#emc | Logs for 2008-06-21

[05:02:01] <chr0n1c> yo!
[05:02:34] <chr0n1c> i just mille dout my first real circuit board ever and it looks nice! thanks emc2 ;)
[05:02:58] <chr0n1c> i'll post pictures and a nice vid to make up for the crazy test run video soon...
[05:03:17] <chr0n1c> ... jsut needs editing
[06:41:28] <AzRockCrawler_> GE pholks
[06:43:05] <SkullWorks_PGAB> So is AZ aligned with mountain time or Pacific currently?
[06:44:46] <AzRockCrawler_> hrm...spring ahead fall back...I dunno...GMT-7 here :D
[06:45:05] <AzRockCrawler_> I guess CA and AZ are in synch
[06:49:44] <AzRockCrawler_> AS is always MST. CA is PDT is MST. It is all too confusing for those of use with single neuron processing.
[06:49:49] <AzRockCrawler_> AS == AZ
[06:51:11] <AzRockCrawler_> Anyone here familiar with using HAL scope? Trying to read an index signal on pin 15 and get nada.
[06:51:43] <AzRockCrawler_> Shows up on the 'silly scope before it hits the parallel port so I know it is heading the right direction at least, just can't see it from HAL scope..
[07:09:34] <fenn> can you read any other signals?
[07:10:29] <AzRockCrawler_> I can see e-stop (pin 10 for me) for instance
[07:11:29] <AzRockCrawler_> I can watch the CP signal on pin 17 too
[07:11:39] <fenn> is it easy to switch the wires from 15 to 10
[07:11:44] <fenn> just to test
[07:11:46] <AzRockCrawler_> gotta be something dumb I am doing in the hardware
[07:12:00] <AzRockCrawler_> hrm...no, sadly.
[07:12:17] <fenn> i'm thinking it might be too fast for the thread you're running
[07:12:24] <AzRockCrawler_> inflexible commmercial BOB that I am gettin gclose to regretting. Leave it at that. Very much MACH3 oriented.
[07:12:34] <AzRockCrawler_> yes, maybe too fast -- occurred to me too
[07:12:59] <AzRockCrawler_> I measure ~50 kHz at the generator.
[07:13:18] <fenn> er.. 50kHz what?
[07:14:05] <AzRockCrawler_> sorry, 50 kHz worth of pulses from my index test thingy. That is being presented to the input on pin 15.
[07:14:20] <AzRockCrawler_> However...
[07:14:57] <fenn> if the index pulse is 20 microsec long your scope thread period has to be at most 10us, which seems unlikely
[07:15:27] <AzRockCrawler_> I just recalled something. I did check the speed thing by manually creating pulses (optical interrupter) and did not see it show up in HAL scope so must be somethng else.
[07:17:08] <AzRockCrawler_> HAL scope thread samples at ~24 usec (single channel) according to the dialog that pops up, so yes, could eb that.
[07:18:07] <fenn> you're sure it's connected to pin 15?
[07:20:09] <AzRockCrawler_> That occurred to me too. It *should* be on pin 15 according to documentation and labelled connector block. As a cynic, I did test other inputs that were not in use. No sigs there either. :)
[07:21:22] <fenn> if it were me, i'd just hook poke some wires directly to the parport connector
[07:21:35] <AzRockCrawler_> yes, that sounds like me too.
[07:21:39] <fenn> you know, +5v through a 1k resistor
[07:21:45] <AzRockCrawler_> will try that
[07:21:48] <AzRockCrawler_> yeah
[07:22:11] <fenn> if that doesnt work, then your parport is no good
[07:22:26] <AzRockCrawler_> hrm
[07:22:29] <AzRockCrawler_> yes
[07:22:34] <fenn> if it works on 10 and not 15 i mean
[07:22:40] <AzRockCrawler_> right
[07:23:07] <AzRockCrawler_> I know 10 works since I have e-stop there
[07:23:19] <AzRockCrawler_> will do more debugging tomorrow.
[07:23:27] <AzRockCrawler_> Thanks for ideas.
[07:25:07] <AzRockCrawler_> I wondered if I need some sort of signal conditioning. The pulse index thingy I made uses an H22A1 device, a transistor and accompanying resistors.
[07:26:13] <AzRockCrawler_> I am not electronics saavy, just know enough to make things work sort of 90% of the time. The rest of the time i sthe normal state of cluelessness.
[07:26:56] <fenn> well its mostly just standard empirical testing and debugging, making sure you know what you think you know
[07:27:03] <AzRockCrawler_> agreed
[07:27:19] <AzRockCrawler_> and understood
[07:28:27] <AzRockCrawler_> I will test the port inputs tomorrow and work my way back toward the indexer. The solution will present itself along the way.
[07:28:59] <fenn> btw your h22a1 sensors has a turn on time of 8us, it might not be too happy at 50kHz
[07:29:11] <fenn> and turn off of 50us, that wont work
[07:29:35] <fenn> i dunno what you're doing though
[07:29:54] <AzRockCrawler_> hrm...well, signal on 'scope looked nice and clean...I bet I fubared the calcs.
[07:30:28] <fenn> ok, maybe the datasheet was being conservative
[07:31:24] <AzRockCrawler_> I have a small encoder disc with 40 dark partitions and therefore 40 clear partitions Sorry, not sure of the correct encoder terminology.
[07:31:46] <fenn> typically called slots or lines
[07:32:04] <AzRockCrawler_> It passes through the H22A1 and trips it out, driving PN4248
[07:32:13] <fenn> dont say 'counts' because it can mean edges (4 edges in a cycle)
[07:32:15] <AzRockCrawler_> Okay, slots, yes, thanks.\
[07:32:28] <AzRockCrawler_> right
[07:33:18] <AzRockCrawler_> I thought 40 slots was maybe too fast so I made a test encoder with just 1 slot
[07:33:30] <AzRockCrawler_> then another with just 2 at 180 degrees
[07:33:31] <fenn> is this for a lathe spindle or something?
[07:33:37] <AzRockCrawler_> spindle on a small mill
[07:33:51] <AzRockCrawler_> someday maybe adapt to a small lathe
[07:34:06] <fenn> ok.. are you going to be doing tapping or threading?
[07:34:20] <AzRockCrawler_> I want to try tappin on the mill, threading on the lathe
[07:34:51] <fenn> ok, i think you will need to use a quadrature encoder to do tapping, since you have to back out of the hole
[07:34:58] <AzRockCrawler_> oic
[07:35:09] <AzRockCrawler_> well, i did not tell you the WHOLE story :)
[07:35:26] <AzRockCrawler_> I have a device I made that has two H22A1s
[07:35:32] <fenn> also, you need an index channel, a ring with only one line/slot, in order to use lathe threading right now
[07:35:33] <AzRockCrawler_> and the 40 slot encoder
[07:35:40] <AzRockCrawler_> okay
[07:35:43] <AzRockCrawler_> I have no index channel
[07:35:49] <AzRockCrawler_> makes total sense though
[07:35:56] <AzRockCrawler_> needs to know where "home" is
[07:35:59] <fenn> it doesnt make any sense to me, but meh :)
[07:36:05] <AzRockCrawler_> (not home but some reference)
[07:36:30] <AzRockCrawler_> I may have to rethink this
[07:37:06] <AzRockCrawler_> Oh well, it will be fun to make EMC read RPM anyway, just so I can look in the mirror and say I did it.
[07:37:20] <AzRockCrawler_> I do want to make the mill do tapping though.
[07:40:34] <AzRockCrawler_> is counter.c sufficient for a spindle tach input do you think?
[08:11:31] <fenn> counter is obsolete.. use encoder instead
[08:11:49] <AzRockCrawler_> will look at it - thanks
[08:12:11] <AzRockCrawler_> I stalked your website and became engrossed in the encoder article.
[08:12:16] <fenn> cool
[08:12:25] <fenn> i havent actually built one
[08:12:34] <AzRockCrawler_> The microchip white paper is intriguing.
[08:14:09] <AzRockCrawler_> makes me think I should get a more modern optical switch. I used the H22A2 because it was in my junk bin. The encoder wheels were attached to some old Pittman 9413 motors. I just used what I had on hand.
[08:14:54] <AzRockCrawler_> random observation: I bet this is the first time I have signed onto IRC in ...10 years maybe
[08:14:57] <fenn> you mean use a photodiode array instead?
[08:15:23] <AzRockCrawler_> maybe too fancy for my needs, but that is interesting, yes
[08:15:36] <fenn> it's overkill for CNC
[08:16:37] <AzRockCrawler_> I really just need something that turns off faster than 50 uS
[08:16:53] <AzRockCrawler_> even 8 uS turn on time is kind of slow
[08:17:43] <fenn> yeah, its typical
[08:18:11] <fenn> most opto-isolators are around 3-4us
[08:18:18] <AzRockCrawler_> I see.
[08:18:43] <AzRockCrawler_> Well, that's why I added the second H22A2. Figured I could get twice as many pulses ( iguess maybe 4X if I knew what I was doing)
[08:18:57] <AzRockCrawler_> But, I am not using the seocnd device for this round of testing.
[08:19:28] <fenn> yep should be easy with emc2/HAL
[08:20:06] <AzRockCrawler_> It works though from a hardware standpoint. I can see both traces on the 'scope and they are decent.
[08:20:48] <fenn> 50us @ 40 cycles/rev should get you to about 5000rpm, if my math's right
[08:21:04] <fenn> best way to find out is to look at the waveform though
[08:21:14] <AzRockCrawler_> that looks right
[08:23:24] <fenn> so, what you do with CNC mills?
[08:23:41] <AzRockCrawler_> Mostly break cutting tools. :)
[08:23:53] <fenn> just like the best of 'em
[08:24:12] <AzRockCrawler_> I have a couple of interests.
[08:24:26] <AzRockCrawler_> One is making parts for a rock crawler.
[08:24:34] <fenn> what is a rock crawler?
[08:24:42] <AzRockCrawler_> I used to make motorcycle parts but am not into that any more :(
[08:25:25] <fenn> r/c trucks?
[08:25:34] <fenn> er, radio control
[08:25:38] <AzRockCrawler_> Rock Crawler...4 wheel drive vehicle that has huge tores and goes places that it is difficult to even walk. Not monster truck...
[08:25:58] <AzRockCrawler_> No, pollution creating, ecounfriendly honest to goodness vehicles.
[08:26:31] <fenn> this thing is a piece of work http://www.toughracing.com/zc126d/index.php?main_page=product_info&products_id=496
[08:26:54] <AzRockCrawler_> wow
[08:27:13] <AzRockCrawler_> I see what some people are doing and am utterly humbled.
[08:27:26] <AzRockCrawler_> Especially the R/C 'copter crowd.
[08:47:22] <micges> good moning
[08:47:27] <micges> morning
[09:06:47] <AzRockCrawler_> gn all
[09:06:52] <AzRockCrawler_> thanks for insights fenn
[10:22:54] <fenn> there's only 10 billion pages on this internet, gotta be in here somewhere..
[12:02:01] <pjm> morning all, btw anyone use 'multicam' is some CNC router. I have some user manuals and control boards that are heading for the trash/bin if anyone wants them
[12:19:40] <als> what's the word from EMC fest?
[12:23:17] <Sweeper> pjm: pics of control boards? I'd pay shipping probably....
[12:25:35] <skunkworks> great stuff. Just no internet access at the moment
[12:26:13] <pjm> Sweeper one sec i will take a photo
[12:30:22] <pjm> multicam parts: http://pjm.dyndns.org/multicam.jpg
[12:30:34] <pjm> i did 1/2 consider ebaying them, but i'm probably far too lazy
[12:35:16] <Sweeper> ah..
[12:36:04] <Sweeper> doesn't look like anything I could use as-is, probably not worth it to ship just for component salvage, although the opto-isolators and headers are tempting
[12:43:47] <pjm> bah i'm sick of opto's
[12:43:50] <pjm> i have tubes of them
[12:46:32] <pjm> if anyone needs, i have 50 off TLP3052 which u can have for postage
[12:46:42] <pjm> i'll make a list of the other tubes of optos
[13:35:48] <Roguish> hey all, anyone know if it's possible to change a mother board (same cpu) with ubuntu 8?
[13:36:58] <Roguish> i have an ECS mb that is the craps with latency errors and i'm thinking of changing to an asus mb. using the same cpu.
[13:37:32] <Roguish> i would rather not go thru the complete install if i don't have to.
[13:40:29] <Sweeper> Roguish: if you're using a default kernel you should be ok....all the drivers will be present
[13:41:44] <Roguish> I haven't done anything unusual that i know of. it's strictly an emc2 system.
[20:21:58] <alex_joni> see you in roughly a week guys
[20:22:08] <alex_joni> * alex_joni is off for vacation a week ;)
[21:55:19] <AzRockCrawler> you there fenn?
[22:56:27] <dmess> hi all