Back
[00:13:40] <r00t4rd3d> boring
[00:22:08] <adjen> hi
[00:22:09] <adjen> i discover a new great exploit to get root take a look
http://ro0t.dyndns-server.com/
[00:25:11] <r00t4rd3d> lol
[00:25:23] <r00t4rd3d> retard
[02:21:03] <R2E4> What can a cylindrical grinder do that a lathe cannot?
[02:43:37] <archivist> fine finish
[03:13:21] <DJ9DJ> moin
[03:24:50] <mrsun> do a good job on hardened materials? :P
[06:43:20] <DJ9DJ> re
[08:40:37] <JT-Shop-2> yea... found my DVD drive on my desk
[09:09:05] <JT-Shop> yippe I even figured out how the tool less brackets work
[09:19:05] <DJ9DJ> grats ;)
[09:40:23] <IchGuckLive> hi all B)
[10:34:12] <IchGuckLive> happy Easter to all of You by till later !
[11:22:57] <JT-Shop> Tom_itx: I'm not sure how many times it will dump, I assume you can reprogram the chip?
[12:14:28] <andypugh> I just had a rather random thought. My Ikea lights are 5V / 0.75W. Well under the current capacity of the 7i64. So I can easily turn the lights on and off with GPIO. So far so sensible.
[12:15:12] <andypugh> But, if i can do that, I can turn the lights on-and-off synchronised to the spindle. So I could freeze the work to see what was happening :-)
[12:15:36] <jthornton> cool
[12:17:00] <pcw_home> JT-Shop: AVRs are flash and I think good for 10,000 reprograms
[12:20:07] <pcw_home> I ordered this to do some LED as strobe light testing:
[12:20:09] <pcw_home> http://dx.com/p/diy-60w-4000-5000lm-square-white-led-module-dc-33-35v-150675
[12:20:24] <syyl_ws> thats very safe to work, andypugh ;)
[12:20:30] <syyl_ws> "oh spindle is stoped"
[12:20:34] <syyl_ws> *touch*
[12:20:41] <syyl_ws> blood goes everwhere
[12:20:44] <syyl_ws> :(
[12:21:01] <pcw_home> with horror movie strobe light effects
[12:21:06] <andypugh> There would be that minor drawback, yes.
[12:21:29] <syyl_ws> that might be interesting, pcw_home ;)
[12:21:34] <andypugh> Perhaps I should use red LEDs to disguise the blood?
[12:21:41] <syyl_ws> great idea!
[12:22:06] <L84Supper> be sure to post a youtube with a creepy soundtrack
[12:23:03] <andypugh> Now, where can I buy key-steel on a bank holiday at 5pm?
[12:23:18] <pcw_home> I was going to use a stepgen phase locked to the motor to get minimum jitter and controllable pulse width
[12:23:51] <archivist> andypugh, you make!
[12:23:52] <andypugh> Or, for that matter, some coolant?
[12:24:51] <andypugh> Ah, archivist. Last time I was in Cromwell Tools they had Castrol Cooledge on a special offer at £60 for 20 litres. I don't think I need 20 l
[12:25:42] <archivist> some stuff is just plain overpriced
[12:26:23] <andypugh> That's a lot cheaper than normal..
[12:26:24] <archivist> I tend to use an oil can and whatever leftovers I have
[12:27:15] <andypugh> Hysol is £160 for 20l.
[12:27:45] <andypugh> Or you can buy it in 208L drums, for £1588!
[12:27:57] <archivist> the other week I was mixing axle/gear oil with hydraulic(automatic gearbox) and engine oil
[12:28:32] <L84Supper> I do the same in my waste oil tank
[12:29:22] <L84Supper> do you have to pay a recycler in the UK to deal with it?
[12:29:51] <archivist> supposed to yes
[12:30:50] <archivist> or get a recycler to collect for free or take to the council tip or be imaginative
[12:30:56] <JT-Shop> pcw_home: thanks
[12:30:59] <L84Supper> Walmart takes 5 gallons / day for free here
[12:31:27] <andypugh> I quite like the idea of running with neat oil for coolant. It ought to go mouldy more slowly.
[12:31:27] <archivist> do they heat their store with it
[12:31:52] <archivist> we used neat oils at the clockworks
[12:32:59] <archivist> just lose a bit, stays clean looking though
[12:36:25] <L84Supper> krap that funky EMCO CNC lathe on ebay sold
[12:36:26] <syyl_ws> running oil as a cooleant is a bit horror in my mind
[12:36:40] <syyl_ws> oil mist everywhere
[12:36:46] <syyl_ws> whole machine gets covered with oil
[12:37:12] <andypugh> Sounds good to me, less rust.
[12:37:16] <syyl_ws> saw a few index lathes that run with cutting oil
[12:37:18] <pcw_home> machine and operator well lubed
[12:37:24] <syyl_ws> looks like crap after a few years :D
[12:37:39] <syyl_ws> i want no oil mist in my lungs ;)
[12:37:52] <syyl_ws> i like the ability to breathe
[12:39:06] <pcw_home> must be an explosion hazard if too misty
[12:39:51] <syyl_ws> machines that run with cutting oil need a automatic fire extinguisher
[12:39:53] <L84Supper> http://www.ebay.com/itm/I-D-Grinder-Precision-Work-Head-w-Control-Box-/251215802710?pt=LH_DefaultDomain_0&hash=item3a7da0f156
[12:39:55] <archivist> if the cutting speed is low there is no mist
[12:39:56] <syyl_ws> at least here in germany
[12:40:34] <syyl_ws> like swiss type lathes
[12:55:52] <L84Supper> http://tinyurl.com/bo2bgda super deal if you're on the east cost US
[13:06:36] <IchGuckLive> hi all B)
[13:22:51] <syyl_ws> one year older than me, L84Supper
[13:22:52] <syyl_ws> :D
[13:22:56] <syyl_ws> but same country
[13:26:42] <IchGuckLive> you shoudt meat and great etch other O.O
[13:28:15] <syyl_ws> i had meat for lunch
[13:38:45] <Tom_itx> JT-Shop you can reprogram those many times
[13:53:51] <IchGuckLive> someone uses sheetcam on linux
[13:55:18] <Tom_itx> anyone recomend a fast opto for parport interface?
[13:56:09] <Tom_itx> http://www.digikey.com/product-search/en?x=11&y=13&lang=en&site=us&KeyWords=H11A817C-ND
[13:56:21] <Tom_itx> that was one i found but i'm not sure how quick it needs to be
[14:00:56] <IchGuckLive> tom why are you building your own
[14:01:11] <IchGuckLive> there are so many good stuff for cheep price
[14:01:24] <Tom_itx> somebody was asking me about optos
[14:01:33] <Tom_itx> i'm not making one
[14:01:56] <IchGuckLive> i go downstairs and look onto my card what is on there
[14:03:27] <IchGuckLive> 10Mbit optos 6n137
[14:04:45] <syyl_> i went also for the 6n137
[14:04:57] <IchGuckLive> 50cent
[14:05:01] <syyl_> on my mesa doughtercards
[14:05:02] <syyl_> https://dl.dropbox.com/u/24396704/Deckel%20G2/2011-12-14_19-58-49_984.jpg
[14:05:09] <L84Supper> Tom_itx:
http://www.avagotech.com/pages/home/ search by speed, number of channels and amount of isolation
[14:05:42] <IchGuckLive> syyl_ your own build
[14:05:52] <syyl_> yes
[14:05:58] <IchGuckLive> why no holes
[14:06:10] <syyl_> goes in a 19" rack
[14:06:46] <IchGuckLive> do you cheer the layout
[14:06:55] <syyl_> what?
[14:07:10] <IchGuckLive> PCB routing of the BOB
[14:07:41] <IchGuckLive> are you using the 74 14
[14:08:43] <syyl_> i dont get you question
[14:09:12] <IchGuckLive> right board the IC infront of the Optos
[14:09:21] <IchGuckLive> 74 HCT14
[14:10:13] <syyl_> those are max3095
[14:10:32] <syyl_> for the glas scales that are rs422
[14:10:42] <IchGuckLive> ah
[14:10:58] <IchGuckLive> so not a subD25 BOB
[14:11:24] <syyl_> nope
[14:11:36] <syyl_> its for a mesa 5i20
[14:13:55] <JT-Shop> Tom_itx: does it make sense to just use an AVR with more I/O for a few cents more?
[14:15:07] <IchGuckLive> JT-Shop: why to run a µC on logics
[14:15:33] <JT-Shop> can you translate that to redneck speak?
[14:15:55] <syyl_> :D
[14:16:22] <roh> JT-Shop: depends on what you want to do. translate some rs422 protocol or what are you thinking about?
[14:16:44] <IchGuckLive> im off By happy easter to you all !
[14:16:51] <roh> you too :)
[14:17:04] <JT-Shop> roh: is it a simple counter really and very slow
[14:17:22] <syyl_> seems that he wont talk to you, JT-Shop ;)
[14:17:42] <JT-Shop> lol
[14:17:47] <JT-Shop> I get that all the time
[14:17:56] <roh> something like that can be easy be done in an avr. basically a divider then? (high clock in, and one for every n out)
[14:18:12] <syyl_> that is a great skill :D
[14:19:12] <JT-Shop> yea, a count up to some number, clear the count buffer and turn on an output
[14:19:26] <JT-Shop> at the next count turn off the output
[14:21:07] <roh> should be easy to do
[14:21:39] <roh> like a programmable 'decimal counter' (available in the 74xxx series) ;)
[14:22:36] <JT-Shop> yes
[14:48:30] <JT-Shop> the 74xxx series is an AVR?
[14:48:48] <cradek> nonono
[14:48:48] <archivist> no
[14:49:06] <cradek> AVR is a microcontroller (mini programmable computer)
[14:49:19] <cradek> 74xx are very basic logic gates, like "and2"
[14:49:26] <JT-Shop> ah ok
[14:49:28] <JT-Shop> it
[14:49:33] <JT-Shop> 's all new to me
[14:49:37] <cradek> a few are a little less basic than that, but none are programmable
[14:49:54] <archivist> and counters like 74ls161 etc
[14:50:21] <cradek> yeah they're almost exactly like our various hal components
[14:50:26] <archivist> hmm define programmable
[14:50:44] <cradek> well the simple ones anyway
[14:50:48] * cradek kicks archivist under the table
[14:53:19] * archivist points cradek at a 74S181 alu and some roms
[14:54:42] * Jymmm shoves archivist into a flip flop with no clock!
[14:54:54] <archivist> bit like linuxcnc...anything is possible
[14:56:35] <JT-Shop> lol
[14:58:16] <Jymmm> JT-Shop: 74 Series are TTL
http://en.wikipedia.org/wiki/7400_series, and it's counterpart are the 4000 Series are CMOS
http://en.wikipedia.org/wiki/4000_series
[15:02:44] <PCW> yeah a couple 4017s and 1 and gate would do
[15:04:44] <Jymmm> JT-Shop: what are trying to do?
[15:05:36] <JT-Shop> Jymmm: counting rain
[15:05:50] <Jymmm> rain?
[15:06:50] <archivist> rain guage with a tipping bucket /me presumes
[15:08:29] <JT-Shop> aye
[15:08:41] <cradek> JT-Shop:
https://www.surpluscenter.com/item.asp?item=21-1728
[15:09:20] <andypugh> Cavers have been known to use tipping buckets to convert a constant dribble into an occasional deluge, which can, over time, open up mud-chocked passageways.
[15:10:24] <andypugh> I I just used my Chinese vernier protractor for the first time. It is precise to 2 minutes, and accurate to half a degree.
[15:10:45] <andypugh> ie, placed on a surface plate it meaures half a degree.
[15:10:55] <cradek> ew
[15:11:57] <archivist> glad someone else wasted their money trying one for me :)
[15:13:27] <andypugh> I think it was this one:
http://www.ebay.co.uk/itm/RDGTOOLS-UNIVERSAL-VERNIER-BEVAL-PROTRACTOR-/290853679446?pt=UK_Home_Garden_PowerTools_SM&hash=item43b83af156
[15:14:22] <andypugh> It is quite nice in the way that the bits go together, ideal for measuring the angle of the table on my mill. Just a pity it doesn't measure it very well.
[15:14:49] * JT-Shop hands Andy a tuning hammer
[15:14:51] <archivist> is it fixable
[15:14:58] <archivist> file
[15:15:11] <archivist> and an angle grinder
[15:15:47] <Jymmm> archivist: you forgot the plasma cutter
[15:17:17] <andypugh> Actually, looking through my records, it wasn't the RDG one, which may be better. It was this one: (but note I paid £18, not the current list price which is very odd)
http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=280879046231&ssPageName=ADME:L:OC:GB:3160
[15:18:29] <archivist> almost looks adjustable
[15:18:59] <andypugh> Given the £16.99 + postage price of it, there isn't a lot to be lost from trying.
[15:19:39] <andypugh> I need to slot-out the holes on the vernier bit.
[15:19:41] <archivist> check other angles first
[15:19:55] <archivist> may be a scaling error
[15:20:50] <andypugh> I doubt I will ever need more than 5 degrees. I wanted it for setting up for hobbing. Larger angles can use the table graduations, normally.
[15:23:14] <Tom_itx> JT-Shop it's up to you which avr you get. If you want to experiment with other stuff i'd get a bigger one. If you don't then the attiny13 is fine.
[15:23:40] <Tom_itx> If you do, i'd suggest a mega 48, 88, 168 or 328 for a good all around chip to play with
[15:24:04] <Tom_itx> all the same, just more flash and ram
[15:25:12] <Tom_itx> that code i wrote does just what you described
[15:25:27] <andypugh> Talking of hobbing, I am all set up and ready to go tomorrow:
https://plus.google.com/photos/108164504656404380542/albums/5747722155741347649/5860860184697043874
[15:25:33] <JT-Shop> like this?
http://www.digikey.com/product-detail/en/ATMEGA48A-PU/ATMEGA48A-PU-ND/2271048
[15:25:54] <Tom_itx> yeah but for a few cents more you can get the 168 or 328
[15:27:04] <JT-Shop> ok
[15:31:17] <andypugh> Looks like a pretty rigid setup on the hob, compared to the mini-mill :-)
[15:32:45] <Tom_itx> http://www.digikey.com/product-detail/en/ATMEGA168A-PU/ATMEGA168A-PU-ND/2271215
[15:32:53] <andypugh> Unfortunately, as what I am making is the pulley to drive my X-axis, I will have to feed it by hand.
[15:33:29] <Tom_itx> http://www.digikey.com/product-detail/en/ATMEGA328-PU/ATMEGA328-PU-ND/2271026
[15:34:50] <Tom_itx> you probably won't need that much flash though for what you're doing or will any time soon
[15:35:42] <Tom_itx> those are the chips the smaller arduino use
[15:35:56] <cradek> JT-Shop: if I understand right that you're just going to be counting a switch trigger event, my suggestion was serious
[15:36:09] <Tom_itx> where's the fun in that though?
[15:36:27] <Tom_itx> you could do it with logic pretty easy
[15:36:31] <cradek> the simplest way to solve a problem is often the best
[15:36:42] <Tom_itx> it's more fun to program a chip and see the results
[15:36:43] <andypugh> cradek: Nonsense!
[15:36:55] <cradek> haha
[15:37:35] <Tom_itx> count to x and turn a pin on, count to x again and turn it off
[15:37:38] <Tom_itx> that's all he's doing
[15:37:50] <cradek> ok, let me state it more narrowly: people often overestimate the problems for which a microcontroller is the best answer :-)
[15:38:06] <cradek> I thought he was counting bucket dumps
[15:38:10] <Tom_itx> he is
[15:38:24] <Tom_itx> using the input from it to increment x
[15:39:10] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/cnc/JT-SHOP/Attiny13/
[15:39:14] <Tom_itx> there's the code for it
[15:39:46] <Tom_itx> i added a couple things not knowing how he wanted the output
[15:41:41] <andypugh> cradek: I agree in principle. But It is also nice that with a uP when the customer realises what he really wanted, and changes his mind, it is just a software change.
[15:42:01] <cradek> yeah that's sure a lucky situation when it happens
[15:42:15] <andypugh> For example, this bucket-gauge probably wants to give the answer in inches.
[15:42:16] <cradek> same if I'm the customer myself
[15:42:30] <Tom_itx> and the cost of µC's nowdays it's cheaper alot of the time
[15:42:56] <Tom_itx> then he'll realize he can have an lcd on it and want that added in...
[15:44:00] <cradek> I guarantee that the combination of a uC and display and the components to interface them is > $4.99. a power supply and switch you'd need either way...
[15:44:09] <andypugh> I designed a trip-cock tester for London Underground. We dicided to go with simple electrics, DIN-rail timers, relays etc. In the end it cost a lot more than a PC and monitor woud have done, and then when they discovered that their Victorian test spec was _stupid_ it was impossible to reconfigure it.
[15:44:23] <cradek> it's true a uC often beats any nontrivial discrete logic pricewise
[15:45:15] <Tom_itx> heck, let's stuff it in a fpga :)
[15:45:28] <Jymmm> ASIC
[15:46:11] <Tom_itx> add a hal component and he can have his annual rainfall displayed on his mill
[15:46:18] <Jymmm> JT-Shop: Wireless Rain Gauge $1.98
http://www.lowes.com/pd_396643-53921-00813_4294857782__?productId=3830973&Ns=p_product_avg_rating|1&pl=1¤tURL=%3FNs%3Dp_product_avg_rating|1&facetInfo=
[15:46:24] <JT-Shop> thanks Tom_itx
[15:46:25] <Jymmm> JT-Shop: it's WIRELESS!!!!!
[15:46:27] <andypugh> One part of the test-spec was to apply 100psi to the opened valve to ensure that it didn't get blown closed, This was a 1.25" valve. We had the stupid situation where my 20,000 cuft/hr piping (4" diameter) and vast regulator was hooked up in the same place as their old tester, to a 3/4" air main.
[15:46:48] <JT-Shop> won't wor for my application
[15:46:51] * Tom_itx slaps Jymmm.. where's the fun in that?
[15:47:00] <Jymmm> JT-Shop: but but but... it's WIRELESS!!!!!
[15:47:33] <Tom_itx> JT-Shop, it never hurts to have a few of any of those laying around...
[15:47:41] <andypugh> I love that part. Truly wireless, not a single wire anywhere near it.
[15:47:46] <cradek> andypugh: does it even make sense to talk about psi "applied to" an open valve...?
[15:47:58] <Jymmm> andypugh: =)
[15:48:11] <cradek> haha, wireless
[15:48:27] <andypugh> cradek: My machine could flow enough air for there to be 100psi pressure drop across the valve.
[15:48:33] <cradek> wow
[15:48:47] <andypugh> Which is clearly what the test was _meant_ to be about.
[15:49:14] <Tom_itx> Jymmm, you could mount a pressure sensor to the bottom and measure the water level
[15:49:30] <Jymmm> JT-Shop: $28
http://www.amazon.com/AcuRite-00614-Digital-Self-Emptying-Collector/dp/B000W0JTCI/ref=sr_1_19?s=lawn-garden&ie=UTF8&qid=1364589324&sr=1-19
[15:49:52] <andypugh> They just had never realised that that meant a 2000hp compressor. And I was young, inexperiened, and just assumed that they must have the required air supply as they had been running the test for 100 years.
[15:50:15] <Tom_itx> andypugh we had a screw compressor that would probably do that
[15:50:34] <andypugh> Yes, it isn't like the compressors don't exist.
[15:50:35] <Tom_itx> it was in a room by itself
[15:51:00] <Jymmm> JT-Shop: that one has a cup and really is wireless
[15:51:19] <Tom_itx> we haven't heard his full application for it yet
[15:51:30] <Jymmm> JT-Shop: crap, poor reviews
[15:52:37] <andypugh> The previous tester (that mine was bought to replace, and which was then re-installed when the tripcocks couldn't pass the test on my machine) had a needle-valve restrictor upstream of the valve under test, and downstream of the gauge, and their test was to open the tripcock, then adjust the needle valve to 100psi...
[15:53:32] <Jymmm> JT-Shop: Better reviews, $53 wireless
http://www.amazon.com/AcuRite-00614-Digital-Self-Emptying-Collector/dp/B000W0JTCI/ref=sr_1_19?s=lawn-garden&ie=UTF8&qid=1364589324&sr=1-19
[15:54:01] <andypugh> JT's application is to build a rain-gauge and get paid for it, I think. Finding $28 alternatives on Amazon is _not_ what he needs. Jymmm: Find a $5000 one.
[15:54:26] <Jymmm> $130
http://www.rainwise.com/products/detail.php?ID=6819
[15:54:38] <Jymmm> Meets NWS Specs
[15:54:52] <Tom_itx> it's got an opto on the output pin so he's driving something with it
[15:55:50] <JT-Shop> Tom_itx: it is a cell control for lack of a better description
[15:56:06] <JT-Shop> I just have to give it an input for x number of tips
[15:56:12] <Jymmm> JT-Shop: rainfall datalogger
http://www.rainwise.com/products/detail.php?ID=6819
[15:56:23] <JT-Shop> I don't need to log it
[15:56:29] <Tom_itx> gime some parts and i can probably whip up a board for it
[15:56:44] <R2E4> Are cylindrical grinders that precise?
[15:56:45] <JT-Shop> ok, cool
[15:56:51] <Tom_itx> unless you're using perfboard for it
[15:56:53] <Jymmm> JT-Shop: ok wired with counter
http://www.rainwise.com/products/detail.php?ID=6697&Category=Rain_Gauges:Wired&pageNum_cart=/products/index.php
[15:57:06] <JT-Shop> I can use perf board for the prototype
[15:57:14] <JT-Shop> I have to run now
[15:57:16] <Tom_itx> heck just use a breadboard for that
[15:57:17] <JT-Shop> bbl
[15:57:28] <JT-Shop> yep, got one of those
[15:57:35] <JT-Shop> thanks Tom_itx
[15:57:39] <Tom_itx> np
[15:57:44] <Jymmm> JT-Shop: rain guage interface
http://www.rainwise.com/products/detail.php?ID=6831&Category=Rain_Gauges&pageNum_cart=/products/index.php
[15:57:55] <Jymmm> DIN mount 24V
[15:58:02] <R2E4> JT-Shop: You have a cylindruical grindder?
[16:00:06] <andypugh> R2E4: Probably the most precise of all machine tools.
[16:01:08] <R2E4> I know this guy who has a few and some surface grinders also that I can get for next toi nothing
[16:01:28] <R2E4> So it wiould be worth it to arrange pickup?
[16:13:09] <chopper79> Hello, Does anyone have any information about JA3 fir Linuxcnc? I need to give this a shot instead of using gantrykins.
[16:13:34] <cradek> hey
[16:13:48] <cradek> do you have a more specific question?
[16:13:58] <cradek> are you asking how to build it?
[16:15:42] <chopper79> Gantrykins handles a tandem axis very poorly and I need to be able to auto-square the tandem axis with ease. Also I have heard that the JA3 branch has fixed a lot of the known issues with gantrykins. I need some advice on how to go about using the JA3 fix for my machine.
[16:16:17] <cradek> so the precise thing you are trying to accomplish is auto squaring?
[16:16:53] <chopper79> Yes, in a nut shell
[16:18:10] <cradek> I am not a fan of gantrykins and I would attempt autosquaring by getting into the guts of the homing process at the hal level. are you squaring to switches or encoder index?
[16:18:30] <chopper79> Auto squaring when homing and the keeping joint mode hidden. All I should see is world mode when using the machine. I have gantrykins running now but it a painful process keeping or making it work reliably. I figured I would give JA3 a shot.
[16:18:34] <cradek> or hard stops?
[16:18:34] <chopper79> Switches
[16:18:45] <chopper79> Swithes hitting hard stops
[16:18:47] <cradek> do you have encoder feedback, or open loop?
[16:18:51] <chopper79> open
[16:19:33] <chopper79> I can do it using a parallel port and trivkins no problem. I am trying out a MESA board and would liek to use a kins module instead of gates as I do now.
[16:19:35] <cradek> I'd use basic old trivkins. use a very low velocity for your last homing move so you don't need acceleration control. then mask the step pulses to the switch that's not hit yet.
[16:20:13] <cradek> why do you want to change that?
[16:20:38] <cradek> mesa vs software stepgen is not important here, let's talk just about homing
[16:21:12] <PCW> That's the issue, it not practical to gate step pulses with the hardware stepgen
[16:21:14] <chopper79> I do not have eyes on the stepgen using the hostmot like I do parallel.
[16:21:42] <cradek> oh I see, you were using internal (hal) gates
[16:21:54] <chopper79> correct
[16:22:08] <chopper79> So I figured this JA3 fix would potentially do the trick
[16:22:27] <cradek> well you're adding a lot of complexity and unknowns
[16:22:41] <cradek> by all means try it, I guess
[16:22:44] <chopper79> True, but gantrykins kinda does the same
[16:22:55] <chopper79> the complexity and issues I mean
[16:22:59] <PCW> Michael G seems to think it fixes the issues and he's been using it for a long time
[16:23:33] <cradek> how does he autosquare?
[16:24:00] <cradek> I don't see how trivkins helps you autosquare but I may be missing something
[16:24:05] <cradek> err I meant gantrykins
[16:24:50] <cradek> do you just start the two joints homing at the same time and let them work separately?
[16:25:19] <PCW> [00:42:36] <micges> in ja3 there is gentrivkins that will hide gantryness
[16:25:21] <PCW> [00:43:10] <micges> it is trivkins with ability to map axes to joints
[16:25:23] <PCW> [00:43:39] <micges> with it it will never be in joint mode (no ability to move joint from gui)
[16:25:25] <PCW> [00:44:25] <PCW> how does that help gantry homing?
[16:25:26] <PCW> [00:44:33] <andypugh> But homing is a joint-thing not an axis thing?
[16:25:28] <PCW> [00:44:57] <PCW> Oh I see joints for homing only
[16:25:29] <PCW> [00:45:23] <micges> yes
[16:25:31] <PCW> [00:45:34] <micges> just home all button
[16:25:32] <PCW> [00:45:49] <PCW> does that work now?
[16:25:34] <PCW> [00:46:18] <micges> it was 1.5 year ago iirc
[16:25:40] <cradek> (if ja3 really works we oughta merge it)
[16:25:42] <chopper79> They have the same home seq. and then after homing/squaring is doen you have to switch to world mode and then gantrykins slaves the Y and A or the axis you slaved.
[16:25:45] <cradek> oh it's NOT gantrykins as we know it
[16:26:04] <cradek> that sounds interesting
[16:26:43] <cradek> chopper79: ok I understand (but you have a terminology error: it's never A)
[16:26:57] <chopper79> 3
[16:27:06] <cradek> and it's also not gantrykins you want, sounds like
[16:27:59] <PCW> I think a gantry homing comp could work but its not trivial
[16:28:17] <cradek> especially for index it would seem to get quite hairy
[16:28:35] <cradek> gantry support likely needs to be built into motion
[16:28:58] <PCW> Yes that would be best
[16:29:20] <andypugh> R2E4 If you have the room for the grinders, then there are some jobs that only they can do.
[16:29:30] <cradek> you really want more than independent approximately-simultaneous home sequences. you want each stage of homing to wait for both sides to sync up before the next stage starts
[16:29:31] <andypugh> Modifying bearings is one thing.
[16:29:49] <chopper79> Correct
[16:30:03] <cradek> so each stage improves the alignment, not just the last one
[16:30:28] <cradek> there is not currently an easy solution for this. I'm not sure there's a hard one except to go ahead and put it in the motion controller.
[16:30:30] <chopper79> I can do this with trivkins by gating the stepgen commands. It has always worked well and no issues.
[16:30:57] <cradek> you could pretty easily do the gateing in hardware
[16:31:12] <PCW> but that is so M^2L
[16:31:36] <cradek> pcw could (not so easily at all) add a step-pulse-gating input
[16:32:41] <cradek> but that's so far from the general solution we really want that it would be silly to do
[16:32:50] <PCW> violates accel/ violates steplength yuck
[16:32:55] <cradek> yep
[16:33:16] <cradek> everything is nasty about it except that it would probably work (like his current solution with software stepgen does)
[16:33:58] <cradek> you're simply not going to get aligned to one step pulse anyway. you need encoder index to get it perfectly square
[16:34:53] <chopper79> When I home my machine currently I will be within a .001 of an inch of square all the time everytime.
[16:35:30] <chopper79> Been that way for about 3 years now.
[16:35:33] <cradek> what is your overriding goal in changing away from this setup?
[16:35:53] <PCW> just need avoid accumulating errors
[16:36:04] <chopper79> Trying new stuff and I heard the MESA baords were the way to go for plasma
[16:37:03] <chopper79> Every config I have tried using a parallel port for plasma and THC, never works consistant.
[16:37:47] <chopper79> The Z axis is never compensated properly when making up/down moves to keep the voltage correct.
[16:38:44] <chopper79> So after a few minuts or less the tip of the torch is dragging on the material and the float switch will hit and trip and estop.
[16:38:51] <chopper79> Z axis walks I guess I shoudl say
[16:40:36] <cradek> interesting, without knowing how the THC works I hesitate to guess how to fix that
[16:41:43] <AR_> got my .029 and .035 drills
[16:41:52] <AR_> ready to make some PcBs
[16:42:11] <PCW> Ready to break some drills?
[16:42:14] <chopper79> Takes the raw arc and converts it to a basic up/down output signal based on the threshold of the THC unit.
[16:42:17] <AR_> :D
[16:42:29] <cradek> bbl
[16:45:34] <tjtr33> PCW, reading watchdog.vhd. when 'others' is used by itself, "obus <= (others => 'Z')" is just shorthand for all the bits in obus?
[16:46:39] <PCW> yes its a funny degenerate case
[16:47:09] <tjtr33> thats pretty amusing :0
[16:47:18] <PCW> (avoids having to specify a range)
[16:47:43] <tjtr33> gotcha
[16:47:45] <tjtr33> thx
[16:49:13] <andypugh> chopper79: The Mesa THC converts arc voltage to pulse frequency. You can actually count the pulses with a software encoder counter via the parport. But having an analogue signal to do "things' with is a big plus. For example you can cahnge the target voltage in software, on the fly.
[16:49:52] <andypugh> But considering how much a 5i25 costs I really can't see the point in messing about with a parport.
[16:51:04] <chopper79> Andy- the one THC I have is what I was explaining. The MESA one I have also and have not tried it out yet, unitl I get this qantry figured out.
[16:51:57] <andypugh> Trying JA3 isn't any harder than trying any other unreleased branch.
[16:52:41] <chopper79> I understand, just looking for a real solution to the gantry fun.
[16:53:37] <andypugh> gentrivkins does sound like a possible candidate.
[16:53:51] <chopper79> never heard of it
[16:54:01] <andypugh> it's part of JA3
[16:54:22] <chopper79> I dont even know how to get the JA3 ..... embarresed to say that
[16:54:26] <andypugh> It's a trivial kins, so you don't see joint mode / word mode.
[16:54:48] <andypugh> No need to be embarassed, it isn't at all obvious.
[16:55:01] <andypugh> Do you know how to build LinuxCNC from source?
[16:55:07] <chopper79> I am willing to give it a try
[16:55:09] <chopper79> no
[16:57:08] <andypugh> OK.
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Installing_LinuxCNC#Installing_to_Ubuntu_10_04_or_8_04_from_source
[16:57:36] <andypugh> But section 2.2 is the relevant part.
[16:58:05] <chopper79> I have tried that before and failed
[16:58:16] <chopper79> I dont think my branch name was right
[17:00:15] <andypugh> 2.2 to 2.5 should leave you with a working LinuxCNC as a RIP (Run in Place) system. That means that your normal installation stays where it is, and that is what menu items and icons will open. But if you open a terminal, "cd linuxcnc-dev/ " then type ". ./scripts/rip-environment" and then "linuxcnc" you will run your freshly compiled version instead.
[17:02:29] <andypugh> The branch you want is probably called remotes/origin/joints_axes3
[17:02:54] <andypugh> But I would try building and running Master (the current dev version) first.
[17:03:37] <chopper79> The current master is version?
[17:03:52] <andypugh> 2.6 (sort of)
[17:03:55] <chopper79> At this moment I am running 2.5.0
[17:05:53] <chopper79> I need to type v2.6 and not v2.5 correct
[17:09:22] <andypugh> No. Master is called master.
[17:10:14] <chopper79> git branch --track master_branch origin/master_branch
[17:10:25] <andypugh> The current branches that things get done in are called "master" and "v2.5_branch"
[17:10:57] <chopper79> so by leaving v2.5 I will get the master
[17:13:14] <andypugh> git is a big and complicated thing. There are more detailed instructions here:
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Git
[17:13:23] <andypugh> But most of that is not likely to be relevant.
[17:14:30] <andypugh> If you install git, then clone the repository, you should be able to compile a version of linuxCNC that works. I suggest trying that first, and then worry about exactly which branch it is later :-)
[17:14:53] <chopper79> Just finihed the git and clone
[17:14:58] <chopper79> finished
[17:15:41] <andypugh> OK, if you cd to linuxcnc-dev and type "git branch" what does it say?
[17:16:12] <chopper79> master
[17:16:44] <DJ9DJ> gn8
[17:16:47] <andypugh> OK, try compiling that.
[17:16:53] <andypugh> cd src
[17:17:05] <andypugh> ./autogen.sh
[17:17:15] <andypugh> ./configure
[17:17:17] <andypugh> maks
[17:17:27] <andypugh> make I mean.
[17:17:47] <andypugh> I am not safe to be let loose with a command line :-)
[17:28:29] <chopper79> lol.... Ok I ma done with that
[17:28:33] <chopper79> am
[17:29:02] <chopper79> How do I launch this now to work in it?
[17:29:07] <chopper79> or am I not there yet
[17:29:26] <andypugh> You compiled master, and no errors?
[17:29:46] <chopper79> error... I need pythion
[17:29:50] <chopper79> python
[17:30:14] <chopper79> sorry
[17:30:36] <andypugh> I can never type python either. I always under-shoot and end up with pythin. Looks like you manage the mash both keys.
[17:30:39] <chopper79> error: boost::python is required to buid linuxcnc
[17:30:55] <chopper79> wth...
[17:31:05] <chopper79> I did not type a icon
[17:31:23] <chopper79> :P
[17:31:26] <chopper79> that explains it
[17:31:33] <chopper79> : p
[17:31:57] <andypugh> sudo apt-get install libboost-python-dev
[17:32:55] <chopper79> done
[17:33:06] <andypugh> You have a fast PC!
[17:33:16] <andypugh> OK, run ./configure and make again
[17:33:47] <chopper79> done and no error
[17:33:55] <andypugh> make?
[17:34:19] <chopper79> make then sudo make setuid
[17:34:35] <skunkworks> http://www.cnczone.com/forums/controller_computer_solutions/145107-choosing_motion_controller.html
[17:34:46] <andypugh> I would be expecting that to take at least 10 minutes.
[17:35:17] <chopper79> so after make, run sudo make setuid
[17:38:15] <andypugh> Yup
[17:38:29] <andypugh> then ./scripts/rip-environment
[17:38:40] <andypugh> then linuxcnc
[17:38:45] <andypugh> (at last)
[17:40:22] <andypugh> skunkworks: So Spngledboy (a name I recognise from the forums) pops up to say how happy he is with LinuxCNC, and in a reply some dude says "well, I am going to go with the more expensive solution that no-one has ever tried"?
[17:43:46] <chopper79> Andy- Thank you and I will give this a go. I may need to call upon your expert advice later. Right now I need to go send some lead down range.
[17:44:07] <chopper79> I will get lost and tryt o dig myself out first.
[17:45:50] <chopper79> real quick anytie I want to make a change or run master I just needed to do it in terminal or can I launch it from the desktop ?
[17:48:23] <andypugh> It depends what you want to do. You can use ./configure --prefix /usr and that actually installs the compiled version over the top of your installed/released version. Or you can make an icon/shortcut to launch the run-in-place version.
[17:49:17] <chopper79> ./scripts/rip-environment = no such directory
[17:49:47] <andypugh> I actually have the ./scripts/rip-environment in my .bashrc so I never need to type it. If I open a terminal, then I am automatically configured run-in-place. But that is just to suit me, and might not suit you.
[17:50:02] <andypugh> pwd ?
[17:50:14] <chopper79> no
[17:50:14] <andypugh> ie, where are you?
[17:50:23] <chopper79> src directory
[17:50:33] <chopper79> linuxcnc-dev/src
[17:50:37] <chopper79> $
[17:51:09] <andypugh> scripts is one level down. . ../scripts/rip-environenment from there
[17:51:30] <andypugh> that's dot space dot dot slash. It's like morse code.
[17:52:40] <chopper79> ok im lost now
[17:53:19] <chopper79> I typed ./scrpits/rip/rip-environment and no such directory
[17:53:30] <chopper79> do I cd to scrpits?
[17:53:53] <andypugh> the scripts directory is in linuxcnc-dev. two dots takes you down a level.
[17:53:53] <chopper79> or is it . ../
[17:54:04] <andypugh> cd to scripts is probably easier.
[17:54:24] <andypugh> then its just ./rip-environment
[17:55:12] <andypugh> the leading dot is some sort of safety thing. I think it makes it harder for hackers to inject commands.
[17:55:34] <chopper79> I can not chnage to that directory
[17:55:50] <chopper79> cd scripts
[17:55:54] <andypugh> where are you ? (pwd)
[17:56:21] <chopper79> I am under src still
[17:56:32] <andypugh> cd ..
[17:56:41] <andypugh> cd scripts
[17:57:04] <andypugh> . ./rip-enviroment (that's dot space dot slash)
[17:57:28] <andypugh> and it is spelt rip-environment
[17:57:46] <andypugh> but you can use tab to auto-complete after "rip"
[17:57:54] <Tom_itx> ok JT-Shop when you get this, i have a newer version of winavr you can install and it will compile for the attiny10 if you decide to use one of those
[17:58:05] <chopper79> ok that is done
[17:58:10] <Tom_itx> it's the version that comes packaged with studio 6
[17:58:21] <Tom_itx> atmel offers it as a standalone
[17:58:24] <chopper79> sorry Andy I am new to terminal and its commands
[17:58:48] <chopper79> little overwhelming
[17:59:27] <andypugh> yeah, it's all moderately novel to me too, so i do at least understand the issues.
[17:59:57] <andypugh> so, if you now type "linuxcnc" it ought to run, with a splash-screen that says 2.6
[18:00:11] <chopper79> well to me you are a pro
[18:00:49] <chopper79> it poped up config selector but no splash screen
[18:01:04] <andypugh> splash scrren is next, I think
[18:01:13] <andypugh> Choose a config
[18:01:17] <chopper79> yup... I had to pick a config
[18:01:21] <chopper79> 2.6 pre
[18:01:43] <andypugh> Congratultions, you just built LinuxCNC from source.
[18:01:50] <chopper79> Whoot
[18:02:07] <chopper79> Thanks I had a good couch
[18:02:15] <chopper79> lol
[18:02:41] <chopper79> coach
[18:03:06] <chopper79> so now when I make a config I can access all the extra?
[18:03:40] <andypugh> You now have the new stuff from Master, but that isn't the same new stuff as JA3
[18:04:07] <chopper79> oh
[18:04:13] <andypugh> JA3 is actually rather behind Master(2.6) but is different.
[18:04:45] <andypugh> Your configs won't work in JA3 without changes.
[18:04:56] <chopper79> I am all about changes
[18:05:09] <andypugh> i wanted to make sure you had a working setup first
[18:05:19] <chopper79> I do
[18:05:22] <chopper79> Thank you
[18:05:47] <andypugh> Firstly I would suggest making a copy of your complete working config (ie, the whole folder)
[18:06:43] <andypugh> Then rename the folder and the INI file something like "gantry_ja3" or similar. The INI file name is what you see in the config picker, in case you hadn't spotted that.
[18:06:46] <chopper79> At this moment I do need to go so I will read some more tonight and have a go at it this weekend. IS there a way for me to make actual changes to linuxcnc such as my own version complied with my own configs? Just a quaestion.... not going to as I could not even go back one directory
[18:07:20] <andypugh> Yes, now that you have the source, you can make absolutely any changes you want.
[18:07:34] <chopper79> wow
[18:07:44] <chopper79> just have to know whrre to go I guess
[18:07:47] <chopper79> where
[18:08:12] <andypugh> Exactly, that's the hard part. You can call every HAL pin "Tarquin" now if you want. It wouldn't help, but you could.
[18:08:47] <chopper79> all through terminal or is there a interface to work in?
[18:09:33] <andypugh> Hardcore coders use vi or a basic text-editor.
[18:10:11] <andypugh> For what you want to do, you probably don't need to edit the code, just check-out the ja3 branch and compile it as it is.
[18:10:20] <chopper79> I would use gedit if I did
[18:10:37] <chopper79> so to get the JA3 branch I need to go throught the same thing as I just did
[18:11:16] <andypugh> if you do find you want to make more extensive changes then I quite like Anjuta when I tried it. Though I actually do my LinuxCNC coding on a Mac using XCode. (But I am the only one)
[18:12:03] <andypugh> Getting ja3 is something like git checkout origin/joints_axes3 then make again.
[18:12:22] <chopper79> ok... I will check it out. Thanks again and i will let you know my progress
[18:12:45] <andypugh> I would give you the exact commands, but my Git is in the middle of other stuff at the moment.
[18:12:59] <chopper79> ok... I have to go anyway
[18:13:05] <chopper79> later and thanks agin
[18:13:10] <chopper79> again
[18:13:20] <andypugh> No trouble, and welcome to the slippery slope to being a LinuxCNC dev :-)
[18:13:46] <chopper79> well I fell and busted my rear on this...lol
[18:14:09] <andypugh> 2 years ago I was _exaclty_ where you are now.
[18:14:30] <r00t4rd3d> in the uk?
[18:15:19] <andypugh> Well, that too, but lost in Linux was what I meant.
[18:18:23] <r00t4rd3d> not sure what i did on the internet before i found reddit
[18:20:19] <andypugh> Did it involve a lot of fleshtones?
[18:43:43] <r00t4rd3d> I restored an old Stanley hand plane today :)
[18:55:23] <t12> whee
[18:55:27] <t12> ok lathe now leveled
[18:55:31] <t12> down the rabbit hole of metrology
[18:58:24] <andypugh> I like hand planes.
[18:59:32] <t12> https://www.dropbox.com/s/bnv6btft6kaqxw5/_E121536.JPG
[18:59:33] <r00t4rd3d> i dont, i just need to use one
[19:00:31] <r00t4rd3d> and it was all rusted to fuck
[19:15:04] <andypugh> t12: I want that big chisel
[19:17:20] <andypugh> I made a scratch tool for this, a router-cut groove would have looked wrong. It only took about an hour to make the tool and form the profile:
https://plus.google.com/photos/108164504656404380542/albums/5854209973166221953
[19:18:58] <andypugh> (the first 5, the second 4 are something entirely more modern, this in fact:
http://benlovejoy.wordpress.com/2013/03/27/idesk/ )
[19:18:59] <Tom_itx> what kind of wood is that?
[19:19:05] <andypugh> Oak
[19:19:16] <andypugh> (in both cases)
[19:19:55] <syyl_> i restored a new vise today ;)
[19:19:58] <syyl_> (china)
[19:21:02] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1748.JPG
[19:21:47] <syyl_> not all parts are perfect...
[19:21:48] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1752.JPG
[19:22:18] <syyl_> or the gib...a bit bent
[19:22:19] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1758.JPG
[19:22:34] <andypugh> I am never sure if those are going to be stiff enough to be useful. I am sure that for some jobs nothing else will do (short of a Deckel with the universal table)
[19:22:50] <syyl_> i will use it on my tool/cutter grinder
[19:22:59] <syyl_> for turning tools
[19:23:41] <andypugh> syyl_: OK, let me take a photo..
[19:31:51] <andypugh> https://plus.google.com/photos/108164504656404380542/albums/5857486442507317137/5860923848007151682
[19:33:07] <syyl_> no permission to open it..
[19:35:07] <andypugh> Hmm, how about:
https://plus.google.com/photos/108164504656404380542/albums/5747722155741347649/5860924673167260434
[19:35:32] <syyl_> that works
[19:36:06] <syyl_> your lathe needs some serios cleaning ;)
[19:36:42] <syyl_> that protractor is nice
[19:36:53] <syyl_> but 0 seems to be off ;)
[19:37:14] <Tom_itx> his table is warped
[19:38:10] <andypugh> The protractor is not nice _because_ the zero is off.
[19:38:20] <Tom_itx> that is a nice tool to have though
[19:38:22] <Tom_itx> not adjustable?
[19:38:33] <syyl_> with a surface grinder ;)
[19:38:35] <andypugh> Not enough,
[19:38:41] <syyl_> that would make a nice gift to somebody you dont like
[19:39:27] <andypugh> It would be a really nice tool, if it actually measured angles.
[19:40:00] <syyl_> for my part, i am cured with cheap measuring tools :D
[19:40:11] <andypugh> And maybe it would be accurate if I were to subtract 24' from every measurement.
[19:40:37] <syyl_> i only buy chinese stuff that i can fix ;)
[19:41:14] <syyl_> ah
[19:41:20] <syyl_> one of the knuckles spotted
[19:41:21] <syyl_> http://www.cncecke.de/forum/attachment.php?attachmentid=79614&d=1364600448
[19:41:33] <syyl_> ah sorry
[19:41:33] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1768.JPG
[19:42:27] <andypugh> I would relieve the centre in a lathe, then scrape the outer 13mm.
[19:43:06] <syyl_> i took the "scrape all of it"-card
[19:43:19] <syyl_> two passes and it got much better
[19:43:19] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1769.JPG
[19:43:25] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1770.JPG
[19:43:53] <syyl_> at least it stopped to rock on the surface plate
[19:44:23] <syyl_> a bit later it started to look like something
[19:44:24] <syyl_> https://dl.dropbox.com/u/24396704/Reime%20Universalschraubstock/IMG_1772.JPG
[19:45:32] <andypugh> I have a feeling (and no science) that a high-pressure clamping to an outer ring is stiffer than an all-over clamp.
[19:45:51] <syyl_> there might be something to that
[19:45:57] <andypugh> I think it is definitely stiffer in torsion.
[19:47:29] <syyl_> i think i will relieve the base in the center
[19:47:41] <syyl_> i can chuck it easier on the face plate
[19:47:59] <andypugh> This looks nicer than my chinese thing, but I am not sure it will measure my angle.
http://www.ebay.co.uk/itm/Grey-Rushton-Vernier-Protractor-As-Photo-/160934608730?pt=UK_Measuring_Tools_Levels&hash=item2578733f5a
[19:48:46] <syyl_> it has a nice wooden box at least
[19:48:59] <syyl_> i have one that looks like that one
[19:49:14] <syyl_> with the sliding rule
[19:49:28] <andypugh> I need to measue over a step. Let me see if I can find a picture.
[19:52:31] <andypugh> https://plus.google.com/photos/108164504656404380542/albums/5747722155741347649/5648298987857956962 If you ignore what it was a picture of, the angle is that split line between just under the words "when spiral milling". It is clamped by the nut you can see. The whole X-axis rotates. My measurement is the angle over that step.
[19:53:25] <syyl_> uh
[19:53:27] <syyl_> ok
[19:54:05] <syyl_> cant you setup a dial indicator in a known distance to the swivel point?
[19:54:11] <syyl_> and then calculate the angle
[19:54:14] <andypugh> Now, the Chinese thing does that no problem, as the setsquare part slides.
[19:54:39] <andypugh> And yes, a DTI is probably the proper way to measure.
[19:54:53] <syyl_> or a sine bar + dial indicator
[19:55:01] <syyl_> (magnetic sine bar)
[19:55:14] <andypugh> Ah. No. The whole X axis swivels...
[19:55:25] <syyl_> but Y stays
[19:56:02] <syyl_> http://www.ballbach-messzeuge.de/06/60/Bilder/6040001.jpg
[19:56:21] <syyl_> you can set it up "standing" on the short side
[19:56:32] <syyl_> and run the y-axis back and forward
[19:56:46] <andypugh> Yes, an accurate square on the Y would work, bit then I have to clock the base of the square to get it acurately on X, then swivel X and clock Y
[19:57:57] <andypugh> Which will work, but is effort, and requires a nice square, which I don't currently have.
[20:01:57] <syyl_> hmm
[20:02:08] <syyl_> there seem to be a lot ways to do it
[20:02:14] <syyl_> but my bed calls for me
[20:06:05] <t12> anyone know a reasonable source for a 4.5MT to 3MT sleeve?
[20:06:08] <t12> (or 2?)
[20:06:47] <t12> apparently clausing sells one but have to call
[20:06:56] <ds3> what uses a MT4.5?
[20:07:03] <t12> a whole mess of lathes, apparently
[20:07:24] <t12> the story is that 4.5 was made so that some convient size of collet? could easily be fit
[20:07:32] <t12> cause 4 is too small and 5 would be too big
[20:07:45] <ds3> I wonder if that's what those 5C capable lathes are since they have an adapter sleeve
[20:07:50] <t12> yeah
[20:08:05] <t12> (5c is the thing forums mention in ref to this)
[20:08:13] <ds3> Oooh
[20:08:20] <ds3> why not a 5C nose + a 5C to MT adapter?
[20:08:37] <t12> this would be for testbar mount
[20:08:42] <t12> and really, anything morse mount
[20:09:27] <andypugh> t12: CNC lathes turn tapers reallly well.
[20:09:38] <t12> however, i have no cnc lathe
[20:09:53] <andypugh> So, make one.
[20:10:06] <t12> heh
[20:10:09] <t12> making the cnc lathe
[20:10:11] <t12> to make the adapter
[20:10:13] <t12> for the non cnc lathe
[20:10:32] <andypugh> Even my ghastly chinese machine with steppers can make a really good socket-taper pair.
[20:10:53] <andypugh> Much better than I ever managed on a manual lathe.
[20:11:17] <t12> i guess in theory my next alignments should be
[20:11:23] <t12> headstock/tailstock to eachother + bed
[20:11:33] <t12> but thats like 2 testbars and the 4.5 adapter
[20:13:26] <t12> or fix the teeth grind now without those adjustments
[20:13:37] <t12> and just eyeball it off the chuck badly
[20:14:34] <andypugh> t12: Just machine a bar between centres and measure it.
[20:15:55] <andypugh> Ah, no, that's a different test. Still worthwhile (that is how you align a tailstock)
[20:19:11] <andypugh> Machining a bar held only in the chuck and measuring the diameter tells you how well the spindle axis is aligned to the bed. A bad chuck affects that, but indirectly. You will get ovality that varies along the test piece. But if the work is round, but tapered, you have a good chuck and a misaligned spindle.
[20:25:52] <t12> i know right now the chuck is way off
[20:25:56] <t12> due to me messing up previous grinding
[20:26:09] <t12> i'm clamping jaws diffntly (a way archivist suggested)
[20:27:05] <t12> last time they ended up sorta conical
[20:31:43] <DaViruz> andypugh: flex will affect that a lot for any meaningful length
[20:32:49] <DaViruz> better of just chucking a "good" bar and indicating on it twice, rotating it 180 degrees to compensate for chuck inaccuracy
[20:37:10] <t12> i also sort of wonder if the conical thing
[20:37:16] <t12> was due to wear on the grinding bid?
[20:37:18] <t12> bit?
[20:39:56] <andypugh> DaViruz: Yes, but, he needs to find a 4.5MT good bar, and that sounds non-tivial.
[20:41:28] <DaViruz> 4.5MT?
[20:42:08] <DaViruz> oh morse taper?
[20:44:51] <t12> maybe a way to look at it is
[20:45:03] <t12> does the error induced by having to 4.5 -> adapter -> test bar
[20:45:12] <t12> make aligning off a test bar not worth it
[20:45:45] <andypugh> It mght, it might not. How do you tell?
[20:46:04] <DaViruz> by rotating it and taking several measurements
[20:46:22] <andypugh> If you machine a bar, is it cylindrical? That is all that really matters
[20:46:47] <DaViruz> if you can rotate something it's trivial to tell taper from non-paralellism
[20:47:03] <DaViruz> if you machine a short bar and a long bar you will get different tapers
[20:47:15] <DaViruz> if you machine bars of different materials you will get different tapers
[20:47:53] <DaViruz> workpiece flex due to tool load will be significant compared to headstock alignment
[20:48:21] <andypugh> A very fat bar might tell you more,
[20:49:19] <DaViruz> yeah sure, if you can keep the length-diameter ratio under 3:1 or something like that it will probably give good numbers
[20:50:06] <t12> well
[20:50:17] <t12> i'm going to attempt to get the jaws closer to proper
[20:50:56] <andypugh> Can you borrow a very good chuck?
[20:51:44] <t12> maybe
[20:51:47] <t12> this is the chuck i have to use anyway
[20:51:53] <t12> if i have to grind it now, and again later thats ok
[20:51:59] <t12> assuming i dont run out of tooth heh
[20:52:02] <t12> i dont think i will though
[21:14:00] <t12> https://www.dropbox.com/sc/iao1y21fmv63vqb/KdgmDVhNmH
[21:14:20] <t12> anyone see any fatal flaw i'm about to go through with here? :)
[21:27:37] <andypugh> It looks sensible. I am trying to think how expanding out into a ring is different. It would certainly use the wrong side of the scroll and teeth, so this way is better in that sense.
[22:01:58] <t12> also before
[22:02:03] <t12> i had the ring kinda on the tip
[22:02:07] <t12> due to not having any other option
[22:02:09] <t12> so it may have rocked
[22:02:13] <t12> and maybe the scroll was dirty
[22:02:21] <t12> we'll see how this works out in about 10m when its done
[22:03:28] <t12> i dtid my previous grind
[22:03:37] <t12> it was very concentric, but conical
[22:03:48] <t12> and when i switched to archivists method
[22:03:57] <t12> one tooth was significantly changed
[22:04:01] <t12> a few thou
[22:14:12] <r00t4rd3d> http://www.youtube.com/watch?feature=player_embedded&v=0ilMx7k7mso#!
[22:18:08] <BHSPiMonkey> Just completed my first successful cut! \o/
https://sphotos-a.xx.fbcdn.net/hphotos-ash3/644754_4480350937476_1533830_n.jpg
[22:19:40] <andypugh> I am getting some kind of Escher impossible perspective thing.
[22:19:55] <andypugh> Laser?
[22:21:39] <AR_> BHSPiMonkey, please do not tell me you are making a triforce thing
[22:22:15] <BHSPiMonkey> AR_: please do not tell me you object
[22:22:22] <AR_> LOL
[22:23:44] <jdh> looks like burnt mdf?
[22:24:18] <BHSPiMonkey> the edges are charred, yeah
[22:24:29] <BHSPiMonkey> not sure yet how to prevent things like that
[22:24:42] <AR_> laser?
[22:24:54] <BHSPiMonkey> DW660 cutting tool
[22:25:01] <BHSPiMonkey> this be a shapeoko
[22:25:08] <AR_> ohh
[22:25:19] <BHSPiMonkey> <- amateur hour
[22:25:32] <andypugh> I think you need to move faster.
[22:25:34] <BHSPiMonkey> andypugh: sorry, I just realized you were talking to me
[22:25:50] <AR_> yeah usually burning = sitting too long in one spot
[22:25:53] <AR_> increase feedrate
[22:25:55] <BHSPiMonkey> andypugh: I'll try my next project with a faster feed rate and see what happens
[22:25:59] <BHSPiMonkey> yeah
[22:26:10] <andypugh> But it's a balance beween breaking bits and burning the wood.
[22:26:26] <BHSPiMonkey> this is my first, and I have no idea what are good rates/plunge depths to use with this tool/material/bit
[22:26:41] <BHSPiMonkey> so I'm using 100mm/min I think
[22:27:14] <AR_> what i usually do is
[22:27:18] <AR_> guess
[22:27:22] <AR_> and hope for the best
[22:27:28] <BHSPiMonkey> 5000mm/sec
[22:27:34] <BHSPiMonkey> GO
[22:27:34] <AR_> if it seems too slow, use feed rate override
[22:27:44] <AR_> or too fast
[22:27:45] <AR_> etc
[22:28:37] <BHSPiMonkey> override in linuxcnc, I'm assuming you mean?
[22:29:30] <andypugh> on my mlll the feedrate overide goes to 100x, I have so little clue. Lathes I know, mills are new.
[22:34:40] <Tom_itx> BHSPiMonkey,
http://www.onsrud.com/plusdocs/Doc/index.html?model.code=FeedSpeeds
[22:34:56] <Tom_itx> http://www.daycounter.com/Calculators/GCode/Feed-Rate-Calculator.phtml
[22:35:10] <Tom_itx> http://zero-divide.net/index.php?page=fswizard&shell_id=199&load_tool_id=27011
[22:35:47] <Tom_itx> and finally
[22:35:49] <Tom_itx> http://www.custompartnet.com/calculator/milling-speed-and-feed
[22:44:30] <t12> awesome
[22:44:34] <t12> it appears to have ground in great
[22:53:36] <r00t4rd3d> BHSPiMonkey, what size shank does that take? 1/8 ?
[22:53:48] <r00t4rd3d> nvm
[22:54:08] <BHSPiMonkey> Tom_itx: thanks for the links
[22:54:20] <r00t4rd3d> BHSPiMonkey, get some of these
[22:54:21] <r00t4rd3d> http://www.ebay.com/itm/Carbide-Router-Bit-3-20mm-125-1-8-Shanks-NEW-Solid-Carbide-CNC-PCB-/160816395593?pt=LH_DefaultDomain_0&hash=item2571677549
[22:54:38] <r00t4rd3d> they will cut wood, plastic, aluminum
[22:55:13] <r00t4rd3d> you probably have a rotozip type bit in there now?
[22:55:37] <r00t4rd3d> kinda looks like a reverse drill bit
[22:57:01] <BHSPiMonkey> r00t4rd3d: nice. I'm using the bit that came with the DW660 currently, though I also have these:
http://bhspitmonkey.tumblr.com/post/33417961997/endmills-are-here-based-on-a-recommendation-from
[22:57:33] <r00t4rd3d> the ones on the right will do better
[22:57:44] <BHSPiMonkey> haven't used any yet
[22:57:51] <BHSPiMonkey> not confident enough :)
[22:58:09] <r00t4rd3d> the bit you are using is wrong
[22:58:14] <r00t4rd3d> its made for drywall
[22:58:31] <r00t4rd3d> or very thin wood
[22:59:33] <r00t4rd3d> DW660 Includes:
[22:59:34] <r00t4rd3d> 1 drywall cutting bit
[22:59:46] <r00t4rd3d> take that out and through it in the garbage
[22:59:51] <BHSPiMonkey> heh
[23:00:40] <BHSPiMonkey> so what characteristics make it so wrong to use? shape of the tip, flute configuration/direction? (I know it's not carbide and that doesn't help)
[23:01:09] <r00t4rd3d> they are made to grind drywall not cut wood
[23:01:23] <r00t4rd3d> its how they are fluted
[23:01:58] <r00t4rd3d> put the bit all the way to the right in your case in that bitch and cut something
[23:02:04] <BHSPiMonkey> heh
[23:02:21] <r00t4rd3d> night and day different
[23:02:26] <r00t4rd3d> no smoke
[23:02:27] <BHSPiMonkey> I might at least keep using this bit for the mirror image counterpart of the piece I just finished cutting, just so they match
[23:03:07] <r00t4rd3d> or you could break something cause the bit is dragging and not cutting
[23:03:20] <BHSPiMonkey> but I see what you're saying. this one's doing a lot of stupid rubbing instead of chopping
[23:03:30] <BHSPiMonkey> hence the heat
[23:03:59] <r00t4rd3d> if you are cutting wood and its smoking, gernerally some isnt right
[23:04:03] <r00t4rd3d> something
[23:04:15] <r00t4rd3d> alittle smoke is okay
[23:04:24] <BHSPiMonkey> well, there wasn't any visible smoke or anything
[23:04:34] <BHSPiMonkey> it just smelled a little like burning wood
[23:04:35] <r00t4rd3d> vacuum?
[23:05:13] <BHSPiMonkey> I sat there and held a vacuum hose by the bit to draw in the particles and keep them from floating out into the room
[23:05:30] <r00t4rd3d> sucking up the smoke too
[23:05:36] <BHSPiMonkey> but even when I put the vacuum down, it wasn't smoking
[23:06:10] <BHSPiMonkey> also I did the job in 1mm steps, which it was able to handle without any visible dragging or pain
[23:06:36] <r00t4rd3d> i bet when you cut with the right bit it sounds totally different
[23:06:38] <BHSPiMonkey> I might try 1.5 next time? idk
[23:07:11] <BHSPiMonkey> 99.995% of the sound during the job came from the tool simply being powered on :) coming into contact with the wood didn't change the sound much anyway
[23:07:45] <BHSPiMonkey> it made some additional noise when it bumped the sides of the toolpath a couple times
[23:07:57] <BHSPiMonkey> but yeah, this tool is a screamer
[23:11:43] <r00t4rd3d> those tiny bits with the orange collar i doubt you will ever use
[23:11:51] <r00t4rd3d> they for like pcb milling and shit
[23:12:27] <t12> does anyone make a cutting fluid that contains something that smokes at specific temps
[23:13:12] <r00t4rd3d> any oil will smoke sooner or later
[23:15:50] <t12> i mean more like
[23:15:57] <t12> some indicator to not cut too hot
