#linuxcnc-devel | Logs for 2016-11-13

[13:42:05] <andypugh> pcw_home: Do you think that serial encoder no-wrap should be on by default?
[13:42:21] <andypugh> How common are serial single-turn encoders?
[13:44:38] <pcw_home> I think it should be off by default (as it is now)
[13:45:43] <andypugh> no-wrap off == wrap-on == default?
[13:46:38] <pcw_home> single turn absolute are fairly common (resolvers, Hall based magnetic encoder like AM55XX s, no gear/battery backed SSI/BISS etc)
[13:46:55] <andypugh> OK
[13:47:04] <pcw_home> Yes wrap-on= default as it is now
[13:48:27] <pcw_home> There is a new no gear/no battery multiturn absolute encoder from Avago
[13:49:15] <pcw_home> uses the Wiegand effect to generate power when turned
[13:50:29] <andypugh> You know that i am going to have to look that up on Wikipedia?
[13:50:41] <pcw_home> I had to :-)
[13:53:41] <andypugh> Which linked to the Barkhausen effect, which I have heard of, but had forgotten what it is.
[13:55:07] <andypugh> I was planning on a geared secondary resolver to make my lathe Z-axis absolute. But a non-volatile encoder might be neater.
[13:59:54] <andypugh> Hmm. http://www.mouser.co.uk/ProductDetail/Broadcom-Limited/AS38-H39E-S13S/?qs=0ns%252b%2f2Cp7jGQfrKW74VZZA%3d%3d Not _terribly_ expensive, but I think I will stick with plan A (I want to see if it works)
[14:13:24] <andypugh> pcw_home: Those encoders look like something that anyone should consider for future conversions. Homing is tedious.
[14:13:45] <andypugh> (and LinuxCNC now supports homing to absolute encoders)
[14:15:21] <andypugh> Currently the absencoder module uses “gray” as the unitstring to say that the encoder is gray-coded. Was that something I decided to do, or is that something from SmartSerial (the absolute encoder module basically creates dummy smart-serial devices)
[14:17:38] <pcw_home> Pretty sure that was added for absolute encoders befre is was all folded into the sserial stuff but its been a while
[14:19:33] <andypugh> I am thinking to add a “flags” field instead, because setting the unitstring seems like a kludge
[14:24:04] <andypugh> It arrived with absolute encoders. I can’t see any reason for a smart-serial device to use graycode, so I assume that none of them do?
[14:34:25] <pcw_home> No, none do
[14:35:58] <andypugh> OK, then I will alter how I do it.
[14:53:02] <andypugh> pcw_home: OK, I think that I have it done, but have no hardware to test with. Should I push to a new branch for testing?
[14:54:07] <pcw_home> Yes please do that, then I can test tomorrow (no encoders at home)
[15:24:42] <KGB-linuxcnc> 03andypugh 05andypugh/wrapless_encoder 73e6860 06linuxcnc 10docs/man/man9/hostmot2.9 10src/hal/drivers/mesa-hostmot2/abs_encoder.c 10src/hal/drivers/mesa-hostmot2/sserial.c 10src/hal/drivers/mesa-hostmot2/sserial.h Hostmot2 Absolute Encoders: Add a flag to inhibit encoder wrapping. * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=73e6860