#linuxcnc | Logs for 2015-05-09

Back
[00:18:58] <zeeshan> finally home
[00:24:14] <Tom_itx> crap, i gotta remap some pins in order to swap boards
[00:24:26] <zeeshan> ??
[00:24:41] <Tom_itx> changing the spindle pwm source
[00:24:54] <zeeshan> youve been working on that thing too long :P
[00:25:10] <Tom_itx> naw, it's been sitting for 2-3 weeks now
[00:25:21] <Tom_itx> haven't had time to mess with it
[00:25:31] <Tom_itx> besides i got it working with the C6 board
[00:25:31] <zeeshan> why not
[00:25:34] <zeeshan> cnc is priorty in life
[00:25:43] <Tom_itx> now i'm hooking up the 7i47S for spindle control
[00:26:06] <Tom_itx> that and a couple relays should do it
[00:26:39] <Tom_itx> i might have to re-time the delays on the relays since i'm switching those too
[00:26:49] <Tom_itx> but they should be pretty close
[00:36:41] <zeeshan> ah cool
[02:09:58] <Deejay> moin
[02:10:20] <XXCoder1> yo
[03:51:05] <RyanS> does cross table (x, y) size 700 x 210mm, x travel 450mm give enough data to tell what the max length of part i can face is? (i dont have it yet)
[03:52:24] <RyanS> i need to flycut 150x12mm 800 mm flat steel
[03:53:13] <RyanS> so i guess do in 2 sections
[06:17:26] <jthornton> don't forget to add the diameter of the fly cutter to the travel
[06:46:48] <SpeedEvil> Only as much as the cutter doesn't hit stuff
[06:51:39] <archivist> and if you over hang the setup
[06:53:39] <_methods> i hate fly cutters
[06:53:52] <_methods> they always scare the shit out of me
[06:55:31] <archivist> duck :)
[06:56:41] <jthornton> I have a nice balanced 6" one but it still flexes too much so I'd rather use my 2.5" face mill
[07:00:10] <_methods> i like to keep my hands behind my back while i run them lol
[07:00:27] <_methods> i know a guy that lost 2 thumbs on 2 diff occassions to fly cutters
[07:00:34] <jthornton> do you use a stick to press the start button?
[07:00:38] <_methods> the 2nd time he said not again
[07:00:44] <jthornton> I guess not
[07:00:46] <_methods> lol
[07:00:52] <jthornton> out of thumbs
[07:00:57] <_methods> no shit
[07:01:09] <_methods> no it was more of an exclamation of holy shit not again
[07:01:31] <_methods> an exclamation of incredulity
[07:01:36] <jthornton> heh
[07:01:38] <_methods> is that even a word
[07:01:41] <jthornton> I can see that
[07:01:43] <jthornton> dunno
[07:01:51] <_methods> it is
[07:01:55] <_methods> holy shit hahah
[07:02:17] <_methods> the state of being unwilling or unable to believe something.
[07:04:24] <_methods> so needless to say i'm extremely cautious around fly cutters
[07:05:19] <_methods> we used to use "slip steel" blades in the shapers when i used to work for an architectural wood working place
[07:05:24] <_methods> that's some scary stuff too
[07:06:55] <jthornton> damn glade wants to use GTK3 and I tried that with some issues
[07:06:58] <_methods> http://www.woodweb.com/knowledge_base/Antique_unsafe_shaper_knives.html
[07:07:10] <_methods> dependency nightmares
[07:07:32] <SpeedEvil> Shapers are nice and sane.
[07:07:34] <_methods> so you have stuff using old gtk and new gtk
[07:07:35] <jthornton> yea, sucks
[07:07:39] <SpeedEvil> When everything is clamped down just right.
[07:07:59] <_methods> yeah normal shaper blades are cool as long as a moron doesn't load them lol
[07:09:30] <SpeedEvil> The problem is that 99.9% of the time you might do it right.
[07:09:32] <jthornton> I have a mill calculator, G code generator that I wrote in Ubuntu 10.04 and that uses GTK2
[07:09:35] <SpeedEvil> Are you sure about the .1%?
[07:09:51] <jthornton> now I'm on sneezy and it uses GTK3, it being Glade
[07:09:58] <_methods> yeah i always try to stand behind something solid when i see a new guy about to do a shaper/moulder setup
[07:10:22] <_methods> When these knives were used in the sixties, mill workers wore lead aprons to protect their manhood. I've never seen one unscathed or without a story. - See more at: http://www.woodweb.com/knowledge_base/Antique_unsafe_shaper_knives.html#sthash.GHtr3G8P.dpuf
[07:10:26] <_methods> lol
[07:12:43] <SpeedEvil> _methods: Plus, shit happens.
[07:12:45] <SpeedEvil> http://en.wikipedia.org/wiki/NOAA-19#Damage_during_manufacture
[07:16:11] <_methods> yeah that too lol
[07:16:49] <_methods> i had to get out of woodworking, too many body parts disappearing in that line of work
[07:17:11] <SpeedEvil> Metalworking at least until carbide went comparatively slowly
[07:17:18] <_methods> i still do my own woodworking stuff but no more production woodwork for sure
[07:20:11] <_methods> jthornton: is sneezy the latest and greatest debian or something?
[07:20:45] <_methods> i wish they would just use a damn number system for their os
[07:20:58] <jthornton> dunno, the LinuxCNC live CD is based on debian wheezy
[07:21:09] <_methods> yeah i'm on wheezy
[07:21:18] <jthornton> but you can run 2.7 on Ubuntu 10.04 as my plasma is doing that
[07:21:22] <_methods> i thought jessie was the latest
[07:21:44] <_methods> i have 10.04 on my old little pcb mill
[07:21:54] <jthornton> I only went for wheezy because of genie version was much newer and more stable
[07:22:24] <_methods> well i try and stay on stable for most things
[07:22:35] <_methods> if you get ahead or behind it's always issues it seems
[07:22:50] <_methods> harder to find solutions to issues
[07:24:24] <jthornton> well saving my mill G code generator in Glade 3 didn't seem to mess it up
[07:25:20] <_methods> well that's good thing
[07:26:35] <_methods> yeah i use the hell out of your pocket generator
[07:27:20] <jthornton> which one is that?
[07:28:19] <_methods> one i got off the linuxcnc script page
[07:28:57] <jthornton> the one Sammel did? Pocket.py?
[07:29:09] <_methods> yeah that one
[07:29:13] <_methods> oh i thought you did it
[07:29:35] <_methods> i thought your name was in the readme
[07:29:37] <_methods> hmm
[07:29:49] <jthornton> yea, he took my code and added to it to make pocket py
[07:30:48] <_methods> well i use it a lot when i'm too lazy to go upstairs and program something lol
[07:30:51] <jthornton> I made the framework for using Tkinter and widgets and jepler showed me how to send it to Axis
[07:30:57] <_methods> and too lazy to write pocket by hand
[07:31:18] <_methods> which is 99% of the time lol
[07:32:13] <jthornton> CAM is so overrated and cumbersome for most things
[07:33:56] <jthornton> I need to incorporate his code into my mill G code generator
[07:40:55] <jthornton> damn Glade 3.12.1 is buggy
[07:48:08] <_methods> oh yeah?
[07:48:44] <jthornton> yea, trying to type in a frame name and it keeps putting the cursor at the end after every key stroke
[07:49:07] <jthornton> so if your trying to type between something you have to keep putting the cursor back
[07:52:26] <_methods> oh that's convenient
[07:52:33] <_methods> i love it when they add helpful things lol
[07:52:34] <jthornton> yea
[07:54:20] <archivist> how does crap like that get past any sort of quality control
[07:55:15] <archivist> like the disappearing scroll bar on win 8
[07:55:31] <jthornton> so if I put the cursor at the start of a word and type a letter it puts the letter at the end of the word
[07:55:42] <jthornton> how nice
[07:56:17] <jthornton> oh I highlighted the letter to type over it and it put the corrected letter at the end lol
[07:56:37] <archivist> needs a verbose with expletives bug report :)
[07:57:02] <jthornton> all they will say is why are you using an out of date version of Glade
[07:58:52] <jthornton> lol a grid property is orientation... it's a grid how can it be horizontal or vertical?
[07:58:59] <archivist> same thing I hate about boost if it is allowed to infect a project
[08:05:02] <socara> oh
[08:13:02] <jthornton> Warning: cannot create instance of abstract (non-instantiatable) type `GtkBox' self.builder.add_from_file(current_path + "data/mill.glade")Segmentation fault
[08:13:06] <jthornton> nice
[08:20:08] <_methods> oooh seg fault
[08:20:10] <_methods> that's nice
[08:26:53] <jthornton> according to this page https://glade.gnome.org/sources.html you can get glade for either GTK2 or 3
[08:28:01] <jthornton> I'm beginning to not like Glade any more
[08:30:18] <jthornton> oh I found glade-gtk2 in the synaptic package manager
[08:48:52] <mozmck> jthornton: what distribution are you using?
[08:57:26] <jthornton> distribution?
[09:06:10] <pcw_home> Debian Wheezy, Jessie
[09:06:11] <pcw_home> Ubuntu Precise, Trusty, Vivid etc
[09:06:51] <jthornton> Debian Wheezy
[09:07:08] <jthornton> from the LinuxCNC LiveCD
[09:36:26] <zeeshan> jthornton: you should post some pics of your macros :P
[09:37:56] <zeeshan> a couple macros that i want for the mill: facing macro (generates spiral or zigzag path based on work piece face, and cutter diameter), chip break drill cycle for bolt circles
[09:50:14] <CaptHindsight> https://www.kickstarter.com/projects/1598272670/chip-the-worlds-first-9-computer good one more year before people start trying to run Linuxcnc or recrapware on this
[09:50:42] <archivist> a bit tight on memory though
[09:50:55] <zeeshan> hi archivist
[09:51:00] <zeeshan> i found my first gear project..
[09:51:01] <zeeshan> :-)
[09:51:09] <archivist> oops
[09:51:12] <zeeshan> http://i130.photobucket.com/albums/p252/turbozee84/D3EB63F3-691B-43CC-91B8-960AEEFCB38F-4335-00000553821375A1_zpsddabb053.jpg
[09:51:21] <zeeshan> need to make the hypoid rack and pinion gear
[09:51:27] <zeeshan> andd...
[09:51:39] <zeeshan> http://drive.google.com/uc?export=view&id=0B-s8Xj1_LdfyYWMwdTduT0x2Uk0
[09:51:42] <zeeshan> the spline on the top :P
[09:52:17] <archivist> photophukkit where a white screen is the image of the day
[09:53:10] <archivist> I see something a a couple of nanoseconds then it goes blank with a loading timer
[09:53:16] <zeeshan> lol
[09:53:36] <zeeshan> http://www.kijiji.ca/v-tool-other/guelph/tool-and-die-machinist-tools/1070570279?enableSearchNavigationFlag=true
[09:53:40] <zeeshan> finally someone with 246 blocks
[09:54:24] <CaptHindsight> I have 2 4 6 8 blocks :)
[09:58:04] <archivist> another photofukkit example http://www.collection.archivist.info/archive/grabs/Screenshot-24.png
[10:00:27] <recon_lap> wonder if anyone can help me setup my first cnc job on a large machine. got a few questions. like if i have a 5mm slot I want to carve , what would be a suitable bit size?
[10:01:16] <recon_lap> 5mm r x 18mm depth
[10:03:42] <Tom_itx> you want room to do a final pass and to acomodate tool deflection so i'd use something smaller than 5mm
[10:03:47] <Tom_itx> 4mm if you can get them
[10:03:57] <Tom_itx> i don't generally deal with metric
[10:04:04] <archivist> rather deep, there will be some spring of the tool, decide on accuracy and finish needed, go for the largest most solid slot drill, 4.8 maybe
[10:04:57] <Tom_itx> take several passes
[10:05:03] <Tom_itx> until you reach depth
[10:05:14] <Tom_itx> then take a final cleanup on the wall
[10:05:55] <archivist> have a lot of lube to clear chips
[10:07:12] <Tom_itx> http://zero-divide.net/index.php?page=fswizard&shell_id=199&load_tool_id=27011
[10:07:15] <recon_lap> lube?
[10:07:22] <archivist> coolant
[10:07:22] <Tom_itx> that might help whit cuter speeds & feeds
[10:07:33] <Tom_itx> flood coolant or air mist
[10:07:47] <recon_lap> this will be hardwood plyboard 18mm beach
[10:08:01] <Tom_itx> beech is pretty hard too
[10:08:07] <Tom_itx> but no coolant on wood
[10:08:09] <Tom_itx> maybe air
[10:08:10] <recon_lap> sry, thats sort of important info
[10:08:46] <archivist> you still need to get chippings out, air/vacuum
[10:08:49] <Tom_itx> i use 20-30 ply beech hardwood on occasion
[10:08:51] <recon_lap> just have a extractor hood , it's a shopbot cnc
[10:09:42] <recon_lap> looked like 8 or 9 ply to me, did not count though.
[10:11:15] <recon_lap> ok, next question, I got a 20mm circler hole (18mm dept), should I select it as a pocket and route all the material out?
[10:11:37] <recon_lap> and what bit size would you recommend?
[10:11:54] <archivist> often makes sense so you dont have the middle fly out
[10:13:18] <Tom_itx> yeah start from the center out
[10:15:58] <recon_lap> I got some smaller holes to drill , 10mm 4mm and 3mm , should I use 3 appropriately sized bits and do 3 separate drilling passes with tool changes?
[10:16:39] <recon_lap> and I take it I should just use regular wood drill bits?
[10:17:03] <Loetmichel> re from buying grocerys... and the bodyshop hasn't finished my car... now i have to commute monday and maybe tuesday with the 50cc bike ... one hopes it dosent rain. Why cant they be true t their promises?
[10:40:15] <jthornton> zeeshan, is has everything in one app
[10:41:14] <archivist> recon_lap, router bits as you are side cutting not drilling
[10:42:12] <jthornton> http://ibin.co/2193k2vcYUfV
[10:43:34] <jthornton> zeeshan, http://gnipsel.com/files/g-code-generator/ is the latest one, but I'm working on it now
[10:43:50] <jthornton> some things work some do not, they are labeled if they don't work IIRC
[10:44:20] <recon_lap> archivist: I asking to clear up these details, on large wood CNC jobs, do you only use side cutting bits, or do you match the bits to the current job, like plunge bits for screwholes and do tool changes.
[10:45:14] <jthornton> renamed link http://gnipsel.com/files/mill-g-code/
[10:53:24] <archivist> recon_lap, depends on the job, machine and ease or not of tool changing
[10:54:35] <recon_lap> archivist : lol, well whats considered best practice ?
[10:55:18] <recon_lap> like I could just use a 2.8mm bit for the whole job ?
[10:55:53] <archivist> only if you are willing to wait, would be a lot slwer using a small bit
[10:56:01] <archivist> slower
[10:56:57] <recon_lap> ok, and what about plunge drill bits? recommended or to be avoided ?
[10:57:30] <archivist> it is a choice you have to make between speed, accuracy and a pain in the a changing tools if you dont have auto tool changer
[10:58:27] <archivist> a plunge/normal drill bit is better for a plain circular hole
[10:59:06] <recon_lap> archivist: thx, thats the type of info I looking for.
[10:59:42] <Loetmichel> recon_lap: you could "drill" it this way: https://www.youtube.com/watch?v=2wS_URQsyTU
[10:59:50] <recon_lap> I've only got 2 drill hole sizes and 2 bit sizes required for this job.
[11:00:14] <recon_lap> so 4 tool changes, not that much I would think
[11:00:47] <Loetmichel> or like this (careful. LOUD, long amuminium tube) https://www.youtube.com/watch?v=UEd4LCJ3uWk
[11:05:17] <Tom_itx> recon_lap, also use the shortest tool that will suit the job for less deflection
[11:07:04] <recon_lap> will keep that in mind, unfortunately the place I going to do this has not given me a list of available bits, I'm just trying to figure out the software and the job before I end up on the machine atm.
[11:12:38] <Tom_itx> you better know what tools you have before you plan the job
[11:13:03] <Tom_itx> if you program a tool you don't have you'll have to redo it
[11:18:54] <recon_lap> Tom_itx: yep, I know, but since all this is relatively new to me , the more things I can practice before ending up on the machine the better. and as I only have access to vcarve demo I'll have to reprogram it all anywas when I get to the machine.
[11:22:05] <recon_lap> think I've got the first full tool path done, just cant see if my pocket cuts are starting from the inside to the outside.
[11:53:05] <MrHindsight> archivist: there's a traveling Job Shop show that was here last week, it was interesting talking to all the gear shops about custom machines, tools and hobs
[11:53:34] <MrHindsight> one place said they had over 700 custom hobs
[11:54:59] <archivist> that is a lot of money invested
[11:58:07] <MrHindsight> had some great samples on display
[11:59:04] <MrHindsight> 2ft dia ring 1/4" thick with helical teeth on the ID
[11:59:29] <MrHindsight> OD - ID was ~ 1/4"
[11:59:40] <MrHindsight> tiny teeth as well
[12:08:47] <archivist> you do see the good bits at shows
[12:16:24] * JT-Shop wonders why he is even working on a pocket program... he has never machined one
[12:21:25] <Tom_itx> soon as you don't you will need one
[12:30:16] <JT-Shop> lol
[12:30:45] <Tom_itx> ok, new bitfile generated... now all i gotta do is rewire it to match
[13:10:15] * JT-Shop thinks cleaning out the plasma water table is much more fun than a coolant tank any day
[13:15:44] <zeeshan-mill> is there a cheap and easy way to hook up a rtd sensor or some other temp sensing ability ?
[13:15:47] <zeeshan-mill> through 7i77
[13:16:04] <zeeshan-mill> 2of em
[13:16:20] <Jymmm> dallas chip?
[13:16:47] <zeeshan-mill> wassat
[13:16:57] <Jymmm> https://www.sparkfun.com/products/245
[13:16:59] <pcw_home> well a 7I87 would give you 6 channels of +-10V analog in
[13:17:16] <Jymmm> zeeshan-mill: DS18B20
[13:18:17] <zeeshan-mill> imhoping it would be as simple as hooking two wires up :P
[13:18:24] <Jymmm> zeeshan-mill: You can even have 400 connected together
[13:18:55] <Jymmm> zeeshan-mill: each one has a unique serial number
[13:21:38] <Jymmm> zeeshan-mill: and they are programable
[13:38:22] <CaptHindsight> zeeshan: how many do you need hooked up only ~187?
[14:18:28] <witnit> measuring across ground and encoder input on 7i33 is giving me 1-2v is this normal?
[14:19:45] <pcw_home> If its in differential mode, yes
[14:19:51] <witnit> every encoder channel on 7i33 is giving this reading except the last encoder b and b-not
[14:20:41] <witnit> b and b-not on enc3 both read 0v
[14:21:03] <pcw_home> for differential encoders the important measurement is between A and /A or B and /B
[14:21:52] <witnit> I cant seem to get it to read my encoder, but my desk oscope and voltmeter seem to register the pulses fine
[14:22:34] <pcw_home> if theres no difference between B and /B, the encoder or wiring are faulty
[14:22:45] <witnit> is there any reason why just the enc3 b and /b are reading 0 to ground?
[14:23:11] <pcw_home> bad encoder or wiring?
[14:23:17] <witnit> no wires attached to card
[14:23:24] <witnit> just reading from pin to pin
[14:23:33] <witnit> or pin to ground rather
[14:27:28] <pcw_home> I had forgotten, the 7I33 has no pullups on the inputs so basically
[14:27:30] <pcw_home> any voltage between 0 and 5V is possible if they are unconnected
[14:28:04] <witnit> ahh, okay, I just cant seem to get the scope to read my encoders pulses
[14:28:12] <pcw_home> (due to leakage from the RS-422 receivers)
[14:29:58] <pcw_home> if the are differential, make sure the 7I33 is jumpered for differential inputs
[14:29:59] <pcw_home> then check between A, /A and B,/B with a voltmeter (as you twiddle the encoder to get all 4 states)
[14:33:46] <witnit> thx pc
[14:33:54] <witnit> bbs!
[14:34:50] <Tom_itx> man, it's raining buckets here
[14:35:13] <pcw_home> send some here
[14:35:22] <_methods> keeps acting like it's going to here then the sun comes out lol
[14:35:27] <Tom_itx> http://www.intellicast.com/National/Radar/Current.aspx?animate=true&location=USKS0523
[14:35:33] <_methods> so much for the tropical depression lol
[14:35:34] <Tom_itx> looks like it may be short lived though
[14:35:46] <pcw_home> its raining drought here
[14:36:17] <Tom_itx> couple years ago the docks at one of our major lakes were completely dry
[14:36:28] <Tom_itx> like 50-75 feet out
[14:36:34] <Tom_itx> maybe more
[14:37:01] <Tom_itx> i'm sure they're full now...
[14:38:34] <jthornton> rain has stopped here and the sun is shining for now...
[14:39:06] <pcw_home> Just want to be able to have a garden this year
[14:39:30] <Tom_itx> at least it's early enough in the day the tordado threat is less
[14:40:06] <Tom_itx> yeah JT-Shop, your radar looks clear
[14:41:41] <jthornton> I'm just going for a food plot for the animals...
[14:42:35] <pcw_home> Is that deliberate?
[14:42:58] <Tom_itx> heh it was like those buckets at the waterpark, it dumped now it's all done. let's just hope it's not filling up again
[14:47:48] <pcw_home> We feed a lot of ornamentals and citrus to the deer
[14:47:49] <pcw_home> even agapanthus which is supposed to be deer resistant
[14:47:51] <pcw_home> I guess the deer dont read the Sunset garden guide
[15:01:08] <witnit> ok, so just to make sure im not doing this wrong.. I can clearly see pulses with my desktop scope and voltmeter, yet the scope in linuxcnc never shows any noise, ticks, jumps or anything
[15:01:49] <witnit> is there anyway to ensure that the card is even talking to the scope?
[15:03:21] <witnit> alright, i realized watchdog has bit. does this disable my ability to read encoders?
[15:03:27] <pcw_home> no
[15:04:02] <pcw_home> but its a symptom of serious RT trouble
[15:04:37] <witnit> thats due to the pc im on i assume, im just using it for setup
[15:05:55] <pcw_home> also make sure you measure between the A and /A and B and /B the 7I33 inputs dont "see"
[15:05:56] <witnit> Isnt there a way i can just apply a couple volts and see the scope move?
[15:05:56] <pcw_home> voltages relative to ground if jumpered for differential mode
[15:06:08] <pcw_home> nope
[15:06:26] <pcw_home> well you can but you have to apply it differentially
[15:07:14] <pcw_home> thats why you need to check that youe are really getting differential signals to the 7I33 (by measuring _across_ the inputs)
[15:07:49] <witnit> im confused, i thought i just need a and b
[15:08:11] <zeeshan> do you know what a differential signal is?
[15:08:18] <witnit> I thought i did
[15:08:33] <zeeshan> like A should be +5 for example
[15:08:38] <zeeshan> and \A should be -5
[15:08:41] <witnit> right
[15:08:48] <zeeshan> if that doesnt happen, the filters will neglect the signal
[15:08:52] <witnit> but i dont need to use \A do i?
[15:08:56] <pcw_home> well /A would be 0
[15:09:24] <zeeshan> sorry 0 :P
[15:09:25] <Tom_itx> is it necessary to run a latency test every version update?
[15:09:48] <zeeshan> witnit, if you set it up as single ended
[15:10:01] <zeeshan> then you dont need to do that.. and i think is what youre trying to do
[15:10:04] <witnit> All i felt like i have done in the past was +5v, 0v, a,b
[15:10:13] <pcw_home> 1. If you have differential encoders ist better to use them in that mode (better noise immunity)
[15:10:15] <pcw_home> 2. the 7I33 needs to be jumpered to match the encoder type
[15:11:07] <pcw_home> Tom_itx: probably more kernel related than linuxcnc version
[15:11:25] <witnit> http://www.beisensors.com/pdfs/H25-optical-incremental_encoder.pdf
[15:11:43] <zeeshan> pcw_home: please make a temperature interface board!
[15:11:50] * zeeshan would like to know enclosure temperature
[15:12:00] <zeeshan> its cookin in the garage
[15:12:28] <witnit> I can rewire for all 4 a,a/, b,b/ will be back thanks
[15:13:04] <pcw_home> remember to jumper the 7I33 for differential encoders
[15:22:21] <skunksleep> I could not get much more than 10ft with non differential encoders at the speeds I was running
[15:22:50] <skunksleep> Switched them all to diff
[15:24:46] <pcw_home> Yeah single ended TTL level signals (~1V of noise immunity) are not
[15:24:47] <pcw_home> good around motor drives unless pretty slow so they can be filtered
[15:27:15] <zeeshan> pcw_home: have you been following my temperature chamber?
[15:27:29] <pcw_home> if your encoder signals are slow enough so that you can set the encoder filter
[15:27:30] <pcw_home> to about 20 usec you can do OK with single ended
[15:27:47] <pcw_home> A little but have been pretty busy
[15:28:09] <zeeshan> i really want your opinion on it
[15:29:06] <zeeshan> http://i.imgur.com/N1whRmt.png -------- middle block is heating block, bottom block is cooling block which will have fluid circulating at a temp of roughly -20C
[15:29:41] <Tom_itx> pcw_home just to make sure on these optos, if i want an active high, i connect +5v to the OptoOut+ and connect the OptoOut- to the device + and GND the other side of the device?
[15:29:47] <zeeshan> im wondering if by trigging the heating elements off, and then adjusting the flow of coolant through the cooling block
[15:29:59] <zeeshan> if you think ill be able to hold -5C pretty steady
[15:30:06] <pcw_home> Tom_itx: right
[15:30:12] <Tom_itx> ok thanks
[15:30:36] <zeeshan> or i could use a solenoid which just turns on / off
[15:30:39] <zeeshan> rather than duty cycle adjust
[15:32:31] <witnit> ahh nothing at all guys
[15:34:19] <pcw_home> does this temperature control just hold a setpoint or does it need to ramp quickly?
[15:34:56] <witnit> I just dont understand, yesterday I only used a and b. everything worked fine. today i go to finish up and it only counts one way and then wanders the other direction. I check the wiring and encoder, everything seemed fine. now i have added other a/ b/ tried moved all the jumpers, and nothing even shows on the scope
[15:35:40] <pcw_home> did you measure between A and /A with a voltmeter?
[15:35:44] <witnit> yeah
[15:35:55] <pcw_home> what do you get?
[15:36:01] <witnit> but it just reads the same as what the card is always putting out
[15:36:04] <witnit> about 1-2
[15:36:16] <zeeshan> pcw_home: having a linear temperature rate of change would be nice
[15:36:26] <witnit> the only channel on the card that picked up anything on the scope was the channel that read 0v
[15:36:35] <zeeshan> but right now most important is holding the set point
[15:36:41] <zeeshan> within a celcisu
[15:36:45] <pcw_home> did you move the encoder to check the other state
[15:36:50] <pcw_home> ?
[15:36:51] <witnit> yah
[15:37:06] <witnit> it just holds solid
[15:37:11] <pcw_home> it never changes?
[15:37:18] <witnit> but if i disconnect the card it goes up and down
[15:37:39] <pcw_home> sounds like a wiring error
[15:37:41] <witnit> no, the inputs on the card are outputting 1-2v constantly
[15:38:01] <pcw_home> with the encoder connected?
[15:38:04] <witnit> thats all i can read when its attached
[15:39:17] <witnit> ok, if i measure from the cards ground to any encoder input on the card except for ENC3 B and B/ it will read 1-2v. with or without an encoder connected
[15:39:58] <pcw_home> those dont mean anything without an encoder connected
[15:40:44] <pcw_home> what is the encoder model number?
[15:43:09] <pcw_home> (I ask because there are different output options)
[15:46:29] <pcw_home> zeeshan: it seems like you would want a single heat/cool plate, is there some reason they are separate?
[15:46:30] <pcw_home> (or maybe I dont understand what you are trying to do)
[15:47:07] <pcw_home> is there any external variable heat load?
[15:48:24] <witnit> H25-E-F1-SS-1250-ABZC-8830-LED-M18
[15:51:51] <pcw_home> hmm 8830 is not listed in their pdf for an output type
[15:53:42] <Spida> zeeshan: what are you planning to use this for?
[15:55:14] <witnit> i will go doublecheck the encoder label
[15:56:28] <zeeshan> pcw_home: we have a chiller that can hold -20C
[15:56:33] <zeeshan> thats why i wanna circulate fluid through it
[15:56:41] <zeeshan> and the reason why its 2 seperate blocks is for easier machining
[15:56:51] <zeeshan> Spida: experiment apparatus
[15:57:38] <witnit> 8830-LED is correct
[15:58:22] <Deejay> gn8
[15:58:59] <pcw_home> Well then I dont know what the output type is since 8830 is not listed in the PDF
[16:00:33] <pcw_home> I would think that yo would want the heat and cooling plate to be integrated
[16:00:35] <pcw_home> so they have the same thermal constant to the controlled point
[16:01:03] <zeeshan> yes
[16:01:39] <zeeshan> i forsee a huge failure
[16:01:43] <zeeshan> or it magically works :P
[16:02:51] <pcw_home> you might also need a "integrator" = low pass thermal filter if you use bang-bang vs proportional control
[16:03:32] <zeeshan> i could do p control
[16:03:34] <pcw_home> but I am _not_ a thermal engineer
[16:03:42] <zeeshan> p control of flow rate
[16:03:46] <zeeshan> and p control of current
[16:03:49] <zeeshan> (for heating)
[16:04:22] <witnit> ah, it seems to be this unit, http://www.engineering.uiowa.edu/sites/default/files/ees/files/NI/pdfs/00/57/DS005799.pdf
[16:04:57] <witnit> I suppose I will gather up the rest of my encoders, put everything back the way it was and test awhile
[16:05:01] <furrywolf> didn't get much at yard sales today... a few 100ft 12/3 cords with twistlocks, a 6" snatch block, and a broken air nailer.
[16:08:07] <pcw_home> if it indeed has a DS8830 , it has 5V differential outputs but not clear what the encoder power supply needs to be
[16:08:09] <pcw_home> in any case, if you do not read either +2 to +5V or -2 to -5V across A, /A and B,/B encoder outputs, something is wrong
[16:08:18] <furrywolf> blew out a wheel cylinder in my truck, which is an annoying project.
[16:08:42] <zeeshan> nice furrywolf
[16:08:43] <zeeshan> :P
[16:08:44] <pcw_home> didnt want to be fixing brakes today?
[16:08:56] <zeeshan> what its pissing brake fluid?
[16:08:57] <witnit> yeah pcw, im thinking i better go back to the sho and gather my encoders
[16:09:08] <furrywolf> yeah, it's got a puddle forming under one of the back wheels.
[16:09:14] <zeeshan> =/
[16:09:17] <witnit> oooh furry, got a whole truck load of this at a garage sale http://www.estorefixtures.net/heavydutypalletrack/images/pallet-rack_lock.jpg
[16:09:21] <zeeshan> hone and new oring
[16:09:23] <zeeshan> !
[16:09:37] <furrywolf> zeeshan: $15 or something for a brand new one. it's the labor getting it out that sucks.
[16:09:47] <pcw_home> some of those encoders need 12V or so even if they have 5V outputs
[16:10:15] <pcw_home> that needs to be determined maybe by a call to BEI
[16:10:18] <witnit> yes i was thinking maybe, but why does it appear to be perfectly normal on my voltmeter
[16:11:37] <pcw_home> normal in what sense (if you dont get +2 to +5 _and_ -2 to -5 across A,/A in different encoder states something is wrong)
[16:12:29] <witnit> but that doesnt explain why my card read a,b yesterday and flatlines today
[16:12:41] <witnit> atleast it was counting yesterday
[16:13:32] <witnit> i should atleast be able to gather something from it seeming how it did count yesterday both directions
[16:15:45] <pcw_home> no but if you dont do that measurement, all bets are off
[16:16:19] <witnit> yeah i did that though
[16:16:28] <witnit> and it just held steady at 1-2v
[16:16:52] <pcw_home> so it cannot be expected to work
[16:17:25] <furrywolf> I need to build spindle encoders, but I have a lot of other projects to do first, and money to spend...
[16:17:58] <witnit> welllppp to the shop i need to go and gather my other encoders :)
[16:18:34] <pcw_home> 1V across A and /A means a bad encoder or wrong power or wrong wiring
[16:24:01] <furrywolf> I guess today's project is dicking with air compressors... that doesn't take any money. bbl.
[16:32:46] <zeeshan> furrywolf: is your truck rusted
[16:32:52] <zeeshan> calipers are 2 bolts and 1 brake tube line
[16:33:04] <zeeshan> like a banjo usually which doesnt really snap :P
[16:35:40] <furrywolf> WHEEL CYLINDERS.
[16:35:44] <furrywolf> I didn't say calipers.
[16:36:11] <furrywolf> there are no calipers on my truck. :P
[16:39:35] <furrywolf> see, back before you were born, there was this device called a "drum brake"...
[16:39:42] <furrywolf> and my truck has four of them.
[16:42:41] <cpresser> witnit: perhaps they need pullups to +5V
[16:45:27] <pcw_home> They have differential line drivers so that should not be needed
[16:46:24] <pcw_home> but you should get about 3V or -3V across A and /A
[16:47:14] <pcw_home> (and both polarities if you move the encoder and never less than about 2V)
[16:47:26] <zeeshan> what ghetto ass technology is that
[16:47:27] <zeeshan> :P
[17:00:58] <furrywolf> my truck has 4-wheel drum brakes, one of my subarus disc front drum rear, the other subaru 4-wheel disc, and the van 4-wheel disc plus two drum brakes for the e-brake. because ford sucks.
[17:01:14] <furrywolf> oh, and my truck also has one band brake.
[17:02:22] <furrywolf> how's that for an assortment? :)
[17:06:17] <furrywolf> I took apart and carefully cleaned and greased my leaking pressure regulator. now it leaks ten times worse.
[17:07:40] <zeeshan> lol
[17:08:38] <MrSunshine> furrywolf: good work!
[17:09:26] <furrywolf> I couldn't get the valve seat out. I'm guessing the problem is under it...
[17:11:09] <SpeedEvil> furrywolf: Duct tape.
[17:11:20] <SpeedEvil> furrywolf: If you wrap it hard enough so that you exceed the gas pressure.
[17:11:37] <SpeedEvil> I've done this with a pressure washer hose.
[17:11:51] <furrywolf> that's not easy, since it leaks under the handle, and it's a hollow handle that screws down over the whole regulator...
[17:11:58] <SpeedEvil> ah
[17:13:19] <furrywolf> I'm not sure why it's leaking... I'm guessing the inlet valve must be leaking, and it's venting out the overpressure vent.
[17:13:33] <furrywolf> it's not a well-designed regulator.
[17:15:38] <furrywolf> it seems like if I screw it down all the way the leaking stops, or at least slows enough I can't hear it. this may be good enough. lol
[17:17:22] <furrywolf> I'm not actually using the regulator - I just got sick of hearing the compressor cycle...
[17:18:06] <SpeedEvil> My compressor does that too.
[17:18:15] <SpeedEvil> - all my fittings leak
[17:18:26] <SpeedEvil> Well - enough of them
[17:20:01] <furrywolf> yes, and being sick of that, I've sealed all my fittings well. the couplers are fancy ones with 6 balls instead of 3 or 4, and all joints are done using the fancy pink gas pipe thread tape.
[17:21:15] <furrywolf> and none of my hoses are wrapped in duct tape. :P
[17:28:46] <zeeshan> this fucking pedal is making me upset
[17:28:47] <zeeshan> lol
[17:29:57] <zeeshan> http://i.imgur.com/TCS3zWn.png
[17:29:59] <zeeshan> to make this
[17:30:12] <zeeshan> i need to buy a piece of 12x9"x3/8" chunk of aluminum
[17:30:17] <zeeshan> mill the contour out
[17:30:18] <zeeshan> and bend it
[17:30:31] <zeeshan> feels like an utter waste of material
[17:33:17] <SpeedEvil> Well, you could always make it without bending.
[17:33:18] <SpeedEvil> :)
[17:33:23] <SpeedEvil> 12*9*4
[17:33:35] <zeeshan> lol
[17:33:38] <zeeshan> out of billet? :P
[17:33:40] <zeeshan> no!
[17:34:13] <SpeedEvil> 3d print it, and then lost-PLA casting
[17:35:51] <zeeshan> i wanna prototype it first
[17:35:57] <zeeshan> i made it out of cardboard -- looks close!
[17:36:23] <furrywolf> please tell me that's a clutch pedal not a brake pedal? :P
[17:36:38] <zeeshan> it is
[17:36:42] <zeeshan> (a clutch pedal)
[17:36:56] <furrywolf> that's the type of part you make a bunch of at once, all nested into each other.
[17:37:02] <furrywolf> so there's minimal waste
[17:37:13] <zeeshan> yes but i wanna protoytype
[17:37:23] <zeeshan> suggest me a material other than aluminum
[17:37:24] <furrywolf> if you get one from 12x9, how many do you get from, say, 12x20? ten? :P
[17:37:33] <zeeshan> that i can machine easily , is 3/8 thick
[17:37:37] <zeeshan> and i can bend 33.5 deg after
[17:37:51] <zeeshan> furrywolf: i wouldnt be suprised if i got at least 8
[17:38:09] <furrywolf> each prototype can be cut out of the same larger chunk of metal. you don't need a full piece of metal for each one, unless you start with a barely large enough piece of metal.
[17:38:39] <furrywolf> so start with your 12x20, cut one out near the end, test, refine... cut the next one out of the same piece of metal a few inches further in, test, refine... etc
[17:39:24] <furrywolf> and then get a new piece of metal for the 9th prototype.
[17:41:42] * furrywolf needs to invent a right-angle anderson sb50
[17:42:55] <SpeedEvil> zeeshan: Why can't you make it from ~1"*3/8" and then simply bend it
[17:42:56] <furrywolf> I'm thinking I can notch the housing, strip the wire too far back and pre-bend it uninsulated before crimping, then glob it in glue of some form.
[17:43:05] <furrywolf> SpeedEvil: edgewise bends are hard.
[17:43:06] <zeeshan> SpeedEvil: bending isnt the problem
[17:43:19] <zeeshan> im just being annoying by the fact so much material will be wasted
[17:43:22] <SpeedEvil> furrywolf: Well, yes. But it's aluminium and 3/8" thick
[17:43:37] <zeeshan> i can bend 3/8 on my v die on the press
[17:43:39] <SpeedEvil> And the bent bits look ~3/4
[17:43:42] <zeeshan> its only 1.5" wide
[17:44:14] <furrywolf> zeeshan: he's trying to tell you to bend it on-edge.
[17:44:22] <zeeshan> i dunno what that means
[17:44:27] <zeeshan> im making this out of a flat sheet
[17:44:29] <zeeshan> and bending it at one spot
[17:44:39] <zeeshan> furrywolf: if i keep 2 parts and flip em
[17:44:52] <zeeshan> then i get 2 symmetric square pieces i can extract ouf the waste material! :)
[17:44:58] <furrywolf> so the bend is 1.5" thick by 3/8" wide, not 1.5" wide by 3/8" thick.
[17:45:05] <furrywolf> I'm telling him this is a much harder bend to do.
[17:47:16] * zeeshan is lost as to what you guys are talkin about, i thought the bend was obvious :P
[17:47:29] <furrywolf> grrr
[17:47:43] <furrywolf> you're lost, and if you don't figure it out, I'm going to lose it. :P
[17:48:18] <furrywolf> SpeedEvil wants you to start with a piece of flat bar stock 1.5" wide, and do it entirely with bending, including bending it in the other plane than the bend you have planned.
[17:49:40] <zeeshan> lol that'll be some crazy hard bend
[17:49:51] <furrywolf> yes. which is what I said. lol
[17:50:02] <zeeshan> but hey, it'd save material :P
[17:50:13] <furrywolf> is this project to get some special shape, or to save weight?
[17:50:23] <zeeshan> speciual shape
[17:50:33] <furrywolf> if saving weight, consider making it wider but with lightening holes. this sells well with ricers.
[17:50:57] <zeeshan> http://i.imgur.com/ImTQXCd.png
[17:51:00] <zeeshan> this makes me happier
[17:51:09] <zeeshan> 4x6.25" pieces will be useful
[17:51:39] <furrywolf> now rotate it 45 degrees and cut them out of an 8ft stip. :
[17:51:40] <furrywolf> :P
[17:51:52] <zeeshan> actually
[17:51:58] <zeeshan> that is what my flat pattern was suggesting
[17:52:42] <furrywolf> what will your new shape do that the stock shape didn't?
[17:52:55] <zeeshan> ive moved the whole clutch pedal assembly 4" forward
[17:53:01] <zeeshan> to allow to mount the clutch master inside the casr
[17:53:03] <zeeshan> *car
[17:53:04] <furrywolf> ah
[17:53:10] <furrywolf> making room for overly-large engine still?
[17:53:11] <zeeshan> so i gotta compensate to bring the clutch pedal back to stock location
[17:53:23] <zeeshan> this is only ap roblem for rhd cars
[17:53:29] <zeeshan> lhd cars dont have this problem
[17:53:40] <zeeshan> people have put really tiny clutch masters that dont move enough fluid
[17:53:49] <zeeshan> to extend the slave right, and have clutch issues at high rpm
[17:53:53] <zeeshan> disengagement issues
[17:53:57] <furrywolf> swap to a cable clutch. no master needed at all. :P
[17:54:10] <zeeshan> have you seen the side of a lsx clutch ? :P
[17:54:12] <zeeshan> *size
[17:54:20] <furrywolf> both of my subarus have cable clutches. :)
[17:54:21] <zeeshan> it weighs like 40lb by itself
[17:54:36] <furrywolf> my truck has a nice big clutch too
[17:54:44] <furrywolf> no hydraulics. :P
[17:54:48] <zeeshan> haha
[17:54:50] <zeeshan> poor you!
[17:54:59] <zeeshan> your left leg must be bigger
[17:55:19] <furrywolf> nah, the right leg gets plenty of exercise with the non-power brakes, remember? :P
[17:56:13] <zeeshan> if i angle this guy
[17:56:14] <furrywolf> and ever since I put the 38.5" swampers on it, the non-power steering gives you a good upper body workout, too!
[17:56:18] <zeeshan> i can cut both pieces out of 6" flat bar
[17:56:19] <zeeshan> nice!
[17:56:23] <zeeshan> that'll make stuff cheaper
[17:56:46] <zeeshan> i dont know how you drive with no power brakes
[17:56:48] <zeeshan> thats kinda scary
[17:56:56] <zeeshan> do you know the pedal ratio?
[17:57:07] <furrywolf> drum brakes need less pressure than disc brakes, as they're self-servoing
[17:57:11] <zeeshan> i know the older trucks had really long pedal
[17:58:19] <furrywolf> before the '70s non-power brakes were standard...
[17:59:15] <furrywolf> hell, I was working on a '76 the other day that had non-power brakes.
[18:04:17] <zeeshan> i guess youd need like 9:1 pedal ratio
[18:06:33] <furrywolf> you could manufacture a drum brake that needed almost no pressure to apply... it just would be poorly controlable.
[18:07:37] <furrywolf> the design is self-servoing... as you apply some force, the force is added to by the force of the vehicle stopping, resulting in more force.
[18:08:11] <SpeedEvil> For when you absolutely need to do uncontrollable drifts.
[18:08:35] <furrywolf> if you design this too strongly, they lock up and stay locked up until you come to a stop. this is a bad thing. so they design them to need a good amount of pressure... but not an unreasonable amount.
[18:10:21] <furrywolf> it has to do with how the two shoes are linked... when you push on the pedal, one shoe is pushed into the drum. the rotation of the drum pushes the shoe into the other shoe, applying it for you. now, it's more complicated than that since they have to work both forwards and reverse, but in reverse the same thing happens, just with the shoes swapped.
[18:11:19] <furrywolf> this is why the two shoes share only a single attachment point to the backing plate
[18:35:41] <zeeshan> did you fix it!
[18:36:15] <furrywolf> ?
[18:37:02] <zeeshan> the leaking cylinder
[18:37:17] <furrywolf> no. I haven't touched it. I've been working on air compressor stuff.
[18:37:52] <zeeshan> car is more important!
[18:38:29] <furrywolf> truck doesn't get used much...
[18:39:00] <zeeshan> o
[18:39:11] <furrywolf> and has other issues preventing its driving, like having a tarp over the shredded top held down with a bunch of scrap metal, and only firing on 4 of 6 cylinders.
[18:39:43] <zeeshan> drive it into a wall
[18:39:45] <zeeshan> problems solved
[18:40:04] <furrywolf> it'd have to be a very solid wall. otherwise it'd win. :P
[18:40:21] * furrywolf thinks the bumper weighs about as much as a smartcar
[18:41:19] <furrywolf> hrmm, those little gocarts are apparantly 1600lbs... not quite that much.
[18:42:35] <furrywolf> new top is about $300, but I just don't have spare money right now.
[18:42:47] <furrywolf> and I already have the new spark plugs... just have too many projects.
[18:43:17] <furrywolf> I left it idling for a few hours with the heater cranked up full to dry the carpet after the top shredded, and it tends to foul the #1 and #6 plugs if you let it idle too long.
[18:44:06] <zeeshan> nice
[18:44:07] <zeeshan> lol
[18:44:12] <zeeshan> is it missing that bad?
[18:44:55] <furrywolf> one cylinder is dead, one misses if you give it any throttle. it's happened before, just needs new plugs.
[18:45:02] <furrywolf> it burns a lot of oil and fouls plugs routinely
[18:45:27] <furrywolf> it'll be 50 years old soon. I don't worry about the little things. :P
[18:45:50] * furrywolf bets most new vehicles won't survive that long
[18:46:07] <zeeshan> http://i.imgur.com/fi5SZGZ.png
[18:46:13] <zeeshan> WTF pedal
[18:46:15] <furrywolf> with normal driving it needs plugs about every 5k miles. not a problem.
[18:46:19] <zeeshan> why you showing me a stress concentration
[18:46:56] <furrywolf> because that's a good thing to know? because you asked it for one?
[18:47:13] <furrywolf> the pushrod attachment is the peak stress point
[18:47:17] <zeeshan> well thats the stock hole size
[18:47:19] <zeeshan> 8mm.
[18:47:29] <zeeshan> so they musta had a stress concentration there too. i kept the pedal ratio the same
[18:47:31] <zeeshan> forces are the same
[18:47:58] <zeeshan> i have a hard ti mbelieving that they'd use some sort of weldable aluminum (lets suppose 6061-t6) with a yield strength of 40000psi
[18:47:58] <furrywolf> yes. stress concentrates at the fulcrum on a lever, and the hole doesn't help either.
[18:48:15] <zeeshan> and the localized stress concentration is like 6000psi from failure
[18:50:28] <zeeshan> i might make this out of steel
[18:50:48] <furrywolf> are you stressing it the right way? I'd expect the concentration to be on the other side of the hole
[18:51:15] <zeeshan> for a hole
[18:51:17] <zeeshan> it should be the same
[18:51:24] <furrywolf> the front of the hole should only have compression, the back tension
[18:51:40] <zeeshan> http://www.colorado.edu/MCEN/MCEN5023/str_conc.jpg
[18:52:07] <zeeshan> i feel like its just a Bs error
[18:52:07] <furrywolf> the front is only trying to stop the hole from getting pinched closed, purely compression. the back of the hole is under tension and pushing against the pin for the pushrod
[18:52:09] <zeeshan> cause of meshing
[18:52:28] <zeeshan> yea thats how holes work :P
[18:52:41] <furrywolf> you don't just have a hole. you have a hole with a pin through it, acting as the fulcrum on a lever. :P
[18:52:50] <furrywolf> the pin is applying substantial forces
[18:53:10] <furrywolf> that image you just pasted looks absolutely nothing like your circumstances. :P
[18:53:26] <zeeshan> i think its a mesh error
[18:53:33] <pcw_home> and overall bending force
[18:53:34] <zeeshan> it'll prolly change by changing the mesh size
[18:54:16] <furrywolf> you ARE modelling the stresses with a pin in that hole fixed in place, right? :P
[18:54:27] <zeeshan> my boundary conditions are: 100lb force at bottom of pedal
[18:54:32] <zeeshan> fixed hole where pin goes
[18:54:38] <zeeshan> and fixed displacement for pivot
[18:55:09] <zeeshan> (since it's free to rotate about one axis
[18:58:30] <pcw_home> any reason not to make it a bit wider at the pivot and narrower elsewhere
[18:59:24] <furrywolf> grrr. I'd swear that somewhere I have a short extension for anderson sb50s, but I can't find it.
[18:59:28] <zeeshan> the fork that goes there is limited in width
[19:00:25] <furrywolf> a fixed hole might compute differently than a fixed pin in a hole. does fixing the hole stop all sides from moving, while a pin only stops the back side from moving?
[19:00:56] <zeeshan> the onyl differnce i see between them is not account for friction
[19:00:59] <zeeshan> otherwise it's the exact same
[19:01:02] <pcw_home> depth I mean (perhaps on the non fork side)
[19:01:12] <zeeshan> pcw_home: limited by fork length
[19:01:50] <furrywolf> yes. as he just said, on the non-fork side. :P
[19:02:22] <zeeshan> ??
[19:02:26] <zeeshan> the pivot isn't a stressed area
[19:03:31] <furrywolf> ... the fork pivots on a pin in the hole. he means on the back side of the hole, opposite the side the fork is on.
[19:03:37] <pcw_home> not the main pivot but the cylinder pushrod pivot
[19:03:50] <zeeshan> you can make it bigger yea
[19:04:00] <zeeshan> but the problem isn't there..
[19:04:32] <zeeshan> http://i.imgur.com/jrw4BJv.png
[19:04:37] <zeeshan> redid it with 150lb force at pedal
[19:05:10] <witnit> bah you were right as usual pcw, grabbed another linear encoder, and a rotary both worked fine in the same placement
[19:06:08] <zeeshan> looks like im gonna have to machine a custom fork
[19:06:10] <zeeshan> or make this out of steel
[19:06:10] <furrywolf> you're _sure_ that a fixed hole does the same thing as actually putting a pin in the hole and letting it figure that?
[19:06:17] <zeeshan> furrywolf: yes
[19:06:23] <zeeshan> you never put an actual pin in there
[19:06:56] <furrywolf> hrmm. I'd just expect to see the most stress about 60 degrees from where it shows it.
[19:07:19] <zeeshan> where it's showing the stress makes since
[19:07:21] <zeeshan> *sense
[19:07:55] <zeeshan> actually, lets see the deformed image
[19:08:25] <furrywolf> how do you figure? that's where I'd expect not much stress at all. the pin is pushing against the very back of the hole, not the sides. the sides should have very little stress. the back is both trying to be pulled apart, and mushed by the pin.
[19:09:00] <zeeshan> the point of contact where the pin is contacting is in compression
[19:09:15] <zeeshan> 90 degrees from that constant point should be where itll go in tension
[19:09:29] <zeeshan> constant = contact
[19:09:41] <furrywolf> 90 degrees should be bending. 0 should be tension, 180 compression...
[19:09:58] <zeeshan> 90 degress around the axis of the hole
[19:10:21] <furrywolf> oh well, if you can make it bend it until it actually starts breaking, what's it look like? :P
[19:11:19] <furrywolf> I'm not a mechanical engineer, so I may well be wrong... but that stress just looks wrong to me.
[19:11:26] <pcw_home> If its all tension from the pushrod, why not the top of the hole?
[19:11:40] <furrywolf> also, the pushrod can move freely up and down. it's only constrained in one direction.
[19:12:03] <furrywolf> I'm not a mechanical engineer, but I've seen a lot of broken levers. :P
[19:12:17] <furrywolf> and they never break where that shows it
[19:12:25] <zeeshan> http://gyazo.com/68cf7070f68af66fcbe9802239cacda8
[19:12:35] <zeeshan> its a compound angle
[19:13:23] <furrywolf> ohh, it's 3d. your last one showed it flat. lol
[19:13:35] <zeeshan> :P
[19:13:36] <pcw_home> so making it wider would help :-)
[19:13:39] <furrywolf> but that's even more wrong!
[19:13:49] <furrywolf> because the pushrod will deflect sideways.
[19:14:07] <zeeshan> this is one scenario
[19:14:11] <furrywolf> the stress in that direction is concentrated in the weld to the pivot bushing, not at your hole.
[19:14:25] <furrywolf> your fixed hole needs to only be fixed in one axis, and free to move in the other two
[19:14:26] <zeeshan> imagine the pushrod fronen, and pedal pivot frozen, and you applied 150lb of force
[19:14:46] <zeeshan> there would be a twist , bend and normal force
[19:15:04] <furrywolf> it's correctly showing it for if the hole were fixed in 3 dimensions. yours is not.
[19:15:18] <furrywolf> the pushrod can move up/down and side/side. in fact, it probably even has a ball nose.
[19:15:18] <zeeshan> mine is fixed in 3 dimensions :P
[19:15:24] <furrywolf> no, it's not.
[19:15:40] <pcw_home> Ahh that assumes the fork/pushrod are much stiffer than they are likely to be
[19:16:14] <furrywolf> I've never seen a master cylinder than fixed the pushrod from moving around
[19:16:15] <furrywolf> that
[19:16:22] <pcw_home> so probably bogus
[19:16:23] <zeeshan> youre right its free to move in y and Z
[19:16:33] <zeeshan> fixing in in X only
[19:17:12] <zeeshan> hah
[19:17:15] <zeeshan> it got even force
[19:17:16] <zeeshan> *worse
[19:17:39] <zeeshan> 121000 psi!
[19:18:22] <furrywolf> ideally you'd model a pin with a ball joint, a pushrod of the approximate length, and another ball joint, and then fix that point in 3d space.
[19:18:23] <zeeshan> http://gyazo.com/0053df38ee57635bcbe7d739627e73f7
[19:18:34] <zeeshan> thats free to move in Z and Y
[19:18:49] <furrywolf> that's still wrong. you fixed the wrong axis or something.
[19:19:19] <pcw_home> yeah I think that assumes the fork cant twist
[19:19:50] <furrywolf> stop it with the whole "fixed hole" thing and model a linkage. :)
[19:19:52] <zeeshan> http://i.imgur.com/6iULWj5.png
[19:19:59] <zeeshan> lol you dont do that
[19:20:13] <zeeshan> i knew removing the constraint from the y and z would make it worse
[19:20:21] <zeeshan> cause now the X has to support all the load
[19:21:01] <furrywolf> you say "you don't do that". you're objecting to making a more accurate simulation of the actual forces.
[19:21:20] <furrywolf> and from how quickly you're pasting these, it's obviously not taking excessive time to compute, so... :P
[19:21:21] <zeeshan> furrywolf: you always simplify.
[19:21:39] * furrywolf simplifies the world to a sphere, and flies a plane at a constant 100ft altitude
[19:22:45] <furrywolf> your simplification isn't reflecting the actual linkage...
[19:23:28] <pcw_home> I'll bet if you look on the other side you will see the stress riser on the top instead of the bottom of the hole
[19:23:39] <furrywolf> also, you can't fix the entire radius of the hole in place. a pin only exerts force on, at best, 1/4 of the back of the hole. that pin isn't a press fit, and every one I've seen could only be described as a loose fit.
[19:24:24] <furrywolf> when you get it right, you'll see a second stress concentration at the weld between the pedal and the pivot bushing
[19:24:33] <pcw_home> which is what you would expect if the fork cannot rotate
[19:25:00] <SpeedEvil> furrywolf: exceptions I've seen are when a steel pin is forced into a greased undersized wooden hole. But...
[19:25:38] <furrywolf> SpeedEvil: when you find a car where the clutch pedal is made of greased wood with an undersized hole...
[19:28:31] <furrywolf> pcw_home: I think he's not going to get an accurate result until he models a pin and its constraints.
[19:29:04] <furrywolf> the way it's still showing the bending being at the fork instead of at the pivot well means it's still way off base.
[19:29:09] <furrywolf> weld
[19:30:39] <zeeshan> furrywolf: you're making me upset
[19:30:40] <zeeshan> lol
[19:30:43] <furrywolf> zeeshan: what's it look like if you don't fix the hole at all, and instead fix only the upper pivot from moving?
[19:31:05] <furrywolf> actually, never mind, that'll show a different stupid failure.
[19:31:35] <furrywolf> the simulation may be right in one way... it is quite likely this thing will break if you stomp on it. :P
[19:31:38] <zeeshan> =P
[19:31:54] <furrywolf> 3/8 by 1.5 does seem a bit thin for how much force you're putting on it.
[19:32:04] <zeeshan> thats what the stock pedal is :/
[19:32:14] <zeeshan> maybe the problem is im overestimating the force
[19:32:20] <furrywolf> is the stock pedal aluminum or steel? and do you put 150lbs of force or 15lbs of force on it?
[19:32:22] <zeeshan> theres no way i was applying 150lb for my clutch pedal
[19:32:24] <zeeshan> that'd destroy me
[19:32:29] <zeeshan> i think its more like 50lb.
[19:32:44] <furrywolf> I think even my truck is less than 50.
[19:33:11] <zeeshan> im asking my buddy to take his weigh scale
[19:33:18] <zeeshan> and put it between his foot and his pedal and tell me the number
[19:33:19] <zeeshan> lol
[19:33:35] <furrywolf> lol
[19:34:15] <furrywolf> I have a 70lb postal scale, but last I tried it it didn't turn on. probably needs a battery. but it's all plastic, and thin plastic at that... I'd need to sandwich it between two blocks of wood to keep it from just cracking.
[19:34:23] <postaL> that's a big scale.
[19:35:16] * zeeshan waits :p
[19:35:19] <furrywolf> is there room to re-engineer it? say, use a longer pivot tube, with the pedal straight and not bent, and a second tab welded to the tube for the pushwod?
[19:35:21] <furrywolf> pushrod
[19:36:35] <zeeshan> haha
[19:36:38] <zeeshan> he's like "wtf"
[19:36:42] <zeeshan> hes trying it
[19:36:51] <zeeshan> furrywolf: yea it looks like ill hjave to make that section wider
[19:36:55] <zeeshan> and use a custom fork
[19:38:33] <Crom> well the reprap is moving
[19:38:33] <furrywolf> you say "yeah", but that has nothing to do with what I asked. :P
[19:38:54] <Tom_itx> pcw_home you around?
[19:39:36] <zeeshan> the pedal position can't be changed
[19:39:39] <zeeshan> the pedal can be made thicker
[19:40:12] <pcw_home> yeah
[19:40:18] <Tom_itx> pcw_home, on the ADC on the 7I47S, i've hooked it up with 10V on the Analog+ and GND on Analog-. with a command speed of S5000 M3 i get a pulse from the Analog out then it goes low
[19:40:36] <furrywolf> so there's no room at all to the right of the assembly?
[19:40:41] <Tom_itx> i tried both the 1.4khz and 5khz frequencies
[19:41:37] <zeeshan> whatcha mean?
[19:41:40] <zeeshan> "To the right?"
[19:41:46] <zeeshan> please refer to this image haha so i dont get confused
[19:41:57] <zeeshan> http://i.imgur.com/TCS3zWn.png
[19:43:15] <furrywolf> ... to the right. of the car.
[19:43:24] <zeeshan> you see the image
[19:43:28] <zeeshan> you see how its attached
[19:43:30] <zeeshan> :-)
[19:43:33] <pcw_home> Hmm if you have 10V and enable (active low) and PWM you should get output
[19:43:38] <zeeshan> its a matter of making the pedal thicker
[19:43:45] <zeeshan> and making a custom fork
[19:43:50] <zeeshan> or............
[19:44:00] <zeeshan> throwing this all in the garbage!
[19:44:05] <furrywolf> ok, if you want to start with that image. extend the pivot tube out the right side of that box, and attach the pedal there. inside the box, just have a tab down to the fork. but do you have room there?
[19:44:48] <Tom_itx> maybe enable is high
[19:44:51] <Tom_itx> lemme check that
[19:45:06] <Tom_itx> can i invert that if i need to?
[19:45:10] <zeeshan> no room
[19:45:15] <zeeshan> itll hit steering shaft
[19:45:22] <pcw_home> enable has to be low (yes set the invert option)
[19:45:43] <Tom_itx> i might have to reverse that, i'll check
[19:45:54] <pcw_home> I dont remember the PWM polarity off hand
[19:46:15] <pcw_home> but it may need to be inverted also
[19:46:19] <Tom_itx> i think i inverted that on the other board and left it that way for this one
[19:46:19] <zeeshan> i was really hoping to avoid making a new Fork
[19:46:20] <zeeshan> ;/
[19:46:28] <zeeshan> but its relatively easy to machine
[19:47:55] <furrywolf> make a second plate that bolts onto your pedal, and goes over the fork?
[19:48:22] <zeeshan> i was thinking that
[19:48:29] <zeeshan> trying out the .5 thick plate first
[19:48:56] <furrywolf> does the stock pedal have the sideways bend in it, or is that a result of your new location?
[19:48:57] <zeeshan> looks like .5" plate works
[19:49:06] <zeeshan> it has the bend sideways
[19:49:10] <zeeshan> but its not extended 4" back
[19:49:17] <zeeshan> that makes a pretty big difference :/
[19:49:20] <furrywolf> and is the stock pedal alu or steel? I know every vehicle I've own has been steel...
[19:49:21] <Tom_itx> fixing /ENA fixed that
[19:49:23] <Tom_itx> thanks
[19:49:25] <zeeshan> aluminum
[19:49:41] <zeeshan> w/ a 1/2" plate
[19:49:54] <zeeshan> with 100lb force on the pedal , im seeing 35959 psi
[19:49:57] <zeeshan> thats better!
[19:52:41] <zeeshan> pedal is .82 lb not bad
[19:52:56] <furrywolf> that's pretty light.
[19:53:07] <zeeshan> if i didnt have to weld a spacer thing
[19:53:12] <zeeshan> i coulda used 7075
[19:53:19] <zeeshan> and not have to worry about any of this nonsense! :P
[19:53:49] <zeeshan> friend just got back to me
[19:53:50] <zeeshan> 60lb
[19:53:53] <Tom_itx> ok with pwm set to 5khz i had to half the scale but i haven't hooked the motor up yet to check if it's at the commanded speed
[19:54:09] <Tom_itx> it's 'in the ballpark'
[19:54:56] <furrywolf> make it from good alloy steel. machine it out of a leaf spring. :P
[19:55:12] <XXCoder1> Make it from titanium'
[19:55:14] <pcw_home> it should linear with 1% at 5 KHz
[19:55:18] <zeeshan> okay at 60lb
[19:55:19] <Tom_itx> pcw_home, if the pwm is inverted, it will run fast with a slow commanded speed right?
[19:55:22] <zeeshan> theres a factor of safety of 2.5
[19:55:23] <zeeshan> good enough
[19:55:28] <pcw_home> yes
[19:55:30] <Tom_itx> ok
[19:55:36] <Tom_itx> that's easy to check
[19:56:03] <Tom_itx> i'm getting nearly 10v at a S5000 which is about what i want
[19:56:17] <Tom_itx> just need to check the low end
[19:56:46] <furrywolf> zeeshan: what software are you using for all of this anyway?
[19:57:01] <zeeshan> ansys
[19:57:05] <pcw_home> the OP Amp is rail-rail but not perfect of course
[19:57:13] <Tom_itx> this sure makes a cleaner install than the other board
[19:57:17] <Tom_itx> yep
[19:58:40] <zeeshan> yea i just did a simulation on th stock pedal
[19:58:43] <zeeshan> and its way less stressed
[19:58:50] <zeeshan> that 4" extension makes a big diff
[19:58:56] <zeeshan> ;/
[19:59:04] <zeeshan> looks like stock factor of safety was around 2.5 as well
[19:59:30] <furrywolf> that's good, then. yours is just as strong as stock.
[20:00:27] <furrywolf> ansys looks ungodly expensive.
[20:02:27] <zeeshan> thank god for student license!
[20:02:37] <zeeshan> an alternative i use a lot is abaqus
[20:02:40] <zeeshan> thats a bit reasonable.
[20:02:46] <zeeshan> abaqus standard that is
[20:03:48] <pcw_home> man, even the sheep would not eat these bran muffins
[20:03:50] <pcw_home> swear they are made of sawdust soaked in fruit juice
[20:03:50] <furrywolf> anything that, rather than you buying, they only let you select to have a sales rep contact you, is overpriced.
[20:04:11] <zeeshan> fea is a joke to program
[20:04:23] <zeeshan> the hard part is how to generate the mesh
[20:04:29] <zeeshan> im sure theres tons of open source software out there to do that
[20:05:08] <furrywolf> I've generated meshes on 3d objects for 3d rendering purposes...
[20:05:12] <SpeedEvil> A ~1" tube on the rod, press-fitted into a very undersized hole in the Al might work OK
[20:06:06] <SpeedEvil> Or threaded at the end, with a nut to clamp it on.
[20:06:13] <SpeedEvil> (and a safety wire)
[20:08:22] <witnit> anyone in Massachusetts? looks like the perfect little starter shop for sale
[20:08:26] <witnit> http://boston.craigslist.org/gbs/tls/5017115442.html
[20:08:45] * furrywolf is nowhere near massachusetts
[20:08:54] <furrywolf> and I have no money.
[20:08:55] <witnit> oh furry you could use all that stuff too!
[20:08:59] <furrywolf> other than that... :P
[20:09:24] <zeeshan> furrywolf:
[20:09:27] <zeeshan> problem is solved
[20:09:30] <witnit> it might be a little high price, but it looks rather complete, takes a long time to gather all the right pieces
[20:09:31] <zeeshan> using my old design :]
[20:09:42] <furrywolf> what was the problem?
[20:09:59] <zeeshan> well it was twisting and bending too much right?
[20:10:01] <zeeshan> so just needed gussets
[20:10:48] <zeeshan> bah.
[20:10:56] <furrywolf> ... you do know several people here suggested making it wider, right? :)
[20:10:57] <zeeshan> its cause it was set to 60lb not 100 :/
[20:11:04] <zeeshan> back to crap
[20:37:26] <antcat> anyone knowing about the STATIC inline double tpGetScaledAccel(TP_STRUCT const * const tp,TC_STRUCT const * const tc)
[20:38:04] <antcat> if (tc->term_cond == TC_TERM_COND_PARABOLIC || tc->blend_prev) { a_scale *= 0.5;}
[20:39:09] <antcat> why the a_scale multipled by 0.5
[21:18:30] <furrywolf> well, I shouldn't be spending money, but I just got craigslist tools. heh.
[21:18:37] <zeeshan> 43693psi with both tabs; 77658psi with one tab. <-
[21:18:38] <zeeshan> er
[21:18:40] <zeeshan> what did you get
[21:19:03] <furrywolf> got a paslode impulse framing nailer, battery, charger, non-mar tip, 6 fuel cells (although only 2.5 of them seem to be full), and a box and a half of nails for $25.
[21:19:18] <furrywolf> I would have haggled, but he even drove them here.
[21:19:21] <zeeshan> fuel cells? :]
[21:19:50] <furrywolf> that's what paslode calls them.
[21:19:58] <furrywolf> they're little butane canisters with an inflated price tag
[21:20:21] <furrywolf> it's an older model, but as long as it works...
[21:21:09] <zeeshan> cool
[21:21:17] <zeeshan> youre addicting to craigslist finds
[21:21:22] <zeeshan> and flea markets!
[21:21:26] <furrywolf> I think the battery is toasted. yay nicd.
[21:21:36] <furrywolf> no. I do not go to flea markets. I do not like flea markets.
[21:22:53] <furrywolf> in theory, flea markets are where random people take their junk to sell it, like a big garage sale. in reality, flea markets are populated entirely by dealers who travel to flea markets to try selling their overpriced crap because they want to run a thrift store but don't want to deal with pesky things like a business license or paying taxes.
[21:23:57] <zeeshan> ;/
[21:24:11] <t12> sweet stuff shows up at fleamarkets if you are on your game
[21:24:19] <t12> i got a set of brown and sharpe toolmakers vises
[21:24:20] <t12> for $5 ea
[21:24:33] <Tom_itx> pcw_home, could i need a pullup on the opto out? on the /ENA i had to set it to acive low but i've got a relay on the pin. I reversed the relay leads but it still enables the relay all the time
[21:24:45] <t12> things like boxes of small C clamps for a buck each
[21:24:56] <furrywolf> the flea market here is the same dealers, every single time.
[21:24:58] <t12> more obscure useful tools often show up
[21:25:03] <furrywolf> I usually leave empty-handed.
[21:25:06] <t12> depends on the market really
[21:25:19] <furrywolf> they know they're useless and not going to sell anything, so they charge you to get in the door.
[21:25:30] <Tom_itx> pcw_home, currently the Optoout - is set to GND and the Optoout + goes to the relay with the other leg of the relay to +5v
[21:25:30] <furrywolf> just to make sure they get your money even when you realize it's all dealers with overpriced crap.
[21:26:45] <furrywolf> I always slip in through the back, just so I don't get pissed off at spending money to not get anything. lol
[21:28:40] <furrywolf> http://www.ebay.com/itm/PASLODE-IMPULSE-IM-300-75N-FRAMING-FRAMER-POWER-NAILER-CHARGER-NO-BATTERY-FAIR-/251836131067 that's the one I got, but I got a battery, the blow-mold case, the non-mar tip, the 2.5 fuel cells, and 1.5 boxes of nails.
[21:30:10] <furrywolf> the charger doesn't seem to like the battery, and it's got corrosion by the contacts, so I'm guessing mine didn't actually come with a battery either. will need to rebuild it.
[21:33:26] <HSD> So before I go too far down this road... Anyone ever put LinuxCNC on a chrome box? Am I wasting my time?
[21:33:39] <furrywolf> chrome box?
[21:34:04] <furrywolf> heh, the charger declares the battery fully charged... putting a load on it, it drops from 6.5v to 2v in about 3 seconds.
[21:39:59] <furrywolf> I guess I need to try rigging up some contacts for it and trickling it for a while, but it's probably toast.
[21:40:16] <furrywolf> I might get enough of a charge into it to test that the rest of the gun works, however, and thus it's worth fixing the battery.
[21:42:36] <HSD> http://www.amazon.com/ASUS-CHROMEBOX-M004U-Deluxe-4Gb-Memory/dp/B00K048H7E/ref=sr_1_4?ie=UTF8&qid=1431224429&sr=8-4&keywords=google+chromebox
[21:42:51] <HSD> 4Gb memory. I'
[21:43:09] <HSD> m most concerned about it being able to do the RT extensions without issue
[21:54:36] <furrywolf> you can boot the livecd and run the latency test
[21:55:14] <furrywolf> how do you plan on connecting to your machine?
[21:56:00] <furrywolf> the realtimeness mostly matters for parallel port connections, but that seems to lack a parallel port.
[21:56:12] <zeeshan> F
[21:56:23] <furrywolf> if you're planning on using one of mesa's ethernet cards, the realtimeness doesn't matter as much, and just about any system will work.
[22:04:48] <zeeshan> realtime is alawys a good thing! :P
[22:04:52] <zeeshan> for motor control!
[22:06:03] <zeeshan> i finally have a simple pedal
[22:06:08] <zeeshan> that doesnt fail miserably.
[22:08:06] <zeeshan> no idea how im gonna bend 5/8 thick aluminum
[22:08:06] <zeeshan> :/
[22:10:57] <furrywolf> that seems to have grown again.
[22:11:04] <zeeshan> it is what it is :P
[22:11:17] <zeeshan> if i want a simple 2d piece
[22:11:25] <zeeshan> with no welds on it except at the bushing thing
[22:11:32] <furrywolf> what about using 3/8, and welding a second tab from the right side of the pivot tube to the middle of the 35 degree bend?
[22:11:34] <zeeshan> that i just bend, i gotta do this
[22:12:36] <zeeshan> yea im gonna try a rib combination now
[22:12:45] <zeeshan> if it fails, then i know i have a design that works
[22:12:53] <zeeshan> just wanted something to fall back onto
[22:14:22] <furrywolf> the second tab would just be flat, no bends, no holes, but would need two more welds.
[22:22:25] <Tom_itx> would PID help a condition where the spindle undershoots at low speeds but is pretty close at higher speeds?
[22:23:11] <zeeshan> yes
[22:23:22] <zeeshan> itll try to correct it :P
[22:23:58] <zeeshan> maybe not :/
[22:24:02] <zeeshan> now that i think about it
[22:24:06] <zeeshan> that sounds similar to my stiction issue
[22:24:06] <Tom_itx> but will it overshoot on the high end then?
[22:24:31] <Tom_itx> it's not very close under 500 rpm
[22:24:57] <zeeshan> yea thats what im thinking
[22:25:48] <Tom_itx> well at least the 7i47S is working
[22:26:12] <Tom_itx> i removed that board you sent and added a couple relays to try this 7i47s
[22:26:33] <Tom_itx> i think either one will work ok
[22:27:08] <Tom_itx> i had to do some rewiring since the dac is on certain pins on the mesa card
[22:40:59] <zeeshan> too much haxin
[22:41:02] <zeeshan> =D
[22:42:20] <HSD> furrywolf, ah yes. for some reason I thought it had one. Must have been one of my other machines. guess I'm back to the drawing board.
[22:43:06] <HSD> furrywolf, hmm... mesa's ethernet cards?
[22:43:35] <HSD> could be an issue unless it comes in micro pcie style boards
[22:58:36] <furrywolf> mesa sells cnc control boards with an ethernet connection
[22:59:27] <zeeshan> i give up on the ribs idea
[22:59:37] <zeeshan> they either hit something
[22:59:40] <zeeshan> or they\re not big enough
[23:00:07] <zeeshan> i remember seeing a vbending die at one of the presses at school
[23:00:12] <zeeshan> maybe i can press it there
[23:00:15] <zeeshan> we have a 250 ton press
[23:00:27] <zeeshan> !
[23:03:08] <HSD> http://store.mesanet.com/index.php?route=product/category&path=83\
[23:03:10] <HSD> got it. thx
[23:16:57] <Crom> ugh trying to find 3.5mm pitch terminal blocks is not fun and they are freaking expensive. $5.70 for 10 pieces
[23:26:05] <Tom_itx> http://www.ebay.com/itm/20pcs-3-5mm-Pitch-2-pin-2-way-Straight-Pin-PCB-Screw-Terminal-Blocks-Connector-/281396246592?pt=LH_DefaultDomain_0&hash=item418485f840
[23:26:30] <Tom_itx> http://www.ebay.com/itm/20-pcs-8-pin-3-5mm-Screw-Terminal-Block-Connector-Pluggable-Type-Green-/231328654134?pt=LH_DefaultDomain_0&hash=item35dc432336
[23:43:26] <Crom> looking more at these http://www.ebay.com/itm/261016427718
[23:43:34] <pcw_home> Tom_itx: The enable outputs should only be on when commanded by linuxcnc
[23:43:36] <pcw_home> if its always on ( + connected to - ) make use you dont have it wired backwards
[23:43:47] <pcw_home> make sure
[23:43:54] <Tom_itx> i figured it out but i had to add a pullup
[23:44:12] <pcw_home> a pull up should not be needed
[23:44:23] <Tom_itx> it's how i'm using it that its required
[23:44:35] <pcw_home> OK
[23:44:48] <Tom_itx> by itself the /ENA is fine but i'm also using a relay to drive the inhibit on the driver board
[23:45:12] <Tom_itx> it gives me a little deadband to switch from FWD to REV and not blow out the relay contacts
[23:45:20] <Tom_itx> just a few ms
[23:46:23] <Tom_itx> if i tried adding PID to this, what terms should i start adjusting?
[23:46:59] <pcw_home> velocity loops are usually just P and I
[23:47:20] <Tom_itx> what sort of numbers should i start with?
[23:47:48] <Tom_itx> i generally start with P and get it to overshoot then add I
[23:47:58] <Tom_itx> after backing P off a bit