#emc | Logs for 2009-10-11

Back
[00:14:28] <oPless> oh he left
[00:14:51] <i-pink> hii
[00:15:07] <i-pink> i want to kiil the HAL file
[00:15:14] <i-pink> *kill
[00:15:35] <oPless> I'd suggest he use the svga driver instead
[00:16:56] <i-pink> way is soooo complicated! to add encoder!!!!!!!!
[00:26:24] <BJT_Shop> * BJT_Shop listens to Dave Brubeck...
[00:27:12] <BJT_Shop> http://www.youtube.com/watch?v=DUDtFdnn9oQ
[00:28:24] <tom3p> i learned that song in the 70's at the Old Town School of Folk music ;)
[00:28:52] <BJT_Shop> I love it
[00:29:19] <BJT_Shop> that and "little boxes"
[00:29:32] <tom3p> tickytacky
[00:30:02] <BJT_Shop> http://www.youtube.com/watch?v=AN3rN59GlWw
[00:30:09] <BJT_Shop> lol yep
[00:32:02] <BJT_Shop> did you listen to Telstar?
[00:32:38] <BJT_Shop> or any of the other Ventures songs
[00:33:02] <BJT_Shop> Green Onion I think is the best
[00:35:02] <BJT_Shop> * BJT_Shop ponders why the R key is the dirtiest one on this keyboard
[00:38:47] <tom3p> haha i was thinking musical keyboards and wth is the R key !?
[00:39:30] <BJT_Shop> on the keyboard for my plasma torch out here in the shop
[00:39:40] <BJT_Shop> the r is black
[00:40:29] <BJT_Shop> music, wine and internet in the shop... now I only need central heat and air
[00:40:43] <BJT_Shop> maybe a pottie too :)
[00:55:50] <ftkalcevic> Is there a watchdog like hal component? My pendant can generate a heart beat, and if it stops I want emc to stop.
[01:00:46] <tom3p> yes, look for 'charge pump', frequency in keeps an output valid
[01:02:38] <ftkalcevic> Isn't that the other way around? Charge pump generates a keep alive signal?
[01:03:35] <spasticteapot> Is it possible to make a router by attaching an XY table to a big old drill press?
[01:03:48] <spasticteapot> This is probably a dumb question, but one I feel is worth asking.
[01:04:34] <andypugh> Yes, though it won't be the best router in the world
[01:04:45] <ftkalcevic> I've read the bearings in the drill press aren't designed to take sideways forces of routing.
[01:04:51] <spasticteapot> I assumed as much.
[01:04:51] <andypugh> And the spindle speed might be low
[01:04:52] <BJT_Shop> might be the worst
[01:05:20] <spasticteapot> I've been wondering if I could save $$$ by buying a broken router and swapping the XY table to something homemade.
[01:05:54] <andypugh> Have you seen that all-mdf gantry router kit?
[01:06:12] <spasticteapot> I have, actually.
[01:06:46] <spasticteapot> I'm told that the cost of an adequate spindle is quite high, and using a router is a bad idea.
[01:07:31] <andypugh> There are some fairly cheap spindles (including water cooled) on that eBay
[01:07:37] <tom3p> router better than drill press. drill press never meant to have work move laterally, router yes.
[01:08:14] <spasticteapot> The problem with routers seems to be very limited Z travel - most top out at about 2".
[01:08:35] <andypugh> Though I can't imagine that a drill press would have anything other that deep-groove balls or taper rollers, and they are both perfectly good for lateral loads.
[01:08:49] <tom3p> oh, dont count on the internal z travel of the router, just use it to spin the tool.
[01:09:01] <andypugh> You take the plunge mechanism off.
[01:09:52] <andypugh> ftkalcevic: The charge-pump HAL component is designed to create a keepalive signal
[01:10:00] <spasticteapot> tomp3: I meant that the bit is only 2" from the bottom of the router.
[01:10:17] <spasticteapot> If I move it down any further, whatever I'm working on is bonked by the bottom of it.
[01:10:46] <andypugh> That is a problem for any router system, you need to use longer tools
[01:10:53] <ftkalcevic> Yes. I was looking for a watchdog component to monitor a keepalive signal and stop emc if the keepalive stopped.
[01:11:15] <andypugh> Ah, sorry, I misundrstood
[01:11:38] <ftkalcevic> No problem. comp is simple enough. I'll whip one up.
[01:11:55] <spasticteapot> andypugh: Plan A is to buy a mill. This costs a buttload, and most mills are HUGE.
[01:12:09] <andypugh> Just use an encoder and trip out if the speed drops....
[01:12:19] <spasticteapot> Plan B is buy a spindle. See first flaw of plan A.
[01:13:23] <andypugh> I think that an encoder component can run in counter mode on only one channel. So if our watchdog stops the encoder velocity will go to zero. Job done :-)
[01:14:04] <ftkalcevic> Sounds simple. Thanks.
[01:16:49] <andypugh> Bear in mind that the encoder is the only component I know at all well, and when all you have is a hammer every problem looks like a nail.
[01:18:18] <AchiestDragon> 4lb lump hammer ?
[01:19:31] <BJT_Shop> I always tell mechanics to place their hand between the object and the hammer so they have a good feel for the force applied
[01:19:46] <AchiestDragon> then you need to learn new tricks like pannel beating and blacksmith techniques with it
[01:20:14] <tom3p> yes, the charge_pump component is an enabled output, sorry. you need some detect circuit.
[01:21:02] <BJT_Shop> perhaps an encoder input and use the velocity
[01:21:45] <andypugh> This is a write-only medium, right?
[01:22:18] <AchiestDragon> no read/write but not modify
[01:22:34] <AchiestDragon> once written
[01:22:56] <tom3p> i agree, use the velocity, configure as 1 channel
[01:22:57] <ftkalcevic> WORM - Write Once Read Many
[01:23:13] <tom3p> FISH, first in stays here
[01:23:37] <tom3p> ( stuck stack )
[01:25:26] <AchiestDragon> no a stack is sequential read back ,, correct term would be a data log
[01:25:49] <BJT_Shop> goodnight folks
[01:25:56] <tom3p> gnite,
[01:26:04] <tom3p> log, yeh that describes it
[01:26:26] <AchiestDragon> would say real time other than theres often data lag
[01:26:38] <AchiestDragon> nn BJT_Shop
[01:32:07] <i-pink-cnc> hii
[01:32:13] <i-pink-cnc> i add this code
[01:32:23] <i-pink-cnc> http://www.pastebin.ca/1611015
[01:32:58] <i-pink-cnc> to the end of the standard_pinout.hal
[01:33:25] <i-pink-cnc> and i get this error when i start the EMC
[01:33:58] <i-pink-cnc> standard_pinout.hal:93: pin 'spindle-phase-a' does not exist
[01:35:18] <i-pink> andypugh, AchiestDragon, tom3p, BJT_Shop, ftkalcevic, spasticteapot - someone can help me with that error?
[01:40:40] <ftkalcevic> Line 14 in the paste bin is saying connect "spindle-phase-a" and "parport.0.pin-12-in" to "spindle-index". "spindle-phase-a" and "parport.0.pin-12-in" must be pins. "spindle-phase-a" is not. It is a signal.
[01:41:08] <ftkalcevic> The fix? Depends on what you want to connect?
[01:41:10] <andypugh> My goof, I guessed that
[01:41:40] <andypugh> The aim is to have a quadrature encoder where index and phase-a are the same p=port pin
[01:41:51] <i-pink-cnc> but the pins is OK
[01:42:20] <ftkalcevic> Just remove the spindle-phase-a from that line.
[01:42:40] <ftkalcevic> I find it easier to think in terms of signals rather than pins.
[01:42:45] <i-pink-cnc> net spindle-index spindle-phase-a <= parport.0.pin-12-in
[01:42:45] <i-pink-cnc> net spindle-phase-b <= parport.0.pin-13-in
[01:42:53] <ftkalcevic> Connect each pin to the signal
[01:43:24] <i-pink-cnc> net spindle-index <= parport.0.pin-12-in
[01:43:26] <andypugh> So put the spindle index and spindle-phase-a on seperate lines, with the same p-port pin?
[01:44:06] <ftkalcevic> Yep. That's what I do.
[01:44:23] <i-pink-cnc> i want to get the encoder signal from pins 12-13
[01:45:26] <ftkalcevic> I'd create signals encoder_in_a and encoder_in_b. Then net the pins to the signals.
[01:45:36] <i-pink-cnc> i make it like this
[01:45:49] <andypugh> Worth mentioning that this isn't actually a spindle, it's the z-axis motor
[01:45:51] <i-pink-cnc> #net spindle-index spindle-phase-a <= parport.0.pin-12-in
[01:45:51] <i-pink-cnc> net spindle-index <= parport.0.pin-12-in
[01:45:51] <i-pink-cnc> net spindle-phase-b <= parport.0.pin-13-in
[01:46:03] <i-pink-cnc> yes
[01:46:24] <tom3p> the syntax is net signalname pinThatSignalIsConnectedTo <opt> nextPinThatSignalIsConnectedTo ...
[01:47:30] <andypugh> Put the first line on 2 lines
[01:48:00] <i-pink-cnc> now is ok?
[01:48:01] <i-pink-cnc> # net spindle-index spindle-phase-a <= parport.0.pin-12-in
[01:48:01] <i-pink-cnc> # net signalname pinThatSignalIsConnectedTo <opt> nextPinThatSignalIsConnectedTo ...
[01:48:01] <i-pink-cnc> net spindle-phase-a spindle-index <= parport.0.pin-12-in
[01:48:01] <i-pink-cnc> net spindle-phase-b <= parport.0.pin-13-in
[01:50:37] <i-pink-cnc> what happen here, now i get this error
[01:50:50] <i-pink-cnc> standard_pinout.hal:95: pin 'spindle-index' does not exist
[01:51:07] <andypugh> You need to do it on 3 lines
[01:51:21] <i-pink-cnc> how??
[01:51:35] <andypugh> net spindle-index <= parport.0.pin-12-in
[01:51:55] <andypugh> net spindle-phase-a <= parport.0.pin-12-in
[01:52:21] <andypugh> net spindle-phase-b <= parport.0.pin-13-in
[01:52:35] <andypugh> (probably I am guessing)
[01:53:07] <i-pink-cnc> mmm i try now
[01:54:00] <Dave911> At the risk of being nominated for the most stupid question of the day award - what is the point of having a "simulator" compile of EMC2 when I can run it with EMC2 configured with
[01:54:02] <Dave911> stepper outputs? And what is the point of wiring the Z axis motor to the spindle encoder inputs?? (perhaps two stupid questions ... :-) )
[01:54:18] <i-pink-cnc> no, i get error
[01:54:29] <i-pink-cnc> standard_pinout.hal:99: pin 'parport.0.pin-12-in' was already linked
[01:54:55] <andypugh> I am out of my depth now
[01:55:34] <i-pink-cnc> depth = ?
[01:55:47] <andypugh> It isn't actually a spindle encoder in this case, but it is in the halfile I stole the syntax from
[01:56:00] <Dave911> gotcha
[01:56:15] <AchiestDragon> Dave911: first on ,, so you can dry run the gcode on a machine other than the one attached to the cnc ,
[01:56:25] <AchiestDragon> on / onee
[01:56:28] <AchiestDragon> -e
[01:56:29] <i-pink-cnc> i try in now with this confiruration
[01:56:37] <i-pink-cnc> # net spindle-index <= parport.0.pin-12-in
[01:56:37] <i-pink-cnc> net spindle-phase-a <= parport.0.pin-12-in
[01:56:37] <i-pink-cnc> net spindle-phase-b <= parport.0.pin-13-in
[01:56:55] <andypugh> forget the first line
[01:57:15] <andypugh> just keep spindle-phase-a and spindle-phase-b
[01:57:20] <i-pink-cnc> no i disable it
[01:57:42] <i-pink-cnc> and now the EMC WORK!!!
[01:57:43] <andypugh> Ah, OK
[01:58:14] <i-pink-cnc> i try now to get signal from legs 12-13
[01:58:21] <Dave911> So the simulator is really relevant for servo driven machines .. ? I have my PC removed from my lathe and I am running it on my desk...... but I am using step and dir outputs - so no Mesa boards etc
[01:58:33] <andypugh> So now try changing the line that says net spindle-index encoder.0.phase-Z to
[01:58:53] <andypugh> net spindle-phase-a encoder.0.phase-Z
[01:59:34] <andypugh> You should be able to see spindle-phase-a and spindle-phase-b in hal meters (or halscope)
[02:00:28] <AchiestDragon> the simulator will let you dry run the gcode on a machine without the real time kernel can be usefull to test your gcode so you know the tool for example is not driven into the machine bed or vice before transfering the gcode to the machine that has the cnc attached to run it on there
[02:01:09] <AchiestDragon> to then find theres a hole in the table and a big cutout in the vice if you got the gcode wrong
[02:01:24] <Dave911> OK... >>on / onee -e What does this mean?
[02:01:24] <i-pink-cnc> i get signal in pin 12, but pin 13 not work
[02:02:18] <andypugh> Actually, you should see it in parport.0.pin-13-in
[02:02:40] <andypugh> Have a look at the actual pin
[02:02:51] <i-pink-cnc> actual pin?
[02:02:59] <tom3p> look at the integrators manual 9.1.4 http://linuxcnc.org/docs/EMC2_Integrator_Manual.pdf for how 'net' works, yes try a meter on pin 13
[02:03:01] <AchiestDragon> typo corrections ,,, "first on" should of read " first one " on/one is what i should of put meaning replace on with one but got one as onee so -e to remove the extra e
[02:03:07] <i-pink-cnc> i see with the hal scoop
[02:03:32] <andypugh> which signal are you looking at in halscope?
[02:03:47] <AchiestDragon> <
[02:03:50] <AchiestDragon> opps
[02:04:15] <AchiestDragon> <all text> / probablay subject to translation errors
[02:04:36] <andypugh> Me telling i-pink is a good case of the blind leading the blind, you know.
[02:04:39] <jst_home__> jst_home__ is now known as jst_home
[02:05:55] <i-pink-cnc> ok, now the pin 13 work too!!!!!
[02:06:17] <AchiestDragon> your doing a good job andypugh will be an expert soon at that rate
[02:06:30] <tom3p> i-pink-cnc: both 12 and 13 are always inputs, so if 12 works and 13 doesnt... then bad wiring or damaged port. be careful, check with a voltmeter if halmeter shows no change.
[02:06:46] <tom3p> i-pink-cnc: cool
[02:07:08] <andypugh> OK, do you see encoder.0.position changing as you spin the disc?
[02:07:13] <Dave911> ArchiestDragon> Thanks for the explanation of everything .. ! I need to learn the sim setup also I guess. AndyPugh> I saw your note about another way to get axis positions into G code .. :-) Thanks - I'll try that also .... I think andypugh needs a bravo for patience!
[02:07:43] <i-pink-cnc> no no, the gain it was on minimum... ;-)
[02:07:44] <Dave911> I'll bug out from you hal setup session. :-)
[02:08:02] <andypugh> Did you see the third way to get axis positions? Even better?
[02:08:45] <Dave911> andypugh> Yep I read it just a minute ago - I'll try it and see.. Thanks!
[02:08:53] <andypugh> No good if you are already using G92, but if you do G92 X0 Y0 Z0 then an immediate G92.2 you can then read the axes out of 5211 onwards
[02:09:25] <i-pink-cnc> now how i connect the motor???
[02:09:33] <Dave911> I think that will work fine then. A one line solution is very good!
[02:09:45] <andypugh> Do you have a H-bridge circuit?
[02:10:19] <i-pink-cnc> not yet..
[02:10:31] <i-pink-cnc> but i have fluk
[02:10:43] <i-pink-cnc> multimeter
[02:10:46] <andypugh> I think you need to do that, you can't drive the motor direct from the parallel port
[02:11:23] <i-pink-cnc> i know..
[02:11:30] <AchiestDragon> http://www.whipy.demon.co.uk/h-bridge-sch.pdf
[02:11:31] <andypugh> That circuit uses PWM from a p-port pin to do motor speed, and another to do direction
[02:12:42] <i-pink-cnc> WTF WOW how many transistors...
[02:13:06] <i-pink-cnc> i have L298 is dual H-bridge
[02:13:25] <andypugh> It's a H-bridge controller I think
[02:13:43] <i-pink-cnc> the L298, yes..
[02:14:15] <i-pink-cnc> is 5 min to build H-bridge with him.
[02:14:24] <AchiestDragon> the one i posted on that link has pwm speed , direction and a break also its good to 15A with the mosfets used , more if bigger mosfets are used and sutable heatinks ,,, will do up to 1.5A motors without heatsinks
[02:15:03] <i-pink-cnc> 15A WOW
[02:15:23] <i-pink-cnc> the L298 is ok for me
[02:16:48] <andypugh> You linked to http://emergent.unpy.net/projects/01142347802
[02:17:18] <i-pink-cnc> this site is empty!!
[02:17:37] <andypugh> Which in a slightly cunning way pulses the "enable" line with a PWM signal to control speed
[02:18:01] <tom3p> ? http://emergent.unpy.net/projects/01148303608
[02:18:14] <i-pink-cnc> real problem is how to get DC up/down output from the EMC
[02:18:44] <AchiestDragon> tom3p: that link works here
[02:19:23] <i-pink-cnc> how to make the H-bridge
[02:19:56] <tom3p> you have the hbridge, the l298 is an hbridge
[02:19:57] <i-pink-cnc> but i dont know what is the configur of the HAL file for him!!
[02:20:13] <i-pink-cnc> tom3p, iknow
[02:21:08] <i-pink> i am here
[02:21:54] <AchiestDragon> http://www.st.com/stonline/books/pdf/docs/1773.pdf page 6/13 fig 13
[02:22:06] <i-pink-cnc> flkjglkd
[02:22:17] <i-pink-cnc> and i here again
[02:22:26] <AchiestDragon> wb
[02:22:45] <tom3p> for emc configs for that l298 from jepler, look in your files .../configs/etch-servo
[02:22:49] <andypugh> You probably need to add a PID function to your HAL file
[02:23:47] <andypugh> As tom3p just said, you should be able to get clues from the etch-servo config
[02:24:08] <andypugh> Too late here, I am quitting myself
[02:24:27] <tom3p> no! sorry, the configs are stepper, it changed from servo to stepper at some date
[02:25:04] <i-pink-cnc> ok
[02:25:53] <i-pink-cnc> but i want to get a DC output signal from the EMC
[02:26:59] <tom3p> i-pink-cnc: i'm still reading the config. he used the term YDIR and i thiught it was stepper. i think its is servo. you must read
[02:27:36] <i-pink-cnc> YDIR = ?
[02:27:51] <tom3p> i-pink-cnc: he drives 2 servos with this config. YDIR is like CW/CCW.
[02:28:06] <i-pink-cnc> servo is DC motor & encoder
[02:28:28] <i-pink-cnc> what is ?? CW/CCW.
[02:29:09] <tom3p> yes, it is compatible. he has dc motor and encoder. he sets direction ( path thru h bridge ) and uses PWM to set speed.
[02:29:11] <AchiestDragon> cw clockwise ,, ccw counter clockwise
[02:29:33] <i-pink-cnc> OK
[02:29:39] <tom3p> i-pink-cnc: what is your native language, i know a few words
[02:30:03] <i-pink-cnc> is heb
[02:30:21] <tom3p> well, sorry, i never had hebrew schematics or books ;)
[02:30:31] <tom3p> but you do fine
[02:30:32] <i-pink> Hebrew
[02:31:18] <tom3p> do you see his config sets direction and speed?, and reads position from the encoder. its very much setup for your hardware
[02:31:41] <i-pink> i need to go to UK to few mounts to learn english...
[02:31:46] <tom3p> but you only need 1/2 of what he's done
[02:32:08] <i-pink> yes
[02:32:11] <i-pink> i need
[02:32:19] <i-pink> UP / DOWN
[02:32:35] <i-pink> pwm up / pwm down
[02:32:41] <tom3p> thats CW/CCW, turn your screw like a clock... does it go up or down?
[02:32:59] <i-pink> this is my configure
[02:33:35] <i-pink> i have 2 wire, one is up and the other is down
[02:34:04] <tom3p> where are these 2 wires?
[02:34:27] <tom3p> one wire does nothing in electricity
[02:34:45] <i-pink> is from the H bridge
[02:36:55] <tom3p> L298 has In1 In2 Ena Out1 Out2 Vs and ground. which is Up?
[02:37:56] <tom3p> maybe you have a board that uses an L298, and has other labels. then you need the book for that pcb.
[02:38:31] <i-pink> If you want to rotate the motor forward you give a logic 1 in the first wire, but if you want to turn it back you gives logic 1 to the second wire
[02:39:12] <i-pink> this is how my electronic work
[02:39:37] <i-pink> now i need to get the UP / down from the EMC
[02:39:39] <tom3p> ok, you have some kind of driver board. you need the info for that board, and you need to drive it from emc. thats called interfacing. it is not just an l298.
[02:40:03] <tom3p> you have an up and a down, what about speed?
[02:40:14] <i-pink> i have all the info about him ( i make it)
[02:40:29] <i-pink> i need only to drive it from the EMC
[02:42:19] <i-pink> the speed is very slow Because the gear
[02:44:02] <tom3p> ok, use XDIR for up down ( XDIR into an inverter give UP and /UP which you can use as DOWN). so, far you could use jepler's config, but you added a chip to your design. next, you need to control speed.
[02:45:28] <jimbo655> setting FERROR and MIN_FERROR ? should FERROR be set for the lowest error? FERROR = 0.0005 and MIN_FERROR = 0.010
[02:45:39] <spasticteapot> Is anyone here using a variable-speed router as a spindle motor?
[02:45:44] <spasticteapot> If so, what are they cutting?
[02:45:44] <tom3p> or no chip, and use HAL to create the down
[02:46:20] <i-pink> i think to make it in this way but is make damage to the motor
[02:47:12] <i-pink> i want it in 2 separate wire
[02:47:40] <i-pink> and i get up off down
[02:48:16] <tom3p> i-pink: make sure you dont turn on both sides of hbridge at one time. it would be 2 separate wires. they would always be differerent, only one would be true at any time
[02:48:45] <i-pink> ok
[02:48:57] <tom3p> off = enable, another input to the the device
[02:49:28] <i-pink> what?
[02:50:06] <tom3p> look at the ENA pin http://www.st.com/stonline/products/literature/ds/1773.pdf
[02:50:18] <i-pink> is 00=off 01=up 10=down 11=not work
[02:50:35] <tom3p> 11 = brake
[02:51:11] <i-pink> yes i know it
[02:51:56] <i-pink> and is ok to me
[02:51:57] <tom3p> that is different from 1 wire up and 1 wire down, it is more that just up & down. ok, i see your design now
[02:52:43] <tom3p> i know the speed is small, but do you want any control over speed?
[02:54:22] <i-pink> no
[02:54:34] <i-pink> the speed is not importent
[02:55:01] <i-pink> on of is ok
[02:56:27] <i-pink> if the EMC send the control via PWM is ok
[02:56:54] <i-pink> but is not impotent
[02:58:12] <tom3p> EMC is meant to be in position and on time. it does that by reading position and regulating speed. so its not important to you, but is important to EMC.
[02:58:38] <i-pink> ok
[02:58:58] <i-pink> from zero
[02:59:41] <i-pink> i have DC motor, what i need to do in the EMC HAL file to control it
[03:03:46] <i-pink> Because I was building the Driver of the motor, it's less matter to me how build it , and more important for me to understand what the EMC send to me
[03:04:13] <tom3p> list the inputs to your board, list the outputs to emc. begin with the etch-servo and see what is missing , work to fill in whats missing. i can guess that your board wants up, down and motorvoltage ( the speed ). emc needs encoder A B and optionally Index. emc gives Enable and PWM and DIR( can be fixed rate i suppose )
[03:05:06] <tom3p> you can read what the EMC gives you in etch.hal file
[03:05:30] <i-pink> where i found it?
[03:06:05] <tom3p> i will paste
[03:06:14] <i-pink> ok
[03:08:12] <i-pink> i paste my hal file
[03:08:49] <i-pink-cnc> http://pastebin.com/f83a0498
[03:09:27] <tom3p> http://pastebin.ca/1611443
[03:10:53] <i-pink> mmm ok i see
[03:11:25] <LawrenceG> jymm, http://imagebin.ca/view/LA2KdU.html
[03:11:29] <tom3p> your file is for 3 steppers, you need ( as i recall ) a file for 2 steppers(XY) and one servo (Z).
[03:12:33] <i-pink> now i need to use only half of the etch.hal file
[03:13:29] <tom3p> and half of yours, good going
[03:13:34] <i-pink> i am right?
[03:13:37] <i-pink> hehe
[03:13:41] <i-pink> omg..
[03:13:42] <tom3p> i think so
[03:14:10] <i-pink> i feel like a blond woman
[03:14:56] <i-pink> (but i am woman with braun hair)
[03:17:43] <tom3p> if you can run EMC wit the etch-servo config, then you have the 3 files neccesary for servo, and can use them to create your new hybrid
[03:18:53] <i-pink> i try now to run the EMC with the etch-servo config
[03:20:21] <tom3p> wow i didnt know clockwise counterclockwise was a bad analogy because some Hebrew and Arabic clocks actually progress in anti-clockwise direction.
[03:20:22] <i-pink-cnc> no,
[03:20:28] <i-pink-cnc> i get this error
[03:20:30] <i-pink-cnc> standard_pinout.hal
[03:21:24] <i-pink-cnc> opss
[03:21:27] <i-pink-cnc> insmod: error inserting '/usr/realtime-2.6.24-16-rtai/modules/emc2/trivkins.ko': -1 File exists
[03:21:27] <i-pink-cnc> standard_pinout.hal:3: exit value: 1
[03:21:27] <i-pink-cnc> standard_pinout.hal:3: insmod failed, returned -1
[03:24:12] <tom3p> you've already got some of emc running, i dont know how or why, but soem parts are already running
[03:24:14] <i-pink-cnc> i think to jump from the roof is easier
[03:24:46] <tom3p> only for a short time, then it is difficult again
[03:24:51] <i-pink-cnc> what already running??
[03:25:48] <i-pink-cnc> maybe is not only the hal file and we need to more files
[03:25:57] <tom3p> 'File exists' means its already been put in place by EMC, and cannot be put inplace again. are you already running EMC in another window?
[03:27:08] <i-pink-cnc> it can load this "loadrt trivkins"
[03:27:18] <i-pink-cnc> what is trivkins?
[03:27:55] <tom3p> trivial kinematics ( a simple xyz cartesian machine as opposed to a complex twisty robot arm )
[03:28:24] <i-pink-cnc> if i disable this line?
[03:28:34] <tom3p> kinematics is the relation between 'joints' ( like you arm ) and the left right motion that could be called 'X axis'
[03:28:45] <tom3p> no, dont hack
[03:29:39] <i-pink-cnc> i disable the line
[03:29:53] <i-pink-cnc> but now i get this error
[03:29:56] <i-pink-cnc> insmod: error inserting '/usr/realtime-2.6.24-16-rtai/modules/emc2/motmod.ko': -1 File exists
[03:29:56] <i-pink-cnc> standard_pinout.hal:4: exit value: 1
[03:29:56] <i-pink-cnc> standard_pinout.hal:4: insmod failed, returned -1
[03:30:08] <i-pink-cnc> what is motmod.ko?
[03:30:45] <tom3p> check if you are running emc first. if so, exit and re-run with etch-servo. if not, and its not too difficult, reboot and try to run it. ( not a good suggestion, but a clean start )
[03:31:16] <tom3p> motmod.ko is a kernal moduile that moves motors
[03:31:48] <spasticteapot> What do you call those doohickeys with a pin on the end and a dial that you use to center things on a lathe?
[03:31:56] <tom3p> you are having trouble starting because you have already started.
[03:32:01] <tom3p> indicator
[03:32:17] <tom3p> and the term is thingamajig
[03:32:54] <i-pink-cnc> it's c'not load 2 modules
[03:33:14] <i-pink-cnc> the trivkins and the motmod
[03:36:53] <i-pink-cnc> on.. my problem is to open the EMC with the etch.hal configuration
[03:38:35] <tom3p> i can only suggest restarting the computer, i dont know how to remove the files so they can be loaded.
[03:39:37] <i-pink-cnc> if you try to load the etch.hal configuration in your computer is work?
[03:43:07] <tom3p> yes i have it running now
[03:43:28] <tom3p> i chose it from the tree menu and it ran
[03:44:48] <i-pink-cnc> ok, you can send to me the folder /home/youruser/emc2/configs/etch-servo
[03:45:23] <i-pink-cnc> make a tar.gz / zip
[03:45:57] <tom3p> your problem is you began enc and aborted. some things did not clean up. rebooting is ugly but will get you a clean start. there are soem tricksm but i cannot find them.
[03:46:17] <tom3p> i pasted one, will pastebin the other 2 now
[03:46:43] <i-pink-cnc> i wont all the folder
[03:47:11] <i-pink-cnc> and upload it to here
[03:47:18] <i-pink-cnc> http://www.speedyshare.com
[03:47:20] <tom3p> ok, i'll send to filebin, not here
[03:47:31] <i-pink-cnc> here http://www.speedyshare.com
[03:47:59] <i-pink-cnc> is 6- 7 file, pastebin not help me
[03:49:01] <i-pink-cnc> make right click on the folder , then create archive
[03:49:20] <i-pink-cnc> and upload it to http://www.speedyshare.com
[03:52:01] <tom3p> http://www.speedyshare.com/530206508.html
[03:52:15] <i-pink-cnc> thank youuuuu!!
[03:52:44] <tom3p> best of luck, good nite
[04:13:21] <i-pink> vhh
[04:13:27] <i-pink> hii*
[06:55:15] <jymm> http://www.geekalerts.com/ikey-submersible-bluetooth-keyboard/
[09:37:31] <jymm> SWPadnos: Amazon sells Electronic components, including surface mount.
[13:23:21] <andypugh> Does G-code have a unary minus?
[13:23:40] <andypugh> I am having trouble with G1 X[-#1]
[13:24:21] <SWPadnos> no
[13:24:28] <SWPadnos> G1 X[0-#1]
[13:24:34] <andypugh> "Bad Number Format". However in MDI (debug, -#1) works fine
[13:24:47] <SWPadnos> it prints the -, then the variable
[13:24:54] <SWPadnos> try it with a negative number and see what you get
[13:25:06] <andypugh> Thanks SWP
[13:25:14] <SWPadnos> sure
[13:25:27] <andypugh> And 0- is neater than my idea of -1 *
[13:27:08] <SWPadnos> either should work, so you can choose the one you like the best
[13:31:10] <tomp> tomp is now known as tom3p
[13:58:26] <jthornton> Benny just joined the forum...
[14:50:43] <AchiestDragon> great sometimes dont you just hate google ,,, like "break pipe fitting dimentions " does return 57,300 results all of whom would sell you the fittings but none tell you the size of the thread
[14:53:49] <AchiestDragon> phone calls dont help eather ,,, want to know the size of the thread on a 3/8's break pipe fitting ,, give how many do you want ,, will be wanting a box of about 500 but need to know the size ,, hmm its the standard size for break pipe ,, yea i gather that do you know what it is , no will have to get back to you on that ....
[14:56:04] <AchiestDragon> so cant cad the hole unitll i know the size
[15:15:45] <alex_joni> AchiestDragon: http://en.wikipedia.org/wiki/National_pipe_thread
[15:16:38] <AchiestDragon> found one ,, ty 12mm x 1mm pitch for the metric fitting nuts
[15:17:29] <AchiestDragon> although great that most places dont know the sizes of what there trying to sell you
[16:29:17] <andypugh> ArchiestDragon: You might find what you are looking for better with a search for "Brake pipe thread dimensions". I am depressed that 57,300 web page writers can't spell
[16:30:10] <andypugh> In other news, I have a curious bootstrap problem, I need to solder a new magentic switch into my soldering iron....
[16:31:52] <fenn> you only have one soldering iron?
[16:34:52] <oPless> you could always have a vindaloo and breathe on it ;)
[16:36:23] <MattyMatt> a 600mm linear servo (used). what sort of positional accuracy could I expect off that?
[16:40:55] <MattyMatt> without temperature compensation :p
[16:41:22] <MattyMatt> (that's too much analog circuitry for me)
[16:45:27] <andypugh> MattyMatt: What sort of actuator? Ballscrew? Belt? linear or rotary encoder?
[16:46:06] <AchiestDragon> andypugh: yea you need a soldering iron to fix a soldering iron
[16:47:28] <andypugh> I might have a portable gas-powered one, but I can see me using Mole-Grips, a blow-torch and one of the spare tips.
[16:48:17] <MattyMatt> andypugh, it doesn't say, but it looks like the ones with acme or ballscrew and plastic potentiometer track
[16:48:32] <MattyMatt> fleabay :)
[16:49:06] <MattyMatt> http://cgi.ebay.co.uk/LINEAR-DRIVE_W0QQitemZ350260957549
[16:49:52] <AchiestDragon> yea got a gas one also ,,but have had the weller iron for over 20 years ,, its had 4 new elements , about 50 tips and at least 2 new handles , and a couple of new cables in that time though
[16:53:07] <AchiestDragon> still original psu though
[16:53:31] <andypugh> MattyMatt: I can't see if it has a motor or not, if it has an encoder or not, or even if it is electric or pneumatic
[16:54:30] <andypugh> Weller TCP with the clever curie-temperature tips? The King of soldering irons, and why I am looking to spend £20 on a new switch for an iron worth £19
[16:54:38] <AchiestDragon> unlike the cartaker and his broom
[16:56:09] <AchiestDragon> yea there woth the money are the weller TCP ones used to go though irons at a rate of 1 per year or so untill i got mine
[16:57:04] <AchiestDragon> like it used to get turned on at about 8:30am and off at about 5pm every working day of the year
[16:57:16] <MattyMatt> I think it's a nippon bearings BG actuator, if it's from the same place as last lot I saw in Cardiff. recirculating ballscrew
[16:58:46] <MattyMatt> dunno if it's worth a punt. I think it is but what I really dunno is if gbp20 will kill me
[16:59:48] <AchiestDragon> a full tcp iron and psu is still about £80
[17:01:03] <MattyMatt> £, I can c'n'p it, but I can't type it :p. NV reintroduced a bug that was eliminated from xorg 4 years ago
[17:01:26] <AchiestDragon> the antex manis ones at about £20 only last about 12 months , the cheaper ones you see for about £4 or so would only last a couple of days given the same sort of use
[17:01:58] <MattyMatt> my irons rarely die unless I abuse them
[17:02:52] <MattyMatt> having a range of sizes is good. a needle point antex. a fat 40W Weller, and one of their soldering guns is what I call a set
[17:03:49] <MattyMatt> I only have a needle point atm, as I'm building up from scratch again
[17:04:03] <AchiestDragon> with the cheap ones its not the fact they die , its the fact that if left on for a period of time the handle plastics melt where the element fastens on and they fall apart
[17:04:42] <MattyMatt> leaving them on for a period of time is the major abuse :)
[17:04:42] <andypugh> If I wanted an actuator, I would punt £20 on that eBay one.
[17:04:53] <andypugh> It's cheap for what it is
[17:05:43] <MattyMatt> I don't deperately want an actuator, unless it can be a whole X or Z on this router
[17:06:25] <AchiestDragon> not when your in a porduction/repair or prototyping enviroment where there given almost constant use all day then its not , its what is needed from them
[17:08:32] <AchiestDragon> ie a cheep unbrandded electric drill at say £10 from a diy store may suit a diyer to put up a bookshelf ,, but would not last a day with a contracter say screw fixing plasterboard pannels all day long
[17:09:39] <MattyMatt> yeah it's true, but with 2 cheap ones you've got one to repair the other one :)
[17:11:11] <MattyMatt> like with the head changer idea I've got. one cheap draper 'dremel' from Wilkos in each head, and they can cut all day at 25% duty cycle
[17:11:21] <MattyMatt> with 4 of them
[17:11:40] <MattyMatt> and I've still spent less than one quality router
[17:12:37] <AchiestDragon> there comes a point were its not economical ie a 20p screwdriver ,, does the job but because of the qualaty the tip wairs at a rate a contractor would use 1 a day ,,, where as say a £7.50 one may only last 2 to 3 years but is a cheaper option than the 20p per day , its not abuse of them its just there spesification is rather low
[17:12:45] <andypugh> My dad was known to use 3 angle-grinders in rotation when dressing some of the bigger gears he had to deal with. (like 1/8" off the drive face of every tooth of a 3 foot wide pinion...
[17:15:19] <andypugh> But to drag us back on to topic. I intend to add a joypad to my lathe/mill combo. Has anyone any ideas for button allocation?
[17:16:04] <AchiestDragon> tool quality is somthing that is worth the expence ,, a good tool should last a lifetime , not break the frist time you over stress it slightly or wair after a couple of days heavy use
[17:16:32] <MattyMatt> I've lost all my tools more than once. it's heartbreaking
[17:16:54] <MattyMatt> it's like losing a part of your body
[17:16:56] <AchiestDragon> is no fun
[17:17:29] <AchiestDragon> whats a pain is that i have some special tools replacing those is a real pain
[17:17:35] <MattyMatt> I lost my grandads, which is even worse. all his gauge blocks etc
[17:18:19] <MattyMatt> they were WW2 inches tho, not 25.4
[17:18:50] <AchiestDragon> and its almost imposible to get imperial sized tools now although snapon still do them
[17:19:44] <AchiestDragon> the last snapon socket i bought a 36mm one cost £48
[17:20:55] <MattyMatt> eek
[17:21:25] <AchiestDragon> if it breaks thay replace it for free nomatter how mutch abuse you give it
[17:22:29] <MattyMatt> I bought a load of TRS-80s at 2 grand each, thinking they would last me for lifetime. I didn't understand moore's law
[17:22:31] <AchiestDragon> only used it 4 times so far
[17:22:38] <andypugh> Can you tell the difference between a 25.4mm inch and a 25.399956mm inch?
[17:22:57] <MattyMatt> with gauge blocks you can >:)
[17:23:36] <renesis> bullshit
[17:23:41] <andypugh> At 0.000005mm breathing on them will change the thickness y as much
[17:24:14] <andypugh> I think renesis put it better than me :-)
[17:24:20] <AchiestDragon> lol
[17:24:26] <MattyMatt> ah yeah I got my dp 3 places wrong, yeah sorry
[17:25:33] <andypugh> I now wonder if MattyMatt is the enigmatic Bidder1
[17:25:34] <MattyMatt> that's 5nm, not 5um
[17:25:48] <MattyMatt> yes, I am bidder1
[17:26:31] <andypugh> Depending on the internal support arrangments it might make a good Z-axis all by itself
[17:27:03] <AchiestDragon> got a item number for that
[17:27:07] <MattyMatt> I was thinking that, but it's too long. it would make a nice X with one normal rail
[17:28:09] <MattyMatt> ooh. I won it
[17:28:58] <AchiestDragon> need to source a couple of rotary tables that could be modded to stepper drive in order to do a 5 axis machine
[17:29:09] <MattyMatt> sherline?
[17:29:26] <AchiestDragon> a 10" and 8" should do
[17:30:17] <AchiestDragon> atm just more interested in specs to ensure the rest of the build is going to take them
[17:32:20] <AchiestDragon> so can make something allong the lines of this http://sharpdimension.com/CNC_Milling/5-Axis-CNC-_Mill_Turn_Table.jpg to fit onto the bed of a 3 axis machine
[17:32:37] <MattyMatt> w00t, well I'll probably use this linear as a Z, with 200mm hanging off the top for my head changer
[17:34:03] <MattyMatt> so you need 3 of them?
[17:34:28] <MattyMatt> or would you just have bearings at one end?
[17:34:33] <AchiestDragon> one end would just be a bearing block
[17:35:38] <AchiestDragon> but need to work out the size that assembly is going to take so i can base the travel limits of the 3 axis machine and bed to take that
[17:36:34] <AchiestDragon> i need a 6 to 8" dia aria in a 5 axis configuration
[17:37:28] <AchiestDragon> and there will probablay be a 8" 3or 4 jaw lathe chuck for workhoding on that
[17:37:31] <MattyMatt> are you going to cast the cradle?
[17:38:29] <AchiestDragon> no was thinking of using a short lenth of steel channel section about 20mm thick for that
[17:38:47] <AchiestDragon> short as in 8 to 10"
[17:40:13] <AchiestDragon> may need extra bracing brackets welding to it to give it extra strenth , but casting one is out of the question for me
[17:41:06] <MattyMatt> that would be a mighty casting to do in a backyard furnace
[17:42:13] <MattyMatt> biggest I've seen done is 60lb of iron
[17:42:57] <AchiestDragon> my uncle used to own a foundry they did aluminium brass gunmetal , bonse etc ,, but he sold it a few years back , the comany that took it over closed it down and flattend the place for the land
[17:42:59] <MattyMatt> 8" tables are out of my league, if they are properly specced ones
[17:43:07] <andypugh> I used http://www.axminster.co.uk/product.asp?pf_id=453684&name=rotary+table&user_search=1&sfile=1&jump=44
[17:43:17] <AchiestDragon> miffed as wanted to buy it but did not know he was selling it at the time
[17:44:51] <andypugh> But I still ended up having to change all the worm bearing arrangements to make it work reliably with a stepper (needle roller inboard and a pair of back-to-back angular contacts outboard near the motor/coupling
[17:45:48] <andypugh> If I was making a 5th axis assembly like that one illustrated I would be very tempted to use toothed belt, though enormous toothed-belt pulleys are rather expensive themselves
[17:46:37] <andypugh> I am assuming that you don't have a harmonic-drive sized budget
[17:46:38] <MattyMatt> that was actually why I asked about the casting :)
[17:47:17] <AchiestDragon> its totaly inpractical for me to setup a "backyard" furnice here as i dont have a sutable back yard ,, its a semi style flat here one up one down , i have the front garden the flat below has the back , cant have a shead or and furnice in the front garden realy without a lot of objections its also a "conservation aria" so got extra planning restrictions in force here also
[17:48:11] <andypugh> You could pioneer the "attic furnace" or even the "bedroom furnace"
[17:48:19] <AchiestDragon> the reason i was going to use a pair of rotary tables is there already calibrated as in number of turns per deg on the drives
[17:48:23] <MattyMatt> or the fireplace furnace
[17:48:26] <AchiestDragon> drives / shafts
[17:48:34] <MattyMatt> is it an old house?
[17:49:16] <MattyMatt> mmm, molten iron, upstairs flat, wooden floor >:)
[17:49:20] <andypugh> There are cheap rotary tables on eBay. just accept that you will be heavily modifying
[17:49:39] <AchiestDragon> yea,, council , they ripped out our nice coal powered fireplace / heating 2 years ago and fitted the totaly uneconomical gas system i have to put up with now
[17:49:41] <andypugh> Yeah, what could possibly go wrong?
[17:50:40] <AchiestDragon> like can nolonger make horse shooes wile keeping warm in the winter
[17:50:51] <andypugh> The drawback with any conventional rotary table is going to be backlash, which will lead to chatter in 5-D routing applications.
[17:50:57] <MattyMatt> where are you building this machine?
[17:51:38] <AchiestDragon> will be buinding it to fit in the 6foot square brick outhouse that used to be the coal bunker
[17:51:51] <AchiestDragon> where it will live and work
[17:52:26] <AchiestDragon> thats the total usable workshop space i have avalable
[17:56:44] <AchiestDragon> a garage would be ideal , but £12.5 a week rent for one that has no power and no way to get power to it other than a £2500 supply connection charge is not practical eather
[17:58:45] <MattyMatt> I'm building a 25ft by 14ft one for a friend. he's got his rolloer shutter up now, I can start on the roof
[17:58:55] <AchiestDragon> andypugh: then looks like i am going to need to design a rotary table worm drive gear with antibacklash
[17:59:28] <MattyMatt> you can buy them, but they are frightfully expensive
[17:59:42] <AchiestDragon> sort of going to be hard to do that without a lathe to make them on
[18:02:35] <andypugh> The tables are adjustable, but they tend to get stiff at zero backlash, and the zero-backlash setting isn't all that constant round the rotation.
[18:03:40] <andypugh> You might be able to find a compromise, or you might be able to bias it with a big rubber band.
[18:04:29] <MattyMatt> or nurse it constantly, if you've left enough room for an operator
[18:05:47] <andypugh> Keep an eye out for an even cheaper one of these?
[18:05:48] <andypugh> http://cgi.ebay.co.uk/HARMONICDRIVE-CS-20-50-2A-GR-SP-HARMONIC-DRIVE-NIB_W0QQitemZ370266892812QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item5635a06a0c&_trksid=p3286.c0.m14#ht_600wt_941
[18:06:56] <andypugh> http://en.wikipedia.org/wiki/Harmonic_drive
[18:08:36] <AchiestDragon> yea bit pricey
[18:08:46] <MattyMatt> if you can cut gears from spring steel you can make your own
[18:09:09] <AchiestDragon> http://cgi.ebay.co.uk/HEAVY-DUTY-12-HORIZONAL-VERTICAL-ROTARY-TABLE-FREE-P-P_W0QQitemZ260489398706QQcmdZViewItemQQptZLH_DefaultDomain_3?hash=item3ca660c9b2&_trksid=p3286.c0.m14 ok so 12" but not a bad price for the table
[18:12:20] <AchiestDragon> dam as per usual see just what i need when i dont have the cash
[18:12:22] <AchiestDragon> http://cgi.ebay.co.uk/Rotary-Table_W0QQitemZ150378336318QQcmdZViewItemQQptZLH_DefaultDomain_3?hash=item23033f503e&_trksid=p3286.c0.m14
[18:14:31] <MattyMatt> you've got 4 days and 20 hours to rob a bank
[18:15:40] <MattyMatt> it's not coaxial, so it's beautiful
[18:15:49] <MattyMatt> ^BUT it's
[18:16:07] <AchiestDragon> and not long afte that to get the £120 it would cost me in petrol to collect it
[18:16:26] <MattyMatt> full version. it's not coaxial so it'll need more space, but it's beatiful anyway
[18:17:15] <AchiestDragon> yea
[18:22:16] <AchiestDragon> could always sell the ex , although i think she may object
[18:23:37] <AchiestDragon> although would have other advantages as i would get the house back then and have space
[18:26:10] <MattyMatt> no kids to share a hobby with?
[18:26:31] <MattyMatt> "here son/daughter, your first lathe"
[18:26:45] <MattyMatt> "I'll come and show you how it works"
[18:27:09] <AchiestDragon> well could buy it for kildet for xmas , but some how think he would complain that its a bit too heavy to attach to the lego mill
[18:27:11] <MattyMatt> ex : "she's only 2"
[18:29:04] <MattyMatt> one thing I'm working towards is milling & ginding moulds for injection molded plastics. custom lego parts are on the agenda
[18:30:15] <AchiestDragon> was thinking allong the lines of making lego compatable parts from aluminium
[18:30:34] <MattyMatt> me too :)
[18:31:29] <MattyMatt> those long technik struts with the holes in them. they could be 2ft long in ali
[18:31:32] <AchiestDragon> although one of the inportant factors with lego is the type of plastic as the flex is inportant for a correct tight fit
[18:31:58] <AchiestDragon> yea
[18:32:14] <MattyMatt> yeah, which is why cheap substitue lego won't do
[18:32:44] <AchiestDragon> its a abs type plastic they use ,, plastic to alli would work but custom alli to alli wound have some fit problems
[18:33:09] <MattyMatt> I thought if there's a market, it will be from students, who can use superglue
[18:33:11] <AchiestDragon> megablox use an abs thats close enough
[18:34:48] <AchiestDragon> old hp printerplastic thats abs is about right ,,, abs you grind up , heat to 240c in an oven to melt and you can pour into a pre heated alli mould (preheat to 80C) then let it cool , part done , but fumes are a bit toxic
[18:35:04] <AchiestDragon> so can do it in the kitchen oven
[18:35:39] <MattyMatt> neat :) I was planning to use a brass piston and suck the air out before heating
[18:36:10] <AchiestDragon> keep meaning to have a go to see if the level of fumes is acceptable for outdoor /well vented use
[18:36:35] <MattyMatt> I would keep it airtight anyway if possible
[18:36:50] <AchiestDragon> although was thinking of just using that way to create solid blocks of abs for further machining to needed shapes
[18:36:51] <MattyMatt> it will fill the mould much better too
[18:37:27] <MattyMatt> ah, pile it up in the ingot mould, and microwave it :)
[18:37:34] <AchiestDragon> with all the recycle logos you would be suppised at the amount of it used on packaging like bottle tops
[18:37:55] <MattyMatt> hose from the vent from the microwave to outside
[18:38:00] <AchiestDragon> other than the fumes it is a ideal reusable material
[18:38:39] <AchiestDragon> toaster oven is a better option than microwave ,, abs has no water content so would not heat up in a microwave
[18:39:08] <AchiestDragon> have a toaster oven i use for bulk smd desoldering
[18:39:34] <AchiestDragon> but would need another to preheat the moulds in
[18:39:45] <MattyMatt> ah cool. so will I when momma gets fed up of her latest SS toaster
[18:40:26] <MattyMatt> heat it up directly in the mould
[18:40:47] <MattyMatt> top it up when the first load is flat
[18:41:28] <AchiestDragon> bulk smd desoldering is fun ,,, place motherboard in toaster oven , cook for 5 to 10 mins , remove turn over and tap and you have a pile of smd components and a empty board
[18:41:55] <AchiestDragon> putting it back together takes a bit longer though
[18:43:01] <AchiestDragon> you need to remove the none sm connectors first as they tend to melt in the process
[18:43:58] <MattyMatt> I saw a nice vid of a component placer. it sucked the component up, then slid it around a square milled pit to get it centred on the nozzle
[18:44:42] <AchiestDragon> pick and place form tape or trays removes the need to do that
[18:44:57] <MattyMatt> slow but thorough
[18:45:32] <AchiestDragon> just sucks it up and places as its picked up centered in the first place
[18:45:59] <MattyMatt> yeah that would need precision feed of the tape
[18:46:28] <AchiestDragon> yea the machines that take tapes or trays are
[18:46:46] <MattyMatt> this was a homebrew one, so I was impressed
[18:48:09] <MattyMatt> the tapes had mixed components, so he probably made them by hand
[18:56:20] <jymm> SWPadnos: http://www.geekalerts.com/transistor-cufflinks/
[18:57:37] <MattyMatt> 240C to liquify ABS? that's rather higher than the polythene most reprap appear to use
[18:59:35] <AchiestDragon> http://www.whipy.demon.co.uk/ram.pdf well first robot part assembly done
[19:00:30] <AchiestDragon> 120mm stroke pneumatic ram
[19:02:09] <jymm> NOW we're talkin.... 38,000,000 Candlepower flashlight http://www.geekalerts.com/maxblaster-the-handheld-sunbeam/
[19:03:20] <ds3> nice
[19:03:29] <AchiestDragon> sweet
[19:05:08] <fenn> candlepower is a stupid measurement
[19:06:17] <AchiestDragon> yea should be in magnitiude astronomical term sun =1
[19:06:31] <fenn> it measures intensity, not light output
[19:06:47] <fenn> so a focused light would have more candlepower than a defocused light
[19:09:30] <AchiestDragon> yea but magnitude in astro terms is relative to photon count at the detector so higher birghness lowere mag value ,,,mag 1 is the brightness of the sun , mag 9 to 12 is the lowest limit of the human high , like a candle at 1 mile ,
[19:10:47] <AchiestDragon> its an inverse log scale
[19:12:12] <MattyMatt> at last. I've won a drill press. It's not quite what I wanted tho. Clarke lever press with collet for B&D
[19:12:32] <MattyMatt> it's better than hand held tho
[19:12:39] <AchiestDragon> yea
[19:13:09] <MattyMatt> collar, not collet :)
[19:15:16] <MattyMatt> I'm still gonna bid on the Meddings with a dead motor. Warrington is drivable from here
[19:15:35] <AchiestDragon> the problem i find with the little one i have is if you are drilling alli block and say drill a 6mm hole then want to widen it to a 12mm hole in oder to get the 12mm drill bit in the chuck you need to move the table out of the way and drop it as the 12mm drill is longer than the 6mm but also longer than the "z" range of the drill meaning that you have to manualy try to center the work again
[19:15:45] <MattyMatt> that looks seriously heavyweight, and is going for scrap price so far
[19:17:00] <AchiestDragon> found that making the stepper couplings ,, end result is the cupplings ended up off center
[19:17:26] <jymm> I think I like it... http://www.artlebedev.com/everything/optimus-tactus/
[19:17:40] <MattyMatt> jymm, look at the price
[19:18:09] <MattyMatt> we'd all love a screen in every key :)
[19:18:36] <jymm> MattyMatt: what price?
[19:18:56] <jymm> MattyMatt: what key? there are no keys on it
[19:19:02] <AchiestDragon> 1692.31 US$ for the other one with keys on
[19:19:05] <MattyMatt> of the optimus. the last one was $1300
[19:19:23] <jymm> MattyMatt: Ok, and?
[19:19:33] <MattyMatt> ah tactus is just a touchscreen?
[19:19:39] <MattyMatt> * MattyMatt looks
[19:19:57] <AchiestDragon> yea its also a concept design so not in production yet
[19:21:01] <AchiestDragon> so no price on that ,, could actualy be cheaper , should be arround $300 but bet they will keep it above the price of the other for a year or so yet if it was out
[19:21:59] <MattyMatt> the concept is "can we get a touchscreen LCD the shape of a keyboard?" and the answer is no, and when they are available you'll get them direct from Samsung or whoever
[19:23:38] <MattyMatt> it will happen eventually (Next Generation :-) ) but it's not mature enough for mortals yet
[19:24:04] <MattyMatt> except on the Nintendo DS
[19:24:34] <MattyMatt> which could run emc, I think
[19:24:47] <MattyMatt> it's naturally real time kernel
[19:25:47] <jymm> SWPadnos: Finally... http://www.geekalerts.com/back-to-the-future-flux-capacitor-replica/
[19:26:06] <AchiestDragon> the answer is yea , custom lcd pannels have been arround for some time , in porduction qanities theres not going to be a lot of price diference ie quantities of 10,000 or more ,, custom costs for setup are arround £4000 the point is if there asking $1600 for a keyboad there not going to sell a enough each year , as they would rather sell 10 with a $1000 per item profit than sell 10,000 for a $10 profit each so they could but it would
[19:26:07] <AchiestDragon> hurt there wallets as they dont beleve in a fair price
[19:26:26] <MattyMatt> IBM could have waded in with digital paper real keycaps, but they've sold both businesses
[19:26:34] <jymm> It's not a keyboard, it's Functional Art
[19:27:01] <MattyMatt> jymm, don't believe the hype. it's a sodding keyboard :)
[19:27:41] <andypugh> MattyMatt: Go for the Meddings drill , they are proper bits of kit
[19:28:19] <AchiestDragon> keyboard with a so called "desgner logo" if a big manufacture like sony made it then in a year or so they would be selling for about £100 rrp
[19:29:19] <andypugh> Plus that drill is a rarity, it actually has a chuck guard still fitted!
[19:29:39] <MattyMatt> I have 2x Model M. modern replacements are ~£100, or genuine reconditioned got for ~£55
[19:30:09] <MattyMatt> I found one at the dump :)
[19:30:13] <MattyMatt> this one
[19:30:31] <andypugh> A friend of mine was after a Model M, then I suggested he try the new Apple keyboards, and he totally changed his mind about whathe wanted.
[19:31:20] <andypugh> The Apple ones are a love or loathe thing, very short action.
[19:32:15] <AchiestDragon> still prefer the old keyboards with proper keyswitches rather than the modern membraine ones ,, there better to acutaly use and type on ,, these moddern ones are not as posative and tend to lead to too many typeos
[19:32:38] <MattyMatt> I've got a thinkpad which could persuade me to like short travel eventually
[19:32:55] <AchiestDragon> the tuch ones tend to hurt the ends of you fingers if you have to type a lot of text in a day
[19:33:39] <MattyMatt> I developed a rolling typing action on the Spectrum, but that was rubber for comfort :)
[19:34:53] <AchiestDragon> the zx80 was a pain ,, after you typed in the max 1k of code on the basic spectrum you needed to take a break for a day to let your fingers recover
[19:35:33] <MattyMatt> yeah zx80 was worse than playing guitar
[19:36:31] <MattyMatt> when I started playing guitar I had callouses to thick they ground holes in my shift and A key
[19:37:13] <AchiestDragon> this is a cherry (linux) keyboard (with the pengwin logo rather than the windos one) not too bad good by todays standard but have better the old dell one i find better for lots of typing
[19:37:13] <MattyMatt> cheap membrane keyboard
[19:39:33] <MattyMatt> those holes I wore in my keytops suggests that ABS needs conditioning like chocolate, it needs to be mixed constantly while molten
[19:39:36] <andypugh> The software co I used to work for had ZX80s (and spectrums) plugged into some sort of host system with a proper keyboard that did something clever with memory so you could edit in one system, run the ZX80, watch it crash then try again
[19:40:21] <MattyMatt> was it Software Projects?
[19:40:43] <andypugh> No, the much less famous M C Lothlorien
[19:41:19] <MattyMatt> the interface could well have been my design
[19:41:24] <AchiestDragon> unfortunatly the good keyboards with individual contact switches are dificult to get for pc's ,, the last one i saw was about the time of the pc xt with the 5pin din plug ,, not seen one with a at type ps2 connector
[19:41:28] <andypugh> (I am assuming you are not _the_ Matt Smith)?
[19:41:44] <MattyMatt> the one and only
[19:42:03] <fenn> matt smith #127
[19:42:09] <andypugh> Manic Miner?
[19:42:25] <MattyMatt> and I still beat the new Doctor Who on google \o/
[19:42:47] <MattyMatt> that one yeah
[19:43:47] <AchiestDragon> oh you may know graham (kenny) everett then
[19:44:24] <MattyMatt> nope. bruce everiss from microspot/imagine
[19:45:14] <MattyMatt> graham everett doesn't ring a bell, although I'm running through the grahams I knew back then
[19:45:29] <AchiestDragon> saw kenny yesterday ,, he did a lot of amega stuff ,, not to mention ps1,2 games and others
[19:45:39] <MattyMatt> grob was graham roberts (panda alarms)
[19:46:09] <AchiestDragon> psignosis amongst other places
[19:46:13] <MattyMatt> I was well out of it by the time amiga and psx were hot
[19:46:52] <MattyMatt> he isn't one of the pair that took over Jester is he?
[19:47:04] <AchiestDragon> i did h/w back then never realy got into software writing
[19:47:06] <AchiestDragon> no
[19:47:28] <AchiestDragon> "brain in a jar" was the main one he talks about
[19:47:57] <AchiestDragon> hes doing java web apps now contract working indipendantly
[19:48:19] <andypugh> I do wonder what happened to the Lothlorien guys, they seemed immensely old when I was 15, so must be ancient now. Roger Lees, Mike Cohen, Geoff Street
[19:49:24] <MattyMatt> it is a young man's game, although in retrospect I shouldn't have lost heart at 19 because 16 yo were catching up :)
[19:51:21] <AchiestDragon> yea althoug i now find myself getting the current generation of coders to learn asm seems they find writing inificiant c code the way to go and would rather compain the machines too slow for it
[19:51:29] <pjm> wow, Matt Smith, amazing. I still play MM and JSW !!! I'm in awe!
[19:51:57] <MattyMatt> they got 4 mentions on telly this week :)
[19:52:39] <AchiestDragon> ie they woud rather use a microconroler with 64k of program space to do what i used to do in 1k and find they have speed issues also
[19:52:42] <pjm> i showed my son what games i used to play when i was around his age, and he is addicted to MM, of course i'm playing it on a 48k speccy rather than a pc-emu
[19:52:53] <fenn> C is too hard just use UML
[19:53:25] <AchiestDragon> may as well use python rather than c
[19:53:29] <MattyMatt> I want mixed C & UML diagrams. kdevelop almost does it for C++, but not C
[19:54:04] <fenn> my sarcasm backfired
[19:54:16] <andypugh> Yeah, it was about the age of 19 that I decided that computers were just a tool like lathes, and there would soon not be any real money in working them, so I went into science. That time might be coming, but I could probably have made a fair bit of cash in the intervening 20 odd years
[19:54:18] <AchiestDragon> python rather than c++ rather c is ok
[19:54:28] <MattyMatt> I started with hardware myself. I like schematics :)
[19:54:44] <MattyMatt> UML = code as a schematic
[19:55:05] <andypugh> I like Labview
[19:55:36] <andypugh> And all the ECU software on the cars I work on is coded directly in Simulink
[19:55:40] <AchiestDragon> did a number of cpu designs ,, nothing better than playing arround with microcode
[19:56:24] <MattyMatt> VHDL is the level I'll work at for chip design
[19:56:39] <andypugh> So the documentation we get for the ECUs _is_ the source code, as a 4000 page PDF file!
[19:57:32] <MattyMatt> lovely
[19:58:00] <AchiestDragon> shame realy the p4 and newer have the programable microcode that the alpha cpu had , the shame is that using microcode you can define your own instructions , and the most that gets done is a perdefined set of extentions get loaded at boot and it gets left at that
[19:58:58] <MattyMatt> it would cause too much grief to mess with that, unless you used a 386 distro
[20:00:09] <AchiestDragon> yea but the point is on the alpha you could load a optimised instruction set for each active task ,,, the p4 implimietation is limited
[20:00:50] <MattyMatt> what was that taiwanese one with the totally definable set, transmeta?
[20:00:56] <AchiestDragon> so besides compiling the code you could optimize the instruction set for it also
[20:02:25] <MattyMatt> gotta think reusable these days
[20:02:36] <MattyMatt> that's why I use C, not x86
[20:02:41] <AchiestDragon> dec made the alpha a opensource design , after compaq bought out dec they sort of crushed it , it was a far better cpu than even the current X86 chips
[20:03:49] <MattyMatt> Yeah I can rememebr looking at the Alpha Screamer mobo, but it was too exotic
[20:04:00] <AchiestDragon> but seems that the compilers and o/s's today would need major chages to support tec like that , also support for true parrallel is not easy in current ones its just not scailer in that respect
[20:04:59] <andypugh> Transmeta wasn't Taiwanese (I was randomly reading about it a few days ago)
[20:05:14] <MattyMatt> if you could rapidly switch between codes, or have one code per core, it could work nice
[20:05:51] <andypugh> Transmeta Crusoe was VLIW
[20:06:47] <MattyMatt> oh yeah it used JIT
[20:07:06] <AchiestDragon> well more inportantly would be dynamic multiprocessor support ,, but seems ms wont develop it as it would cost$ with no incentive , and linux alhough some provision is provided the industry is still geared up to systems that dont practiacly support it
[20:07:34] <MattyMatt> AMD had a plan for that, but they've gone quiet lately
[20:07:46] <MattyMatt> 8 dissimilar cores
[20:07:55] <AchiestDragon> although as the trend is now towardes multicore cpus then we may see some advances
[20:08:28] <MattyMatt> AMD supported that fpga coprocessor, but sales must have been disappointing
[20:08:53] <AchiestDragon> fell out with amd after the k6
[20:09:59] <AchiestDragon> as every amd machine i had at that time would crash at about a 24hour interval ,, dew to some mistiming or cpu bug ,, not overclocked eather
[20:10:12] <MattyMatt> the plan was, develop co-processor apps on the fpga kit ($4000 thing that went in one socket of dual mobo) and then later AMD would produce custom chips from the same VHDL
[20:10:54] <AchiestDragon> although i am not a big intel fan all the pc's other than the laptop here have p4 xeons in them earher single or dual cpu boards
[20:10:57] <MattyMatt> this 4450e is rock solid
[20:11:32] <MattyMatt> quiet too
[20:13:58] <AchiestDragon> this is a dual 2.8ghz twin xeon , got another 3 of these , and one single 2ghz one in current use , another one that i removed form service (needs hdd) and a p4 laptop as main ones , with a couple of 1.2ghz via mini itx systems that i use for low power control
[20:14:49] <AchiestDragon> and a number of other systems in bits kicking arround the place
[20:16:32] <MattyMatt> http://www.drccomputer.com/drc/products.html
[20:17:09] <MattyMatt> them's the ones, although the old ones didn't have their own dram slots
[20:18:26] <AchiestDragon> fpgas are kool for some respects but they cant beat what you could do on the same amount of silicone with fixed asic design speed or power consumption wise
[20:18:58] <MattyMatt> true, but these are off the shelf
[20:19:34] <AchiestDragon> true ,, and you cant beat the flexibilaty of a fpga
[20:19:40] <MattyMatt> and AMD were making noises about small runs with these in their 8 different core design
[20:20:28] <MattyMatt> their equiv of Larabee
[20:20:51] <AchiestDragon> was amd that started the fpga development ,,, there LCA chips (logic cell array) back in the 80's they sold that section off it became xilinx
[20:20:59] <MattyMatt> but they are going ball-out for "more x86 cores than intel" atm
[20:22:00] <AchiestDragon> amd have a good development team , shame there silicone production is not as good as it could be
[20:22:48] <AchiestDragon> intel have a good silicone porduction but there development team dont seem to come up with the right concepts
[20:23:48] <AchiestDragon> the amd29000 was a realy good cpu also quashed by market forces demanding m$ compatable o/s use hence forced x86
[20:24:06] <AchiestDragon> i am not a big fan of the x86 arcitecture
[20:24:15] <MattyMatt> tell that to motorola
[20:25:29] <AchiestDragon> why the ppc core is gaining strenth where its needed now anyway ,, and the cell type ppc stuff has its own market almost tied up with the games consoles
[20:26:01] <MattyMatt> maybe their'll be a new Amiga :)
[20:26:19] <andypugh> I think Apple have proved that you can switch CPU architectures without too much trouble, they have done it twice now, including an endian-switch
[20:26:55] <MattyMatt> if you control the OS you can, and more manufacturers should
[20:27:18] <AchiestDragon> m$ withdrew suport for non x86 cpus about the time that dec did the alpha that sort of killed the alpha cpu ,, another realy exelent cpu design that made the x86 look like a steem engine on a airoplain
[20:27:54] <MattyMatt> DEC shoulda made a VM
[20:28:18] <MattyMatt> it's easy with hindsight :)
[20:28:21] <AchiestDragon> linux is inehrantly multiplatform if it ever beats m$ then we may see a lot better cpu designs arround
[20:28:47] <MattyMatt> linux made the arm netbook possible
[20:28:59] <MattyMatt> ignoring Win CE :)
[20:29:18] <AchiestDragon> but for the time intel and ms have a monopoly that needs breaking in order for better machines to be developed
[20:30:37] <andypugh> Hmm, make it 3 times Apple have switched CPUs, the iphone uses ARM, and it basically MacOS (you should see the hundreds of pages of GPL licencing info in the info bit)
[20:31:35] <MattyMatt> the internet appliance is becoming a reality. so far they are normal laptops but many users wouldn't notice if it didn't run Win apps
[20:31:39] <AchiestDragon> not a big apple fan , over priced under powered and pay for the name
[20:32:29] <MattyMatt> apple 2 wasn't underpowered >:)
[20:33:01] <AchiestDragon> did you know that apple got the 68000 cpu in the clasic apple made with bits removed in order to get the price down ,, actualy cost them more to have the bits removed than it would of just to buy the chips with it in ,, they later did that
[20:33:09] <MattyMatt> ST was a quarter of the price tho, so that was an easy decision
[20:33:19] <MattyMatt> ^of the Mac 512
[20:33:35] <AchiestDragon> the st was good had a st1024fm
[20:34:36] <AchiestDragon> shame atari lost out to commodore ,, did not like commodor as every machine they produced was incompatible with the last
[20:35:06] <MattyMatt> oh amiga was good too
[20:35:15] <MattyMatt> but I never had one
[20:35:43] <AchiestDragon> atari made a big statement on the st relese that they where extreamly sorry that the st would not be able to run the atari 400 / 800 software on
[20:35:52] <MattyMatt> any amigas that came into SP were kept well out of my hands. I was on the naughty step
[20:36:02] <AchiestDragon> lol
[20:37:09] <MattyMatt> I had an 800. the most overengineered machine ever :)
[20:37:56] <AchiestDragon> just seems a shame that the best tech doesent always win the race ,, like the whole vhs beta issue ,,, v2000 was the best tech wise , but just so badly marketed that it never got the chance
[20:38:26] <MattyMatt> you could but porn on V2000
[20:38:30] <MattyMatt> buy
[20:38:51] <MattyMatt> that seemed to be the main market
[20:39:57] <AchiestDragon> v2000 tapes wher same size as vhs and the v2000 tapes where double sided like cassets turn it over and you got another 2/4 hours recode time on other side of tape , the heads had pizzo allignment for tracking also
[20:40:49] <MattyMatt> yeah the kids put a V2000 cassette in a piano key VHS and broke it on that BBC4 prog
[20:41:00] <AchiestDragon> would of been good if philips actualy bothered to alling the heads a bit better in the factory then would of saved them going becasue of constant stress on them
[20:41:15] <andypugh> Thought experiment: If you take the spools out of a VHS casette and swap them round so the tape is upside down, what happens to the picture?
[20:42:11] <alex_joni> it stops working
[20:42:24] <andypugh> No, I have done this :-) It still works
[20:42:34] <AchiestDragon> sound bars on video and no sound because of soundtrack /video positioning on the tape
[20:42:52] <andypugh> Yes, no sound is one (less interesting) effect
[20:43:29] <AchiestDragon> suprised you could get the video to lock on the tracking though
[20:43:34] <alex_joni> backwards?
[20:43:40] <AchiestDragon> yea
[20:43:51] <AchiestDragon> and upside down
[20:43:53] <MattyMatt> http://www.bbc.co.uk/iplayer/episode/b00n59t4/Electric_Dreams_1980s/
[20:44:20] <andypugh> Upside down is the fun bit.
[20:44:49] <MattyMatt> if it's both upsiode down and backwards, the stripes would be in the same direction
[20:45:14] <andypugh> It occured to me that it should happen after a TV program where Tim Hunkin swapped the read heads for felt-tip-pens and pulled some paper through to demonstrate the helical scan
[20:46:02] <AchiestDragon> the cctv time lapse ones are good if you get the tracking right you can timelaps a 480 hour shot on a 3hour tape , and play it back , just about the right speed to wach a plant grow
[20:46:21] <pjm> andypugh ah u remember the 'secret life of machines' series - that was excellant
[20:46:43] <tom3p> still available on line
[20:46:54] <pjm> yeah i have them all on my popcorn-hour
[20:47:12] <pjm> its a damn pity stuff like this isnt made properly any more
[20:47:20] <tom3p> http://www.secretlifeofmachines.com/
[20:47:24] <pjm> one of my favorites is the 'lathe fax machine'
[20:47:36] <andypugh> Indeed, marvellous program. I remember the one where Tim and his sidekick (who's name will come back to me, as I vaguely know him from RobotWars) made a fax machine by snchronising their Colchester Student Lathes
[20:47:58] <andypugh> I am not the only one then.
[20:48:15] <pjm> hah that one has to be the best of the series i reckon
[20:48:21] <andypugh> I liked the way they just casually assumed that everyone had a lathe to try it with :-)
[20:48:32] <pjm> just think, if they had VFD's they could have PLL'd the drum speeds together
[20:48:39] <pjm> andypugh dont they then??
[20:48:42] <MattyMatt> that's pretty much how the original fax machines worked. drum scanners are still used for top quality scanning
[20:49:09] <pjm> i was talking to a mate the other day, and in fact we were wondering how most people get on without machine tools
[20:49:20] <alex_joni> http://www.youtube.com/view_play_list?p=860C057349A252D8
[20:49:23] <andypugh> You could just fit the lathes with synchronous motors and let the mains handle the phase lock
[20:49:37] <andypugh> Yeah, how else do you peel carrots?
[20:49:49] <MattyMatt> manual horizontal hold knob
[20:50:06] <AchiestDragon> fax is like a sstv format ,, a slower speed format that tv but esentaly the same timing structure ,, even the baird flying spot scanner tv system use the same principle
[20:50:33] <andypugh> Ah! Yes, Rex Garrod. He made the Brumm car from the TV series
[20:50:53] <pjm> yeah! that is featured in one of the secret life progs too
[20:51:32] <AchiestDragon> by the same token you could porbablay convert a bitmap to nc and get a simlar result
[20:52:37] <AchiestDragon> theres software to do fax , sstv and others to bitmap already for linux
[20:52:57] <tom3p> alex_joni: the last entry isnt from secret life, but it is an edge finder for a mill.
[20:53:06] <AchiestDragon> bitmap / usual graphics formats
[20:54:13] <AchiestDragon> saw an interesting one couple of days back will find link again
[20:54:49] <AchiestDragon> http://www.usbcnc.com/index_gallerie.html first image top left
[20:55:41] <AchiestDragon> done on acrilic by looks clear but where cut shows as white so looks fine on a dark background
[20:58:41] <AchiestDragon> http://www.usbcnc.com/index_machine2.html the idea of that machine he has made i think looks ok also ,, cant make my mind up if to do something on the lines of that or stick with the desgin i have been playing arround with
[20:59:06] <pjm> andypugh the other excellent 'secret life' is on tape recording where they use scotch tape and rust to record on - genius ;-)
[20:59:26] <AchiestDragon> just for a 2foot by 1foot by 1foot tavel machine it would take a lot of space
[20:59:46] <andypugh> Ah yes. They were brilliant at trying stuff you felt sure couldn't work
[20:59:49] <MattyMatt> will a diamond dust mill point etch glass?
[21:00:14] <AchiestDragon> sould of though so MattyMatt
[21:00:33] <andypugh> Yes and no. You will probably have problems with the fact that the actual point doesn't turn much...
[21:00:44] <MattyMatt> it's omething to try , with the 54 accecories I got with the draper dremel
[21:01:45] <MattyMatt> felt under the glass. that's a trick I got from a site with a swiveling castor wheel arrangement
[21:01:51] <AchiestDragon> yea andypugh thats something that has puzzeled me when i see a 5axis macine with a ball cutter on the end doing final surfacing
[21:03:15] <MattyMatt> I bid on the meddings
[21:03:31] <MattyMatt> not too hopeful tho
[21:05:13] <AchiestDragon> they seem to be only cutting on the axis point did try a bit of pcb with a ball nose , it marked the surface but was not till it went of the edge of the work that it droped moving back it then started to cut properly
[21:05:14] <MattyMatt> and I can't afford to restore it to full glory right now, if it needs more than a rweind
[21:06:00] <MattyMatt> I need a drill press to make lathe. I don't want to need a lathe to fix my drill press :)
[21:06:27] <AchiestDragon> ok so a lot of backlash but the tool presure should of been enough to cause it to mill into the work rather than ride ontop
[21:07:06] <andypugh> I would guess that a new motor might be all that you need.
[21:07:10] <AchiestDragon> and like theres more than 3kg of tool presure as i have a weight on the z axis to enure it
[21:07:24] <andypugh> In fact I have one spare from my mini-lathe.
[21:07:26] <MattyMatt> tool flexure?
[21:07:45] <MattyMatt> work flexure?
[21:07:46] <AchiestDragon> used the weight to counteract the backlash
[21:07:53] <andypugh> (I went 3-phase and VFD so the original single phase motor is redundant)
[21:08:13] <MattyMatt> w00t. how much?
[21:08:40] <AchiestDragon> no all ok tool did not sink into work ,, cutting face at base of ball not good enough to plough down though it
[21:09:23] <tom3p> re: "Axis will not display toolpaths if a file containg a subroutine contains an error. " how is this a bug?
[21:09:45] <andypugh> If you pick it up you can have it, but I only have a motorbike so can't easily deliver.
[21:10:02] <AchiestDragon> 3kg i balanced a 3kg vice ontop of the moving setction of the z axis
[21:10:52] <MattyMatt> I can use momma's car on the rare occasions she doesn't :) that means effectively thursdays or when she's in scotland like now
[21:11:00] <AchiestDragon> removes the z backlash anyway but seems that tool is the problem
[21:11:05] <MattyMatt> till tuesday, when I know if I have a meddings
[21:11:27] <andypugh> Where are you? (I am in Basildon)
[21:11:59] <andypugh> You wouldn't have to be very far away for the petrol to cost more than a motor
[21:12:05] <MattyMatt> wrexham
[21:12:26] <MattyMatt> oops not me, that was for a search. I thought it was you in Wrexham
[21:12:30] <AchiestDragon> not far from me
[21:13:01] <MattyMatt> I'm in Wallasey. yeah basildon is too much petrol, but thanks anyway :)
[21:13:49] <AchiestDragon> only problem it is for me is i only get 200miles on a £64 tank full of petrol
[21:15:22] <AchiestDragon> is 15 miles south of wrexham
[21:15:29] <andypugh> I thought you meant that you were in Wrexham, which would have been OK as I am bringing a van to Llanidloes in a couple of weeks for an Enduro race
[21:17:19] <MattyMatt> I could meet you there, if the spectator's car park fee isn't more than the motor's worth :)
[21:17:44] <andypugh> It's an enduro, you would have to pay people to spectate
[21:18:20] <MattyMatt> neat, what day is it on?
[21:18:33] <andypugh> There are hundreds of motors on eBay though, and some probably never even attract a bid
[21:18:45] <MattyMatt> yeah, it's not a worry
[21:18:47] <andypugh> Sunday 25th
[21:19:11] <MattyMatt> I'll try and fix the one that's on it first
[21:20:17] <MattyMatt> a 1000W motor from a hoover would suit it OK if the rpm is right
[21:21:47] <MattyMatt> a trifle underpowered for the full capacity of that drill, but fine for me for now
[21:21:53] <andypugh> I doubt that it would be, vacuum cleaner motors are normally brushed DC motors running on AC
[21:22:40] <MattyMatt> I thought they were all brushless, to avoid dust contamination
[21:22:54] <MattyMatt> to survive it, rather
[21:22:59] <AchiestDragon> depends on the make of vac
[21:24:16] <MattyMatt> ah well I'm sure I can weigh in the copper and get a working one from somewhere
[21:24:18] <andypugh> I haven't taken apart any vacuum cleaner made in the last 30 years, so I might be wrong.
[21:25:03] <andypugh> But I would rather expect it to be of the "very fast and low torque" school of power. which isn't ideal for a drill.
[21:26:49] <MattyMatt> yeah. upright cleaners are ready fitted with a small pully tho, so it's a painless experiment
[21:27:45] <andypugh> There will be a step pulley on the existing motor, you really want a motor with the right shaft size to suit that.
[21:27:59] <MattyMatt> yep
[21:28:20] <MattyMatt> is Meddings american?
[21:28:33] <MattyMatt> it has that look
[21:28:40] <andypugh> I don't think so
[21:29:13] <MattyMatt> "Meddings Machine Tools are one of an elite few machine tool builders remaining in the UK. "
[21:30:03] <MattyMatt> smart
[21:48:41] <MattyMatt> I've got this linear drive, and there's a couple of 500mm supported bars I'm eyeing. I think I might be building the mk2 before the mk1 here
[21:49:12] <andypugh> It is so often the way.
[21:49:16] <MattyMatt> 500mm isn't long enough for my Y tho
[21:50:18] <MattyMatt> unless I switch to moving gantry
[21:51:17] <MattyMatt> which would let me use the Y motion for the head changer
[21:52:04] <andypugh> You are going for an auto tool changer then?
[21:52:28] <MattyMatt> not immediately :)
[21:52:43] <andypugh> I was thinking it seems a little over-zealous
[21:53:49] <MattyMatt> I will design my dremel holder with a flat base plate tho, so I can use it off the cnc
[21:54:38] <andypugh> Is it worth using an industrial die-grinder instead of a dremel?
[21:55:38] <MattyMatt> dunno yet. I'll destruction test this 19.99 dremalike and report empirical evidence
[21:56:02] <andypugh> There are air die grinders on ebay for half that
[21:56:06] <MattyMatt> I'll carve enough wood to sell to buy a new one first
[21:56:31] <MattyMatt> no compressor yet, although my friend has one for his garage
[21:57:32] <AchiestDragon> the last "dremel" like tool i got lasted 1 hour before both the bearings and motor and speed control failed , nothing like a test substitiute for an actual dremel
[21:57:39] <andypugh> http://cgi.ebay.co.uk/ELECTRIC-DIE-GRINDER-WITH-LONG-SHAFT-Debur-Sand-Weld_W0QQitemZ130302842359QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item1e56a795f7&_trksid=p3286.m63.l1177#ht_1202wt_754
[21:57:54] <andypugh> Looks like it might do better, but not at 240V
[21:59:24] <MattyMatt> yeah it's a never ending upgrade cycle :)
[22:00:10] <andypugh> You might as well see how the Ur-dremel does first.
[22:00:49] <AchiestDragon> the actual dremel i have now got has the standard dremel threaded mounting point the thread is about 18mm dia quite course not a standard size so its a pain to find a mounting clamp ,,, the proxon units with the 20mm mounting look a better option and look upto the task
[22:01:14] <MattyMatt> 54 tools to break. 53 I did the first cutting disk attaching it
[22:01:21] <andypugh> eBay sellers seem to be pushing Kress versions
[22:01:34] <AchiestDragon> if i had known about the dremel mounting i would of gone for the more expencive proxon unit
[22:01:41] <MattyMatt> mine has the same, with a nut. I was going to drill a 19mm hole in 3mm steel
[22:01:43] <AchiestDragon> may still get one later
[22:02:47] <andypugh> The Kress one is 1050W! In fact it looks like an angle grinder without the angle
[22:03:15] <MattyMatt> when I've got 4 heads, I can have one of each >:)
[22:03:55] <MattyMatt> I will use an actual angle grinder in one, for a laugh
[22:04:40] <AchiestDragon> well i got a 20mm mounting clamp that will take the proxon ,, i have some bits of 20mm pipe cut a bit to lenth and a slit down one edge , fits the mounting , if i filed off the thread arround the dremmel it would just fit in the pipe and the clamp would compress the pipe to fit ,, would do the job other than it would mean i could not then use the dremel with other accessoires
[22:04:42] <MattyMatt> that's why it's 4 heads, to leave room in the limits for a 4" wheel
[22:04:50] <AchiestDragon> so decided not to go that route
[22:05:56] <andypugh> http://www.cuttingedgecnc.com/dremelmount.htm might be interesting.
[22:06:43] <AchiestDragon> to be honest i have had a few of the dremel like tools and not realy liked there abilaties ,, but the actual dremel i got last month am most impressed with
[22:07:54] <AchiestDragon> using one of the little cutting disks i have managed to cut a pice of 20mm dia 2mm thick steel pipe with with not a lot of porblems and without breaking the disk in the process
[22:08:00] <AchiestDragon> by hand i have to add
[22:09:24] <andypugh> I would say that was a job for a hacksaw.
[22:09:30] <AchiestDragon> although i did break a few cutting a aluminium sat jack motor casting in half with it to make the mounting i am using atm
[22:09:57] <AchiestDragon> could not manage to use a hacksaw with having a broken finger
[22:10:38] <AchiestDragon> should not of realy been doing it anyway i guess the doc would of frowned
[22:13:44] <AchiestDragon> ie doc" have you been resting it " A:" yes only managed to build a cnc machine prototype "
[22:13:54] <MattyMatt> ah, if you don't use them, you lose them
[22:14:21] <MattyMatt> resting it just means "don't break it again"
[22:14:34] <AchiestDragon> not that i bothered going to hosp in first place when i did break it
[22:15:10] <AchiestDragon> still not 100% but another week or so and it will be fine
[22:15:28] <MattyMatt> 8 weeks for bones to knit, iirc
[22:16:46] <AchiestDragon> as for the cnc well not the best construction i have done , will be getting dismantled at some point leaving me with bits back in my now almost empty pile of scap
[22:18:52] <AchiestDragon> like i usualy say ,, sometimes i build things and they work sometimes i build things and they end up a pile of scrap , although usualy it starts life as a pile of scrap in the first place so nothing lost and at the end of the day i had somethign to keep me going
[22:23:38] <AchiestDragon> i keep looking at the 6foot 19" rack in the halway here , theres 4 nice aluminium extrusions arround the edge and a cast aluminium top and bottom to it ,, keep thinking the rails would make nice frame bits and the casings a good base for a cnc , just cant bring my self to dismantle it and cut it up to make a cnc from
[22:24:35] <AchiestDragon> but it looks like i could make a nice strong frame from it
[22:24:44] <oPless> all you need is a set of z-axis runners :)
[22:27:20] <MattyMatt> build it into a 12U slot, so you can still use the rest of the rack
[22:27:46] <AchiestDragon> yea ,, the rack only cost me £40 and a hefty £75 shipping flebay ,,, its now redundant to me as servers i find are too noisy , have thought of cutting it to half hight but its still full of kit that may get used
[22:28:12] <AchiestDragon> idea of having a cnc is so i can make frontpannels for kit to fill it with
[22:29:05] <AchiestDragon> theres still 4 5u hight servers in it with raid drive arrays
[22:30:49] <AchiestDragon> the servers are redundant and only 12 4.8gb scsi drives still in , was going to use the server cases for radio kit and make new front and backpanels for them
[22:31:26] <MattyMatt> a fridge would be quieter for cooling, although that many drives will always be noisy
[22:33:16] <AchiestDragon> it holds 64 drives total ,, but gave up with raid as its cheaper noe to get 1 1tb sata for the cost of getting a couple of matching 4.8gb ultrascsi drives
[22:33:24] <andypugh> I have been working on that sort of thing today. It ocurred to me that one of those Python Gcode composers for panels would be good, with library profiles for BNC, D-subs (front and back mount) etc which neatens up the grid would be good.
[22:33:26] <AchiestDragon> noe / now
[22:34:15] <andypugh> But for my application a 1-off 9 pin and 15 pin and a bit of code to make an (arbitrary) square array of BNC sockets was all I needed
[22:35:10] <AchiestDragon> what i have started to do in autocad is to generate parts libs you can then just add parts as a block where you want them on the drawing and that bits done ,
[22:35:24] <andypugh> I am using a server for my CNC controller. Nice and easy to rackmount. I think it is unusual in having a parallel port though.
[22:35:47] <MattyMatt> I'm doing that in blender
[22:36:53] <AchiestDragon> working with solids you generate the cutout profile as a solid and save that as a file , add it as a block and subtract it from the solid pannel it gives you the cutout in a form that you can then export the sloid to mastercam
[22:37:18] <andypugh> I was just coding in pure "G" :-)
[22:38:02] <AchiestDragon> blender is a porgram that you hate and love at the same time ,,, the gui is a pain , but its main issue is scail ,, blender is a relative coordinate system
[22:38:26] <MattyMatt> just remember what unit you're working in
[22:38:49] <MattyMatt> it has made me go all-metric :)
[22:39:33] <AchiestDragon> the trick is i found with blender is to first draw the flat mesh in cad then inport it to blender to edit , when you have it edited export again and it normaly holds its scale
[22:39:57] <MattyMatt> if my simulator works well, you can feed it G-code, and it'll return the 3d object that creates
[22:40:07] <AchiestDragon> or draw a referance grid in cad and use that as like a ruler in blendr
[22:42:24] <andypugh> Blender looks interesting, but as I have a Maya install on this computer that I have barely looked at, I probably won't bother installing it any time soon.
[22:42:46] <MattyMatt> steep learning curve with blender
[22:43:03] <andypugh> (Maya is a free download for non commercial use from Autodesk, by the way)
[22:43:12] <MattyMatt> it's a couple of weeks solid work before you have the basic mesh editing skills
[22:43:52] <andypugh> Yeah, I tend to use Autodesk Inventor. 2 years as a Cad-jockey with that makes me unwilling to switch to anything else
[22:44:08] <AchiestDragon> hmm http://www.whipy.demon.co.uk/teedee2.pdf thinks this robot is going to take some time to cad all parts ,, just outlining aprox ram positions for it
[22:44:38] <andypugh> What is it?
[22:44:58] <andypugh> Are you planning to build some sort of Millipod?
[22:45:04] <AchiestDragon> a spider type robot ,,, 24 rams 3 per leg
[22:46:53] <AchiestDragon> last attempt was a bit of a desaster dew to the speed of the electric jacks giving it about a 0.5mph top speed and a 1kw power consumption will find pic of that
[22:47:54] <AchiestDragon> http://fc09.deviantart.com/fs21/f/2007/290/f/e/tinybug_in_the_garden_oct2007_by_AchiestDragon.jpg only 6 legs on last one
[22:48:18] <oPless> oooh hexapod :)
[22:48:32] <AchiestDragon> 10 foot foot to foot 120kg , with a 20kg payload
[22:49:00] <MattyMatt> I would double up the thigh pistons and use the diffential to move legs forwards, backwards
[22:49:02] <andypugh> Which CAD package are you using?
[22:49:12] <AchiestDragon> autocad 2005
[22:49:32] <andypugh> Can you get Inventor? It has mechanical simulation built in.
[22:49:54] <andypugh> And despite being by the same company, is far more pleasant than AutoCAD
[22:50:26] <AchiestDragon> happy with autocad ,, although do have solidworks someware but the location of the disk illudes me atm
[22:50:48] <MattyMatt> that big and no saddle? :)
[22:51:05] <andypugh> With respect, if you are happy with AutoCAD it can only be because you don't know any better. It's horrid.
[22:52:11] <MattyMatt> it's not a vehicle, officer, it's my mount
[22:52:25] <AchiestDragon> 20kg is not enough paylad capacaty to carry someone , i did it as a chalange to see if i could get it to work to the specs they gave out for the euromars rover
[22:53:07] <MattyMatt> ah yeah. legs is better for Mars than wheels
[22:54:07] <AchiestDragon> they use wheels on all the designs i have seen ,, but that robot would of cleard the headge and been able to walk over the car witout having to go around like a wheeld or tracked desgin would have to
[22:54:12] <MattyMatt> they should compromise. legs with roller skates
[22:54:23] <AchiestDragon> yea
[22:54:35] <andypugh> Though I might have been put off excessively by Mechanical Desktop. I spent days cursing that. You would draw a profile to extrude, and it would keep telling you that a node or a line was inadequately contrained, but give no clue which node or line. So you would have to look at it like a logic puzzle to figure it out. Then if you were wrong you would get a "constaint is incompatible with anothe constraint" Argh!
[22:55:58] <AchiestDragon> i did 3d design using autocad at collage many years back now (acad v10) its second nature to use acad to me like some still write in g code direct rather than cad to g code converion
[22:57:00] <MattyMatt> it's what you're used to. I use blender because Im doing games in it too
[22:57:20] <AchiestDragon> took about 2 hours to cad the parts for that ram in autocad
[22:57:57] <andypugh> I would reckon on 20 mins in Inventor.
[22:58:52] <AchiestDragon> did play with blender couple of years back and yea it is good ,, but once you have not used it for a couple of months you forget all the commands and its like starting again leaning it
[22:58:59] <andypugh> Inventor was Autodesk's attempt to streamline AutoCAD, throwing away Autolisp, and having only one way to do things rather than menu, command line and command button for almost everything.
[23:00:36] <MattyMatt> http://cgi.ebay.co.uk/MIDORI-LP-300F-10K-LINEAR-26-POSISTION-SENSOR-CNC-qty2_W0QQitemZ180417303557 good price in US if you're doing pnumatics
[23:00:50] <AchiestDragon> i still do a lot in comand line in acad ,, will be mid december when i am over your end of the world next , need to pick up my graphics tablet from a mate there, that speeds up acad use by about 60% over a mouse
[23:01:50] <MattyMatt> doesn't help with blender, afaict
[23:02:17] <MattyMatt> now one of those 6dof controllers would, but not for engineering stuff
[23:02:22] <Himanshu12> hey anyone have experience Pico Systems Brushless PWM Servo Amp?
[23:04:31] <AchiestDragon> on tinybug ( the 10foot robot ) i used some old 10k mil spec rotary pots ,, they have solid steel shafts and brass bushes for the pivot joints /position feedback in one go (old valve kit ones ) cheap and take the load , the new pots you get use cheap cast "fudgemetal" shafts at best and steel bushes
[23:06:16] <AchiestDragon> 2 per pivot joint so you get the loading right and 2 pots in parrallel removes a fair bit of "noise" / inacuarcy in the read value
[23:07:33] <MattyMatt> nice. I expect you could graft new tracks onto steel shafts if you needed to repeat that
[23:07:35] <AchiestDragon> got a bulk load of them for the job (new old stock 20years old all packed and seald in mil spec bags )
[23:07:51] <andypugh> Himanshu12: At least one of the chaps on the mailing list (Jon Elson) is from Pico systems. I can't remember if I have seen him here.
[23:07:51] <Himanshu12> hey anyone have experience WITH Pico Systems Brushless PWM Servo Amp?
[23:08:52] <Himanshu12> andypugh: thanks.... i just want to know how much power it can deliver . it is rated 120V 20AMP.. so whts in kilowatt.. i want to convert shiffli EMBROIDERY machine
[23:09:03] <Himanshu12> i need 1.5kw to 2kw power ;)
[23:09:18] <MattyMatt> W = V * A
[23:09:46] <andypugh> 120V 20Amp is, simplistically, 2.4kW (But I just multiplied the numbers together)
[23:09:46] <AchiestDragon> the down side with the spider type is theres no real way to make it self contained ,, the compressor is giong to have to be seperate meaning a ambilical air supply to it
[23:09:59] <Himanshu12> thanks MattyMatt.. that i know.. but in practicle it gives that much power???
[23:10:58] <MattyMatt> you'll have to ask Pico how long you can draw 20A before it melts
[23:11:01] <AchiestDragon> power in W or kW is relative ,,, you cant get more kw out than in end of story
[23:11:01] <andypugh> Simplest answer is to email Jon direct: http://www.pico-systems.com/motion.html
[23:11:09] <Himanshu12> coz its such a small .. thats why i am afraid.. btw i am software guy so please forgive if i ask silly questions abt electronics
[23:11:13] <andypugh> Sorry, mis-paste
[23:11:21] <andypugh> elson@pico-systems.com
[23:12:47] <MattyMatt> we are all software guys here tonight it seems :)
[23:12:50] <AchiestDragon> yea 120v * 20A is the right formular = 2400A 2.4K as long as you need less than 2.4Kw that would do
[23:13:20] <andypugh> And I bet your machine is running at far less than 2kW most of the time
[23:13:21] <Himanshu12> i just need 1.5kw maximum
[23:13:59] <andypugh> There should be plenty of overhead then
[23:14:16] <MattyMatt> does it have a fan already?
[23:14:18] <AchiestDragon> idealy you should have the driver and psu rated at double what its normal load is
[23:15:02] <AchiestDragon> so with a 2.5kw controler and power source you sould look at 50 to 75% of that
[23:15:13] <Himanshu12> Right now i am using TAIWANIS AC Servo Drive RATED OUTPUT 9.0A 1.5kw
[23:15:25] <AchiestDragon> theres another rule of thumb you should follow
[23:15:42] <Himanshu12> Right now i am using TAIWANIS AC Servo Drive RATED OUTPUT 9.0A 1.5kw 3PH 0-200V 0-333hz
[23:16:08] <MattyMatt> is it melting?
[23:16:30] <Himanshu12> ha ha ha.. its a huge box with fan .. and very very thick heat sink
[23:16:44] <Himanshu12> i think heat sink covers more than 70% of the drive
[23:16:46] <AchiestDragon> it depends on on time % of the machine in a 24hour period ,, if the machine is running 24hours a day 7 days a week none stop then the psu and drivers should be rated at twice the load or more
[23:17:08] <Himanshu12> yes its 24/7 operation
[23:17:09] <AchiestDragon> if on office hours ie 9 to 5 then 75% more
[23:17:42] <Himanshu12> and wht is the difference between BLDC and AC Servo in terms of power and accuracy ??
[23:17:46] <Himanshu12> which one is good??
[23:17:48] <AchiestDragon> then you should look at making the power supply and controls rated at double the max normal working load
[23:18:17] <AchiestDragon> ie 1.5kw motor then 3kw controls and power ,
[23:18:41] <MattyMatt> with servos, it's the feedback mechanism that provides the accuracy, the positional sensor
[23:18:45] <Himanshu12> let me check current motor ratings..
[23:19:22] <AchiestDragon> otherwise that leads to a higher fail rate in use
[23:19:46] <Himanshu12> its almost same 1.5kw, 7.16NM 2000 RPM 9AMP 200V
[23:20:17] <Himanshu12> wht abt the power???i mean BLDC and AC Servo??
[23:22:17] <Himanshu12> any other drive suggesttion?? i mean if you have exp and is working fine....
[23:22:56] <AchiestDragon> regardles of the controls the motor will only draw the amount it needs ,, its easy for the motor to draw more than its rated power though if you dont account for that each time that happens then if your controls dont allow for it then you end up with dead controls
[23:23:40] <andypugh> 9A 200V it says. The Pico systems drive is rated to 125V
[23:23:52] <AchiestDragon> dead controls = down time and broken machine that costs $$$ that would of been worth spending the extra on higher rated ones in the first place
[23:24:03] <Himanshu12> but this motor is AC Servo and PICO Drive is BLDC i think
[23:24:11] <Himanshu12> so i need to change motors
[23:24:45] <andypugh> Cheaper to find a more suitable drive I would have said. But email Pico with the motor spec and see what they say.
[23:28:09] <AchiestDragon> tv transmitters for example one i had here used a 26v psu , the rating on the side was 2A ,, but you could pull 4a from it before its shutdown kicked in , if you disabled the shutdown it you could pull 10A form it before it started to show stress short term ,, well over rated but thats what is needed for 24/7 use
[23:29:36] <Himanshu12> hmmm...got it.. do i need hall sensor if i have encoder attached??
[23:29:39] <AchiestDragon> also go for linear psu's not switchmodes
[23:30:33] <andypugh> Himanshu: I don't think so, but I am far from an expert.
[23:31:38] <Himanshu12> see i thought to use Pluto Servo board with pico systems PWM amplifier
[23:32:06] <AchiestDragon> Himanshu12: email and ask ,, if you give them the encoder and motor spec then they will be able to tell you if there controls are suitable or not
[23:32:09] <andypugh> What is wrong with your current drive system?
[23:33:09] <Himanshu12> blown off due to power
[23:33:22] <Himanshu12> now i want to reconfigure it with EMC
[23:34:35] <andypugh> It might still be easier to replace the damaged parts with identical ones, you should still be able to interface with EMC
[23:35:03] <Himanshu12> thanks AchiestDragon: question is Pico Requires 3 commutation signals (commonly known as "Hall" signals) from motor or encoder. and Pluto Servo also requires Encoder signal
[23:35:42] <Himanshu12> System was made around 1995 works on ISA based card named Tech 80
[23:36:24] <andypugh> Ah, ignore that option then
[23:36:49] <Himanshu12> its my friend's machine and for me its something like warm up exercise before creating my own machine...
[23:36:58] <Himanshu12> ;)
[23:37:33] <Himanshu12> can i use same encoder for Pluto Servo and PICO DRIVE???
[23:38:09] <Himanshu12> something like switching or multiplexing or just simple connections...
[23:38:27] <andypugh> They shouldn't both need an encoder. If they both take an encoder input then I think they are not a matched pair
[23:38:29] <AchiestDragon> without reading the spec for the servo and driver i would not be able to say eather way
[23:39:17] <Himanshu12> not much details available but i read somewhere that a fellow using both together.. thats why i decided to try on it...
[23:40:31] <AchiestDragon> beter then to find where you read that and see if you can get more info from there , like what if any problems he had connecting them up
[23:41:10] <andypugh> Ah, looking at it, the Pluto-P has an encoder interface, and so has the Pico board. You could use either, but not both.
[23:41:23] <AchiestDragon> if you can ge an email for him then he may be able to give you extra info about it
[23:42:09] <Himanshu12> i emailed him.... i found it on EMC mailing list but not able to find it again thats why come here
[23:43:00] <AchiestDragon> other option you would have is to find a controler with the power you need for the motor and then look for an encoder to match that controler
[23:44:51] <andypugh> After looking around his website I feel confident that Jon Elson at Pico will be able to answer your questions definitively, whereas we are only guessing.
[23:44:52] <Himanshu12> Pluto Servo is final.. now i am in search of Amplifier only.. and this is a trial base so budget is limited
[23:46:17] <andypugh> The Pluto board looks feasible, but also look at the Mesa stuff.
[23:46:24] <Himanshu12> yup you are right andypugh... but i think its knowlegde sharing atleast benificial for me (ha ha ha)... anyways anyother suggessions abt amplifier compatible with Pluto servo
[23:48:15] <Himanshu12> annd whts the technical name for it so i can do googgling.. liek HBRidge PWM amplifier or wht??
[23:48:26] <andypugh> http://www.mesanet.com/
[23:48:40] <andypugh> The 8I20 is a low cost 2200W 400V three phase torque mode amplifier for synchronous permanent magnet servo motors (Brushless AC servo) up to approximately 3HP. The 8I20 will supply peak currents of 30A.
[23:48:41] <Himanshu12> i mean the technical name of amplifier compatible with Pluto Servo
[23:50:31] <andypugh> I don't know. I suggest you subscribe to the EMC mailing list with your question, where the Pico guy and the Mesa guy can fight for your custom.
[23:51:19] <SWPadnos> the pluto servo outputs PWM
[23:51:43] <SWPadnos> so you can filter that to get 0-10V (or -/-10 with additional circuitry)
[23:51:51] <SWPadnos> or you can directly drive an H-bridge
[23:52:00] <SWPadnos> or you can do something else with it
[23:52:46] <SWPadnos> since it's the EMC2 software in the PC that actually does the PID control and not the pluto board, you can use almost anything as long as it's possible to tune it
[23:53:43] <Himanshu12> hmmm i think Mesa 8I20 is under development
[23:53:53] <Himanshu12> no documents available... no pricing info
[23:55:31] <Himanshu12> hey i thought its the FPGA who does PID control
[23:55:36] <SWPadnos> nope
[23:55:43] <andypugh> Do you need an interface at all? EMC can do PWM, encoder and bridge driving until you run out of pins
[23:56:07] <SWPadnos> the Pluto or other FPGA-based solutions (like Pico and Mesa) do a much better job of it
[23:56:11] <Himanshu12> so performance will be same with Para port or Pluto Servo??
[23:56:12] <andypugh> How many axes do you need to drive?
[23:56:24] <Himanshu12> 4 axis
[23:56:51] <Himanshu12> i am not running out of pins. i have 2 para ports
[23:57:04] <andypugh> You won't have enough pins (or speed) to do 4 axes with the parallel port, you will need an FPGA product.
[23:57:16] <andypugh> It might work then.
[23:57:50] <Himanshu12> can you explain me in short if possible... i mean who will generate PWM ? para port or FPGA?? i think FPGA
[23:57:56] <andypugh> You need 2 outputs and 3 inputs per motor
[23:58:20] <SWPadnos> you can do PWM with either, but the FPGA will be much more "smooth", since it has no latency issues
[23:58:21] <andypugh> FPGA is a better choice. You can do it in software though.
[23:58:28] <SWPadnos> the timebase for the FPGA will be much faster
[23:59:17] <Himanshu12> i have latency time around 13000 and i need max 100khz so i choose pluto servo.. is it fine??
[23:59:18] <SWPadnos> software PWM isn't really suitable for servo axes, unless you want them to be slow or inaccurate, or you have very very good latency measurements from your PC
[23:59:42] <SWPadnos> your PC will not do 100 kHz, and the pluto may not either
[23:59:54] <SWPadnos> if you're talking about a 100 kHz PWM cycle (vs. a 100kHz time slice)