#emc | Logs for 2010-11-18

[00:00:33] <JT-Hardinge> I'm just working out some spreadsheets with feed and speeds for different cutters I have and though I might add something to show surface finish
[00:00:43] <JT-Hardinge> just to help me pick
[00:01:03] <andypugh> Sounds useful
[00:01:31] <andypugh> While you are here, what is it with that Unigraphics postprocessor? How hard can it be?
[00:01:43] <JT-Hardinge> so if I use the middle calc then the part will have at least that surface finish
[00:01:55] <JT-Hardinge> yes, I wonder sometimes
[00:02:03] <andypugh> Yes. Which is good as that is an easier caclulation.
[00:02:52] <JT-Hardinge> seems like I spend more time trying to figure out speeds and feeds for various tools that some charts would be nice
[00:03:37] <JT-Hardinge> as for the Unigraphics fellow everyone can be overwhelmed from time to time and need some hand holding
[00:03:55] <theorb> theorb is now known as theorbtwo
[00:06:51] <JT-Hardinge> I need to think on this a bit...
[00:07:15] <andypugh> It just seems like the (very expensive) software is making a bad job of the task.
[00:08:43] <JT-Hardinge> yes, some over complicate the task and some make it hard to get what you want
[00:09:25] <JT-Hardinge> I use SolidWorks to design machine and it is easy to use but a resource hog
[00:09:25] <andypugh> It doesn't help that he isn't telling us the error message
[00:09:44] <JT-Hardinge> oh yea him lol
[00:10:07] <JT-Hardinge> I tried to tell him to just start simple but he doesn't seem to notice
[00:11:32] <andypugh> Well, he presumably has a solid model. Getting a simple, single, arc is probably a real challenge.
[00:11:33] <JT-Hardinge> basically the intersection of the two arcs will give you the height of the scallop, so that would give you some indication of surface finish
[00:11:58] <JT-Hardinge> not in SolidWorks :)
[00:14:19] <andypugh> I need to learn not to use the elipsis on the forums...
[00:14:45] <JT-Hardinge> is that not a pia
[00:15:18] <JT-Hardinge> ok 1/2 of the feed per tooth is the segement so the height can be calculated from that
[00:15:49] <JT-Hardinge> segment height I meant
[00:16:25] <andypugh> It is chord height isn't it?
[00:21:18] <JT-Hardinge> * JT-Hardinge pulls out the large print edition of the Machinery's Handbook
[01:39:49] <skunkworks> wow - my glasses were so dirty. I was wondering why my laptop screen seemed so dull.
[02:36:44] <mikegg> woa, sweet. who did the colored text for the INI file in gedit with 10.04?
[02:36:47] <mikegg> i love you guys
[02:39:43] <SWPadnos> that should have been there about forever :)
[02:40:17] <SWPadnos> I wonder if we ever added the G-code highlighting. I know one or two people made definitions for it
[02:42:51] <atmega> no g-code defs in mine
[02:42:55] <atmega> no cobol either
[02:43:01] <SWPadnos> gosh darn it!
[02:44:19] <mikegg> the company I was working for that made me be all secret about what we were developing....yeah they went bust
[02:44:33] <mikegg> but not before we built this!
[02:45:28] <mikegg> http://imagebin.ca/view/VPg_C1.html
[02:45:54] <mikegg> the blue box next to it is a 4U rack that contained all the electronics / pneumatics
[02:46:27] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Highlighting_In_Gedit
[02:47:09] <mikegg> sweet
[02:52:14] <atmega> I just upgraded to 10.04, whenever I start emc, my spindle goes on for a second
[02:57:19] <SWPadnos> what hardware?
[02:58:21] <atmega> pport?
[02:58:26] <atmega> what context?
[02:58:51] <atmega> pport -> relay -> AC to spindle
[02:59:41] <SWPadnos> if you put a delay in the hal file (loadusr sleep 5 or something like that), does the spindle stay on longer before shutting off?
[03:00:16] <SWPadnos> I'm thinking the default port state has the output on, and then it gets turned off once emc is started up and the spindle HAL pins are connected
[03:00:23] <SWPadnos> (and the HAL threads start to run)
[03:01:49] <atmega> didnt' stay on any longer
[03:02:43] <SWPadnos> hmmm. you probably need some sort of wait in there
[03:02:53] <atmega> what would set the default state? or is it just 'there'
[03:03:02] <SWPadnos> halcmd just starts programs and keeps going
[03:03:21] <SWPadnos> the parport driver may, when it initializes the port
[03:03:51] <SWPadnos> or it may be the parport itself, going to some default state when the control registers are teiddled
[03:05:11] <SWPadnos> ah yes, -w will wait for the program to finish, so loadusr -w sleep 5
[03:05:30] <SWPadnos> after the loadrt hal_parport
[03:06:31] <atmega> yep, stayed on then
[03:07:16] <SWPadnos> ok, that makes me more confident that my guess is right :)
[03:07:24] <atmega> sounds good to me.
[03:07:35] <SWPadnos> you can live with it, or
[03:07:52] <atmega> I can wire around it... I want an external ON button anyway
[03:07:54] <SWPadnos> change the hardware so it doesn't come on when the port is initialized
[03:08:38] <SWPadnos> while also making sure that it won't be on from the time you turn on the computer until you run EMC :)
[03:09:12] <atmega> before, it would turn on with the computer
[03:09:20] <SWPadnos> oh
[03:09:32] <SWPadnos> then I suggest a charge-pump based external board
[03:09:41] <atmega> yeah, someday
[03:10:30] <SWPadnos> or you could hack hal_parport.c so that it initializes the port to a different value (or alternatively add a load-time parameter to hal_parport that lets the user specify an initial output value)
[03:10:59] <atmega> seems like all off would be a typical start
[03:11:11] <SWPadnos> well, yes and no
[03:11:27] <SWPadnos> 0 is often active in logic circuits, and also with NPN transistors
[03:11:41] <SWPadnos> it's also the higher drive side of a parport, so you want that active
[03:11:55] <SWPadnos> check the port with a meter, see what the voltage actually is
[03:12:25] <SWPadnos> there may also be some outputs that don't change, like one of the inverted ones maybe, where you can move the spindle control
[03:13:29] <atmega> Pin 14-Auto Feed
[04:34:03] <asru> Anybody have any experience with products from Easysupply?
[04:40:13] <KimK> Not I, tell us about them, what are we talking about here?
[04:52:38] <asru> They're selling what look like decent Nema 23 motors, and an inexpensive controller board built around TB6560 controller chips.
[04:53:35] <asru> The board and a 10 amp power supply are selling for under a hundred bucks or so.
[04:54:15] <asru> Problem is that the stuff is sold from outside the states, and I don't trust low-priced Chinese stuff (some of the stuff ships from inside China).
[04:55:44] <asru> The low prices might be the result of smart engineering, . . . or they might be the result of garbagy building.
[05:01:43] <asru> And as long as I'm yacking about prices, people might want to that Rutland has Melin single flute aluminum routers on sale (at about half normal price).
[05:11:46] <KimK> How do these items compare (price and performance) with similar items from Keling? I'm a little more familiar with them, they were at the last CNC workshop in Ann Arbor. http://www.kelinginc.net/ They're in the Chicago area.
[05:12:15] <asru> That's the sort of thing I'm trying to find out <grin>
[05:13:56] <qq-> asru, have an URL for your Easysupply ?
[05:14:25] <qq-> easy-supply.com/ ?
[05:14:58] <asru> Only UrL I have right now is http://stores.ebay.com/EasySupply
[05:15:31] <qq-> thanks
[05:16:28] <asru> As for comparison with Keling, the prices are much, much lower.
[05:19:49] <KimK> OK, thanks for that report. Although Keling already seems pretty low. I'm not sure I want to go scary low. A good friend of mine once advised me, "You don't want the cheapest tire you can buy".
[05:20:39] <asru> In this case, bucks are a serious issue--I need a controller for a college student project I am supervising, and funding is very, very tight right now.
[05:21:11] <asru> And I don't need maximum quality, just something that works reliably.
[05:23:19] <KimK> Ah, college student? Well, why didn't you say so? Let me show you some of the returns and closeouts from the back room, lol. You should ask the sellers that when you call/email them. Got any closeouts or returns? Open boxes?
[05:23:50] <qq-> 3 Axis CNC Router Schrittmotor Steuerkarte NEMA 23 PSU Approximately US $234.06
[05:24:24] <asru> You can find the same in English if your root around the site a bit <grin>
[05:24:41] <asru> As for being a college student, not quite--I'm the instructor <grin>
[05:26:38] <KimK> Schrittmotor Steuerkarte --> Google says --> Stepper motor control card
[05:27:41] <asru> Indeed. However, Easysupply seems to advertise the same stuff in both German and English. It ships from different places--England, Hong Kong, or China.
[05:27:45] <qq-> anyways ebay decided i speak german as i live in switzerland ,...
[05:28:14] <KimK> OK, so it's *for* college students then? Just one up front or one for each student? (Or pair of students?)
[05:28:25] <asru> As long as you speak German, no reason to hunt for the English.
[05:28:47] <asru> Kimk: Just one, since the two kids are working together (and doing very nicely, actually).
[05:28:51] <qq-> asru, no , i don't
[05:29:21] <qq-> french side here
[05:29:28] <asru> In that case, please trust me when I say you can find exactly the same stuff in English if you look around their site (I did).
[05:29:44] <asru> qq: Didn't see anything in French, though.
[05:31:37] <KimK> Good luck with the project, sorry I was not able to help you much.
[05:32:15] <asru> Kimk: Thanks.
[05:33:04] <asru> Kimk: The project will limp along, and may or may not ever be completed. What is important is that both kids have learned one heck of a lot while working on it.
[05:33:32] <KimK> BTW, EasySupply does not seem to organize his eBay store very well, he has everything filed under "other".
[05:34:06] <qq-> asru, see http://tinyurl.com/32nj8x6 maybe it speak english to you ..
[05:34:34] <asru> KimK: And if you really look carefully, you'll find that their English ain't exactly the Queen's.
[05:35:52] <asru> qq: http://cgi.ebay.com/3-Axis-Stepper-Driver-Nema-17-60oz-in-PSU-CNC-Kit-/140467777127?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item20b4882e67
[05:36:40] <qq-> asru, cool . that's in english
[05:36:48] <asru> qq: Ooooops. That one's a Nema 17. But the Nema-23's are there also. I used their search window with the keywods "cnc kit"
[05:37:43] <asru> qq: And both the Nema 17 and 23 kits appear to be using the same control board.
[05:38:28] <qq-> i think so
[05:38:47] <qq-> and US $187.10 seems a good price
[05:39:18] <asru> qq: It's the price that attracted my attention.
[05:39:32] <qq-> hehe
[05:40:12] <qq-> add to it the computer price ..
[05:40:28] <asru> qq: Only problem with that particular kit is that the motors are only 30 oz. in
[05:41:06] <asru> qq: As for computers, I have plenty of computing power available--6 quad core i96's and 19 Sun 100's <grin>.
[05:41:32] <asru> qq: Oooops. That should have been 6 quad core i86's.
[05:42:40] <qq-> btw how many controller may guide a pc ?
[05:43:45] <asru> qq: If you mean how many controllers can a pc handle, it is probably a function of how much CPU power there is, and how fast the buses are.
[05:44:25] <asru> qq: If you mean how many people-controllers can be handled by a PC, well, these are Linux machines, so I would say several hundred <grin>
[05:44:58] <qq-> asru, i meant former
[05:45:36] <qq-> and i thought only one per pc
[05:45:51] <asru> qq: No offense, please, I was joking. I would say that one controller is the limit for just about all single-core machines.
[05:46:26] <asru> qq: However, on multi core machines, the game is quite different, and I think would really depend on the motherboard.
[05:46:28] <SWPadnos> you can run one machine, but with emc the number of motors or other I/O that can be interfaced is limited only by the hardware you attach and the CPU
[05:47:18] <SWPadnos> CPU power isn't very important until you get to very large amounts of I/O, non-trivial kinematics (where there may be matrix math and multiple iterations to get to a solution), etc
[05:47:22] <asru> SWPadnos: With a multicore machine and the right software, it might well be possible to run more than one copy of EMC.
[05:47:27] <SWPadnos> nope
[05:47:48] <SWPadnos> not without some changes, which I suspect are fairly major
[05:48:35] <asru> SWPadnos: I'll stick to my guns on whether or not there might be enough hardware in one box to run more than one copy of EMC if the right software is used.
[05:48:53] <asru> SWPadnos: And, yes, major changes in interrupt handling might well be required.
[05:48:56] <SWPadnos> well the right software would be a modified emc, I guess :)
[05:49:07] <qq-> SWPadnos, means how many parallel port attached to pc ?
[05:49:30] <SWPadnos> you can use as many as you want, but at some point the I/O speed will be too much
[05:49:44] <SWPadnos> you would have to slow down the base thread, and at some point the servo thread as well
[05:50:18] <asru> All depends on bus speeds, board design, and interrupt handling.
[05:50:21] <SWPadnos> asru, it's not a matter of optimization or CPU usage - some parts of EMC are realtime kernel modules, and at the moment there is only one RT domain
[05:50:56] <SWPadnos> well, if you want to design a computer and then stick EMC or something like it on there, then you're absolutely right
[05:51:22] <asru> If EMC were ported to IBM architecture, there are wonders to be seen when you run VM.
[05:51:31] <SWPadnos> but if you want to run on a stock PC, and not write the entire OS, then you have limited choices
[05:51:35] <SWPadnos> IBM mainframes?
[05:52:08] <SWPadnos> in the PC world, VMs provide terribly poor access to hardware
[05:52:10] <asru> Quite limited choices. I wonder if EMC could be compiled to on Sys 390 architecture. Linux can.
[05:52:18] <SWPadnos> they're kind of the opposite of what you need to run a machine
[05:52:39] <SWPadnos> you can probably run the sim mode on it, but you would have no access to hardware
[05:53:07] <asru> You'd have all the hardware access you need. Sys 390 VM is not VMware.
[05:53:15] <SWPadnos> I understand that :)
[05:53:25] <SWPadnos> you'd have to write drivers
[05:53:55] <SWPadnos> if there is PCI access, and also access to inb/outb, then it may be possible
[05:54:06] <asru> VMware is actually a sort of simulator, and trying to get at the underlying hardware is murder. IBM VM was designed to allow access to underlying hardware under
[05:54:10] <asru> controlled conditions.
[05:54:29] <SWPadnos> there are a few x86-isms in the code, but not many (AFAIK)
[05:54:49] <SWPadnos> well, definitely post videos once you get it going :)
[05:55:07] <asru> Not where my current interests lie :)
[05:55:23] <asru> However, if you'd like to discuss how to solve jigsaw puzzles . . . :)
[05:56:33] <boelkmoeller3> (just an offtopic hint: I'm currently packaging emc and related stuff for RHEL6 - It's nearly done. all pkgs built fine and I will test it the next weeks)
[05:56:33] <SWPadnos> no, that's OK. I think I'll go to sleep instead :)
[05:56:58] <SWPadnos> great. if you stick it somewhere on the web, we can put it on the linuxcnc.org site, or link to it
[05:57:03] <asru> Makes two of us. I need to be up fairly early tomorrow.
[05:57:34] <SWPadnos> and it's 1:00 AM here
[05:57:36] <SWPadnos> see you
[05:57:41] <asru> Ditto.
[06:11:37] <elmo401> boelkmoeller3: why the enterprise edition? why not Fedora?
[06:16:00] <elmo401> elmo401 is now known as elmo40
[06:38:18] <boelkmoeller3> SWPadnos maybe i need some webspace for a repo...
[08:22:07] <kvirdi> Hi, I am interested in retrofitting my old sodick milling machines, and would like some information about encoders.
[08:45:51] <KimK> kvirdi: Since I'm still awake, what would you like to know?
[08:46:09] <kvirdi> well ive read that i need a differential encoder to work
[08:46:34] <kvirdi> and im trying to find out if my current setup's encoders will work or not
[08:46:52] <kvirdi> they are panasonic MINAS A series drives and motors
[08:46:58] <KimK> When working around electrically noisy devices, a differential encoder is highly recommended, yes.
[08:47:44] <KimK> Have you got a data sheet on those drives and motors?
[08:47:53] <kvirdi> the model numbers listed on my drives and motors don't match with the information from the data sheet. :(
[08:48:17] <KimK> Nothing on Google?
[08:48:35] <kvirdi> googling has brought me to you guys :)
[08:49:05] <kvirdi> Mr Jon Elson from pico systems has posted about panasonic drives
[08:49:22] <kvirdi> and even had pictures that look very similar to what I have
[08:49:32] <KimK> Is there a Panasonic site for those motors and do they have the right manuals available? (Even if supported as "legacy" devices, etc.?)
[08:50:10] <kvirdi> Nope I have downloaded the information from panasonics website, while very helpful, doesn't fully match what I have
[08:50:25] <KimK> OK, how many wires come off the motor? These are small motors (hand-sized) are they not?
[08:50:31] <kvirdi> the catalogue says my encoder should be an A, B or D type and lists what those mean.
[08:50:37] <kvirdi> My models have a K type
[08:50:41] <kvirdi> 17 wires
[08:50:48] <kvirdi> and they are larger motors
[08:50:54] <kvirdi> between 1.5 to 6 kW
[08:51:32] <KimK> Oh, clearly I was thinking of something else, how large are they?
[08:51:49] <KimK> ANd do you have that link so I can at least see what you are seeing?
[08:52:22] <kvirdi> sure, i am going to hunt for the forum link
[08:52:48] <KimK> With a motor of that size, I suspect that they are differential encoders, since it would be foolish not to, but let's proceed.
[08:53:06] <kvirdi> http://www.mail-archive.com/emc-users@lists.sourceforge.net/msg10923.html
[08:53:19] <kvirdi> still looking for the pictures...
[08:55:03] <KimK> And could I have some numbers from the motors?
[08:55:54] <kvirdi> http://www.cnczone.com/forums/controller_cards/69084-new_drive_panasonic_minas_s_series_motors.html
[08:56:00] <kvirdi> post #7 :)
[08:56:04] <kvirdi> sure
[08:56:34] <kvirdi> motor is model - MDM202K1C
[08:57:02] <kvirdi> this one is a 2kW, 200V motor
[08:57:36] <kvirdi> i also have a similar number motor - MDM152K1C, which is a 1.5kW
[08:57:58] <kvirdi> and MKM602D51, which i have not found information for on panasonics website
[08:59:17] <kvirdi> here is a link to panasonics wesite that has the information relevant to the first two motors - http://industrial.panasonic.com/ww/i_e/25000/fa_pro_acs_e/fa_pro_acs_e/a4.html
[09:01:34] <KimK> Does the nameplate call them "AC Servomotor" or "Brushless Motor"?
[09:02:11] <KimK> Oh, and about how large are they (physically?)
[09:02:35] <kvirdi> AC Servo Motor
[09:02:45] <KimK> OK
[09:03:05] <kvirdi> the 1.5kW motor in front of me is 9.5" x 5 x 5
[09:03:35] <KimK> Any more MINAS numbers besides A ?
[09:03:47] <KimK> A4? A5?
[09:04:11] <kvirdi> ive found the A4 to be similar to what i have, but the motor itself doesnt say
[09:04:36] <kvirdi> and one motor isnt similar to anything i found on the site: MKM602D51
[09:06:02] <KimK> I see MINAS A4, A4P, A4F, and A4N. Have you looked at all of those? (Plus there's MINAS A5 and MINAS E.)
[09:06:53] <kvirdi> yes i have.
[09:06:55] <KimK> BTW, are these geared-head motors or not? (Is there a gear-reduction box at the front of the motor?)
[09:07:04] <kvirdi> no gears
[09:07:31] <KimK> OK. I'm just full of questions.
[09:07:36] <kvirdi> i couldnt find anything that started with MKM in the A5s or E series
[09:07:43] <kvirdi> anything that could help :)
[09:08:13] <KimK> Did you say you had photos? Or similar?
[09:08:50] <kvirdi> no i dont have any right now but this link looks similar to what i have
[09:08:54] <kvirdi> http://www.cnczone.com/forums/controller_cards/69084-new_drive_panasonic_minas_s_series_motors.html
[09:08:59] <kvirdi> post #7 from the top
[09:09:20] <kvirdi> the post also mentions 17 pin encoders like i have :)
[09:09:30] <KimK> You might have some kind of OEM motor that Panasonic custom-made from the MINAS line for a customer. Let's keep going.
[09:09:58] <kvirdi> yes i have thought the same thing, my motors and drives seem custom made
[09:10:32] <kvirdi> or custom packaged versions of their standard products
[09:12:06] <KimK> And your "encoder" (whatever) looks like the five pictures there? And you have the 16 wires plus a shield?
[09:12:49] <KimK> (That's an odd duck if it's that serial-hall business that Jon is talking about.)
[09:13:16] <kvirdi> i have the 1 wires, my father has opened up this before and says thats what it looks like
[09:13:41] <kvirdi> i have the 17 wires*
[09:14:46] <kvirdi> hm seems they are labelled as well, i have never noticed, lets see what they say...
[09:16:02] <kvirdi> well it says JAE, which i think means Japan Aviation Electronics
[09:16:14] <kvirdi> and the rest are lettered A, B, C etc
[09:19:59] <KimK> In the last posting Jon seems to have identified everything, although he does not give details. But he does give a link to his Panasonic MUMS/MINAS to Generic Hall Effect converter ($100). http://pico-systems.com/oscrc4/catalog/product_info.php?cPath=4&products_id=26
[09:20:17] <KimK> Does yours have the 3x5 15-pin white plastic rectangular connector?
[09:21:12] <kvirdi> no
[09:21:31] <kvirdi> mine has two cocentric circles of 5+12 pins
[09:22:42] <kvirdi> hm nice link. i would pay 100 dollars if itll work for me
[09:24:13] <KimK> Are yours metal connectors with a fastening ring, like the "Cannon" (or military) type? Maybe that's why your motors are special? Plastic connectors are sometimes unsuitable in certain applications, oil, chips, heat, etc. Maybe plastic wasn't good enough for somebody?
[09:24:45] <KimK> Wait, you have the correct drives that go with the motors, right?
[09:24:57] <kvirdi> yes i do
[09:25:03] <kvirdi> and yes our cables look like that
[09:25:22] <kvirdi> these are fitted to my sodick cnc milling machines
[09:25:40] <KimK> Then you don't need a converter, your drive already accepts the "strange" motor signals.
[09:26:43] <KimK> And Jon's comment, "The ABZ part of the encoder is totally standard..." is the second piece of evidence we need to say that your motors have differential encoders.
[09:27:15] <kvirdi> ok great
[09:28:49] <kvirdi> does my drive need to pass on this information to my control?
[09:28:49] <KimK> Sometimes encoder outputs from motors must be used directly, sometimes the drive supports the encoder and provides either one or two buffered outputs. So your drive might have the encoder outputs you need, you should look into that. Maybe if you contact the Panasonic rep in your area he can help?
[09:29:03] <kvirdi> ok
[09:29:26] <kvirdi> i am currently in india, and they only support in Chennai, about half a sub continent away
[09:29:54] <KimK> The commutation encoder is only needed by the drive. The position encoder is probably not needed by the drive, but might or might not go through it.
[09:30:49] <KimK> The commutation encoder is the strange one in your motors. The position encoder is, according to Jon, completely standard. And without seeing the motors, I would agree.
[09:31:38] <KimK> If you ever have a drive or motor failure, you may want to consider replacing them as a pair with something more standard.
[09:31:40] <kvirdi> ok so the ABZ is positional and is standard
[09:31:54] <kvirdi> i have more than enough spares of these :)
[09:33:31] <kvirdi> ok so now i am looking at my drive to find out about these possible encoder outputs.
[09:33:33] <KimK> Yes, the A and B are the "quadrature" outputs (90 degrees out of phase) and the Z output (sometimes called the I or Index output) occurs once per revolution. Does your motor nameplate say how many pulses per rev?
[09:34:05] <kvirdi> my drive mentions that the encoder is 5000 p/r
[09:34:24] <kvirdi> another oddity, as panasonic only mentions 2500 p/r
[09:35:02] <kvirdi> the motor nameplate doesn't say anything about the encoder
[09:38:11] <KimK> "Pulses" is somewhat ambiguous in this situation. 5000 seems pretty high, you might have a 1250 line/cycle encoder, which would give you 5000 quadrature counts. If it was really a 5000-line encoder, that would give you 20,000 quadrature counts, and that seems unlikely to me.
[09:40:41] <KimK> 2500 "pulses"(?) could be 2500 quadrature counts, and would be a 625-line encoder. That's reasonable for a "standard" encoder. 2500-lines would be 10,000 quadrature counts and that is possible, but at the high end of what I would expect.
[09:41:13] <KimK> So let me know what you find out, I'd be curious to know if you have such a high-res encoder.
[09:41:24] <kvirdi> ok i will
[09:41:39] <KimK> If it is 5000-lines 20,000-counts, I mean.
[09:41:50] <KimK> That would be really nice.
[09:42:09] <kvirdi> i dont have any more reference regarding this
[09:42:09] <KimK> So how did you find out about EMC2?
[09:42:23] <KimK> And what are you planning to do with it?
[09:42:38] <kvirdi> well i have been interested in changing my original PC based controllers
[09:42:51] <kvirdi> currently they are win95 delta tau controllers
[09:43:11] <kvirdi> the software is really picky and the PCs are starting to wear
[09:43:11] <KimK> Oh, yes. What kind of a machine is it?
[09:43:30] <kvirdi> Sodick 450 and 600 VNCs
[09:43:38] <kvirdi> i have 5 total
[09:44:15] <kvirdi> ive had them for close to 10 years and work well for us
[09:44:38] <kvirdi> we are mold makers and use these machines to machine graphite electrodes
[09:44:53] <KimK> I think you will be really happy with EMC2, but be patient, as it can do a lot, and there can be a steep learning curve in the beginning.
[09:45:22] <kvirdi> the more i learn about it the more positive i get about the possibilities
[09:46:09] <kvirdi> i have been interested to know if anyone has gotten good results in 3D using this software
[09:46:23] <kvirdi> but that seems more dependent on the interface cards than the software
[09:46:36] <kvirdi> im looking into pico and vital systems right now
[09:46:40] <KimK> Yes, a friend of mine has a Sodick sinker with a servo/run spindle. So it can act as a C axis for helical sinking, or act as a milling cutter for (cleanly) machining graphite while submerged under oil.
[09:47:29] <kvirdi> thats great i have some old roboform sinkers i was hoping to tackle if my sodicks work out
[09:47:51] <kvirdi> i have 4 roboform 20s and a very old roboform 400
[09:48:29] <KimK> OK. Be sure to look at Mesa's "Anything I/O" too. http://www.mesanet.com/ You'll need to pick a motherboard (or two?) and as many daughter boards as it takes to do what you need.
[09:48:49] <kvirdi> these seemed too goo to be true
[09:48:56] <kvirdi> good*
[09:49:27] <kvirdi> and so far they haven't responded to my emails :(
[09:50:13] <kvirdi> vital systems people are very helpful, and are on skype most of the day and pico systems seem helpful as well
[09:51:25] <KimK> It was fortunate that I was still awake, in the future you might find more people in this channel if you try during your night, lol!
[09:51:55] <kvirdi> its afternoon right now, i think i will in the future try in my morning
[09:52:07] <kvirdi> which equals US evening
[09:52:19] <KimK> OK, that might work too.
[09:52:27] <kvirdi> you have been very helpful though thanks a lot
[09:52:53] <KimK> Anyway, good chatting with you, maybe we'll meet again here.
[09:53:27] <kvirdi> yup, thanks again bye
[09:53:34] <KimK> bye
[10:06:32] <Bonny> Hello... I just look over http://wiki.linuxcnc.org/emcinfo.pl?Axis_Embed_Video and wan't to try
[10:07:03] <Bonny> But the 'manual' is written for more experienced users
[10:07:19] <Bonny> can someone help me where to put that file (.axisrc
[10:22:13] <Bonny> Nobody to help?
[12:44:05] <jthornton> Newegg has the Atom D510 with CPU for $65
[12:44:58] <jthornton> have all the issues with it been solved as far as EMC goes?
[12:46:14] <psha> jthornton: for me D510 is working fine
[12:46:22] <psha> intel D510MO mb
[12:46:44] <jthornton> cool, I think I'll get one for backup for the Discovery mill
[12:46:57] <psha> but you have to glue additional heatsink on chipset, it's too hot
[12:47:10] <psha> it's working nice without heatsink too
[12:48:29] <psha> but parport is used in output mode only (mill is still without any feedback :( )
[12:51:02] <IchGuckLive> Hi to the USA from Cold Germany -3Deg Cel. this morning
[12:51:30] <IchGuckLive> i wrote a simple inch to metric for a Frend -> http://www.python.org/dev/peps/pep-0008/
[12:51:34] <IchGuckLive> sorry
[12:51:41] <IchGuckLive> http://paste.pocoo.org/show/293048/
[12:51:57] <IchGuckLive> is this worth to take it to the wiki ??
[12:53:03] <jthornton> if your friend needed it someone else surly will
[12:54:25] <IchGuckLive> as we spoke yesterday there is under normal surcomstances no need to convert from inch to metric by the number just give it a G21/G20
[14:42:19] <JT-Hardinge> * JT-Hardinge goes back to figuring out the formula for a segment of a circle
[14:42:38] <atmega> want me to have one of my kids help?
[14:42:46] <JT-Hardinge> please
[14:45:13] <atmega> they are in school for another 6 hours
[14:45:40] <JT-Hardinge> I'll just have to have my grandkids figure it out they don't go to school yet
[14:46:59] <atmega> luckily, my grandkids aren't born yet :)
[14:49:54] <JT-Hardinge> lol, half of the circle links are the same thing with different ads
[14:50:17] <atmega> there is only so much you can do to a circle.
[14:58:43] <JT-Hardinge> * JT-Hardinge gives up on google and goes back to the LARGE PRINT EDITION OF MACHINERY'S HANDBOOK
[15:11:55] <Jymmm> http://www.mathopenref.com/segment.html
[15:13:26] <Jymmm> http://www.mathopenref.com/segmentarea.html
[15:15:14] <JT-Hardinge> Jymmm: not exactly what I need but I think I brute force fingered it out
[15:15:25] <Jymmm> k
[15:15:41] <JT-Hardinge> I knew I should have paid more attention to my math book in school instead of the hot teacher :/
[15:16:22] <atmega> that was so long ago, it wouldn't have mattered
[15:16:28] <atmega> in my case anyway.
[15:19:15] <JT-Hardinge> well I thought I did anyway
[15:21:12] <JT-Hardinge> * JT-Hardinge gives up for now and goes over to the mill to make some chips
[15:32:23] <JT-Hardinge> if anyone can figure it out it would be most appreciated... I know the width of the segment, the radius and need the height of the segment
[15:33:50] <SWPadnos> the "height" of an arc segment?
[15:36:17] <JT-Hardinge> yes, to give the approximate scallop when face milling
[15:37:36] <JT-Hardinge> this is what I came up with but it is wrong $B$17-(SQRT(4*($B$17*$B$17))-(C$13*C$13)/2)
[15:37:54] <JT-Hardinge> b17=radius, c13=width
[15:39:52] <SWPadnos> http://mathforum.org/dr.math/faq/faq.circle.segment.html
[15:40:00] <SWPadnos> http://mathcentral.uregina.ca/QQ/database/QQ.09.06/h/bob4.html
[15:40:15] <SWPadnos> http://mathforum.org/dr.math/faq/formulas/faq.circle.html#segment
[15:44:50] <JT-Hardinge> looks like it might be this one http://mathforum.org/dr.math/faq/faq.circle.segment.html#7 I wish I knew what a theta was so I could get past step 1
[15:47:26] <elmo40> theta is the angle
[15:49:29] <elmo40> in radians...
[15:49:29] <elmo40> for #7 it means: ANGLE = 2 ARCSIN(chord / 2*radius)
[15:49:30] <JT-Hardinge> the angle from the center of the circle to the ends of the segment?
[15:50:03] <elmo40> what is it you are looking for? the size of the slice of pie?
[15:50:19] <JT-Hardinge> the height of the segment
[15:51:09] <JT-Hardinge> so if I have a 0.032 radius tip on my face mill and I take 0.015 cut per tooth how high is the scallop is what I'm after
[15:52:23] <SWPadnos> .015 is the distance you move in the horizontal plane though, isn't it?
[15:53:10] <elmo40> JT-Hardinge: not very much :P your step is smaller then the width of the tip.
[15:53:22] <SWPadnos> ah right. you're moving a .032 radius curve 0.15 over, so you want to know the height above the chord at +/- 0.007.5
[15:53:43] <SWPadnos> this one may be closer then: http://forum.onlineconversion.com/showthread.php?t=8152
[15:54:02] <SWPadnos> you'll have to flip the equations around, since I think they were looking for R
[15:54:46] <SWPadnos> and remember, the scallop height is (height at the tip) - (height at half the stepover)
[15:59:47] <JT-Hardinge> SWPadnos: yes the heigh above the surface at the middle which should be equal to the height of the segment described by the stepover amount and the tip radius
[16:00:04] <SWPadnos> yep
[16:08:57] <rene> rene is now known as Guest3561
[16:27:07] <JT-Hardinge> thanks, I finally got a right answer from my formula... goes to show you with a little help even a guy with the same education level as an Arkansas virgin can finally figure it out
[16:30:27] <alSMT> alSMT is now known as Al_Smt
[17:35:50] <JT-Hardinge> well, I got my two cubits made... time for lunch I think
[17:47:38] <KimK> Has anyone used ClassicLadder's GRAFCET (or sequential) programming? I'd like to try using it (first-time). I think I understand the general concepts, but I'm having trouble using/clicking some of the "non-ladder" widgets, any advice appreciated.
[17:49:32] <skunkworks> I have not used it. The the clasicladder gui in general takes a bit to get used to.
[17:51:04] <KimK> Hi Sam! Specifically, I'd like to know how do you get any of the "joiners" to connect to any the "steps"? I can make steps appear on the diagram, but I can't make any of the "and", "or", "up", "down", joiners appear.
[17:51:25] <KimK> Yes, I've used the ladder part of CL many times.
[17:53:09] <KimK> At this point, I don't care if it runs correctly, I'd just like to stick a bunch of stuff there and see if I can approximate any of the screenshots.
[17:56:25] <SWPadnos> click until it does what you want or it crashes :)
[17:57:12] <KimK> Haha, good advice! Maybe clicking faster would help? Clickclickclickclickclickclickclickclickclick...
[17:57:30] <SWPadnos> it should help you get to one or the other of those paths quicker
[18:52:00] <elmo40> JT-Hardinge: so... how large is the scallop?
[19:11:18] <X704> X704 is now known as WalterN
[19:11:21] <JT-Hardinge> depends on the tip radius and the feed per tooth!
[19:13:37] <ries_> ries_ is now known as ries
[19:33:22] <IchGuckLive> Hi here from germany
[19:33:52] <psha> IchGuckLive: snow, -3 degrees as i remember? :)
[19:34:40] <IchGuckLive> i wrote a litel GUI for conferting Inch Toolpathes to metric for non itelligent Systems (Where ever they are)
[19:35:23] <IchGuckLive> someone in the US i guess you are on Windows maybe and have some simple NC-Code to test with its only G0/G1
[19:36:14] <IchGuckLive> written in Python TK
[19:36:26] <IchGuckLive> compile thru py2exe
[19:38:57] <i_tarzan> base Python has TK ?
[19:39:46] <IchGuckLive> yes
[19:39:50] <IchGuckLive> tekinter
[19:39:58] <IchGuckLive> tkinter
[19:40:06] <IchGuckLive> http://euroschall.de/dino/inch_to_metric.zip here is the GUI
[19:42:07] <IchGuckLive> please woudt be nice if someone coudt test it it is for HAAS distrib tomorrow they asked for it i simply dont no why someone in the states can do this they are paid well i guess
[19:43:10] <atmega> psha?
[19:46:03] <IchGuckLive> so by its late here got to go programmed since 18:00 its now 20:45 got to work at 05:30
[19:47:00] <psha> atmega?
[19:48:03] <atmega> how do get camview-emc to embed in the window? is there a window argument like the mplayer one?
[19:52:06] <psha> -w {XID}
[19:52:27] <atmega> heh, just like the mplayer one :)
[19:52:32] <atmega> guess I should have tried that.
[19:53:14] <psha> i'm too narrowminded to invent another name for Window parameter :-P
[19:53:30] <atmega> mplayer displays my cam video fine, I have yet to get camview to display anything
[19:53:48] <atmega> it will sometimes display a single garbled frame when switching input types
[19:54:01] <psha> what's your cam?
[19:54:09] <atmega> err... V something :)
[19:55:37] <atmega> can't ssh in for some reason so I can't say
[19:58:23] <psha> try to remember
[19:58:43] <psha> it only shows you UNRECOGNIZED formats? or normal like YUV too?
[19:59:18] <atmega> camview shows choices for various resolutions of inverted YUV?
[20:00:54] <psha> i don't know :) it's yours camera :)
[20:01:26] <atmega> first time I ever saw camview, I wasn't really sure what I was doing.
[20:01:54] <atmega> but, building the chain seemed straightforward enough
[20:02:20] <psha> camview-emc tries to further simplify process by generating chain on the fly
[20:02:31] <atmega> yeah, it did the same thing
[20:02:32] <psha> but that may be a bit buggy now until it will be ironed
[20:02:49] <atmega> all black at 640x480, single garbled frames on everything else
[20:07:13] <atmega> does the 10.04 install default to no services running?
[20:28:30] <psha> atmega: what services?
[21:16:48] <psha> atmega: may you try updated version?
[21:17:07] <atmega> I can when I get home... an hour or so.
[21:17:35] <atmega> my new install doesn't seem to have ssh running
[21:18:47] <SWPadnos> you have to install openssh-server. it's not installed by default
[21:18:58] <atmega> whack... is that new?
[21:19:24] <atmega> All 1660 scanned ports on are: closed
[21:19:40] <SWPadnos> no. ubuntu has been that way for a long time
[21:23:13] <psha> atmega: when you'll get hands on it install new libcamunits6 and camview-emc
[21:23:21] <psha> hope this show you correct image
[21:23:29] <atmega> cool, thanks.
[21:25:28] <atmega> can you add some horizontal and vertical edge finding for probing while you are at it :)
[21:29:15] <psha> what's 'edge finding'?
[21:29:33] <psha> http://en.wikipedia.org/wiki/Edge_detection
[21:29:34] <psha> ?
[21:30:48] <seb_kuzminsky> psha: do you mean in machining, or in image processing like the wikipedia page talks about? they're pretty differentn!
[21:31:20] <psha> seb_kuzminsky: i try to understand what atmega means :)
[21:31:28] <psha> i'm not common with machining :)
[21:32:24] <seb_kuzminsky> #emc may not be the most helpful irc channel for microcontroller programming or image processing questions ;-)
[21:33:08] <psha> for image processing i'm on #opencv :)
[21:35:05] <SWPadnos> psha, has there been much progress on Harpia lately? (do youknow)
[21:35:37] <SWPadnos> I never have managed to get any good Hough circle detection
[21:35:49] <psha> SWPadnos: i'm not active there :) only general programming :)
[21:35:54] <SWPadnos> ah, ok
[21:36:02] <psha> i've ended with home breeded circle detection
[21:36:04] <SWPadnos> yet another channel to monitor :)
[21:36:13] <psha> hough cirles were not working good
[21:36:22] <psha> SWPadnos: it's not active channel :(
[21:36:29] <SWPadnos> I was trying to detect a basketball, more or less. the circle detection seemed to be the closest thing in the toolkit
[21:36:47] <SWPadnos> great! that means I might as well not bother joining ;)
[21:37:08] <SWPadnos> I'm kind of waiting for the version that has GPU support, which I think should be out in a few months
[21:37:31] <psha> to detect everything around? :P
[21:37:39] <SWPadnos> heh
[21:38:09] <psha> i've asked robotics people on my work if they are using some special algorithms to detech markers
[21:38:27] <SWPadnos> so I can do automatic image stabilization of 50-100 large images in realtime
[21:38:47] <psha> they said that nothing difficult :) only manual detection without any hough algos
[21:39:03] <SWPadnos> there are corner detectors that are used often, and you would put a fiducial mark on the things you're trying to detect
[21:41:34] <psha> SWPadnos: i've 'fudicial' in my own task
[21:41:44] <psha> but i don't understand what atmega means :)
[21:41:46] <SWPadnos> ?
[21:41:51] <SWPadnos> ah
[21:42:13] <SWPadnos> edge finding in the machine world means locating the edge of a part or fixture with high precision
[21:42:22] <psha> my own one is not true fudical but specially placed dots (4 of them)
[21:42:30] <SWPadnos> you could use image-based edge detection to do it in some cases
[21:42:55] <SWPadnos> yeah, some form of target is normal - a few rings with a crosshair in the middle
[21:43:30] <psha> i have for dots
[21:43:32] <psha> ..
[21:43:32] <psha> .
[21:43:33] <psha> .
[21:43:36] <psha> like this :)
[21:43:40] <SWPadnos> there are others though - a black blob with a certain number of white blobs inside, each of those having a certain number of black blobs in them
[21:43:44] <psha> but nothing difficult
[21:43:48] <SWPadnos> yep
[21:44:09] <psha> but i have not find any reliable way to detect points
[21:44:18] <psha> maybe i have to do canny and find closed arease
[21:44:24] <psha> i'm using hsv filtering now
[21:44:50] <SWPadnos> well, one day when I start messing with it again, I'll let you know if I figure anything out :)
[21:45:35] <psha> but hsv filtering gives nice results
[21:46:01] <psha> i leave only colors closer to selected one in HSV (not rgb) colorspace
[21:48:42] <grommit> psha: I see you've updated camview-emc, should I install it?
[21:49:28] <psha> wait a bit for next one
[21:49:47] <grommit> what are you adding/fixing?
[21:49:48] <psha> i'm trying to fix 'strange' cams like qq-'s one
[21:49:54] <grommit> ah
[21:50:02] <psha> and some improvements alongside
[21:50:17] <psha> like proper format selection
[21:50:32] <grommit> cool
[21:50:39] <psha> so if you'll try it too and say if it's working or failing it would be great
[21:51:19] <grommit> sure. when will it be ready?
[21:52:01] <psha> ~5 minutes
[21:52:38] <psha> grommit: lucid version is ready
[21:52:46] <psha> please install new libcamunits6 too
[21:52:57] <grommit> that was a quick 5 min :-)
[21:53:16] <psha> version 20
[21:53:25] <psha> i don't know how much time build will take
[21:53:32] <psha> there are several hosts building
[21:53:39] <grommit> will libcamunits6 get updated with camview-emc or I have to do it separately?
[21:54:06] <psha> separately (but in one command maybe)
[21:55:37] <grommit> ok, got libcamunits6 0.2.0-5 and camview-emc 0.0.20
[21:56:30] <psha> that's right
[21:57:20] <grommit> by the way, my update manager (in ubuntu) is came up and says there a new camunits-gtk6-dev and camunits6-dev and libcamunits-gtk6 should I install those?
[21:57:53] <psha> not needed
[21:58:12] <psha> only real chages are in libcamunits6, others are just version bumps
[21:58:38] <grommit> are those other packages used? that is, should I keep them around or remove them from my system?
[21:59:22] <psha> -dev may be removed
[21:59:35] <psha> libcamunits-gtk6 is needed
[21:59:55] <psha> it holds gui part of camunits framework
[22:00:24] <grommit> new one seems to work fine
[22:00:34] <grommit> anything specific I should test>?
[22:00:53] <psha> try to set BGR format
[22:01:14] <psha> change format from one to another
[22:01:22] <psha> maybe sizes
[22:02:38] <grommit> bgr works
[22:03:00] <psha> great, thanks for testing
[22:03:04] <grommit> there is one called UNRECOGNIZED which doesn't
[22:03:14] <psha> it's ok
[22:03:25] <grommit> YUV doesn't seem to either (with my camera)
[22:03:28] <psha> i may filter UNRECOGNIZED in camview-emc so they won't be displayed
[22:04:34] <grommit> If I select a format of say 1280x960 and it works, does that mean the camera is actually displaying that resolution? I don't actually know what this cam supports....
[22:04:50] <psha> at least camera reports that it's possible :)
[22:05:12] <psha> but in most cases so large resolution costs you framerate and image quialty
[22:05:23] <Valen> most webcams are 300k pixels (800x600 or so) and fake higher resolutions
[22:05:34] <grommit> YUV reverse byte works, just not YUV.
[22:05:57] <psha> grommit: there are tons of YUV formats )
[22:06:16] <psha> maybe i'll leave RGB only in camview-emc so users wont' bother with it )
[22:07:07] <grommit> I don't really see a difference between the different formats. I guess I would need to get better lighting and be more critical to tell....?
[22:07:34] <psha> no :)
[22:07:58] <psha> RGB, BGR, not working YUV and UNRECOGNIZED are produced from first group of formats
[22:08:08] <psha> first group are 'native' to camera
[22:08:43] <psha> first groups
[22:08:53] <psha> non-native starts with RGB
[22:09:09] <psha> so difference may be seen only between JPEG/MJPEG and YUV
[22:10:12] <KimK> Well, I am about at my wits end here. I have been trying out ClassicLadder's GRAFCET for the first time and I have failed to master two of the icons. If anyone has any advice or can see something I might try I would appreciate any comments. This is just me pasting icons, it doesn't really run, nor do I expect it to. There are no rungs associated with it, and the steps are imaginary. Any advice appreciated. http://imagebin.ca/view/uSJpwIWe.html
[22:15:56] <grommit> I did a screen shot of each different format and have them side-by-side on the screen. No matter the magnification, or how critically I look at them, I can't tell the diff.
[22:16:18] <psha> grommit: there is no difference as i mentioned
[22:17:05] <grommit> I thought you said there was between mjpg and yuv (for example)?
[22:17:12] <psha> yes
[22:17:23] <psha> there _may_ be difference
[22:17:30] <grommit> ah, may
[22:17:30] <psha> but not large
[22:17:49] <psha> for example on slow links mjpeg resolution may be higher
[22:17:57] <psha> since it fits in USB bandwitch
[22:18:00] <psha> and yuv - not
[22:18:14] <grommit> ok
[22:18:24] <psha> and if you camera has good sensor then jpeg will be slightly worse
[22:19:04] <psha> but it seem all webcam sensors are manufactured from recucled cameras so they are not good )
[22:19:24] <grommit> not surprising
[22:20:05] <psha> thanks again for testing :)
[22:20:18] <grommit> no problem, thanks to you for getting it working!
[22:20:18] <psha> i'll test it tomorrow with gspca_zc3xx camera
[22:23:04] <grommit> What is the url of that readme that you did for installing this stuff? I want to install it on a friends machine..
[22:23:20] <psha> wait a bit
[22:23:27] <psha> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Axis_Embed_Video
[22:23:29] <psha> here it is
[22:23:41] <psha> if you'll add notes there about installation it will be great
[22:24:16] <grommit> I will pay attention while doing it on his machine :-)
[22:24:23] <psha> thanks
[22:24:34] <psha> it's difficult for me since i'm dev and not user :)
[22:24:44] <psha> so i may miss some important points
[22:27:43] <andypugh> jepler / PCW: Is there a 5i23 / 7i39 bitfile prebuilt anywhere?
[22:40:00] <atmega> camview-emc isn't opening a separate window now, but nothing shows up inside the axis window
[22:40:19] <atmega> File "/usr/bin/camview-emc", line 54, in __init__
[22:40:19] <atmega> xid = int(options.window, 0)
[22:40:20] <atmega> ValueError: invalid literal for int() with base 0: 'id'
[22:40:43] <psha> show EMBED_TAB_COMMAND line from ini
[22:41:03] <atmega> EMBED_TAB_COMMAND = camview-emc -wid {XID}
[22:41:13] <psha> -w {XID}
[22:42:03] <atmega> cool, got the (lack of) video in axis
[22:42:32] <atmega> Vimicro USB Venu V4L2 cam
[22:42:47] <psha> install last version of libcamunits6
[22:42:52] <atmega> various resolutions of YUV 4:2:2 reverse byte
[22:42:54] <psha> it's 0.2.0-5
[22:44:08] <atmega> lots more options, but none working
[22:44:16] <psha> in camview-emc?
[22:44:21] <atmega> right
[22:44:27] <psha> run camview
[22:44:33] <psha> select input plugin for your cam
[22:44:36] <psha> select opengl rednerer
[22:44:47] <psha> and select RGB output format
[22:47:13] <atmega> still no video, sometimes a single garbled frame
[22:48:27] <atmega> mplayer says:
[22:48:28] <atmega> VO: [xv] 640x480 => 640x480 Packed YUY2
[22:48:28] <atmega> Selected video codec: [rawyuy2] vfm: raw (RAW YUY2)
[22:48:29] <psha> qv4l2 is working nice?
[22:48:45] <atmega> what is qv4l2?
[22:48:51] <psha> small viewer programm
[22:48:57] <psha> based on libv4l
[22:49:05] <psha> same library is used in camunits
[22:49:14] <atmega> don't have it
[22:49:27] <psha> install it please to check
[22:49:42] <psha> if qv4l2 is working that i've broken something
[22:49:44] <atmega> got a package?
[22:49:52] <psha> yes :) ubuntu got it :)
[22:49:53] <psha> qv4l2
[22:50:06] <atmega> VO: [xv] 640x480 => 640x480 Packed YUY2
[22:50:06] <atmega> Selected video codec: [rawyuy2] vfm: raw (RAW YUY2)
[22:50:09] <atmega> erk...
[22:50:19] <atmega> can't find package qv4l2
[22:51:05] <psha> wait a bit
[22:51:50] <psha> do you have libv4l installed?
[22:52:20] <atmega> yes
[22:52:23] <psha> it seem that they don't package qv4l2
[22:52:26] <psha> strange...
[22:52:31] <psha> mplayer is working fine?
[22:52:35] <atmega> yep
[22:52:45] <atmega> well, as well as I would expect for a crappy webcam
[22:53:01] <psha> what's your version of libv4l-0
[22:53:08] <psha> 0.6.4-1ubuntu1
[22:53:09] <psha> ?
[22:53:49] <atmega> how do I tell?
[22:54:05] <psha> dpkg -l libv4l-0
[22:54:28] <atmega> 0.6.4-1ubuntu1
[22:54:54] <psha> YUY2 format...
[22:55:01] <atmega> odd, huh?
[22:55:33] <psha> let me check logs
[22:56:05] <atmega> gotta go get kids... bbl
[23:01:44] <psha> i'll test it tomorrow on wierd cam too
[23:29:08] <jepler> andypugh: I have a pin file for firmware 5i20/SVTP6_7I39.PIN but haven't released an update of the hostmot2-firmware packages containing it yet.
[23:29:46] <andypugh> I have a 5i20 version, but it won't work with the 400k FPGA in the 5i23
[23:35:53] <andypugh> Well, it might work, but the hm2 driver won't let it work.
[23:45:17] <andypugh> http://www.youtube.com/watch?v=UxPzodKQays&feature=related
[23:46:51] <jepler> andypugh: it should be easy for me to build one for 5i23. give me a few minutes..
[23:47:13] <andypugh> That would be very helpful, thanks
[23:47:39] <jepler> andypugh: warning, though: you'll be a guinea pig for some other changes as well
[23:47:49] <andypugh> Hmm..
[23:47:53] <skunkworks> heh
[23:48:49] <andypugh> I am just being lazy really, I could point-to-point wire an existing TPEN6_6.BIT file but it's quite a lot of wires.
[23:51:44] <jepler> http://emergent.unpy.net/files/sandbox/5i23_svtp6_7i39.tar.gz
[23:52:12] <jepler> if it doesn't so much as load, let me know and I'll roll back these unrelated changes and build another one
[23:59:50] <andypugh> It seems to load