guido@p57BD0A5B.dip0.t-ipconnect.de] has joined #emc
[20:24:26] -!- vezza [vezza!~andrea__@host247-160-dynamic.7-87-r.retail.telecomitalia.it] has joined #emc
[20:25:00] <grommit> tried doing these in order in kinematics.hal but it doesn'
[20:25:06] <grommit> doesn't work....
[20:25:17] <grommit> loadrt [EMCMOT]EMCMOT base_period_nsec=[EMCMOT]BASE_PERIOD
[20:25:18] <grommit> loadrt threads name1=cp-thread period1=125000
[20:25:20] <grommit> loadrt [EMCMOT]EMCMOT servo_period_nsec=[EMCMOT]SERVO_PERIOD
[20:25:36] <grommit> I also used name2/preiod2, still doesn't work....
[20:25:45] <grommit> period2 that is...
[20:29:32] <atmega> hopefully the real people will show up for you soon :)
[20:31:32] <micges> grommit: hi
[20:31:41] <micges> grommit: what is your problem?
[20:34:05] -!- PCW [PCW!~chatzilla@] has joined #emc
[20:34:39] <grommit> I am trying to create a third thread that is between the speed of the base and servo therad
[20:34:41] <grommit> thread
[20:34:50] <grommit> Can't get it right...
[20:35:33] <PCW> andypugh you there?
[20:36:56] -!- skunkworks has quit [Ping timeout: 265 seconds]
[20:37:21] <micges> you must creating threads in order from fastest to slowest
[20:37:33] <PCW> (oops wrong channel)
[20:38:15] <grommit> micges: I see that, but how?
[20:38:47] <JT-Shop> http://www.linuxcnc.org/docview/html/man/man9/threads.9.html
[20:39:14] <micges> i motion line create base thread and servo-thread (with middle speed)
[20:39:37] <grommit> I can't do it in EMCMOT because it only see servo and base periods, but if I try to use loadrt threads it complains because the new thread is faster then the servo thread
[20:39:41] <micges> and after that loadrt threads with some name and period1=1000000
[20:39:49] <JT-Shop> grommit: are you putting the loadrt in you hal file?
[20:39:57] <micges> and that thread will be your servo-thread but with different name
[20:40:40] <grommit> so my middle thread will be named servo-thread (instead of charge-pump thread) and vis-versa?
[20:41:03] <micges> yes
[20:41:07] <grommit> ugh
[20:41:09] <grommit> ok
[20:41:10] <micges> this is only issue
[20:41:29] <grommit> that won't be confusing at all ;-)
[20:41:42] <micges> hehe
[20:42:10] <micges> you can use 'alias' to change names
[20:42:30] <grommit> is there a manpage on alias?
[20:42:38] <micges> but I didn't test it with your problem
[20:48:12] <grommit> It works. It's ugly but it works.
[20:48:33] <grommit> I now have servo-thread and real-servothread
[20:49:27] <grommit> thx miges
[20:53:18] <micges> welcome
[20:54:46] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[21:06:11] <Tom_L> anybody awake?
[21:06:46] <micges> sure, what's up?
[21:07:15] <Tom_L> well i've sorta followed emc off and on for a long time but haven't used it
[21:07:32] <Tom_L> i was curious if there was ever talk about porting it to a USB interface?
[21:08:23] <psha> there was talk
[21:08:28] <micges> there was many talks
[21:08:32] <psha> but no work
[21:08:44] <psha> since nobody really needs that
[21:09:00] <psha> i mean among people who may and want to add/fix that
[21:09:14] <Tom_L> i thought it would be handy since so many pc's anymore don't have parports
[21:09:25] <Tom_L> i have somewhat of an interest in it
[21:09:28] <psha> Tom_L: there are tons of issues
[21:09:49] <psha> with USB 1.1 you'll get maximum of 1khz update frequency
[21:10:05] <Tom_L> i'm sure there would be issues
[21:10:50] <psha> and since all realtime usb projects are rotting - there even no basis for this work...
[21:10:59] <Tom_L> i've sorta mentioned it to the guy that wrote LUFA which is a USB framework for avrs
[21:11:10] <psha> surely it may be done atop of standard usb stack
[21:11:25] <psha> but then you'll get no guarantees about signal transmission
[21:14:52] <Tom_L> what sort of update frequencies do you need for microstepping?
[21:15:01] <psha> for stepper drives?
[21:15:08] <Tom_L> yeah
[21:15:13] <Tom_L> or servo for that matter
[21:15:46] <psha> 1khz update rate is 500hz of step pulses (need one frame for lowering signal)
[21:16:30] <psha> even for 1.8 degrees per step you'll get only 2.5 revolutions per second (maximum) without microstepping
[21:16:48] <Tom_L> yeah, that's pretty slow
[21:16:49] <psha> that's top speed
[21:17:15] <Tom_L> well it may not be practical but it was a passing thought i had
[21:17:20] <psha> for usb 2.0 it's a bit faster
[21:17:36] <Tom_L> unfortunately, the avrs aren't full usb2
[21:17:39] <psha> but as i recall chips like ft232bm are usb 1.1
[21:17:50] <Tom_L> the avr32 might be but i doubt it
[21:17:51] <psha> yes, and avrs too
[21:18:16] <Tom_L> they may be usb2 compliant but at 1.1 speed
[21:18:43] <psha> speed is not playing role here
[21:19:20] <psha> you have to utilise not bulk but async mode
[21:19:31] <psha> or interrupt
[21:19:48] <Tom_L> he's been working on async functions for his lib lately i think
[21:21:02] <micges> async mode in rtai/hal driver will be problematic
[21:21:18] <psha> micges: why?
[21:21:40] <micges> async is based on interrputs right?
[21:22:01] <psha> incorrect word
[21:22:04] -!- abcminiuser [abcminiuser!~Dean@2001:700:303:2:258b:3051:6a4c:ef6] has joined #emc
[21:22:06] <psha> it's not async but isochronous
[21:22:07] <Tom_L> i may have picked a bad time to bring it up
[21:22:18] <Tom_L> since he's rather busy right now
[21:22:22] <Tom_L> abcminiuser welcome
[21:22:40] <Tom_L> guys, i invited dean in to hear your thoughts on it
[21:22:52] <Tom_L> since he's the author
[21:22:53] <psha> he'll shame us for our poor knowledge of USB! ;D
[21:23:01] <Tom_L> i doubt it
[21:23:20] <psha> micges: it's not related to sync/async handling of requests and response
[21:23:24] <psha> it's just mode of transmission
[21:23:26] * abcminiuser SHAME!
[21:23:45] <micges> oh, sorry for interrupts then
[21:23:52] <Tom_L> is parport currently the only interface for emc?
[21:23:58] <psha> interrupt is of a same manner - just a type of transfer
[21:24:15] <psha> Tom_L: there is some work towards ethernet (ethercat)
[21:24:28] <Tom_L> well that would be alot faster
[21:24:31] <Tom_L> than usb
[21:25:10] <psha> again - problem is not in speed but in latency
[21:25:30] <Tom_L> they sorta go hand in hand don't they?
[21:25:54] <Tom_L> abcminiuser where are you with your async stuff?
[21:26:11] <psha> not really
[21:26:20] <psha> s/really/exactly/
[21:26:39] <abcminiuser> Tom_L, trying to figure out how it would be useful
[21:26:42] <psha> i believe that gbit and 100mbit latencies are pretty close - but that's only my feeling
[21:26:52] <abcminiuser> I don't want to force everyone to packet everything, due to RAM restrictions
[21:27:03] <abcminiuser> So I prototyped an asynch ringbuffer implementation
[21:27:08] <Tom_L> abcminiuser they seem to think async would be one of the few ways it might work
[21:27:11] <abcminiuser> But it looks to be more trouble than it's worth
[21:28:13] <Tom_L> abcminiuser, are there ANY avr chips that are true USB2 compliant?
[21:28:22] <Tom_L> aka avr32
[21:29:14] <abcminiuser> USB 2.0 compliance != High Speed
[21:29:19] <Tom_L> i know
[21:29:28] <abcminiuser> The USB AVRs are 2.0 complient, but only run at Full Speed
[21:29:32] <Tom_L> i figured you knew what i meant
[21:29:42] -!- motioncontrol has quit [Quit: Sto andando via]
[21:29:44] <abcminiuser> There's the AVR32UC3A chips that are high speed IIRC
[21:31:23] <psha> abcminiuser: may you clarify bits of USB protocol
[21:31:41] <psha> there is isochr transfer mode which has guaranteed latency but may be unreliable
[21:31:59] <psha> i think that's one is interesting (or interrupt mode)
[21:32:46] <abcminiuser> Yes, isochronous have guaranteed latency, but no error retries - good for streaming unreliable data such as video/audio
[21:33:33] <psha> docs state that interrupt also has fixed latency
[21:33:41] <andypugh> EMC supports PCI too, let's not forget that
[21:33:59] <Tom_L> andypugh i've been out of the loop forever :)
[21:34:22] <psha> andypugh: so you may plug second parport! :D
[21:34:25] * psha kidding
[21:34:41] <Tom_L> has anyone developed a PCI card for it?
[21:34:46] <psha> yes
[21:34:49] <psha> mesa
[21:34:53] <andypugh> We need a parport emulator for 5i20 and Motenc.
[21:34:53] <Tom_L> oh nice
[21:34:54] <psha> andypugh: right?
[21:35:07] <andypugh> Aye, Mesa, Motenc, Opto22
[21:35:44] <andypugh> Mesa also do a PCi-E (or -e, I get them mixed up, external cabled) card with 144 io pins.
[21:35:49] <Tom_L> it might not be a viable project for usb
[21:36:04] <psha> abcminiuser: another dumb question: what do you thing about opinoin that ethernet implementation is more simple then usb?
[21:36:17] <Tom_L> i've seen usb based controllers though is what made me think of it
[21:36:34] <andypugh> There is an existing Ethercat library at least.
[21:37:10] <psha> andypugh: that's for 'server' side
[21:37:14] <andypugh> (And I have a pile of servo drives behind me working with emc2 connected via cat5 patch cables, but not using ethernet)
[21:37:57] <psha> for client side some simple board with ethernet would be really great
[21:38:01] <andypugh> The USB controllers put the clever stuff in the controller, and EMC2 is about moving all the clever into the PC where you can see it.
[21:38:22] <psha> i suspect that ethercat hw is expensive...
[21:40:25] <Tom_L> i know alot of commercial cnc's use ethernet
[21:40:56] <Tom_L> but then it's used for a different purpose there
[21:41:06] <Tom_L> although you can DNC with it
[21:41:24] <Tom_L> but you can DNC with rs232 too
[21:41:47] <psha> Tom_L: is not it rs485 just packaged in RJ-45 connector?
[21:42:26] * abcminiuser phone
[21:42:27] <Tom_L> no, what i wired was rs232
[21:43:11] <Tom_L> rather impressed with it's reliability considering the environment it was in
[21:43:44] <Tom_L> but the machines had huge buffers so it could take it's time catching up
[21:44:16] <Tom_L> and unless you were cutting a surface with alot of short linear moves it would keep up pretty constantly
[21:45:53] -!- e3m has quit [Quit: Pull the pin and count to what?]
[21:46:08] <Tom_L> i put a repeater in the line about halfway across the shop on a couple machines
[21:46:21] <elmo40> wouldn't keep up on our machines... we have files that are in the MB in size! That is a LOT of XYZAC movement. Funny thing, though, our machine capacity is only 14"x14"x16".
[21:46:38] <Tom_L> yes, some of these were 500k plus files
[21:46:57] <Tom_L> well much larger too
[21:47:08] <Tom_L> wing profiles etc
[21:47:28] <Tom_L> i never ran 5axis stuff though
[21:47:46] <psha> hm, 100mbit latency (via one fast switch) is living in range 11-120us for 'empty' network
[21:48:02] <psha> at least only paper i've fond states that
[21:49:16] <andypugh> Tom_L: If you have been away, perhaps you haven't seen the 5-axis Cincinatti?
[21:49:26] <Tom_L> no
[21:49:49] <Tom_L> i haven't been in the commercial end for a long time now
[21:50:52] <Tom_L> what cam software do you work with on it?
[21:51:34] <Tom_L> i had smartcam way back when and was familiar with mastercam and surfcam a bit
[21:51:58] <Tom_L> it got replaced with a couple seats of CATIA
[21:53:46] -!- Valen [Valen!~Valen@ppp226-148.static.internode.on.net] has joined #emc
[21:53:57] <andypugh> It's not my machine, but it does show what EMC2 is doing nowadays.
[21:55:43] <andypugh> http://www.youtube.com/watch?v=mxxdq6y8z8M
[21:56:58] <Tom_L> does it run at high speed?
[21:57:37] <Tom_L> that thing should be able to push thru alum at some pretty good rates
[21:58:54] <andypugh> I confess I have only seen it on Youtube, and know very little about it.
[21:59:01] <Tom_L> ok
[21:59:19] <Tom_L> i realize it's utilizing all axis but not at production speeds
[21:59:25] <andypugh> My little toy is the EMC2 gear cutting that you probably get as a side link?
[22:00:06] <Tom_L> helical gear?
[22:00:15] <andypugh> No, the Hobbing one.
[22:00:19] <Tom_L> k
[22:04:09] <Tom_L> those both run in synch?
[22:04:16] <Tom_L> that's a nice feature
[22:05:02] <Tom_L> did you build the head?
[22:05:34] <andypugh> Yes, the head is just a stepper, bearings and a collet holder from eBay.
[22:06:14] <andypugh> EMC2 runs the head as a slave from the spindle encoder, divided down by a factor from a textbox in the control panel.
[22:06:17] <Tom_L> what stepper drivers are popular here?
[22:06:33] <Tom_L> or is there one that stands out over the rest
[22:06:42] <Tom_L> besides gecko
[22:06:46] <andypugh> Mine are from Motion Control Products, but I think I am the only one.
[22:07:21] <andypugh> CNC zone all swear by Geckos, but I think that here they all seem to work fine.
[22:07:24] <Tom_L> i've been lookin over this guy's stuff http://www.probotix.com/
[22:08:10] <andypugh> That driver looks a whole lot like a differently-badged version of the ones I have.
[22:08:29] <Tom_L> what about cost?
[22:09:24] <andypugh> Similar: http://www.motioncontrolproducts.co.uk/products/10/36/msd880_high_performance_microstepping_drive/
[22:10:29] <Tom_L> probably the same thing
[22:11:02] <andypugh> exactly the same spec
[22:12:33] <andypugh> grommit: You know that you can run a charge pump with functions other than "charge_pump" if you need a different frequency. A 50% PWMgen works well.
[22:15:15] <Tom_L> what's the recomended pc for emc? i've got a pile of pc104 boards i considered using for one
[22:15:21] <Tom_L> may be too slow
[22:16:03] <elmo40> high latency is worse then slow cpu... besides, I am sure a 386 is fast enough
[22:16:41] <andypugh> The popular one at the moment is Intel D510MO. $50 for 6k latency is a win.
[22:16:54] <Tom_L> these are via chips iirc
[22:18:06] <psha> Tom_L: D510MO
[22:18:17] <Tom_L> 3.8M file... http://tom-itx.dyndns.org:81/~webpage/SBC/PDF/PCM-9372%20Manual_ed2.pdf
[22:18:27] <Tom_L> these were free though
[22:18:42] <andypugh> Try them then.
[22:18:53] <Tom_L> i had xp on one for a bit
[22:19:00] <Tom_L> and a linux server for a while
[22:19:10] <andypugh> Mesa also do a lot of PC104 stuff, though I am not sure if it is EMC2 compatible.
[22:21:46] -!- micges has quit [Quit: Ex-Chat]
[22:21:53] <Tom_L> http://tom-itx.dyndns.org:81/~webpage/SBC/server/DSC_0005.JPG
[22:22:00] <Tom_L> that was one of em
[22:23:02] <Tom_L> originally used for gps tracking of vehicles
[22:29:45] <elmo40> kinda large for that, eh?
[22:29:54] <elmo40> dones they use 'linux on a chip' for tracking?
[22:30:01] <elmo40> those instant-on distros
[22:33:38] -!- sideone [sideone!~overflow3@mail.traeger.com] has joined #emc
[22:33:48] -!- sideone [sideone!~overflow3@mail.traeger.com] has parted #emc
[22:34:31] <andypugh> We have a bunch of loggers tracking our test fleet, they log hundreds of data channels as well as GPS position, and report it all back to base with cellular data. The loggers themselves run linux. I can ssh in and feel quite at home.
[22:48:27] -!- skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc
[22:53:07] -!- Fox_Muldr has quit [Ping timeout: 276 seconds]
[22:54:20] -!- mhaberler [mhaberler!~chatzilla@host88-93-static.38-79-b.business.telecomitalia.it] has joined #emc
[22:54:22] -!- Fox_Muldr [Fox_Muldr!quakeman@frnk-4d01e75f.pool.mediaWays.net] has joined #emc
[22:56:12] -!- odiug has quit [Ping timeout: 265 seconds]
[23:00:50] -!- LawrenceG [LawrenceG!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc
[23:03:39] <Valen> mornin Tom_L
[23:03:48] <psha> oh!
[23:03:51] <psha> gnight valen
[23:05:05] -!- psha has quit [Quit: zzz]
[23:06:44] -!- grommit has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122310]]
[23:07:11] <Tom_L> hey Valen
[23:07:30] <Tom_L> didn't realize you were here
[23:08:00] <Tom_L> elmo40, no
[23:08:17] -!- mhaberler has quit [Ping timeout: 272 seconds]
[23:09:43] -!- elmo40 [elmo40!~Administr@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has parted #emc
[23:11:51] -!- Tom_L has quit []
[23:14:08] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[23:14:33] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[23:15:58] <Tom_L> their original use http://tom-itx.dyndns.org:81/~webpage/SBC/inside_control.jpg
[23:16:23] <Tom_L> the box would make a good housing for the whole cnc control i think
[23:16:33] <Tom_L> add some vents
[23:16:37] <Tom_L> and fans
[23:17:23] -!- abcminiuser has quit [Ping timeout: 260 seconds]
[23:30:31] <L84Supper> Tom_L : I forget which part number the VIA Twister chipset was but try the latency test on it
[23:31:09] <Tom_L> i doubt it's quick enough to keep up but i might try one
[23:31:25] <Tom_L> it's fairly old equipment
[23:31:40] <L84Supper> yeah, ~8 years
[23:31:56] <Tom_L> i've got half dozen of those boxes though
[23:32:06] <Tom_L> out of ~200 given away
[23:32:29] <L84Supper> but if you keep the video 640x480 it might be good enough
[23:35:15] <L84Supper> the oldest VIA board I tried was a 1GHz C7+ cn400
[23:35:22] -!- adb has quit [Ping timeout: 276 seconds]
[23:36:47] <Tom_L> i have an EPIA MII 10000 that probably would work fine
[23:36:47] <L84Supper> IIRC we had a latency issue once every hour or so that bumped it to 100,000
[23:37:48] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[23:38:08] <Tom_L> it's currently my _itx logon
[23:39:08] <Tom_L> http://tom-itx.dyndns.org:81/~webpage/itx/itx_thumb_index.php
[23:39:12] <Tom_L> there's it's assembly page
[23:40:34] <L84Supper> there is coreboot support for the cn700 and cx700, that should clear up any smi issues if any