#linuxcnc-devel | Logs for 2015-06-05

[07:02:07] <skunkworks> pcw_home, so - instead of using a dac - you take pwm and convert it to +/-10v. What sort of resolution would that be?
[07:02:36] <skunkworks> compared to say a 12bit dac
[07:17:17] <Tom_itx> pcw_home if you get a chance could you check the 7i90 schematics.. i've got a bad pin list to try and narrow down the bad buss switch chip
[07:17:49] <Tom_itx> the bad IO are: 0, 1, 2, 3, 17, 18, 19, 20, 21, 37, 44, 46, 47
[07:18:35] <Tom_itx> i suspect those are all mapped to one chip.
[09:36:46] <JT-Shop> seb_kuzminsky, I still think the Getting Started Guide should be separate PDF and limited to downloading and installing LinuxCNC. After LinuxCNC is installed the rest of the manual is on the machine.
[09:37:36] <JT-Shop> this way a potential user/builder only has a small PDF to download and in the case of Gene print on paper :)
[09:45:03] <pcw_home> skunkworks: our standard PWM is 12 bit plus sign so 13 bits (7I77/7I83 DSPIC based PWM DACs are 14 bits)
[09:50:31] <ssi> pcw_home: I put in an order yesterday and made a note on it that it wouldn't let me order with the 7i42TA in the cart since it was out of stock
[09:50:43] <ssi> hopefully someone will see my note and it'll all show up together :D
[09:52:08] <pcw_home> yeah I saw that, will add to order (but note this will go out next week sometime, still have a fair backlog, but finally catching up)
[09:52:41] <ssi> that's fine, I'm in no rush
[09:52:53] <pcw_home> Tom_itx: I will PDFisze the schematics sometime today and send you a copy
[10:15:14] <seb_kuzminsky> JT-Shop: i like that
[12:19:22] <KGB-linuxcnc> 03John Thornton 052.7_Docs bd5a972 06linuxcnc 10docs/src/Master_Documentation.txt 10docs/src/Master_Getting_Started.txt Docs: move setup chapters to manual * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=bd5a972
[12:28:01] <skunkworks> pcw_home, http://www.practicalmachinist.com/vb/dmg-mori-gildemeister-maho-cnc/fix-your-maho-thread-270525/index16.html#post2563250
[12:28:35] <skunkworks> pcw_home, glad you are seeing the light at the end of the tunnel
[12:51:13] <Tom_itx> pcw_home thanks.
[12:51:54] <Tom_itx> i can follow the header pinouts back to the chips
[12:52:07] <Tom_itx> i could ohm it out but it would take longer
[12:54:34] <cradek> skunkworks: I bet you don't need very many bits for good velocity control. it's not like if you command exactly the right analog value for the move it'll track perfectly...
[13:49:14] <skunkworks> cradek, you kinda proved that with your little lathe running from the printer port
[13:49:24] <skunkworks> 8 levels or something like that..
[13:49:44] <cradek> yeah I bet it had no more than 4 bits, heh
[13:49:48] <cradek> it didn't work very well
[13:50:57] <cradek> even just 10 bits would give you .02v steps
[17:02:47] <alex4nder> hmm
[17:03:23] <alex4nder> anyone have any experiencing with having variable max velocities based on current speed?
[17:04:42] <alex4nder> er, max accel
[17:07:56] <seb_kuzminsky> not i
[18:09:04] <cradek> we sure don't have that. why do you want that?
[18:13:02] <alex4nder> cradek: because I'm trying to optimize rapids with my stepper motor setup
[18:14:00] <alex4nder> I don't need it, but I'm curious about how quick this machine could be in the best case
[20:30:19] <skunkworks> alex4nder: the closed thing I have done is switch from half stepping to full stepping at a certain velocity. (full stepping works better for higher velocities.)
[20:30:26] <skunkworks> *closest
[20:33:42] <alex4nder> skunkworks: ah, yah, that also makes sense