#linuxcnc | Logs for 2012-03-24

Back
[00:18:33] <Nick001> I noticed hardinge used resisters with all the hall effect micro switches - limit switches on the axes
[00:18:56] <Nick001> I just hooked 12v to them and it worked ok
[00:20:47] <pfred1> take a 50 ohm half watt resistor and put it across a car battery and tell me how that works for you :)
[00:23:39] <Nick001> fried of course - these are inline with the 12v - not hooked to gnd
[00:28:14] <ssi> god he's still beating that horse?
[00:28:17] <ssi> it's been almost an hour
[00:30:17] <Jymmm> tenderized bbq horse
[00:33:19] <Jymmm> Though, if I tenderize that long, I'd just have shredded meat
[00:33:44] <ssi> speaking of shredded meat
[00:33:58] <ssi> I need to make one of these
[00:33:58] <ssi> http://www.porkpuller.com/
[00:34:34] <Jymmm> S L O W E S T website ever
[00:34:44] <ssi> watch the video if you can
[00:37:47] <Jymmm> heh, 4 seconds
[00:38:56] <pfred1> Jymmm the slowest website is hosted on a potato
[00:39:34] <pfred1> it ain't easy getting a spud to serve html!
[00:39:47] <Jymmm> ssi: These work really well for lifting/carrying/pulling http://www.amazon.com/Bear-Paw-Meat-Handler-Forks/dp/B003IWI66W and the tips stay sharp too
[00:40:28] <Jymmm> ssi: half the price here https://www.industrialkitchenequipment.com/index.php?main_page=product_info&products_id=1219
[00:40:46] <ssi> nice
[00:41:10] <Jymmm> pfred1: I have no idea what your talking about, but here's a bunny with a pancake on it's head
[00:41:34] <pfred1> Jymmm 01:18 < Jymmm> S L O W E S T website ever
[00:41:43] <Jymmm> ssi: they make good for carving meat too
[00:42:06] <Jymmm> pfred1: Ok, here's a pancake with a bunny on top
[00:42:42] <Jymmm> pfred1: In other words, you still aen't making much sense with what a potato has to do with a slow server
[00:43:06] <pfred1> Jymmm http://totl.net/Spud/gallery.html
[00:44:13] <Jymmm> pfred1: is it real or another hoax?
[00:45:46] <pfred1> Jymmm that may or may not be a real spud server but I've accessed one and it was very slow
[00:46:36] <Jymmm> Heh, well it looks like it got /. so that would make sense.
[00:48:01] <pfred1> Jymmm that page is just a page about it not a page served off the actual spud
[00:48:45] <Jymmm> you know, i have a 64MB DoC, I could probably do the same thing. I think I have a Pentium aroudn here somewhere. Ill have to see what it's wattage draw is.
[00:49:20] <Jymmm> I do have a working 486, but it's in perfect condition so not going to mess with that
[00:51:38] <ssi> make a web server powered by a stirling engine
[00:52:06] <Jymmm> I'd rather make it power a light than a webserver
[00:52:25] <ssi> I want to build a stirling engine that'll sit on top of my imac (which gets quite hot)
[00:52:28] <ssi> and power a fan
[00:52:55] <Jymmm> clean out your imac then
[00:53:20] <ssi> it's plenty clean
[00:53:24] <ssi> it's just full up with hot things
[00:53:28] <ssi> like quad core i7s
[00:53:34] <ssi> and graphics cards
[00:53:59] <Jymmm> in an imac?!
[00:54:01] <ssi> yeah
[00:54:08] <ssi> where'v you been? :P
[00:54:12] <Jymmm> seems a waste as opposed to a desktop
[00:54:24] <ssi> it is a desktop :P
[00:54:38] <Jymmm> no, it's a hybrd lcd
[00:55:03] <ssi> it's a really spectacular monitor stuffed full of a pretty decent computer
[00:55:16] <Jymmm> that overheats
[00:55:21] <ssi> doesn't overheat
[00:55:22] <ssi> just gets hot
[00:55:23] <Jymmm> like ALL aple products
[00:55:30] <frysteev_> its a space heater
[00:55:48] <ssi> hot > noisy
[00:55:53] <frysteev_> i have a macbookpro, the one people used to cook bacon on
[00:56:17] <Jymmm> I watch netflix and the temp on my mbp jumps big time
[00:56:40] <ssi> that's cause silverlight is shitty :P
[00:56:55] <Jymmm> still
[00:57:09] <frysteev_> they just ant you warm and cozy while watching neflix
[00:57:18] <Thetawaves> you may not notice similar heat (even though it exists) on other laptops because they don't use the case as the cpu heatsink
[00:57:52] <Jymmm> Thetawaves: No, I hae a taskbar fan rpm and cpu temp gauge, I know =)
[00:58:06] <Jymmm> 2K rpm and 165f atm
[00:58:09] <Thetawaves> what?
[00:58:17] <ssi> my mbp is nothing like my old vaio in terms of heat
[00:58:21] <ssi> that vaio gave me lap blisters
[00:58:22] <ssi> it was awful
[00:59:46] <Jymmm> Thetawaves: http://i41.tinypic.com/s1q0eg.jpg
[01:01:38] <Thetawaves> so what?
[01:15:03] <ssi> ok turret solenoids are measuring 56mA to operate
[01:15:07] <ssi> I probably should put those on a relay
[01:19:29] <alex4nder> yoh
[01:20:16] <Thetawaves> just use a transistor
[01:20:33] <Thetawaves> you will have to use a transistor if you want to throw a relay too
[01:21:01] <ssi> shouldn't have to
[01:21:10] <ssi> depends on the control coil current
[01:21:16] <Thetawaves> what is the driver?
[01:21:23] <Thetawaves> any micro, and you definitely have to
[01:21:48] <ssi> I'm controlling SSRs no problem with gpio pins
[01:21:57] <ssi> prolly use SSRs to run the solenoids
[01:21:59] <Thetawaves> ssr != relay
[01:22:10] <Thetawaves> ssr has an optocoupler inside
[01:22:12] <ssi> it's got relay right in the name :P
[01:22:13] <Thetawaves> so 20ma
[01:22:27] <Thetawaves> or less
[01:22:31] <ssi> I have a bunch of electromechanical relays, but I just looked at them and they're 24vdc
[01:22:37] <ssi> and I don't feel like dicking with 24V
[01:22:40] <ssi> so SSR it is
[01:23:08] <Thetawaves> what is the on current for your ssrs
[01:23:33] <ssi> pretty low
[01:23:36] <ssi> like 6ma I think
[01:23:41] <Thetawaves> :D
[01:24:18] <Thetawaves> you need a robust driver for coils because of dirty back emf
[01:24:44] <Thetawaves> even with freewheeling diodes it can wreak havoc on small micro controllers
[01:25:44] * Thetawaves has 24v servos
[01:25:45] <ssi> I dunno what used to drive these damn things
[01:25:54] <ssi> probably the 24v electromechanical relays
[01:26:00] <Thetawaves> and i built a 24v 300w hbridge driver for them
[01:26:53] <Jymmm> ssi: https://www.google.com/search?q=ULN2803
[01:27:39] <Thetawaves> weird
[01:27:54] <Thetawaves> i didn't even know an 8 channel variant of the uln2003 existed
[01:28:54] <Thetawaves> wow, uln2803 is 2x as much as the uln2003
[01:29:11] <Thetawaves> not a huge deal in small quantities
[01:29:22] <Thetawaves> 0.25 vs 0.50$ :D
[01:29:36] <Jymmm> Thetawaves: so what?
[01:30:01] <Thetawaves> it's just a habit of being an electrical circuit designer
[01:30:22] <Thetawaves> a lot of cool parts exist at a price
[01:31:21] <Thetawaves> when you design a circuit and try to build a 100 units, negligible price differences such as those can mean real money
[01:41:41] <Valen> also tends to lead to parts that are "just good enough" going in rather than a more reliable part
[01:42:01] <Jymmm> Valen++
[01:42:31] <Valen> then the Chinese realise they can replace the brownout detection IC with a piece of wire
[01:43:17] <Jymmm> fuse?
[01:43:32] <Valen> no
[01:43:45] <Jymmm> =(
[01:43:46] <Thetawaves> brownout is where the power falls below acceptable levels
[01:43:47] <Valen> just air wire a component leg over to the mcu
[01:43:50] <Thetawaves> causes random resets
[01:45:12] <Valen> so your mcu rather than having a seconds warning that power is going down and safing everything, or killing the power to the mcu so it can't do anything daft, decides to randomly restart, execute random instructions and generally try its best to kill whatever its attached to
[01:46:03] <Valen> but the chinese manufacturer saved 10 cents per board and pocketed it
[01:46:11] <Thetawaves> the 'glitch' technique used to crack into hardware is based off of this principal
[01:46:34] <Valen> or a torch i got, where they just took all the components off the board and wired the battery direct to the LED
[01:46:46] <Thetawaves> lol
[01:47:24] <Valen> they kept the PCB, even gold plated it, but in the "sealed" end of the torch you could see they just had the led wires poking into where a power conditioning cap was meant to go
[01:47:57] <Valen> oh yeah, and bypass caps are for wimps
[01:48:44] <Valen> (so are products that don't spontaneously lock up)
[01:50:04] <pfred1> consumer grade electronics works great for what it costs
[01:50:21] <Valen> thats the thing it doesnt
[01:50:26] <Thetawaves> yeah i've heard wonderful stories of china vs most other places
[01:50:30] <pfred1> it is amazing PCs even work at all for what htey are and cost
[01:50:34] <Thetawaves> the main difference, the chinese shit works
[01:50:42] <Valen> for an extra 15% they would be reliable
[01:51:17] <pfred1> then I assembled electronics we had two chinese girls that worked there
[01:51:28] <pfred1> I swear they're just naturally gifted assembling electronics
[01:51:54] <pfred1> the boss would only let them stuff PCBs
[02:02:09] <Valen> don't get me wrong, good stuff can be made in china, but QC has to be eagle eyed and always on the job, and preferably not from china
[02:02:36] <Valen> oh yeah, and trusting of no-one
[02:04:03] <Valen> (with all the fake electronics floating about)
[02:07:42] <pfred1> I've lost 2 motherboards to capacitors
[02:08:13] <pfred1> when I figure it out it is never worth recapping them either
[02:08:51] <pfred1> for a little more than what the caps cost I can get a whole new board
[02:09:48] <pfred1> my TV died and I put new caps in it though
[02:10:11] <ssi> I recapped a tek 420 tds
[02:10:13] <ssi> that was worth doing
[02:10:44] <pfred1> it was worth recapping my TV i just replaced 4
[02:11:12] <pfred1> $1,400 TV
[02:11:24] <Valen> motherboards these days advertise themselves with all solid japanese capacitors
[02:11:37] <pfred1> yeah my new one has SS caps
[02:11:47] <tom1987> i want write one msg for wait of encoder input, but i don't know how to make.any help?
[02:11:56] <Jymmm> stainless steel caps?
[02:12:02] <pfred1> solid state
[02:12:08] <Valen> I'm building a joystick for my computer currently
[02:12:10] <Jymmm> it has solicone in them?
[02:12:14] <Jymmm> silicone
[02:12:35] <Valen> I want ~65535 counts over ~40 degrees of travel, any suggestions?
[02:12:45] <Valen> its a ~60k count encoder :-<
[02:12:58] <ssi> clever gearing?
[02:12:59] <Jymmm> Valen: division
[02:13:12] <Valen> gearing up sounds like a bad idea
[02:13:43] <Valen> Jymmm: division?
[02:14:07] <Jymmm> Valen: multiplication?
[02:14:18] <ssi> I don't see how you're going to get 64k counts out of a 60k line encoder in less than a turn without gearing
[02:14:26] <Thetawaves> put the encoder on the motor, the motor drives a pulley, pulley drives your axis
[02:14:57] <Valen> i meant if you were to achieve this resolution with an encoder it will require an encoder of ~60k counts per revolution
[02:15:16] <Valen> IE its impractically expensive for an optical encoder
[02:15:24] <Valen> I can get 12 bit hall effect rotary encoders but they only give 512 counts over the 40 degree range
[02:15:34] <Thetawaves> why not use x counts per rev * pulley ratio = 65k
[02:15:47] <Valen> Thetawaves: its a joystick, not a mill
[02:15:48] <ssi> Valen: my math says 600k
[02:15:53] <Thetawaves> ohhh
[02:16:00] <Valen> ssi: i could well be off by a factor of 10
[02:16:02] <Valen> ;->
[02:16:15] <ssi> 360/40 is 9, 9*64k is ~600k
[02:16:19] <Valen> yeah
[02:16:23] <Valen> i just did the same
[02:16:26] <Valen> your right
[02:16:30] <Valen> 589815
[02:16:58] <Thetawaves> you should just gear up
[02:17:06] <Valen> The only thing i can think of that'd come close would be some kind of ball bearing resolver thing
[02:17:18] <Valen> Thetawaves: i'm trying to minimise the number of moving parts
[02:17:31] <pfred1> Valen what is a ball bearing resolver thing?
[02:17:33] <Thetawaves> but you could save so much money by avoiding 600k count encoders
[02:17:53] <Valen> Thetawaves: the idea is not to use optical encoders, they are not suitable
[02:18:14] <ssi> resolver is likely a good choice
[02:18:18] <pfred1> Valen why aren't they?
[02:18:22] <ssi> number of steps over a range is only limited by your AD
[02:18:37] <Valen> pfred1: because 600k count encoders are kinda expensive
[02:19:02] <Thetawaves> afaik optical is just one way quadrature encoders are made
[02:19:05] <pfred1> I bet if you were slick you could do it wiht an optical mouse
[02:19:33] <Valen> pfred1: possibly but the update rate would suffer
[02:19:42] <Valen> I need absolute positioning at 1khz
[02:19:50] <ssi> some sort of mems device
[02:20:03] <ssi> what's the application anyway
[02:20:09] <Valen> computer joystic
[02:20:20] <Valen> I want one that isnt a pile of expensive plastic crap
[02:20:36] <ssi> instead you want one that's a pile of expensive 600k encoder crap?
[02:20:52] <pfred1> like i said optical mouse
[02:20:52] <Valen> i don't want to use 600k encoders
[02:21:09] <Thetawaves> camshaft positioning uses differential hall effect sensors mounted around a cam, you are only limited to your AD converter
[02:21:23] <Valen> and the sensor sensitivity
[02:21:38] <Valen> http://newall.com/news/?nwid=23 btw
[02:22:14] <Valen> the magnetic hall effect sensors I mentioned that gave 12 bits of resolution for a rotation are done that way
[02:22:26] <Valen> I was planning on using those, but the resolution is too low
[02:22:50] <Valen> although a belt driven 10:1 gear up would look damn sexy in the stick itself lol
[02:22:50] <ssi> one way or another you need multiple rotations over the length of your travel
[02:23:12] <Valen> ssi yeah, thats why I was thinking of making it into a linear measurement problem instead
[02:23:20] <ssi> yeah that makes sense
[02:23:38] <Valen> expensive or challenging sense is my concern lol
[02:23:38] <ssi> can you get a linear encoder with enough counts over a throw that works for you?
[02:23:48] <Valen> thats the question
[02:23:55] <ssi> yes I know, that's why I asked it :D
[02:24:16] <Valen> lesse, we need 60k counts over a distance, what is a "nice" distance for 60k counts?
[02:24:38] <Valen> hmm this is sounding bad
[02:24:40] <ssi> what are glass scales typically... 20um, something like that?
[02:24:54] <Valen> with .01mm (cheap) * 60k = 600mm
[02:25:08] <ssi> yea
[02:25:15] <Valen> my expensive glass scales are .001 which makes it 60mm which is still pretty excessive
[02:25:18] <pfred1> what is that in english?
[02:25:29] <ssi> 20" or so?
[02:25:32] <Valen> 23.622047244
[02:25:40] <pfred1> almost 2 feet
[02:25:43] <ssi> or 2.3" for 60mm throw
[02:25:45] <ssi> still too long
[02:25:55] <Valen> 60mm i could probably swing
[02:26:05] <Thetawaves> Valen, there are very precise differential hall effect sensors
[02:26:09] <Valen> it really is sounding like gearing it up is the "best" answer
[02:26:18] <ssi> Valen: laser rangefinder!
[02:26:24] <Valen> Thetawaves: i guess they cost more than the $3 i was looking at
[02:26:36] <Valen> ssi: sure, if you want to measure it over eleventy million km ;-P
[02:26:42] <ssi> ELEVENTY MILLION
[02:26:52] <Valen> some kind of interferometer perhaps
[02:27:11] <ssi> echolocation
[02:27:17] <Valen> perhaps pfred1s optical mouse could be made to work
[02:27:28] <Valen> but with a twist
[02:27:30] <ssi> resolver is worth exploring
[02:27:41] <Valen> xmit a laser in an X shape onto the sensor
[02:27:52] <Valen> add gobs of processing power, and determine the angle of the X
[02:27:59] <pfred1> Valen see what your cycloptic friend informs you
[02:28:00] <ssi> sounds awful
[02:28:02] <Valen> with a 300K image sensor you should get pretty accurate
[02:28:21] <ssi> shine a light on the shaft and measure the brightness of the reflection
[02:28:24] <ssi> cosine gives you angle
[02:28:26] <Valen> I do like that it should be insensitive to pretty much anything
[02:28:37] <Valen> again accuracy is hard doing it that way
[02:28:46] <ssi> accuracy schmaccuracy
[02:28:56] <Valen> you can't take a shortcut, you need perfect optics and peltier cooled sensors ;-P
[02:29:21] <ssi> I prefer peltier cooled beers
[02:29:31] <Valen> I prefer my beer in the fridge
[02:29:33] <pfred1> I like ales
[02:30:04] <Valen> I was half hoping the sensors they use in those digital calipers would do, but no way that'll work
[02:30:13] <ssi> they're just shitty glass scales
[02:30:20] <Valen> no they use no optics
[02:30:23] <ssi> er
[02:30:25] <ssi> magnetic or something
[02:30:35] <Valen> its all capacative pickup from a PCB under the scale
[02:30:44] <ssi> yea there's the one
[02:30:51] <pfred1> I bought 2 different models from HF one i like the other pair man they suck
[02:31:34] <Valen> dad had one set get the screen smashed
[04:08:45] <Valen> http://www.youtube.com/watch?v=Y2Rnqc6LoRo
[05:48:10] <elmo40> helicopter, such OLD tech... gotta get rid of the very vulnerable rotating wing! This is the new way of thinking. http://www.youtube.com/watch?v=pF0uLnMoQZA
[05:49:38] <Loetmichel> elmo40: MUCH unneccessary weight
[05:50:10] <Loetmichel> and tha efficienca of a fan inside THAT much Flaps and obstacles is at lest doubtful .
[05:50:14] <Loetmichel> and: OLD
[05:50:26] <Loetmichel> -a+y
[05:50:51] <elmo40> more robust
[05:50:56] <elmo40> can take a hit ;)
[05:51:37] <Loetmichel> made one if this myself...
[05:51:37] <elmo40> chopper falls down, this will keep you moving
[05:52:06] <Loetmichel> but much simpeler
[05:53:19] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=6743
[05:54:35] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_view=core.DownloadItem&g2_itemId=4153
[05:55:04] <Loetmichel> but never flown it
[05:55:15] <Loetmichel> to much projects, to less time ;-)
[07:20:37] <A2Sheds> hmm I need a 12HP DC motor driver with -/+ 10V in
[07:25:45] <JT-Shop> I need to get my mind right about this loop-d-loop
[07:27:49] <joe9> i get a lot of warnings like this in linuxcnc build: http://codepad.org/gzmDUQv4 . i think they are ok. but, just wanted to double check.: http://codepad.org/sFQZX2q1
[07:27:56] <joe9> warning: #warning "Attempt to use kernel headers from user space, see http://kernelnewbies.org/KernelHeaders"
[07:28:22] <joe9> i did "make headers_install" in the kernel src directory.
[07:59:40] <frysteev_> moening
[08:28:09] <raynerd> my limit switches have just arrived in the post and I`m not a little confused about how to wire them! Each switch has three terminals. Anyone explain to me in simple terms??!
[08:33:03] <archivist> sounds like change over switches use the common and no or nc depending which way you use them
[08:34:42] <dgarr> examples updates for consideration (2.5): http://www.panix.com/~dgarrett/stuff/ngcgui_examples_update.mbox
[08:50:53] <jthornton> Hi Dewey!
[08:52:49] <raynerd> Sorry, another question. I just had to reinstall linux on my pc and so I saved my config file.
[08:54:04] <raynerd> I`ve used step config wizard and clicked modify and selected my backed up config file, ran through it all and it has correctly made a new folder raynerdCNC and RaynerdCNC.desktop.
[08:54:58] <raynerd> Previously, when I selected RaynerdCNC.desktop, it would open up linuxcnc with the RaynerdCNC settings. Now when I click it, it tells me that it is an "untrusted application launcher" ...bla bla and my only option is to select cancel.
[08:55:31] <raynerd> How do I make it trusted. This never happened last time when I made the intial configuration on my previous install.
[08:55:34] <raynerd> instal*
[08:55:55] <raynerd> nope, I was correct, 2 x l = *install
[08:56:42] <jdhnc> same linux installer?
[08:56:50] <raynerd> yes, same CD
[08:57:44] <jdhnc> perms are probably wrong
[08:57:52] <raynerd> It works OK if I go to linux CNC from the top menu of the desktop and then select raynerCNC configuration from the first splash screen. It just seems to be this shortcut that isn`t working correctly.
[08:58:22] <raynerd> I`m not into linux so no idea what to even try!
[08:58:59] <jdhnc> open a terminal window, ls -alF *.desktop, paste the line that has your emc config
[09:01:35] <raynerd> ls -alF *.desktop - No such file or directory
[09:02:43] <jdhnc> heh, I don't have any either. Wonder what I was thinking of :)
[09:02:59] <raynerd> :-S
[09:03:07] <Mjolinor> scrap it :)
[09:03:24] <jdhnc> I could have sworn that's what they were, but I haven't actually looked at one in years
[09:03:31] <raynerd> lol hello Mjolinor..how do?
[09:03:37] <Mjolinor> good
[09:03:42] <Mjolinor> did seom numebrs
[09:04:01] <Mjolinor> http://homepage.ntlworld.com/dknpower/clock1.jpg
[09:04:08] <Mjolinor> http://homepage.ntlworld.com/dknpower/clock2.jpg
[09:04:40] <raynerd> wow! very nice!!!
[09:04:50] <raynerd> that in brass with the engraving bit?
[09:04:57] <Mjolinor> yup
[09:05:02] <Mjolinor> with a 90 degfree one
[09:05:13] <Mjolinor> I tried it with the point ones and it jsut snapped
[09:05:29] <raynerd> 90 deg ?
[09:05:54] <raynerd> I know what 90 deg is...which cutters are those was really my question.
[09:06:38] <Mjolinor> they are the same as the ones oyu have but the end is 90, did you not see them here?
[09:06:52] <Mjolinor> you didnt stay long enough :)
[09:06:57] <raynerd> nope...
[09:07:11] <raynerd> I would have but you were hvaing your tea!
[09:07:22] <raynerd> do you have an ebay link so I can order some?
[09:07:53] <Mjolinor> you dont want them from where I got them, it took me over 3 months to sort the ebay sale out, they kept getting it wonrg, it got rather nasty actually
[09:08:05] <Mjolinor> the seller was supermario
[09:08:14] <Mjolinor> but that company I go thte last ones fomr have them I think
[09:08:17] <Mjolinor> hang on
[09:09:47] <raynerd> I was quite chuffed with the little part I`d made last night, first proper steps anyway! www.raynerd.co.uk/images/isobel.jpg
[09:10:00] <Mjolinor> http://www.ebay.co.uk/itm/5-Carbide-PCB-Engraving-Bit-CNC-Router-Tool-90-1-/310341406865?pt=UK_Home_Garden_PowerTools_SM&hash=item4841ca1091
[09:10:06] <jdhnc> raynerd: it is in ~/Desktop
[09:10:11] <Mjolinor> :)
[09:10:20] <Mjolinor> did you do the font with inkscape?
[09:10:27] <Mjolinor> or vcarve?
[09:10:49] <Mjolinor> http://www.ebay.co.uk/itm/5-Carbide-PCB-Engraving-Bit-CNC-Tool-90-0-3mm-1-/230666745529?pt=UK_Home_Garden_PowerTools_SM&hash=item35b4cf36b9
[09:11:09] <Mjolinor> two differnet tip sizes there, mine are 0.2 but I hink it needs to be smaller for engraving
[09:11:32] <raynerd> so is the tip size the flat?
[09:11:45] <Mjolinor> yes
[09:11:46] <Mjolinor> the end
[09:12:01] <Mjolinor> one of theose links is a 0.3 the other seems to be 0.008
[09:12:05] <Mjolinor> which is damned pointy
[09:13:04] <raynerd> I wonder if it was too fine, it would chip easily/
[09:13:56] <Mjolinor> but you need it fine if you want an engraved V shape
[09:14:06] <raynerd> yes. true.
[09:14:13] <raynerd> Do you know what a drag engraver is?
[09:14:13] <Mjolinor> with mine the bottom of the slot is 0.37mm with a 0.2 cutter
[09:14:20] <Mjolinor> which is OK but shoudl be better I think
[09:14:26] <Mjolinor> no idea
[09:14:31] <Mjolinor> I suppose its one that doesnt rotate
[09:16:00] <raynerd> I must admit, tooling is something that is confusing me a little. I see a lot use V bits and then small endmills and then there are the deburr style. I see a lot of those but not clear why they are used
[09:16:20] <Mjolinor> so we are members of the same club in that then
[09:16:34] <Mjolinor> think you jsut sue what does the job you want, bugger what it is really for
[09:16:37] <Mjolinor> or called
[09:17:25] <Mjolinor> hmm, fail
[09:17:29] <Mjolinor> :eating:
[09:17:42] <Mjolinor> whops wrong window, I am relaly buggering thsi up
[09:18:48] <raynerd> in that case, I see loads of these very pointed ones that we have purchased. what would there use be? I expect they will be prone to snapping!
[09:22:55] <jdhnc> you could send me some.
[09:24:09] <raynerd> lol
[09:24:29] <raynerd> jdhnc: you do any fine engraving?
[09:24:48] * jthornton finally got that loop to work properly
[09:24:51] <jdhnc> define 'fine'
[09:25:14] <jdhnc> I've only done engraving on plastics
[09:25:44] <raynerd> fine: keen or sharp, as a tool: Is the knife fine enough to carve well?
[09:26:07] <jdhnc> that's highly subjective.
[09:38:21] <Tom_itx> Mjolinor, you could also grind a half round drill bit which looks nearly identical to that
[09:38:44] <Tom_itx> then you can get the angle and depth of cut you want
[09:40:57] <Tom_itx> typically they're used for deep hole drilling in brass etc
[09:41:12] <Tom_itx> http://www.travers.com/skulist.asp?r=s&n=||UserSearch1%3Dhalf-round+drill&q=block+id+41084+and+class+level3+id+30329
[09:41:23] <Tom_itx> you can probably find them cheaper
[09:44:14] <pcw_> Why do I keep getting spam about blueberries?
[09:44:53] <Tom_itx> blueberry pi?
[09:46:25] <pcw_> somehow its about the only spam that gets through the filter...
[09:54:18] <raynerd> What depth would you guys be cutting on a pcb?
[09:55:04] <syyl> so deep, i get the desired with of the insulation path...
[09:55:22] <syyl> (and to get trough the copper layer...)
[09:55:33] <Tom_itx> don't cut much of the FR4 because it's abrasive
[09:56:18] <Loetmichel> pcw: just through the 25µ/50µ copper, not more
[09:56:20] <Loetmichel> ideally
[09:58:12] <Mjolinor> I dont have anything that will grind these bits
[09:58:26] <Mjolinor> I thin kit needs diamond or maybe on an oil stone
[09:58:29] <Mjolinor> unfortunately
[09:58:53] <Loetmichel> oil stone is to wak
[09:58:58] <Loetmichel> diamond it is
[09:59:01] <Loetmichel> weak
[09:59:41] <Mjolinor> I can do ti with an idia stone but it would take the rest of my life to get rid of a chip :)
[09:59:45] <Mjolinor> india stone
[10:00:01] <syyl> get a degussit oil stone, thaths perfect for final honing of carbide
[10:00:09] <Mjolinor> i t would be nice to be able to grind it in the machine
[10:00:19] <Mjolinor> #I dont think these cheap chinese things are centred too well
[10:00:52] <Loetmichel> mju: not the brightest idea. the diamond/TC dust is abrasive as hell... and flies in EVERY corner of your machine
[10:01:08] <Loetmichel> Mjolinor
[10:01:10] <Mjolinor> I check my spindle runout and its only a couple of hundredths of a MM , I checked hte end of the cutter under the microscope adn it is 0.2 mm so to get a flat at the bottom of the cut atht is 0.37 side it can only be the cutter that is not ground accurately
[10:01:36] <Mjolinor> I am jsut adding a small sucker thign from an old vacum
[10:01:54] <Mjolinor> vacuum
[10:01:55] <Loetmichel> Mjolinor: more likely: the 0,2mm are measured without the "free" on the following sinde
[10:02:28] <Loetmichel> meaning: the cutter tip IS excentic
[10:02:32] <Loetmichel> +r
[10:02:40] <Mjolinor> yes
[10:02:42] <Mjolinor> that is what I mean
[10:02:44] <Loetmichel> has to be
[10:02:54] <Mjolinor> would be nice to shape it after it is in the chiuck and spinning
[10:03:04] <Loetmichel> not possibel
[10:03:08] <Loetmichel> wouldnt cut
[10:03:28] <raynerd> just tryed to etch some clear plastic for a keyring but rather than getting a nice frosted etched line, the lines look a little melted!
[10:03:30] <syyl> you have no way to relief the cutter
[10:03:36] <Mjolinor> would do PCBs if I shape the tapered bit and leave the bottom alone I think
[10:03:57] <Loetmichel> ah, relief was the word i was searchig for
[10:04:17] <Mjolinor> it is chisel ended, the 0.2 mm is made by grinding front and back like a normal chisel
[10:04:59] <Mjolinor> if I leave that but narrow the end from the sides with a diamond in a dremmel sort fo thing then it will still cut as it shuodl provided I don't want deep but only narrow adn accurate
[10:05:04] <Loetmichel> that souunds like a spear drill, not a engraving bit
[10:05:21] <Mjolinor> probably
[10:05:32] <Mjolinor> allsorts of names for these thigns that I dont know :)
[10:05:47] <raynerd> what exactly is an "engraving bit" then?
[10:05:49] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8958
[10:05:51] <Loetmichel> that
[10:06:01] <Mjolinor> yes
[10:06:07] <Mjolinor> that si the sort of thing I am using
[10:06:11] <syyl> it will throw up a hell of a burr, Mjolinor ;)
[10:06:18] <Mjolinor> cheap ones off ebay, they only cost about £1 each
[10:06:39] <Mjolinor> I am not spending 10 euros on cutters to break at random while I explore what this thing is cpable of :)
[10:07:00] <syyl> get some descent singlelip engraving cutters from a "normal" supplier and get excitet how good things can work :D
[10:07:13] <Mjolinor> I will do that
[10:07:25] <Mjolinor> once I know about my spindle speeds / loads / feeds etc
[10:07:45] <Mjolinor> this little toy plane motor seems to stand up pretty well so far
[10:08:07] <syyl> pcb milling doesnt take very much torque ;)
[10:08:09] * Mjolinor apologises to any radio control nuts in the room
[10:08:14] <Loetmichel> Mjolinor: btw: the ones in the photo aer ground frpom old broken TC bits
[10:08:25] <Loetmichel> ... by hand witrh a dremel and a dioamond disc
[10:08:31] <Mjolinor> :)
[10:08:41] <Mjolinor> so I need to buy soem diamond disks for my dremmel then
[10:08:42] <raynerd> nice
[10:08:43] <syyl> with 100watts you can even machine steel with cutters below 2mm...
[10:08:59] <Mjolinor> well it shoudl be 160 watts
[10:09:01] <Loetmichel> ... had to practice some, bit after about 20 tor the trash the first one had cut:-)
[10:09:02] <raynerd> I always thought you needed a cutter grinder to make them accurate enough on centre
[10:09:08] <Mjolinor> but I havent had it really grinting yet :)
[10:09:12] <Mjolinor> grunting
[10:09:37] <syyl> 1mm cutter, 1mm deep, 6000rpm and 60mm/min feed in steel will take about 80Watts
[10:09:38] <syyl> :D
[10:09:49] <Loetmichel> raynerd: you can substitute the grindig rig with experiendce... ;-)
[10:10:11] <Mjolinor> see that is the sort of information you cant find on the net
[10:10:12] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=8961
[10:10:21] <Loetmichel> made them like this ;-)
[10:10:48] <raynerd> how do you ensure that the point is perfectly central?
[10:10:51] <Mjolinor> I woudl have thought 6000 rmp is too clow to get a good finish at that feed, if I go over 45 the sides of the V get a bit icky
[10:11:31] <syyl> that parameters where for a 2 flute cutter
[10:11:36] <syyl> and just a example ;)
[10:11:54] <Loetmichel> i still think 6000 is a little low for 1mm
[10:12:01] <syyl> no way
[10:12:07] <Loetmichel> i would go more like 12
[10:12:09] <syyl> experience :)
[10:12:18] <Loetmichel> yes, it can differ ;-)
[10:12:40] <syyl> http://dl.dropbox.com/u/24396704/qtl2_4.jpg
[10:12:50] <syyl> i use cutters down to 0,5mm on that monster
[10:13:01] <syyl> and it only goes up to 6300rpm
[10:13:21] <Mjolinor> kind of hard to do 12k then :)
[10:13:35] <Loetmichel> lokks like it
[10:13:37] <Loetmichel> ;-)
[10:13:41] <Mjolinor> never seen a mechanical frequency doubler
[10:13:51] <syyl> i have..
[10:13:51] <Loetmichel> oh, you CAN buy them
[10:13:53] <Mjolinor> well apart fmr a gearbox
[10:13:58] <Loetmichel> but not cheap
[10:14:12] <syyl> and for the cheap ones
[10:14:27] <Loetmichel> planetary gear iwith SK40 on one side and a ER20 chuck on the other ;-)
[10:14:40] <Mjolinor> but you still need a lock for the outside
[10:14:44] <syyl> yes
[10:14:45] <Loetmichel> yes
[10:14:54] <syyl> those are sold on a regular base
[10:14:59] <syyl> not for heavy machining
[10:15:01] <syyl> but they work
[10:15:16] <Mjolinor> more like a casr gearbox than a bit of workshop tooling
[10:15:19] <Mjolinor> car
[10:15:21] <Loetmichel> but that is designed in a way that the hispeed gear still can bie changend auutomatically, like any other tool
[10:16:12] <Loetmichel> locks somewhre on the spindle in a hole os so
[10:16:14] <Loetmichel> or so
[10:16:49] <Mjolinor> http://eahart.com/prius/psd/
[10:17:01] <Mjolinor> thats a good link for understanding planet gears
[10:17:08] <Mjolinor> and prius gearboxes :)
[10:18:35] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/cnc/pendant8.jpg
[10:18:43] <Tom_itx> i did that with a standard center drill
[10:18:53] <Tom_itx> because that's what i had handy
[10:20:32] <syyl> http://dl.dropbox.com/u/24396704/bonelle_24.jpg
[10:20:47] <syyl> that with a normal 90deg singlelip engraving cutter
[10:20:48] <syyl> :D
[10:20:54] <syyl> nice pendant, tom!
[10:21:02] <Tom_itx> that looks real nice syyl
[10:21:05] <syyl> has all you need
[10:21:08] <Tom_itx> you always do nice work
[10:21:15] <syyl> thank you, sir
[10:21:16] <syyl> :=)
[10:22:09] <Loetmichel> Mjolinor: i make planetary gearts myself
[10:22:18] <Loetmichel> i know more or less how they work ;-.)
[10:22:19] <Mjolinor> for waht?
[10:22:29] <Mjolinor> I mend them a lot in auto gearboxes
[10:22:35] <Mjolinor> or I used to, don't do amny now
[10:22:37] <Loetmichel> cnc drive
[10:22:45] <Tom_itx> syyl what sort of coating did you use for the lettering? just paint?
[10:22:50] <syyl> jep
[10:22:51] <Mjolinor> and I seriously hope I never have to mend the one in my Prius :)
[10:22:57] <syyl> black spray paint
[10:22:58] <Tom_itx> then brushed it smooth
[10:23:01] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=11714
[10:23:05] <syyl> right :)
[10:23:16] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=11705
[10:23:19] <Tom_itx> i use to work at a place where another guy had a small engraving shop
[10:23:32] <syyl> at work we use wax crayon
[10:23:36] <syyl> and a hot air gun
[10:23:45] <Tom_itx> he had a bridgeport (i forget the name of the machine) engraver
[10:23:48] <Loetmichel> syyl:like me ;-)
[10:23:51] <Tom_itx> would do rotary stuff etc
[10:24:05] <Mjolinor> what is that for, the gearbox?
[10:24:10] <Tom_itx> he did engraving for roll stamps etc like for the old oil can lids etc
[10:24:13] <Tom_itx> or plaques
[10:24:21] <syyl> rotrary engraving is always interesting
[10:24:24] <Mjolinor> is ti on the end of a stepper motor?
[10:24:32] <Tom_itx> it was an interesting machine
[10:24:58] <Jymmm> syyl: Tom_itx Black enamel oil based paint. rub in with a cloth into the engravings, wipe off the excess on the surface and let dry.
[10:25:11] <Loetmichel> Mjolinor: yes
[10:25:21] <syyl> sounds to be worth a try
[10:25:31] <Tom_itx> Jymmm, jup.. that's how he did it
[10:25:51] <Jymmm> Tom_itx: Sorta
[10:26:07] <Loetmichel> Mjolinor: its beween the stepper and the ball thread
[10:26:08] <Jymmm> Tom_itx: no sanding required withwhat I said.
[10:26:19] <Tom_itx> he made nice plaques etc as well as engraved stamps
[10:26:29] <Tom_itx> i know
[10:26:36] <Tom_itx> but i saw the brushed finish on syyl's
[10:26:37] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=11639 ... on the right
[10:26:51] <Mjolinor> with a fixed ring gear?
[10:27:06] <Loetmichel> correct
[10:27:09] <Mjolinor> so you are jsut gearing it down so more staeps for same rotate?
[10:27:12] <Jymmm> syyl: Only works on non porous materials of course.
[10:27:12] <Loetmichel> glued in the mount
[10:27:20] <Loetmichel> yes
[10:27:39] <syyl> thats self evident ;)
[10:27:42] <Loetmichel> because the thead hast 10mm/rev
[10:27:44] <Jymmm> syyl: you could clear coat first, engrave, then rub in too.
[10:27:45] <Loetmichel> thread
[10:28:07] <syyl> the enamel based oil paint..
[10:28:19] <syyl> is that the normal oil paint you get in art stores?
[10:28:35] <Jymmm> syyl: In porous materials that is.
[10:28:37] <Tom_itx> any oil based paint probably
[10:28:40] <Mjolinor> so you are not really usign it as a mutli speed gearbox at all, jsut a fixed reduction gearbox
[10:28:47] <Jymmm> syyl: OH HELL NO, that's way too expensive. hang on.
[10:29:28] <Tom_itx> enamel would probably do just as well
[10:29:51] <Jymmm> Tom_itx: no, this is OIL BASED enamel.
[10:30:00] <Tom_itx> marine paint
[10:30:08] <Mjolinor> I am having problems setting up my new siemeens VFD
[10:30:12] <Jymmm> Tom_itx: if you can afford it.
[10:30:17] <Tom_itx> what do they used for raised lettering on business cards?
[10:30:25] <syyl> Mjolinor, a micromaster?
[10:30:35] <Mjolinor> my lathe doesnt start until it gets to about 40 Hz then it races offbut it will run down to 4 Hz or so before it stalls
[10:30:39] <Mjolinor> yes
[10:30:51] <Mjolinor> there are about 10^4 variables in the goddam thign that you can set
[10:30:57] <syyl> right :D
[10:31:01] <Jymmm> syyl: This comes in little half pint cans for like $5, black and white are the only colors... http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&langId=-1&catalogId=10053&productId=202041177&R=202041177
[10:31:17] <Mjolinor> the old inverter woudl start at under 10 Hz but it made the motor run hot and sound crap
[10:31:27] <syyl> ah ok, that type of paint i can get here too
[10:31:36] <Mjolinor> its really smooth with the micropmaster but there are so many variables you shoudl be able to do a degree in the goddam things
[10:31:39] <syyl> thanks Jymmm, will try that out sooner or later
[10:31:45] <syyl> Mjolinor
[10:31:51] <syyl> reset it to factory settings
[10:31:59] <Mjolinor> I did, several times :)
[10:32:00] <syyl> then only set the nessescary things
[10:32:08] <Mjolinor> that is the way I have done it
[10:32:10] <Jymmm> syyl: Yeah, just use old cut up t-shirts to apply.
[10:32:29] <Mjolinor> but I htink there msut be soemthign for helping the motor to start soemwhere, jsut havent found ti yet
[10:32:38] <syyl> you will ned the motor base dates, your inputs, ramps
[10:32:54] <syyl> what does happen?
[10:32:58] <Mjolinor> I think the default ramps seem very slow to me
[10:33:03] <Mjolinor> but maybe its my lathe
[10:33:19] <Mjolinor> it doesnt seem to do anyting , I increase hte speed manually in emc
[10:33:22] <Mjolinor> linuxcnc
[10:33:36] <syyl> if you set the start-ramp to steep it will fail
[10:33:36] <Mjolinor> and it doesnt start till I get to about 40Hz on the VFD display
[10:33:46] <syyl> vfd will stop with an error
[10:33:50] <syyl> strange
[10:34:06] <Mjolinor> so maybe Iwill increase hte ramp till it fails then back it off
[10:34:18] <Loetmichel> syyl: on metal parts i have got good results with "puldervbeschichtung" powder, brushed into the engraving ang then heated to 200°c
[10:34:22] <Mjolinor> I havent loked at the ramps at all but for sure when I hit sotp it takes an age to slow down
[10:34:46] <Mjolinor> so repsumably the start ramp is similar to the stop ramp as default and maybe I am not waiting long enough
[10:34:58] <syyl> the default ramps of the micromaster are very slow
[10:35:03] <Mjolinor> makes me feel scared to be pumping electricity at it and it not going round
[10:35:26] <syyl> the motor should run even at 5Hz
[10:35:39] <syyl> there must be another problem...
[10:35:49] <Mjolinor> was going to read soem more today but I have about 10 documents with 200 pages each, its a lot of reading
[10:35:57] <Mjolinor> it does run at 5Hz fine
[10:36:06] <Mjolinor> provided I come down to 5
[10:36:23] <Mjolinor> get it running at 40 and slwo it down it stalls at about 3Hz
[10:36:30] <syyl> strange
[10:36:49] <syyl> it should start right above 0Hz
[10:36:52] <Mjolinor> think the lathe is tight, it hasn't run for 15 years prior to me getting it
[10:37:12] <syyl> never had that problem, and i set up a few siemens VFDs
[10:38:03] <Mjolinor> I had it running last night at about 600 Hz, I connected the grease gun to the headstock and started pumping, none went in initially then the grase gun collapsed and injected a laod of greese, the lathe slowed right down to about 200 rpm and stayed ther about 5 minutes till the grease moved about I suppose
[10:38:15] <Mjolinor> 600 RPM, not HZ
[10:38:39] <syyl> while i read 600Hz i shrugged for a moment ;)
[10:38:45] <Mjolinor> :)
[10:38:48] <syyl> hmm
[10:38:58] <syyl> spindle with ball/roller bearings?
[10:39:04] <syyl> or plain bearings?
[10:39:06] <Mjolinor> no idea
[10:39:12] <Mjolinor> denford Orac
[10:39:17] <Mjolinor> don't knwo what it has in there
[10:39:35] <Mjolinor> perhaps after 15 years ina wood shed I shoudl explore a bit more deeply before I try to get it working :)
[10:39:44] <syyl> most plain bearings want oil, no grease
[10:40:01] <syyl> and with ball/roller bearings to much grease would also not help :\
[10:40:16] <Mjolinor> well it had grease in before, you can see where it coems out at the back of the spindle, I kept sticking it in there until it came out new
[10:41:26] <Mjolinor> the thing is that the old inverter, the one denford supplied, woudl start it very slowly
[10:41:31] <Mjolinor> the siemens one will not
[10:42:05] <syyl> there should be a parameter
[10:42:15] <syyl> to rise the current in the beginning
[10:42:19] <Mjolinor> spec sheet does not say if it is plain berings or roller
[10:42:29] <syyl> maybe that helps
[10:42:32] <Mjolinor> I personally think plain because of the way it feels if you turn it by hand
[10:44:02] <Mjolinor> http://www.denfordata.com/bb/viewtopic.php?t=420
[10:44:10] <Mjolinor> that suggest taper roller bearings to me
[10:44:19] <Mjolinor> cant think of any other type you can tighten up
[10:45:00] <syyl> that drawing is the headstock of your lathe?
[10:45:04] <Mjolinor> yup, it has taper roller bearings
[10:45:10] <Mjolinor> seems so :)
[10:45:11] <syyl> it says 3207 bearings ;)
[10:45:31] <syyl> angular contact bearings
[10:46:05] <Mjolinor> so for sure it shoudl be smnoother than it is when I turn it by hand
[10:46:13] <Mjolinor> I can feel a dismemberment coming on
[10:46:41] <syyl> :)
[10:49:31] <ssi> ok today I must get 100% of my GPIO working!
[10:49:38] <ssi> and maybe the servo amps :D
[10:49:50] <ssi> aaaand the VFD drive wiring
[10:50:34] <syyl> and i must finish the paintjob on my lathe-bed
[10:50:37] <syyl> or pain-job...
[10:50:40] <ssi> yes
[10:50:49] <ssi> you need to come scrape mine in when you're done
[10:51:02] <syyl> sure
[10:51:06] <syyl> 70eur/hour
[10:51:07] <syyl> ;)
[10:51:10] <JT-Shop> can't get it done yacking on here :P
[10:51:29] <ssi> JT-Shop: actually I might need to, cause I need help with some of it :P
[10:51:55] <JT-Shop> I know, I'm just ribbing you a bit
[10:52:04] <ssi> I made good progress though
[10:52:08] <ssi> over half of the limits work
[10:52:28] <ssi> all the wiring is in place, but it doesn't all work yet
[10:52:33] <ssi> https://p.twimg.com/Aoulml0CMAANQHM.jpg
[10:52:51] <ssi> https://p.twimg.com/AouS1wjCMAAebq-.jpg:large
[10:53:11] <ssi> wire ferrules made all the difference in the world wiring those screwless blocks
[10:54:44] <JT-Shop> I use ferrules on just about everythiing
[10:55:08] <ssi> yeah I think I'll start using them a lot now
[10:55:14] <ssi> since I had to invest in the crimp tool
[10:55:24] <ssi> I'm very very happy with the crimp tool I got though too
[10:55:30] <JT-Shop> looking nice
[10:55:50] <JT-Shop> you get the one that is end crimp or side crimp?
[10:55:58] <JT-Shop> I like the side crimp one better
[10:56:07] <ssi> side crimp
[10:56:12] <ssi> it's the DCT frame from automationdirect
[10:56:38] <ssi> and got a die for regular red/blue terminals as well
[10:56:49] <JT-Shop> I got mine so long ago I can't remember where I got it
[10:57:03] <JT-Shop> mine just does ferrules
[10:57:05] <ssi> I have AMP 59250 crimpers at the hangar, and they're phenomenal, but they only really work well on honest to god AMP PIDG type terminals
[10:57:13] <ssi> and those are way too expensive to use anywhere but the airplane
[11:00:30] <ssi> so I have a laundry list of things to address
[11:01:02] <ssi> one of my hall effects is putting out 12v when not tripped
[11:01:25] <ssi> only one, not sure why... the rest seem to be high impedance when not tripped, and sink to ground when tripped
[11:01:48] <ssi> so I'm seeing 12V at the GPIO. The 7i42 is keeping it clamped to 5v, but that's not really ideal :P
[11:02:14] <JT-Shop> I finally got my taper OD sub for ngcgui working the way I wanted :)
[11:02:21] <ssi> GOOD
[11:02:29] <ssi> now do taper ID :D
[11:03:08] <ssi> second issue... the SSRs all had a common control + terminal
[11:03:21] <ssi> originally at 12v, and they're enabled by pulling the - terminal to ground
[11:03:48] <ssi> I'm controlling them with 5V instead... they switch fine at 5V, and flow about 5mA when on
[11:04:11] <ssi> problem i was having is it seemed like when I'd switch one on, it'd pull the whole 5v bus down to 0
[11:04:25] <ssi> I can't imagine that the 5v I was feeding them would be that high Zout
[11:04:36] <ssi> I thought I was taking that 5v directly from the computer PSU's 5v line
[11:04:39] <ssi> but I'll have to double check that
[11:08:16] <JT-Shop> mine had a blown part on the turret board that was drawing the 24v buss down to a couple of volts
[11:08:35] <JT-Shop> that one messed up my mind for a while
[11:08:41] <ssi> so far I have zero 24v in the system
[11:09:07] <ssi> oh which brings me to issue 3
[11:09:18] <ssi> the two turret control solenoids are 12V and draw about 60mA each
[11:09:23] <ssi> so I can't control those directly
[11:10:36] <ssi> I have the original electromechanical relays from the control, but they're all 24vdc control :/
[11:30:56] <joe9> after you install linuxcnc, do you customize your paths? http://codepad.org/WyYEN6zx error, when I do scripts/rip-environment.
[11:34:12] <joe9> http://codepad.org/whwbK9tt this is what the install file says
[11:35:46] <joe9> should the scripts/rip-environment be customized before use?
[11:36:29] <JT-Shop> no but you run it like it says . ./scripts/rip-envirnment
[11:36:35] <JT-Shop> note the period space period slash
[11:36:56] <joe9> JT-Shop: but, it says that EMC2_HOME=/usr/local; export EMC2_HOME
[11:37:04] <JT-Shop> or are you in some strange version of linux?
[11:37:19] <joe9> i was using zsh, let me try bash.
[11:37:38] <JT-Shop> this ain't a standard install is it?
[11:37:58] <joe9> i followed the instructions in installing linuxcnc.
[11:38:13] <joe9> "./configure, make, sudo setsuid.."
[11:38:19] <JT-Shop> on Ubuntu?
[11:38:31] <alex4nder> hey
[11:38:38] <JT-Shop> hay
[11:38:39] <joe9> no, not on ubuntu.
[11:38:45] <JT-Shop> ok, I know nothing then
[11:38:53] * JT-Shop shuts up now
[11:40:16] <alex4nder> sup
[11:41:19] <JT-Shop> I have a dozen 6" scales and can only find one :/
[11:44:40] <joe9> i should have been using emc-environment and not rip-environment.
[11:55:06] <JT-Shop> your using 2.4?
[11:55:23] <joe9> EMC2 - 2.3.4
[11:55:33] <joe9> let me get to 2.4
[11:59:25] <joe9> JT-Shop: what version are you on?
[12:04:26] <JT-Shop> most if not all my machines are running 2.5 from buildbot
[12:04:56] <joe9> you mean from the git repo?
[12:05:03] <joe9> is buildbot == git repo?
[12:05:20] <JT-Shop> no, buildbot deb
[12:05:25] <joe9> oh, ok. thanks.
[12:05:39] <JT-Shop> http://buildbot.linuxcnc.org/
[12:08:24] <joe9> jt-shop, what is your kernel version?
[12:08:58] <JT-Shop> what's that?
[12:09:03] <joe9> uname -a
[12:09:28] <JT-Shop> I have 5 machines running LinuxCNC just pick one?
[12:10:00] <joe9> the one with the least problems? I am on 2.6.38.8
[12:10:09] <joe9> just wanted to check if you are on a different version.
[12:10:24] <JT-Shop> the closest one is 2.6.320-122-rtai
[12:10:31] <joe9> ok, cool. thanks.
[12:10:38] <JT-Shop> 32 not 320
[12:10:57] <joe9> that is what i thought. thanks.
[12:11:02] <JT-Shop> np
[12:13:09] <joe9> http://codepad.org/ZK8XYO2e do you see these warnings while "make"?
[12:13:20] <joe9> i am checking your logs to see if you do.
[12:13:37] <JT-Shop> I don't have to make, I'm using a deb
[12:13:40] <joe9> but, having not checked a buildbot before, I am searching around.
[12:13:51] <joe9> oh, just apt-get install then.
[12:13:54] <ssi> arg wtf
[12:13:58] <JT-Shop> yea
[12:14:21] <ssi> JT-Shop: so I have these ssr driving pins set up as output, open drain, invert
[12:14:34] <ssi> JT-Shop: the first one works fine to drive the enable relay
[12:14:42] <ssi> JT-Shop: the rest of them trip the relay regardless of the state of the pin
[12:15:51] <pcw_> Measure across the SSR input pins to see whats happening
[12:15:53] <JT-Shop> your using a 7isomething not a 5i20?
[12:16:10] <JT-Shop> Yippie Help is here
[12:16:16] <ssi> 5i23 + 7i42
[12:16:44] <ssi> pcw_: control terminal of the SSR has 5V on the + side, all of them bussed together
[12:16:57] <ssi> pcw_: - side of each is connected to an open drain pin
[12:17:04] <ssi> if I disconnect from the pin, I get 5V across the terminals
[12:17:12] <ssi> if I Connect to the pin, I get 0V regardless of state of the pin
[12:18:02] <ssi> er sorry
[12:18:15] <ssi> not 5v across the terminals, but rather 5v at the + side and close to 5v at the - side
[12:19:04] <pcw_> is 5i23 jumpered for 5V pullups and 5V bus switch mode on the 7I42 conns?
[12:19:37] <ssi> I need to double check
[12:19:42] <ssi> that may be the answer though
[12:20:09] <ssi> with the GPIO pin off (pulled up) it's at 3v3, and the ssr has 5v-3v3 across it, and I guess that's enough to trip it
[12:20:15] <ssi> it's juts off that the #1 relay doesn't behave that way
[12:20:21] <pcw_> it will only swing to 3.3V if not jumpered right
[12:20:22] <ssi> it actually turns off when the gpio is in a high state
[12:20:44] <pcw_> well 1.7 is marginal (some will work and some not)
[12:21:08] <ssi> can I switch those jumpers hot?
[12:21:17] <pcw_> i would not
[12:21:20] <ssi> j
[12:21:21] <ssi> k
[12:21:23] <ssi> I'll brb then
[12:21:40] * JT-Shop is hearing a nap call my name...
[12:23:47] <joe9> anyone willing to commit this patch to the git repo: http://codepad.org/VIwoXJRX
[12:24:32] <ssi> k got those changed... let's check it out
[12:24:43] <joe9> anyone with linuxcnc git repo access in here?
[12:25:21] <skunkKandT> joe9: You might want to ask on #linuxcnc-devel or the devel mailing list.
[12:25:23] <ssi> pcw_: yep, that did it!
[12:25:27] <ssi> once again, pcw saves the day
[12:25:32] <joe9> skunkKandT: ok, will do. thanks.
[12:26:02] <Loetmichel> re @ home
[12:26:29] <joe9> yay, i have linuxcnc working now.
[12:27:14] <pcw_> I need to add a paragraph to the manual about driving 5V referred OPTOS
[12:29:05] <pcw_> so for 5V referred load you want 5V bus switch mode and 5V pullups
[12:33:56] <ssi> makes sense
[12:34:11] <ssi> pcw_: so the other thing I need to do is drive two solenoids that want 60mA at 12v
[12:34:21] <ssi> pcw_: is there any way in hell I can drive one of those with a 120VAC ssr?
[12:34:26] <ssi> I'm thinking probably not, because they're triacs
[12:34:51] <ssi> or scrs or some such sorcery
[12:37:50] <frysteev_> ola
[12:41:07] <mrsun> http://www.youtube.com/watch?v=wzpujoPgOLo&feature=g-vrec&context=G2a6db8fRVAAAAAAAAAQ haha cool =)
[12:41:55] <mrsun> tho my guess is that it wont hold up for very long as its not hardened ways on the screw =)
[12:45:34] <A2Sheds> pcw_ : do you have a DC motor controller for something up to 12HP?
[12:45:37] <pcw_> You can drive the solenoids on once with AC SSRs (but you would have to cycle the power to turn them off)
[12:46:02] <pcw_> Nothing that big
[12:46:25] <pcw_> 12 HP at what voltage?
[12:46:35] <A2Sheds> just checking, reworking a plastic extrusion system
[12:46:59] <pcw_> anything left of the original drive?
[12:47:01] <A2Sheds> I pick it up i the next couple days, probably 500VDC
[12:47:19] <A2Sheds> yeah, just old SCR drive
[12:47:41] <pcw_> well thats probably fine if its OK
[12:47:52] <A2Sheds> too old
[12:47:58] <A2Sheds> the cap dry up
[12:48:06] <pcw_> replace the CAPS
[12:48:08] <A2Sheds> cpas, solder connections etc
[12:48:17] <A2Sheds> I can't spell today :)
[12:49:04] <A2Sheds> http://www.scrcontrols.com/products.asp?productName=MELLTRUM&page=specs
[12:49:15] <frysteev_> im working on getting my machine power today
[12:49:20] <A2Sheds> probably a model 1 or 2
[12:51:16] <pcw_> Most likely the cheapest replacement drive would be a SCR drive as well
[12:51:36] <A2Sheds> I'll see what I find new
[12:53:13] <pcw_> Are you sure its bad? older industrial stuff is often better than new if fixed
[12:53:14] <pcw_> (hermetic power components etc)
[12:53:18] <A2Sheds> http://imagebin.org/204979 not quite a 60 scifi film, but close
[12:53:44] <A2Sheds> the front panel has analog meters
[12:54:03] <pcw_> But does it work?
[12:54:09] <A2Sheds> too oxidized to be reliable
[12:54:29] <A2Sheds> you should know that by now :)
[12:54:41] <frysteev_> A2Sheds: what is that?
[12:55:10] <A2Sheds> control cabinet for a 12hp DC motor and plastic extruder
[12:55:25] <A2Sheds> heater relays
[12:55:32] <pcw_> no, I know that most things are fixable if you understand whats wrong
[12:56:00] <A2Sheds> 15 year old board, I'd have to wave solder it again
[12:56:32] <A2Sheds> anyway, was just wondering if you had a larger driver
[12:58:18] <Loetmichel> HAAACH, great to sit in a warm bathtub after a dys work with angle grinder and stick welder... :-)
[12:58:23] <Loetmichel> days
[12:58:43] <A2Sheds> has the feeling come back to your fingers?
[12:58:51] <Loetmichel> *drip in the Notebookkeyboard....
[12:59:00] <Loetmichel> feeling?
[12:59:05] <Loetmichel> never lost
[12:59:13] <Loetmichel> the angle grinder is new
[12:59:22] <Loetmichel> not really vibrating much
[13:00:19] <A2Sheds> http://imagebin.org/204980
[13:00:50] <frysteev_> oo love the analog meters
[13:01:13] <joe9> my first run of linuxcnc gave me this message: http://codepad.org/J8WNGrT8 and the second run: http://codepad.org/71DlCrDo . can I assume that everthing is ok, now?
[13:01:43] * djdelorie is still anxious to get my cnc running *at all* :-P
[13:01:53] <A2Sheds> it screams reliability :p
[13:02:00] <djdelorie> stupid life! Stop interrupting me!
[13:02:20] <pcw_> A2Sheds you must be used to consumer stuff, 15 years is nothing for industrial electronics unless it in a corrosive atmosphere +- some dry electrolytics
[13:02:46] <A2Sheds> uncoated wave soldered through hole board
[13:03:13] <A2Sheds> I would fly that thing
[13:03:18] <A2Sheds> wouldn't
[13:03:54] <A2Sheds> probably more like25 years
[13:17:02] <ssi> how does one go about testing a prox switch?
[13:17:54] <JT-Shop> I use a quarter
[13:18:25] <ssi> I've tried everything from a wrench up to a 2x2x6" block of steel
[13:18:30] <ssi> can't get the output to change
[13:19:18] <JT-Shop> usually a quarter is fine
[13:19:28] <ssi> so that brings me back to
[13:19:33] <JT-Shop> might be borked or wired wrong
[13:19:35] <ssi> how can I Test it to figure out if the swtich is bad
[13:20:18] <JT-Shop> I've seen proxes that the built in led worked as expected but the actual switch did not
[13:20:37] <ssi> I don't think these have a built in LED?
[13:20:44] <ssi> I can take some pctures of them if you want
[13:21:06] <JT-Shop> I imagine they look like mine more or less
[13:21:20] <joe9> JT-Shop: any comments on this, please?
[13:21:22] <JT-Shop> 2 wire or 3 wire?
[13:21:37] <joe9> my first run of linuxcnc gave me this message: http://codepad.org/J8WNGrT8 and the second run: http://codepad.org/71DlCrDo . can I assume that everthing is ok, now?
[13:21:47] <joe9> jt-shop: ^^ this, please?
[13:21:55] <ssi> 3 wire
[13:21:56] <joe9> just to ensure that it is fine.
[13:22:04] <ssi> they look like fat bolts that are ground off on the end
[13:22:08] <ssi> with wires coming out the other end
[13:22:14] <ssi> they're threaded
[13:22:28] <ssi> they've got 12v on one side, ground on the other, and a signal on the third
[13:22:31] <JT-Shop> joe9: you did something to fix your real time delays between the two runs?
[13:22:39] <joe9> nope, nothing.
[13:22:51] <JT-Shop> ssi: a typical 3 wire prox
[13:23:10] <JT-Shop> joe9: then I assume you still have the problem reported in first run
[13:23:11] <ssi> that's what I'd expect
[13:23:23] <ssi> plus, the fine home limits are also proxes, of the same part num according to the drawing
[13:23:26] <ssi> and they work
[13:23:40] <JT-Shop> 12v 0v and signal might be either 12v or 0v depending on the prox
[13:25:01] <jdhnc> were they already wired?
[13:26:59] <ssi> jdhnc: yeah
[13:27:21] <JT-Shop> ssi: to answer your question I don't know if you can meter them or not
[13:28:11] <JT-Shop> does your machine use those tiny little press in terminal blocks at the slides?
[13:28:29] <JT-Shop> check for a broken wire if they are like mine they break easy
[13:29:29] <ssi> they're amp blocks
[13:29:35] <ssi> 1-582305-9
[13:29:43] <ssi> white things, I can't figure out how to release the wires
[13:29:48] <ssi> but they all look pretty solid
[13:31:23] <JT-Shop> mine are some kind of taper fit pin that you have to grab the fitting with pliers and twist to pull out
[13:32:03] <joe9> JT-Shop: now that I have my linuxcnc built, is there anything that I can do to ensure that everything is fine?
[13:32:04] <jdhnc> and if you are lucky, the fitting sockets don't rip out.
[13:32:38] <ssi> yeah I think these are the same
[13:33:15] <ssi> search for the part number comes up with
[13:33:16] <ssi> Taper Pins and Blocks
[13:33:24] <ssi> anyway, I tihnk the drawing might be wrong
[13:33:26] <JT-Shop> joe9: run the lantency test for an extended period if time
[13:33:31] <ssi> because it has me running power to one of the signal lines
[13:33:33] <joe9> ok, thanks. will do.
[13:34:24] <jdhnc> latency on my last test box hung around 30000 for hours until I copied about 5gb between PATA disk. Jumped to 900000 then
[13:35:05] <ssi> yeah I see what the damned problem is
[13:35:09] <ssi> drawing's screwed up
[13:36:17] <IchGuckLive> Hi all
[13:36:54] <pcw_> jdhnc that may indicate a drive problem
[13:39:10] <pcw_> I've seen latency jump to large number with disk errors
[13:40:05] <IchGuckLive> is the latacy champ pulsing every 10-25 seconds ?
[13:44:11] <ssi> of course it had to be the one wire that was too short to change
[13:44:27] <joe9> "dd if=/dev/zero bs=100M | gzip | gzip -d | gzip | gzip -d | gzip | gzip -d >| /dev/null" is this a good program to run when running the latency-test?
[13:44:40] <joe9> the menu talks about glxgears
[13:45:03] <joe9> which i started too.
[13:45:26] <joe9> and the max jitter with the above dd command and glxgears is around 12157.
[13:45:35] <joe9> for the base thread
[13:45:45] <joe9> and 10,657 for the servo thread.
[13:48:35] <alex4nder> yoh
[13:48:49] <joe9> alex4nder: yo, what's up?
[13:49:02] <alex4nder> joe9: not a lot, going to work on my mill today
[13:49:10] <alex4nder> well, the enclosure at least
[13:49:14] <joe9> is the above latency good enough. now, it is around 12500 for the base thread.
[13:49:20] <joe9> max jitter.
[13:49:34] <joe9> are those similar to your readings?
[13:49:39] <alex4nder> that's better than mine
[13:49:53] <joe9> wow, cool. let me run them for some time.
[13:50:38] <archivist> joe9, open an image in the image viewer and zoom it using the mouse scroll wheel
[13:50:54] <joe9> archivist: i am running glxgears and the massive dd.
[13:51:05] <joe9> along with that latency test.
[13:51:28] <joe9> alex4nder: now, I will open up the boxes and assemble the machine. Is there anything I should do regarding the software stuff.?
[13:51:54] <alex4nder> it depends on how you're driving your steppers.. what did you end up buying?
[13:52:16] <joe9> the kelinginc motors that you use.
[13:52:23] <alex4nder> I don't use keling motors. :D
[13:52:31] <joe9> but, i am not sure if I bought the bigger one's though.
[13:52:41] <archivist> joe9, it is just another test, if it can handle it all well and good
[13:52:59] <alex4nder> joe9: what stepper motor driver did you buy?
[13:53:01] <joe9> gecko 540, I think.
[13:53:04] <JT-Shop> make sure you run the latency test for a couple of hours or moe
[13:53:12] <JT-Shop> more
[13:53:12] <joe9> the one with 4 ports.
[13:53:17] <joe9> JT-Shop: ok, will do.
[13:53:19] <alex4nder> the g540 is easy to configure
[13:53:25] <joe9> yes, the g540
[13:53:25] <alex4nder> just use the configuration wizard
[13:53:30] <joe9> ok, will do.
[13:53:36] <alex4nder> I would recommend you only do one axis at a time.
[13:53:50] <joe9> i am excited. from this point, to getting the machine up and running, will it take a day?
[13:54:09] <alex4nder> uh, honestly?
[13:54:28] <alex4nder> you seem like you like going the 'long way around'. ;)
[13:54:28] <joe9> alex4nder: ok, thanks. connect one motor, test, disconnect, connect the next one to g540?
[13:54:32] <joe9> and so on?
[13:54:44] <ssi> rgh
[13:54:58] <ssi> even with the rewiring I can't make these damn proxes do anything
[13:55:02] <joe9> alex4nder: just for the software stuff. I am a software guy, so, that kinda rubs off.
[13:55:09] <alex4nder> joe9: I'm a software guy. ;)
[13:55:33] <joe9> like to have my linux configured, if you know what I mean.
[13:55:47] <alex4nder> what I did was I picked one axis, and picked where my eelctronics were going to live
[13:56:16] <alex4nder> then I mounted up the hardware to mount the stepper, on the mill
[13:56:40] <alex4nder> then I prepped the stepper, cut the coupler plastic, and mounted it all up
[13:56:42] <alex4nder> adjusted it
[13:57:21] <alex4nder> I did this after I was sure I was wiring them right, by using jumpers on the bench.
[13:57:59] <joe9> ok, thanks a lot.
[13:58:11] <alex4nder> you picked up current set resistors too right?
[13:58:17] <alex4nder> for the motors you're using?
[13:59:00] <archivist> I got X and Y and A running then it was producing work, took a month or few to get around to Z
[13:59:47] <joe9> archivist: one month?
[13:59:48] <alex4nder> haha
[14:00:00] <joe9> oh, ok. you waited.
[14:00:07] <joe9> i misunderstood
[14:00:13] <alex4nder> yah, I spent about a month experimenting and getting it to the point I could do a full run.
[14:00:20] <alex4nder> and trusted it
[14:00:21] <archivist> some months I hand set Z for a while
[14:00:43] <joe9> alex4nder: "current set resistors"? i just have the normal resistors.
[14:00:52] <alex4nder> what normal resistors?
[14:01:18] <alex4nder> I'm talking about any resistor you're using to tell the G540 the current limit.
[14:01:35] <joe9> oh, ok. nope, do not have them.
[14:01:51] <alex4nder> go through and read the g540 manual
[14:02:00] <joe9> ok, will do.
[14:03:59] <JT-Shop> joe9: the current set resistor is just a normal resistor used by the G540 to know how much to limit the current
[14:04:34] <joe9> jt-shop, oh, ok, thanks. i have a lot of them. just need to check the power rating, if that's the case.
[14:04:44] * alex4nder facepalms.
[14:04:52] <joe9> will check the g540 for the specs.
[14:05:58] <joe9> alex4nder: why the "facepalms"?
[14:07:55] <joe9> with all the programs running, top, that dd, glxgears, tail, the max jitter has come up to around 270,000 ns
[14:08:11] <JT-Shop> joe9: no power goes through the resistor
[14:08:20] <JT-Shop> just 1/4 watt is fine
[14:08:36] <joe9> oh, ok. I have more than a bunch of them. thanks.
[14:09:03] <ssi> jogging this damn Z axis by hand with an allen wrench is giving me hand cramps
[14:09:56] <pcw_> run the drives with a 5 v supply
[14:10:16] <ssi> I haven't done anything with the drives yet
[14:10:20] <ssi> they don't have a bias supply even
[14:10:23] <ssi> not sure if that matters
[14:10:33] <ssi> I should have main 90v power though
[14:11:31] <pcw_> I mean unplug the motor from the motor drives s and just run them with a little 5V 2A wall wart or something
[14:12:03] <pcw_> They are just DC PM motors
[14:12:15] <ssi> ahh I see what you mean
[14:14:26] <ssi> wow they're quiet
[14:14:26] <ssi> at least X is
[14:15:17] <ssi> yea Z is too
[14:18:13] <tehDarkAura> wheres a good spot to find rectangle tube for telescoping in the us?
[14:18:43] <ssi> great!
[14:18:51] <ssi> Z prox limits seem to be working
[14:18:54] <ssi> everything except Z coarse home
[14:20:16] <tehDarkAura> i want the support for my z axis to be split so my mill will come apart easily for portability...
[14:20:38] <tehDarkAura> so i figgures a 3x1 id rectangle would be perfect
[14:20:55] <JT-Shop> ssi: do you have an allen that fits a socket?
[14:26:19] <ssi> JT-Shop: no, but running the servos with gosh-darn electricity like pcw suggested works much faster :)
[14:27:00] <Thetawaves> how should i slope the bottom of the pan in my coolant system?
[14:27:34] <Thetawaves> i could use motor but that would be a pain in the ass should i want to move the table
[14:27:46] <Thetawaves> mortar*
[14:28:17] <ssi> beat the low spot with a hammer?
[14:28:40] <Thetawaves> heh
[14:28:55] <pcw_> Move the pan mounting holes?
[14:29:29] <Thetawaves> i've got a flat surface on the desk
[14:29:35] <Thetawaves> but i see what you mean
[14:30:10] <JT-Shop> ssi I rigged up a couple of switches and a power supply to jog mine around
[14:31:03] <alex4nder> joe9: just take your time with the CNC conversion.. there are no real 'instructions' on setting up the taig.. you have to pay close attention to all the details
[14:31:18] <joe9> ok, thanks.
[14:31:22] <alex4nder> you can easily break your brand new mill, and/or your drive electronics.
[14:32:07] <joe9> that scares me.
[14:32:20] <alex4nder> just read everything
[14:32:21] <joe9> will be patient and take it slow.
[14:32:24] <alex4nder> yah
[14:32:36] <joe9> what do you recommend for reading? the g540 manual, and what more.
[14:33:22] <alex4nder> http://www.cartertools.com/mounts.html
[14:34:26] <JT-Shop> joe9: your being paged
[14:35:09] <alex4nder> joe9: http://www.micro-machine-shop.com/lead_screw_conversion.htm
[14:36:19] <alex4nder> pretty much everything on micro-machine-shop.com is worth looking oer
[14:36:21] <alex4nder> +v
[14:36:55] <joe9> alex4nder: ok, thanks.
[14:44:15] <ssi> I gotta pull
[14:44:19] <ssi> -12v out of my computer psu
[14:53:37] <ssi> got it
[14:53:50] <ssi> I'm guessing the bias supply for these servo amps doesn't need to be super stiff
[14:53:58] <ssi> s/guessing/hoping/ :)
[14:58:01] <ssi> now I gotta figure out how the pwm outs work
[14:59:11] <ssi> I think the servo drives want a 0
[14:59:17] <ssi> 0-biased velocity command
[14:59:39] <ssi> but it looks like the 7i49 puts out a positive aout and then a +ena -ena
[14:59:51] <skunkKandT> unless you offset it- 0 command is 0v
[15:00:09] <ssi> yea i'm just not sure how to handle direction
[15:00:13] <ssi> without building some support circuitry
[15:01:34] <skunkKandT> what are the amps?
[15:01:41] <ssi> GE 50 series
[15:01:47] <skunkKandT> not +/-10v input?
[15:01:57] <ssi> yea I think they're +/-10v
[15:02:00] <ssi> or close to it anyway
[15:02:22] <pcw_> 7I49 is +-10
[15:02:23] <skunkKandT> you don't have mesa d/a?
[15:02:23] <ssi> ok nm
[15:02:29] <skunkKandT> heh
[15:02:31] <ssi> I'm just confused about what the 7i49 does :)
[15:02:36] <ssi> I see that it's +/-10v now
[15:02:46] <pcw_> You do need to set PWM mode to 2
[15:03:12] <pcw_> or you will have funny output behavior
[15:03:16] <ssi> that's output-type?
[15:03:39] <ssi> and what's pdm freq?
[15:03:44] <ssi> I'm setting pwm freq to 24000
[15:04:05] <pcw_> yes up down PWM instead of of PWM/DIR
[15:04:21] <pcw_> PDM freq is dont-care
[15:04:39] <ssi> ok
[15:05:11] <ssi> and I believe both enables on both amps are wired together right now, and I just get one enable signal back to my control
[15:05:18] <ssi> should I just pick one enable to use?
[15:06:09] <ssi> it's not actually enable, its pull down to inhibit
[15:06:53] <pcw_> pull down from what voltage?
[15:07:36] <ssi> it's 10.8 right now
[15:07:40] <ssi> might be high impedance though
[15:08:10] <pcw_> thats akward
[15:08:15] <ssi> yes
[15:08:41] <ssi> and no, it's definitely not high impedance :/
[15:08:56] <pcw_> because it need an 'on' switch to disable
[15:09:42] <pcw_> does that run a relay?
[15:10:18] <ssi> no that's a terminal on the drive itself, no relay
[15:10:40] <pcw_> so the amps are enabled by default if enable is unconnected? , thats just wrong
[15:10:52] <ssi> I guess so!
[15:10:57] <ssi> lemme read the book to double check
[15:12:43] <ssi> yeah, it seems to be an active inhibit
[15:12:48] <ssi> it used to be wired to a toggle switch
[15:13:12] <pcw_> maybe theres another enable
[15:13:47] <JT-Shop> mine needed to be pulled down to 0v IIRC to be enabled
[15:14:01] <ssi> oh really
[15:14:15] <pcw_> That makes more sense
[15:14:17] <ssi> this manual is clear as mud about it
[15:14:49] <ssi> jt: how much current does it draw when you pull it down
[15:14:56] <pcw_> maybe its active high "inhibit"
[15:15:00] <ssi> according to the blown fuse in my meter, the answer is >400ma
[15:15:18] <ssi> which seeems wrong
[15:15:49] <pcw_> That does seem wrong (more like you hit power)
[15:16:05] <ssi> ok now I see where it says "closes to run"
[15:16:56] <JT-Shop> ssi: I think I just used a little relay to enable the drives ... let me look
[15:17:15] <pcw_> you can alway use a 50 Ohm or so resistor to ground it when testing so you dont blow and more fuses
[15:17:31] <ssi> considering I don't have anymore fuses to blow at the moment, that's a grand idea
[15:17:36] <JT-Shop> no, it was CR20 that I used
[15:17:41] <JT-Shop> one if the SSR's
[15:17:41] <ssi> CR20?
[15:17:43] <ssi> ah ok
[15:17:58] <ssi> you have DC ssrs in yours?
[15:18:16] <ssi> mine has 11 AC ssrs, and they're all in the lathe, not in the control
[15:18:51] <JT-Shop> yes, they are all in the main power and drive cabinet, a whole slew of them
[15:18:59] <JT-Shop> about 30 or so
[15:19:26] <JT-Shop> they are all driven by the 7i37 outputs
[15:26:58] <ssi> I guess I'm gonna have to jumper them down until I can get some DC SSRs
[15:29:30] <ssi> oh wow
[15:29:34] <ssi> when I pull down inhibit
[15:29:37] <ssi> the monitor goes all wonky
[15:29:43] <ssi> it's sitting right next to the drives
[16:03:28] <ssi> so what does it mean if the servo drive sits at +20V when I'm not commanding a position
[16:04:27] <ssi> the other one seems to sit at 35V
[16:04:35] <ssi> I don't have the motors connected yet,
[16:04:51] <ssi> but I turned my ferror up to 5", and I'm able to jog and make the drives spin up to +100 or -100V
[16:04:52] <pcw_> input voltage is 0?
[16:05:44] <ssi> seems to be!
[16:07:24] <ssi> I apologize for my ignorance... I've never worked with servos before, and these drives seem a bit obtuse
[16:07:33] <ssi> and I'm just very scared of letting the motors loose to break stuff
[16:07:45] <ssi> especially the fragile and unobtainable little couplers
[16:16:58] <Mjolinor> why the hell do manufacturers use non standard things. I am rebuilding a smart engine and need a timing chain. 8mm pitch. Standard is 8mm pitch, 5 mm roller adn 3 mm internal width. If that is not strong enough then use duplex or triplex
[16:17:14] <Loetmichel> ssi: for the first test: decouple the motors from the leadscrews ;.)
[16:17:15] <Mjolinor> Not bloody mercedes they have a "let's make our own chain"
[16:17:48] <Mjolinor> so the timing chain is 8 x 5 x 5 nothing fits except a proper chain at 60£
[16:17:55] <Mjolinor> most annoying
[16:18:00] <Loetmichel> Mjolinor: normal. so they can sell it for exorbitant spare part prices
[16:18:16] <Mjolinor> normal 8mm chain is £30 for 5 metres, this chain is maybe half a metre long and it's £60
[16:18:23] <Mjolinor> robbing gits
[16:18:31] <ssi> Loetmichel: I don't know how easy that's going to be
[16:18:42] <ssi> Loetmichel: and the resolvers are coupled to the screw, not the motor
[16:18:51] <Mjolinor> I'm already in tehre to the tune of £300 and I haven't rebored the thing, just crank and a rering
[16:18:55] <Mjolinor> depressing it is
[16:19:12] <pcw_> ssi: I would first see if you can zero the drives output voltage with its analog input pins shorted together
[16:19:21] <ssi> ok
[16:19:49] <pcw_> there should be a input zero pot
[16:19:55] <Loetmichel> ssi: thats bad
[16:20:12] <ssi> there used to be a gain pot for each axis on the back of the control
[16:20:23] <ssi> but I'm not sure if those were attached to the drives or to the control
[16:20:25] <Loetmichel> maybe you can dial down the max. current to "nondestruct"-levels?
[16:21:13] <ssi> I can't find anywhere on the drives that those would have attached
[16:21:23] <ssi> and they were just wired into the point to point hell that was the logic backplane
[16:22:31] <pcw_> Output voltage with no load may not be terribly useful because the drives likely have a current control PI loop (which may saturate with no motor)
[16:23:22] <ssi> pcw_: by that do you mean that I need to just hook up the motors?
[16:24:34] <ssi> I'd like to figure out how to scale my pwm output so I can limit it... right now i can jog up to 15ipm velocity, and that produces full-scale voltage at the drives
[16:24:34] <ssi> 110V or so
[16:24:34] <pcw_> this is where you short out the inputs (dont short the 7I49 outputs, disconnect it)
[16:24:34] <pcw_> and very carefully enable the drives/ and be ready to pull the plug/enable
[16:24:36] <ssi> aha
[16:24:40] <ssi> I did wire the enable to a toggle switch
[16:25:19] <ssi> so with a shorted input, they should just sit
[16:29:01] <pcw_> they may creep or take off if there a problem
[16:30:17] <pcw_> If motors/tachs are disconnected the voltage out from the amps is meaningless
[16:30:26] <pcw_> (which means this is not time to worry about PWM scaling)
[16:30:28] <ssi> gotcha
[16:30:29] <ssi> jumpering the inputs now
[16:33:13] <ssi> looking good
[16:33:26] <ssi> hooked up the X servo, when I flipped enable the motor started whining a little bit but it didn't move
[16:36:57] <ssi> trying Z now
[16:37:08] <ssi> both look good
[16:37:17] <ssi> is it safe to try to hook up the PWM?
[16:37:25] <ssi> or what sort of things should I double check first
[16:37:54] <pcw_> I would not do that until you determine feedback direction
[16:38:11] <ssi> ok... not sure how to go about that
[16:38:53] <pcw_> can you make a small voltage source (say 1.5v battery divided by 5 or so so 300 mV)
[16:39:07] <ssi> I have an adjustable bench supply
[16:39:25] <pcw_> will it go down to 0?
[16:39:28] <ssi> I believe so
[16:39:30] <ssi> I'll double check
[16:40:27] <ssi> yeah it goes down to zero, and the readout is to 0.1v precision
[16:40:44] <ssi> with fine adjust I can probably dial it in 25mv increments easily
[16:41:54] <pcw_> make sure you are centered between your stops (so you have some time) and briefly connect
[16:41:58] <pcw_> the + power to the drives analog in and see (and write down) which direction the axis moves
[16:44:26] <ssi> ok I have like 120mv on the supply
[16:45:46] <pcw_> that _should_ be about 1.2% of full speed so safe, make sure your lab supply ground is same as the drive input gnd
[16:45:52] <ssi> yeah it's quite slow
[16:46:00] <ssi> moving positive
[16:46:05] <ssi> toward the tailstock, +Z
[16:46:22] <pcw_> so positive in = positive Z
[16:46:24] <ssi> yes
[16:46:30] <ssi> and resolver position is increasing toward +Z as well
[16:47:58] <pcw_> If you have the PWM setup and PID setup the 7I49 analog out should go negative when you move positive
[16:49:25] <pcw_> you probably have to restart EMC so that the command position and actual position start out at the same place
[16:50:56] <pcw_> (dont jog or home or anything))
[16:50:58] <pcw_> then if you move the motors positive, the analog out should move negative (to resist the "error" you are creating)
[16:52:14] <ssi> does it have to be in machine-on for that to happen?
[16:52:53] <pcw_> yes
[16:54:14] <pcw_> your basically running the machine with the feedback loop disconnected so you can see how it behaves before closing the loop
[16:54:22] <ssi> I can't seem to make that happening without it faulting due to follow errors
[16:54:37] <ssi> but I'm basically seeing nothing on the vout
[16:54:50] <pcw_> (to make sure you dont have positive feedback)
[16:55:12] <ssi> and the pid isn't really set up at all
[16:55:15] <pcw_> I though you had 5 inches of FE
[16:55:17] <ssi> it's set to 1/0/0 right now
[16:55:25] <ssi> I do have 5" of FE, but it's not seeming to let me
[16:55:55] <ssi> it was letting me jog for 5" of command before faulting before
[16:56:03] <ssi> but if the axis actually moves, it faults immediately
[16:56:09] <ssi> I didn't change MIN_FERROR, should I?
[16:56:17] <pcw_> yes
[16:57:28] <pcw_> closing the loop is not going to work too well until the resolver inputs are scaled correctly
[16:57:31] <ssi> ok yes
[16:57:39] <ssi> when I go positive, I get negative aout
[16:57:45] <ssi> and my resolver inputs are scaled correctly
[16:57:57] <pcw_> got some resistors?
[16:58:01] <ssi> I've put a dial indicator on the axes and ensured that I get linear travel equal to the delta position on the DRO
[16:58:04] <ssi> sure
[16:58:22] <ssi> oh you mean the 1v rms scaling
[16:58:27] <pcw_> i meant the electrical scale
[16:58:31] <ssi> right ok
[16:58:34] <ssi> I guess I need to get on that
[16:58:43] <ssi> so I need a divider per leg of drive, yes?
[16:58:47] <ssi> and those need to mach
[16:58:48] <ssi> match
[16:59:18] <ssi> and I'm looking for roughly 1/3 scale
[16:59:26] <ssi> I can't remember what overall impedance you said I needed
[16:59:37] <ssi> but 2k2/1k or 22k/10k should do the trick
[16:59:37] <pcw_> Yeah but we will send you a High input version sometime next week to try
[17:00:07] <pcw_> Yeah
[17:00:20] <pcw_> one the inputs (not drive)
[17:00:21] <ssi> for what it's worth, the dros seem to track the axis movement very well
[17:00:39] <ssi> ah I thought you had said that I could scale the drive for simplicity
[17:01:03] <ssi> if I Scale the inputs, do all four networks need to match very closely?
[17:01:14] <ssi> or just match the sin- network to the sin+ network etc
[17:01:17] <pcw_> You can scale the drive but its better to scale the inputs
[17:01:28] <pcw_> Yes we use .1% for all
[17:01:45] <ssi> I won't be able to use .1%, but I can hand select some 10%s :D
[17:02:08] <pcw_> I would just wait for our high input card
[17:02:25] <ssi> well if i try to close the loop as is, what can I expect?
[17:04:48] <pcw_> since the resolver is geared down so much (.020 per electrical turn) you will get repeatable cycle position errors with a 5 mill cycle (20 mills/quadrature) (might scream when you jog)
[17:05:34] <ssi> ok
[17:05:37] <ssi> I can live with that for testing
[17:05:42] <pcw_> s/cycle/cyclic/
[17:07:26] <ssi> I have Z jogging
[17:10:31] <Mjolinor> it can be cured, see a doctor
[17:13:50] <pcw_> you should go through the same exercise for X to make sure the feedback is correct before closing the loop (to prevent runaways)
[17:14:26] <DJ9DJ> gn8
[17:15:07] <pcw_> gn9
[17:18:05] <ssi> yeah working on that
[17:18:11] <ssi> for some reason I can't seem to move the X axis now
[17:18:42] <ssi> confirmed a half volt on the position voltage terminals on the servo amp, but I have zero volts out on the motor leads
[17:18:45] <ssi> and enable is in fact on
[17:22:28] <pcw_> hmm so the drives have just the enable in, analog and tach inputs and motor out?
[17:23:48] <ssi> and a +/-12v bias supply
[17:24:50] <pcw_> you could check pin/pin with Z and see if anythings different
[17:27:00] <ssi> yeah that's what I've been doing
[17:27:03] <ssi> I'm not sure what's changed
[17:28:09] <pcw_> is there a drive fault output, maybe you tripped a fault somehow
[17:31:58] <ssi> yea there's a /FAULTTRIP pin
[17:32:00] <ssi> lemme check it
[17:33:26] <ssi> seems like maybe it's a sinking output...
[17:33:33] <ssi> I see 1V on the Z drive, and 0V on the X drive
[17:33:55] <ssi> JT-Shop: does your machine have the GE PWM50 drives?
[17:34:24] <JT-Shop> no, siemens
[17:34:32] <pcw_> maybe cycle ye old power
[17:35:04] <ssi> oh yeah I forgot that cycling the enable doesn't cycle the power
[17:35:06] <ssi> I'm a brilliant one
[17:39:22] <ssi> no such luck
[17:39:25] <ssi> I'll have to deal with it later
[17:39:28] <ssi> I hope I didn't kill this amp
[17:40:17] <pcw_> the motor and tach are connected?
[17:40:24] <ssi> yeah
[17:41:04] <pcw_> are there drive manuals around?
[17:41:22] <ssi> I'm holding one
[17:41:27] <ssi> going through the troubleshooting section
[17:47:36] <ssi> nothing remarkably useful so far :/
[17:54:10] <mirage335> Is there an efficient FOSS tool for editing ngc files? pcb2gcode has left double isolation traces around some of my ground places, and I want to correct that
[18:01:23] <pcw_> vi?
[18:01:42] <mirage335> No. That is not efficient. :|
[18:02:16] <mirage335> I need something at least visual enough to tell which bits of code produce which traces.
[18:02:32] <mirage335> 12k lines of random text editing != fun
[18:02:45] <pcw_> Depends
[18:04:31] <pcw_> ssi is there a schematic?
[18:05:06] <pcw_> i would be tempted to trace the fault backwards from the fault output
[18:05:46] <mirage335> hmm
[18:05:49] <mirage335> There is a schematic.
[18:06:23] <mirage335> The problem is that pcb will draw isolation traces around the ground plane, as well as isolation traces around the tracks that contain it.
[18:06:34] <pcw_> Sorry that was for ssi
[18:06:38] <mirage335> I don't really think I can fix this in gschem or pcb.
[18:07:31] <pcw_> well i guess you could single step though the gcode around the interesting parts
[18:07:47] <mirage335> That's what I've been doing so far. Painful.
[18:10:16] <mirage335> Thanks for the help anyway. bbl
[18:10:23] <pcw_> I dont quite understand what you mean, what is the ground plane (and why would pcb2gcode know anything about it
[18:25:36] <raynerd> Guys, I was discussing this earlier today on here but never came to a conclusion. To attempt, attempt being the key word, what engravinh bits would you use for etching something like a dog collar tag? - i.e fine written engraving
[18:34:08] <raynerd> :(
[18:58:45] <mirage335> pcw_: Polygons in gschem are my ground plane (bulk copper areas connected to ground).
[18:59:08] <mirage335> pcb2gcode treated them as separate tracks, and drew isolation lines for them
[18:59:52] <pcw_> so its a PCB2gcode issue
[19:00:34] <pcw_> what about the voronoi scheme?
[19:06:23] <mirage335> Hmm...
[19:06:34] <mirage335> Voroni means all the extra copper is removed, right?
[19:11:13] <mirage335> In any case, I have fairly specific requirements for the circuit traces. I need them to be symmetrical as much as possible... this is for an instrument amplifier (normally these are built as a single chip).
[19:11:52] <mirage335> Its very important to keep inductance, capacitance, etc, symmetrical.
[19:37:32] <The_Ball> I have a rust problem in my shed, I get some slight surface rust on bare metal surfaces, so I've started monitoring temperature vs. humidity: http://wigen.net/shed/ At what relative humidity levels does surface rust start?
[19:38:14] <pcw_> is it humidity or condensation?
[19:38:44] <The_Ball> pcw_, good question, I don't know
[19:39:06] <The_Ball> For example on my lathe chuck there is a very fine almost dust like layer of rust
[19:40:08] <The_Ball> I've put in an air condition unit, and I want to use it to control the humidity, but I need a target
[19:44:13] <pcw_> I would suspect if you keep the lathe temp above the dew point you will not have anymore rust
[19:45:15] <pcw_> http://www.model-engineer.co.uk/forums/postings.asp?th=46666
[19:45:28] <joe9> after around 4 hours of running the latency test, the max jitter is around 270,925 ns. Is that good?
[19:45:36] <Jymmm> The_Ball: http://en.wikipedia.org/wiki/Dew_point
[19:45:43] <joe9> with heavy load.
[19:45:49] <Jymmm> The_Ball: http://www.dpcalc.org/
[19:45:55] <joe9> this is with pretty good load.
[19:49:03] <pcw_> Its fairly terrible, I would go through the WIKI latency issues checklist
[19:50:17] <joe9> oh, really. that bad, huh. let me check the wiki then. thanks.
[19:51:18] <pcw_> machines with decent latency like the Atoms are often in the sub 10,000 ns region
[19:52:52] <joe9> ok, thanks.
[19:54:27] <The_Ball> pcw, Jymmm, thanks
[19:57:30] <The_Ball> interesting reading, I've moved a server out into the shed which uses about 150W which might help it seems
[20:00:25] <pcw_> I think the trick is to apply some heat directly to the machine
[20:00:53] <pcw_> so its too warm for water to condense on
[20:02:06] <The_Ball> yeah, I could put some power resistors on the machines and run them a few hours before 0730, that looks like time when the damage happens
[20:42:03] <djdelorie> The_Ball: woodworkers use paste wax to protect iron from rust; either Johnson's Furniture Wax or Butcher's Bowling Alley Wax
[20:42:17] <djdelorie> those two are the only two known to *not* contain silicone, which messes up wood finishes
[20:42:23] <djdelorie> (in case you care ;)
[20:42:59] <cradek> I tried Johnson's on my mill (for under the vise). WD40 works much better for me.
[20:43:31] <The_Ball> djdelorie, ah, not a concern but interesting
[20:43:58] <ssi> cradek: have you ever gotten a fault on one of your HNC's servo amps?
[20:44:15] <djdelorie> silicone is a water repellent, which is normally good, except it's an everything-else-repellent too, so you get bare spots and fish-eye if you try to put a finish over it
[20:44:32] <cradek> I don't think so. I don't even think I remember them having a fault output?
[20:44:43] <ssi> they do have a fault output, but it's not hooked up in the original control
[20:44:54] <cradek> you having troubles?
[20:44:57] <ssi> my X amp isn't doing anything now all of a sudden
[20:44:59] <djdelorie> the wax itself dissolves in hydrocarbons, so it won't repel grease or oil if you later want to oil you machine
[20:45:01] <ssi> I have my Z axis jogging
[20:45:08] <cradek> ouch
[20:45:44] <ssi> I hope I'm just doing something dumb
[20:45:52] <ssi> I dunno what to do if i killed the amp :P
[20:46:08] <cradek> fix it? it's all discrete.
[20:46:24] <cradek> or at least I think it is
[20:46:38] <ssi> yea it is
[20:51:21] <cradek> I have schematics ... somewhere
[20:51:32] <ssi> I've got them here
[20:51:37] <cradek> within 10 miles of where I'm sitting
[20:51:40] <ssi> I'm reading through the manual to see if I missed anything
[20:51:47] <cradek> oh yay! much closer at hand than mine.
[20:51:53] <ssi> book says the inhibit circuit is 10mA sink
[20:53:51] <cradek> how dead is it? I assume you checked all the power supply lines.
[20:53:59] <ssi> yea I've got good power on the big cap
[20:54:08] <cradek> and the +-12
[20:54:12] <ssi> yep they're good too
[20:54:22] <ssi> I haven't checked it since I went to the store
[20:54:29] <ssi> maybe with it sitting powered down for awhile the fault cleared
[20:54:59] <cradek> I think I remember helping someone else debug a fault on one of those
[20:55:10] <ssi> cradek: are you using your PC power supply for the +/-12 bias?
[20:55:18] <cradek> I think we ended up disabling (with sidecutters) a fault detection that wasn't working right
[20:55:22] <cradek> yes
[20:55:29] <ssi> I wasn't sure how much current it requires
[20:55:36] <ssi> cause the PC supply is rated 0.6A for -12
[20:55:38] <cradek> teeny bit, I bet
[20:55:44] <ssi> I'm sure
[20:55:52] <cradek> it probably runs the op-amps and nothing else
[20:55:54] <ssi> probably just to run the amps
[20:55:54] <ssi> yea
[20:56:24] <cradek> I wish the schematics were online
[20:56:37] <ssi> oh hell
[20:56:39] <ssi> I bet I know what it is
[20:56:48] <ssi> you know
[20:56:51] <ssi> monday I'll bring them to work
[20:56:57] <ssi> I can run them through one of the big copiers
[20:57:04] <ssi> and it'll double-side scan a sheaf and email me a pdf
[20:57:07] <ssi> and I'll put it up somewhere
[20:57:30] <ssi> so for overcurrent or overvoltage faults
[20:57:38] <ssi> the inhibit clear is "remove bias voltage"
[20:57:43] <ssi> even though I've shut down the 90VDC power to the thing
[20:57:48] <ssi> I have to shut the computer down to remove bias
[20:57:50] <ssi> and I haven't done that
[20:57:58] <cradek> aha!
[20:58:16] <cradek> can you tell which fault it was?
[20:58:22] <ssi> not sure
[20:58:43] <cradek> I've worked on so many amps my memory is fuzzy about these
[21:02:52] <ssi> heheh
[21:03:52] <ssi> so I have ALMOST enough IO to do everything in two ports
[21:03:59] <ssi> I wanted to reserve a port for front panel
[21:04:19] <ssi> I wired a GPIO for the spindle brake, which I have removed and may not put back in
[21:04:23] <ssi> so I can steal that one back
[21:04:34] <ssi> but I need one for the coolant pump VFD
[21:04:42] <cradek> they sure run out fast, don't they
[21:04:50] <ssi> I might try to use one of the 7i49's PWM outs to run the coolant pump
[21:04:53] <ssi> not sure if that's smart or not
[21:05:10] <ssi> but I would like to use a GPIO for the inhibits on the servo amps
[21:05:24] <ssi> wire them to machine-on
[21:05:31] <cradek> yeah you absolutely must be able to turn the amps off
[21:05:41] <ssi> I have them wired to a toggle switch at the moment
[21:05:44] <cradek> they must turn off when you get a following error for instance
[21:05:47] <ssi> yeah
[21:06:04] <cradek> remember the 7i33 has enable-outs
[21:06:19] <ssi> OH RIGHT
[21:06:20] <ssi> yeah that'll work
[21:06:40] <Jymmm> The_Ball: You would be surprised how much heat even a 40W lightbulb can produce
[21:06:41] <cradek> I think they'll work, not sure if they're isolated differently (or not)
[21:06:48] <ssi> should work fine
[21:06:57] <ssi> just need to pull it down to ground, and the book says 10mA
[21:07:39] <ssi> anyway
[21:07:48] <ssi> I'm gonna go try removing bias supply and see if I can get the damn thing back online
[21:23:00] <ssi> ok it's working now
[21:23:13] <ssi> and I was able to confirm -AOUT for positive travel
[21:45:58] <ssi> cradek: the enables on the amps sit at 10.5V when they're not pulled down... the manual claims that it's 10mA when on
[21:46:10] <ssi> cradek: the 7i49 ENA+- inputs are optocoupled according to the book
[21:46:13] <ssi> can I do that directly?
[21:46:42] <djdelorie> based on those specs, you should be able to short that pin to ground to switch it...
[21:47:05] <ssi> yes, of course
[21:47:12] <ssi> and that's what I've been doing, through a toggle switch
[21:47:26] <ssi> I just want to make sure I can pull it down to ground with the ENA pins on the '49 safely
[21:48:06] <djdelorie> the opto isn't driving the pin enough, or something else is broken?
[21:48:19] <ssi> ?
[21:48:32] <djdelorie> sorry, coming in in the middle of the conversation...
[21:50:15] <ssi> according to the book, there are two enable inputs on the '49, and as long as both of them are low, all the aouts work and all the enables are switched on
[21:59:01] <djdelorie> grasping at straws here - do you have an ammeter ?
[21:59:23] <ssi> haha not anymore, sadly
[21:59:32] <ssi> wait, maybe i do
[22:00:01] <djdelorie> use the ammeter to short that enable input low; it will tell you how many milliamps are required. If the output you're using can provide that, you're good :-)
[22:00:23] <ssi> no, no I don't
[22:00:26] <ssi> dammit
[22:00:34] <ssi> and I think the book lies
[22:00:36] <ssi> book says 10mA
[22:00:41] <ssi> but I'm pretty sure this thing is eating optos
[22:00:44] <djdelorie> the 7I49 book?
[22:00:45] <ssi> so I don't think it's 10mA
[22:00:49] <ssi> no the servo drive book
[22:01:53] <djdelorie> you can add a bigger transistor on an opto output, in a darlington configuration, using the opto as the first stage transistor
[22:03:01] <djdelorie> add a series resistor on the opto to limit *its* current to 10mA or less, let the bigger transistor drive the servo
[22:05:34] <ssi> I have it pulled down to ground through 150R right now
[22:05:39] <ssi> and that's low enough to enable it apparently
[22:05:52] <ssi> but the enable pin is at 3.81v across that 150R resistor
[22:05:55] <ssi> which is 25mA
[22:06:15] <djdelorie> that will let the magic smoke out, sure.
[22:06:37] <djdelorie> heh. you have an ammeter *now* ;-)
[22:06:43] <ssi> yes exactly :)
[22:07:02] <ssi> I guess it's cause the two amps have their enable pins paralleled
[22:07:25] <djdelorie> it would be more than 10 if they weren't, though.
[22:07:31] <djdelorie> based on those numbers
[22:11:07] <djdelorie> that puts the internal pull-up at around 270 ohms, WAY too little resistance for a pullup, IMHO
[22:11:32] * djdelorie smells a short...
[22:11:55] <ssi> lemme check the scheme
[22:12:40] <djdelorie> given the pullup resistance, and the voltage it's pulling to, and the trigger voltage for whatever's watching that voltage, you should be able to calculate a maximum shorting resistance that still enables it...
[22:16:01] <ssi> pcw: I think I may have blown a couple of the optos on the ENA outs on my '49
[22:17:58] <ssi> djdelorie: I'm gonna post the scheme
[22:19:25] <ssi> http://www.prototechnical.com/~imcmahon/pwm50-scheme.jpg
[22:22:05] <ssi> I guess R61 (560Ω) is the "pullup" for that?
[22:22:25] <djdelorie> which pin are you connecting to?
[22:22:30] <ssi> E
[22:22:44] <pcw_> I think the 7I49 optos are 10 mA
[22:22:54] <ssi> pcw_: yeah, that's what the manual says
[22:22:56] <djdelorie> and V+ is... ?
[22:23:05] <ssi> pcw_: and the servo drive manual says that the enable input is 10mA
[22:23:17] <ssi> pcw_: but I neglected to take into consideration that both of them are paralleled :(
[22:23:20] <ssi> djdelorie: 12V
[22:23:26] <pcw_> 25 shouldn't hurt
[22:23:34] <ssi> djdelorie: the open state of pin E is around 10.5V
[22:23:35] <djdelorie> 12c / 560R = 21.4 mA
[22:24:20] <pcw_> but it may not pull low enough
[22:24:42] <alex4nder> evening
[22:24:44] <ssi> pcw_: well, I can put an ohmmeter across the ena+- pins on the optos that I haven't attached this thing to, and when I enable the machine the ohmmeter reads continuity
[22:24:46] <pcw_> But how did you blow your ammeter fuse on enable?
[22:24:50] <djdelorie> er, 12V. 10.5V yeah, the 10.5 is just the 12V times the 560/5600 resistor divider
[22:25:01] <ssi> pcw_: on the enas that I have hooked the damn thing to, I don't read continuity anymore
[22:25:12] <ssi> and I don't know how I blew my fuse on enable... I must have screwed something up
[22:25:37] <ssi> however, I pulled enable to ground through a 150R resistor, and I measured 3.8V across the resistor
[22:25:41] <ssi> so 25mA
[22:26:24] <pcw_> at 3.8V from ground...
[22:26:36] <djdelorie> to turn on Q40 you have to sink enough to bring R160 down to ... 4.7V or so, so E around 4V
[22:26:42] <ssi> and for what it's worth, when I had enable connected to the '49, it pulled that line down to 4.8V or so, which wan't enough to enable the amps
[22:26:58] <djdelorie> that's 6.3V across R61 (560R) or 11 mA
[22:27:24] <djdelorie> barely
[22:27:44] <ssi> I wonder if I should separate the two enable lines from the two amps
[22:27:49] <ssi> and pull each separately through its own opto
[22:28:17] <djdelorie> you're still pushing the limits for that opto output. If you just short E, you're sinking 21.4 mA
[22:28:24] <djdelorie> that's 43 mA when tied
[22:28:47] <pcw_> Yeah that could kill it
[22:29:07] <djdelorie> I think you need a bigger transistor inbetween those
[22:29:15] <djdelorie> at least you'll only be blowing the transistor from now on :)
[22:29:17] <ssi> I can do that if necessary
[22:29:21] <ssi> just is a pain
[22:29:34] <pcw_> (plus whatever zapped your meter)
[22:30:06] <djdelorie> in theory, you could swap out R61 and R160 for bigger valued ones, but I don't know what hFE is for Q40 or what it's driving
[22:30:46] <ssi> I'm pretty sure I have some TIP121's lying around
[22:30:51] * djdelorie assumes you know what a darlington configuration is...
[22:30:54] <ssi> I can probably deadbug one onto a terminal strip
[22:30:55] <ssi> yea I do
[22:31:03] <pcw_> Maybe a relay
[22:31:17] <ssi> relays and me don't get along so well right now
[22:31:51] <pcw_> just saying maybe the transistor drives a relay
[22:31:57] <djdelorie> Rb between the opto and the new transistor should be (12 - 0.7) / 0.010 = 1.2k or more
[22:32:09] <ssi> I don't really have much in the way of relays right now
[22:32:35] <ssi> I have a bunch of 24VDC relays but I don't have a 24V source in the machine anymore
[22:32:47] <djdelorie> max Rb would be 12 / (0.043amps / hFE) I think
[22:33:59] <pcw_> I wasn't suggesting a relay, just that a relay in the drive might explain the high drive current
[22:34:01] <pcw_> on the /enable input input --> transistor --> relay on drive
[22:34:15] <ssi> I don't think so
[22:34:24] <ssi> I posted the scheme for it
[22:34:36] <ssi> but I think that was before you joined :)
[22:34:45] <djdelorie> is this a PWM or a enable-once signal?
[22:34:45] <ssi> http://www.prototechnical.com/~imcmahon/pwm50-scheme.jpg
[22:34:58] <ssi> not sure what you mean
[22:35:07] <djdelorie> how often does the ENA signal change?
[22:35:22] <djdelorie> TIP121's hFE depends on frequency
[22:35:22] <ssi> pretty sure it's tied to machine-enable
[22:35:34] <djdelorie> less than 30 KHz ?
[22:35:43] <ssi> less than 0.0000001Hz?
[22:35:56] <ssi> on before machining, off when closing up for the day?
[22:36:18] <pcw_> on with machine on off when machine off or following error
[22:36:21] <djdelorie> ok, hFE is 3000, so max Rb is 3.3 Mohms
[22:37:10] <djdelorie> so 10k is probably a decent choice, 1mA through the opto and plenty to drive the E pin
[22:37:27] <ssi> yea that's what's in my drawing
[22:37:29] <ssi> putting it up for you
[22:38:03] <ssi> https://p.twimg.com/AozWCHHCEAAgAzc.jpg:large
[22:38:05] <ssi> thusly?
[22:38:17] <djdelorie> that's what I was thinking
[22:39:52] <ssi> I need to double check what transistors I can actually put my hands on
[22:40:48] <djdelorie> I think any transistor should do, you don't need much to get that to work.
[22:40:53] <djdelorie> well, any NPN transistor ;-)
[22:40:55] <ssi> yeah
[22:41:02] <ssi> ok I misremembered
[22:41:06] <djdelorie> biab - I need to work on my controllers for a while ;-)
[22:41:10] <ssi> it's TIP31 and TIP32 that I have
[22:41:24] <pcw_> a MOSFET would work as well from one of your direct FPGA outputs
[22:41:31] <djdelorie> no
[22:41:32] <ssi> trying to avoid using a GPIO
[22:41:53] <djdelorie> with a mosfet, you *also* need a resistor to discharge the gate when the opto shuts off
[22:42:03] <pcw_> right
[22:42:10] <ssi> djdelorie: hFE on the tip31 is only 50
[22:42:16] <ssi> djdelorie: 10k should still work in this case, yes?
[22:43:16] <djdelorie> 1mA * 50 = 50mA so a small Rb would probably be a good idea; 3.3K should work.
[22:43:46] <djdelorie> and hFE is 25 guaranteed for Ic=1A, that's the number to work with
[22:43:59] <ssi> ok
[22:44:15] <pcw_> or MOSFET if you have one
[22:44:24] <ssi> I definitely have BJTs
[22:44:27] <ssi> I don't know that I have MOSFETs
[22:44:48] <djdelorie> 22mA / 25 = 0.9mA or so, so a few mA through the opto is better
[22:45:01] <djdelorie> 10 / 3300 = 3mA
[22:46:09] <The_Ball> Jymmm, "You would be surprised how much heat even a 40W lightbulb can produce", yeah, my small office has three computer in it, gets unbearable after a while
[22:46:10] <djdelorie> with an N-mosfet, you don't need Rb but you do need Rgs, min 2k or so but much bigger is ok
[22:46:47] <djdelorie> when I got the air conditioner for my office, I added up all the electrical stuff in it, and it came to about 1000 watts. That's a *heater*.
[22:47:17] <djdelorie> ok, biab, and this time I mean it ;-)
[22:47:23] <ssi> thanks for the help
[22:47:26] <ssi> I'm gonna try to implement this
[22:47:59] <The_Ball> djdelorie, for sure, not sure how many watts a AC needs to cool 1000 watst on top of that
[22:54:29] <The_Ball> I'm quite impressed though, this is a 7,6m x 7.6m metal shed witha 8KW AC unit. Yesterday was a fairly hot day and it was only using about 1200 watts peak when cooling it: http://wigen.net/shed/
[22:54:57] <The_Ball> The roof has some very thin glass fiber mat insulation, but the walls are bare
[23:14:08] <ssi> SWEET
[23:14:11] <ssi> ok enables are working
[23:14:23] <ssi> plus I'm using one of the opto enables to run my main enable relay
[23:14:25] <ssi> it was on a GPIO
[23:14:36] <ssi> and the ones I thought were dead aren't in fact
[23:14:55] <ssi> both axes jog
[23:14:59] <ssi> they could use some tuning for sure
[23:17:43] <pcw_> Its a velocity mode drive so you will need good sized P, FF1 of one _If_ the units are right
[23:17:45] <pcw_> and probably 0 or very small D
[23:19:09] <ssi> right now it's 0 across the board except for a P of one
[23:19:18] <ssi> but I can see the X axis bounce a little bit as it stops
[23:19:45] <pcw_> thats going to be pretty sluggish
[23:20:55] <ssi> also I wasn't sure what to do with OUTPUT_SCALE and MAX_OUTPUT
[23:21:08] <ssi> the drive wants +/-8.5V, so I assume I need to use one of those to scale the output
[23:21:33] <pcw_> My preference is to set output scale so the the PID loop is normalized
[23:22:28] <pcw_> since this is a velocity mode drive you need to figure out how many IPS is full scale (10V)
[23:23:38] <pcw_> this is affected by the pot settings on the drive
[23:24:11] <ssi> yeah I recall reading that there's a "max speed" adjustment somewhere
[23:25:25] <ssi> ~2.3V is giving me 60ipm
[23:26:15] <pcw_> so roughly 4 ips full scale
[23:26:47] <ssi> sounds about right
[23:27:16] <ssi> trying to convince it to let me go faster than 1ip
[23:27:21] <ssi> 1ips
[23:27:31] <ssi> oh there it is
[23:28:38] <pcw_> so now for example if you set the PWM output scale to 4 (IPS) and FF1 to 1 you would have
[23:28:39] <pcw_> full scale with a commanded velocity of 4 IPS even without the PID loop terms
[23:28:40] <ssi> does 180 nicely
[23:29:18] <ssi> FF is feedforward?
[23:29:37] <pcw_> (I think of FF1 as centering the PIDs operating point about the current velocity)
[23:29:43] <pcw_> Yes FF
[23:29:56] <pcw_> =feedforward
[23:30:02] <ssi> 3ips is 7.15V
[23:30:21] <pcw_> dont slam into the stops
[23:30:46] <ssi> limits won't let me do that thankfully
[23:30:51] <ssi> that's why I made sure I had limits first
[23:30:52] <pcw_> are your limits working?
[23:30:54] <ssi> yep
[23:31:10] <ssi> I know because I just had to manually turn the screw to get Z back off the limit :)
[23:31:16] <pcw_> now is the time to reduce your FE for safety
[23:31:17] <ssi> and there's like 4" of extra bed beyond the limit
[23:31:31] <ssi> yea, about t hat
[23:31:39] <ssi> I tried to reduce it, and I got follow errors immediately
[23:31:44] <ssi> even with them set to 0.1"
[23:32:04] <frysteev_> yo
[23:32:07] <pcw_> well you have a lot of tuning to do
[23:32:24] <ssi> is there a pin or signal where I can look at the current fe as I'm jogging?
[23:32:36] <pcw_> yep
[23:32:40] <ssi> yea I found it
[23:33:00] <pcw_> halscope is what you need now
[23:33:16] <ssi> so it gets up to around .19 while jogging
[23:33:59] <ssi> ok what should I be scoping?
[23:34:01] <frysteev_> is the 96 i/o mesa board supported with emc?
[23:34:33] <pcw_> well consider with a P gain of 1 you need to get behind 10 inches for full speed
[23:34:57] <pcw_> (since the output scale is 10)
[23:35:20] <pcw_> frysteev_ 5I22? yes
[23:35:44] <frysteev_> cool
[23:35:46] <ssi> frysteev_: but it's a million billion dollars
[23:36:11] <frysteev_> well if i have to replace the cpu in the laser cutter thats when im going to get
[23:36:42] <ssi> pcw_: I had monkeyed with output_scale and max_output earlier... let me put them back to 10 which I think was the default on each
[23:37:12] <ssi> now it behaves real crazy
[23:37:16] <ssi> very very sluggish
[23:37:21] <ssi> I'm guessing that's the low P term
[23:37:57] <ssi> setting P to 10
[23:38:16] <ssi> now it's back to approx the responsiveness it was before
[23:38:42] <pcw_> so ramp up P (slowly) until it gets unstable
[23:39:04] <ssi> is there a way to attach it to a control in axis so I can tweak it without having to edit the ini and restart every time?
[23:39:33] <pcw_> yes, its called calibrate or something
[23:39:48] <frysteev_> i like the "or something" menu
[23:39:49] <ssi> aha
[23:41:02] <pcw_> its a little awkward making the values stick as I recall but better than edit/restart/shutdown
[23:41:08] <ssi> how will I know when it's getting unstable
[23:41:14] <ssi> (also should I be scoping something?)
[23:41:28] <pcw_> you can feel/hear it
[23:41:52] <ssi> it's starting to accel faster
[23:42:11] <pcw_> set your FE to 1/2 inch or something so if you lose control it FEs
[23:43:55] <djdelorie> ssi: darlington worked?
[23:44:14] <pcw_> To really get good performance you may have to tune the drives (with the pots and the drive manual)
[23:44:24] <ssi> djdelorie: beautifully :)
[23:44:37] <djdelorie> now if only my problem was that easy to solve :-)
[23:45:28] <ssi> pcw_: up to 90 and it still feels about the same
[23:45:43] <ssi> pcw_: at 60ipm it's pretty good, 240ipm is smooth but the accel and decel feel really slow
[23:46:11] <pcw_> well maybe they slow in the ini file
[23:46:17] <ssi> I set them to 5
[23:46:28] <ssi> but I'm not sure if maybe there's a master max accel that I missed
[23:46:51] <pcw_> that is slow
[23:47:03] <ssi> default was like 2
[23:47:12] <ssi> 5i/s^2 is slow?
[23:47:16] <ssi> seems like that'd be pretty fast
[23:47:21] <pcw_> yes
[23:47:48] <ssi> I'll try 20
[23:48:05] <ssi> yes that's much better
[23:48:22] <ssi> at 4ips and 20i/s^2, it rather hums along :D
[23:49:32] <pcw_> Now you get halscope out and start looking at the FE
[23:50:26] <pcw_> or get some rest , it must be late there
[23:50:43] <ssi> nah, not really :)
[23:51:41] <pcw_> dont mess with it if your are getting tired, thats when bad mistakes happen
[23:52:06] <ssi> yeah, I hear ya there
[23:52:17] <ssi> I'm not tired yet, I'm just struggling to understand how to use halscope :)
[23:52:29] <ssi> I also haven't found any instability yet due to P
[23:53:14] <ssi> ok maybe that's it?
[23:53:22] <ssi> after some moves, servo sounds like it's buzzing?
[23:53:32] <pcw_> well you're a better man than I as I'm getting sleepy...
[23:54:10] <ssi> aha, that was a subtle hint then ;)
[23:54:16] <ssi> I won't keep you, thank you so much for your help
[23:55:09] <pcw_> yes buzzing could be it (but there may also be resolver troubles due to the overdrive (dead spots) these will get in the way of the best tuning
[23:55:28] <ssi> ah right
[23:55:34] <pcw_> and with that
[23:55:36] <pcw_> g'nite
[23:55:37] <ssi> oh I bought some 2% resistors today... not really ideal but maybe I can match them
[23:55:41] <ssi> night :)