Back
[09:48:17] <jthornton> mozmck,
http://gnipsel.com/linuxcnc/toc/index.html
[09:48:42] <jthornton> this is done without frames or server side includes
[09:51:07] <Tom_itx> poplar bluff has alot to be proud of :)
[09:53:13] <Tom_itx> JT-Shop how's the new doc format going?
[10:11:51] <JT-Shop> pretty much done messing with it for now, waiting on a consensus to merge them or not
[10:18:23] <jthornton> I was just showing mozmck that it can be done on the client side without iframe. Server side includes would not work for us as you don't have a server when viewing local html
[10:21:06] <mozmck> hi, jthornton, that looks good!
[10:22:25] <mozmck> And I notice the location bar shows which page you are on, which should satisfy archivist and seb_kuzminsky with regard to sending people links.
[10:22:47] <JT-Shop> hi mozmck I'll combine the expanding collapsing and make another sample
[10:23:02] <JT-Shop> yea, I didn't like the fact that the url was not right clickable
[10:23:27] <JT-Shop> and it has to work as a local copy
[10:23:31] <mozmck> yes.
[11:07:42] <seb_kuzminsky> awesome
[11:08:23] <seb_kuzminsky> that new toc is really nice JT-Shop
[11:11:05] <JT-Shop> thanks seb_kuzminsky
[11:11:21] <JT-Shop> I'll add expanding collapsing to it for a demo
[11:11:58] <mozmck> I'm trying to use motion.spindle-speed-out-abs for something totally non-standard and it almost works :)
[11:13:12] <mozmck> I need a way to send commands to my custom component from Gcode, that take effect at certain points in the path, without pausing motion.
[11:14:07] <mozmck> S codes are what we have been using in "that other software for windows", so I'm trying to do it in linuxcnc the same way if possible.
[11:14:59] <mozmck> The only problem I can see right now is that I need to execute S commands even if the "spindle" (plasma torch really) is off, but motion.spindle-speed-out-abs goes to 0 and stays there if spindle is off.
[11:15:35] <mozmck> Is there a way to make motion.spindle-speed-out* show commanded spindle speed all the time?
[11:19:52] <seb_kuzminsky> mozmck: i'd expect not
[11:20:56] <seb_kuzminsky> you could use m62/m63/m67, and remap s so it does those instead?
[11:21:59] * mozmck is looking those up...
[11:24:15] <mozmck> Hmm, not sure if that would work. We use quite a range of S codes to send various commands as well as numbers.
[11:26:29] <mozmck> So for instance, S3000 to S3255 tells the component to set the tip volts to <s-code> - 3000
[11:27:13] <mozmck> S3100 would set the tip volt set point to 100 volts.
[11:33:01] <seb_kuzminsky> you could do the exact same thing with m67
[11:34:04] <mozmck> I better look closer...
[11:34:22] <seb_kuzminsky> you could use M67 E0 Q100, with motion.analog-out-00 connected to tip voltage and the Q-word appearing on that pin
[11:34:32] <seb_kuzminsky> you have E0 through E3 for different outputs
[11:34:46] <seb_kuzminsky> and you could multiplex them if you had to, like you did with the S-word
[11:35:00] <mozmck> interesting, and I presume those do not pause motion?
[11:35:11] <mozmck> M67 that is?
[11:35:11] <seb_kuzminsky> right
[11:35:41] <mozmck> Ok, I may try that. We can do that instead, we just change our sheetcam post to generate the code we need.
[11:35:48] <seb_kuzminsky> yay :-)
[11:35:52] <mozmck> post = post-processor.
[11:36:09] <mozmck> thanks for the tip!
[11:36:23] <seb_kuzminsky> JT-Shop: how does your nifty toc interact with the toc that asciidoc makes? i'm guessing your toc is somehow added as a post-processing step, after asciidoc generates the html?
[11:37:46] <JT-Shop> it's a js that uses document.write to write the toc to the document when it is loaded
[11:37:56] <JT-Shop> I'll upload the code in a bit
[11:39:10] <JT-Shop> this is the part that inserts the toc, so it would need to be in each page.
[11:39:11] <JT-Shop> <div id="nav">
[11:39:11] <JT-Shop> <script type="text/javascript" src="nav.js"></script>
[11:39:11] <JT-Shop> </div>
[11:40:50] <JT-Shop> the toc is not in the page source but the links are right clickable to open in a new page or copy the link
[11:41:10] <seb_kuzminsky> i feel like a bit of a fossil for not having learned js yet
[11:41:26] <mozmck> seb_kuzminsky: I notice M68 says it breaks blending. does that mean that it will pause motion?
[11:42:46] <mozmck> One problem with the HTML docs is that they reference sections that are not there. In the M67 docs, "See the Integrator’s Manual, Core Components Section, Motion subsection, for more information." But I don't see an Integrator's Manual :)
[11:47:52] <seb_kuzminsky> i think m68 is wrong for you, since the output takes effect immediately, whereas m67 is synchronized with the surrounding motion commands
[11:48:56] <seb_kuzminsky> good point about the busted pseudo-links... the thing that's meant by "Integrator's Manual" is the "Configuring LinuxCNC" part of the main webpage,
http://linuxcnc.org/docs/2.7/html/
[11:49:20] <seb_kuzminsky> or the Integrator Manual pdf doc, which is supposed to have the same info as that section
[11:50:01] <seb_kuzminsky> so that pseudo-link refers to this document:
http://linuxcnc.org/docs/2.7/html/config/emc2hal.html
[11:51:44] <JT-Shop> stuff like this in the User Manual can be fixed Feed Override - which allows you to scale the speed of programmed motions. The default maximum is 120% and can be set to a different value in the ini file. See the Integrator Manual for more information on this setting.
[11:52:02] <JT-Shop> instead of see the x manua there can be alink
[11:52:04] <JT-Shop> a link
[12:06:24] <seb_kuzminsky> JT-Shop: you mean those things can be turned into actual links after the 2.7_Docs branch combines all the docs into one big file?
[12:41:39] <cradek> I bet m67 et al are not well tested in 2.7 yet
[12:48:51] <seb_kuzminsky> i wish i had finished the standalone-motion tester 6 months ago
[12:59:40] <mozmck> I just tested using the MDI and I can enter M67 E0 Q3120 and then G0 X0.5 and the voltage set-point changes only when I run the G0 line.
[12:59:52] <mozmck> So that looks right so far.
[13:00:17] <mozmck> I can use M68 and it works right away as it should.
[13:07:29] <cradek> cool
[13:07:31] <cradek> in 2.7?
[13:07:35] <mozmck> yes
[13:07:53] <mozmck> I'll test it some more on a real machine soon hopefully
[13:09:50] <cradek> that's great
[13:10:05] <seb_kuzminsky> thanks mozmck
[13:10:59] <mozmck> If I tried to run a line and it failed, the M67 did not take effect - even when I ran another line that worked. That is probably correct behavior I think.
[13:11:30] <mozmck> The G0 X1 failed at first because the Y position was past it's soft limits.
[13:15:03] <seb_kuzminsky> it's an exceptional situation if a commanded motion doesn't happen, and it's not clear to me what the M67 should do in that case
[13:15:17] <seb_kuzminsky> skipping it is probably not more wrong than executing it
[13:15:33] <mozmck> It would seem safest to not execute it to me.
[13:16:02] <mozmck> The code quits executing as well.
[15:00:56] <kwallace> OT: I have a clock problem. My Micronta VoxClock went dead. Has anyone fixed one lately?
https://www.youtube.com/watch?v=o48gl67pNp0 . One of the chips is:
http://htfe.com.au/store/product_info.php?products_id=5928
[15:13:39] <kwallace> Hmm, maybe use this as a reference to replace the insides of the clock:
http://www.jsykora.info/talking-clock/
[16:21:11] <cradek> what troubleshooting have you done?
[16:22:45] <pcw_home> most common problem I have with small stuff is electrolytic capacitors drying out in the device or wall wart
[16:23:03] <cradek> or battery corrosion
[16:23:13] <pcw_home> yeah
[16:23:35] <cradek> bet that's 25-30 yrs old (I had one a long time ago)
[16:23:42] <kwallace> I just followed the battery voltage to the board so far.
[16:23:50] <cradek> ICs don't die of old age as much as caps do
[16:23:54] <cradek> what else is on the board?
[16:24:11] <cradek> does the button still work right?
[16:24:13] <kwallace> I'll take a picture...
[16:31:18] <kwallace> http://wallacecompany.com/tmp/IMG_2171-1a.JPG
[16:36:00] <kwallace> The button seems to ground a signal when pressed so seems to work.
[16:36:07] <cradek> did you do the electrolytics already?
[16:37:04] <kwallace> No I haven't touched them yet.
[16:37:42] <cradek> I'd do that and also check all the solder joints with a microscope or loupe
[16:39:35] <cradek> it makes no noises at all?
[16:39:42] <cradek> what's the led usually do? I don't remember it having one
[16:41:06] <kwallace> It doesn't have a hole in the case, so isn't used.
[16:42:33] <cradek> huh, weird
[16:43:03] <cradek> would be nice to know if it's an audio problem or a clock problem
[16:44:40] <kwallace> I'll need to get the scope out, no time now though.
[16:46:02] <kwallace> The led flashes once when I connect the batteries.
[16:46:49] <pcw_home> might check if the 32 KHz oscillator is running (usually a not-fet scope probe will stop it on the input side though)
[16:51:46] <pcw_home> Years ago when we made CPU cards occasionally the 32 KHz TOD clock xtals would fail because
[16:51:47] <pcw_home> the forks got swinging too widely (or they were off center in the case) they would hit the case.
[16:51:49] <pcw_home> These cards could be detected by listening closely, they would make a quiet ping ping ping at
[16:51:50] <pcw_home> ~5 second intervals when the forks hit the case and motion was damped
[16:57:47] <cradek> I have fixed very old digital watches by replacing the crystal
[16:58:03] <cradek> I would have never thought to listen for it!
[17:12:21] <pcw_home> We didnt either, it was more like "what the heck is that noise"
[18:29:09] <mozmck> hmm, I must be missing something. I connected motion.spindle-forward to an output on my 7i92, but when I run M3 in the MDI motion.spindle-forward does not change and natually the output does not turn on.
[18:36:50] <mozmck> Oh, hmm. The docs say, "Spindle speed Sn must be set to a positive non-zero value for M3/M4 to turn on spindle motion."
[18:37:03] <Tom_itx> did you set s?
[18:37:17] <mozmck> Is that new? I don't remember having to do that in 2.4
[18:37:21] <Tom_itx> sry, you figured that out as i was typing
[18:37:43] <Tom_itx> where's it gonna go if S is 0?
[18:37:43] <mozmck> Or 2.3 - whatever my router table is running.
[18:38:19] <mozmck> Well, for plenty of router tables and etc, it's either on or off - S is irrelevant
[18:39:09] <Tom_itx> maybe they tweaked the logic a bit on newer versions
[18:39:26] <mozmck> Could be.
[18:40:00] <Tom_itx> my sherline was on / off for the longest time but i still specified a S word
[18:40:05] <Tom_itx> just from habbit
[18:40:32] <Tom_itx> habit*
[18:42:10] <mozmck> Well, looks like the gcode I've been running has an S1000 with the M03, so I guess I have as well. I mostly use sheetcam and I guess it puts it in there with the emc post
