Back
[00:23:06] <Tom_itx> ok i can confirm the bit file is loading. the functions are remapping like the .vhd file shows
[00:23:14] <Tom_itx> with the dmesg listing
[00:23:41] <Tom_itx> i compared the original with the modified and they are different
[02:07:26] <_DJ_> moin
[03:57:37] <Loetmichel> mornin'
[03:58:13] <archivist> is it?
[03:58:37] <Loetmichel> <- MUCH better is ... pain is gone, only a running nose left... ;-)
[03:58:49] <Loetmichel> over here it is... more or less ;-)
[03:59:00] <Loetmichel> [10:41] <Loetmichel> mornin'
[04:00:01] <archivist> I need to find a cd to try the latency test on a couple of pcs
[04:00:33] <archivist> or...find the one I used last time
[04:01:57] <Loetmichel> burn one?
[04:06:24] <archivist> I have to find a blank to do that, or go out and get some
[04:28:43] <archivist> one silly pc wont boot from cd.....
[04:40:16] <Loetmichel> hmmm
[04:40:41] <Loetmichel> cd and dvd blanks is a thing one should have in stock anyways ;-)
[04:41:13] <Loetmichel> and should my stock be empty i go to the desk of my wife and snatch some
[04:41:15] <Loetmichel> :-)
[04:42:32] <GuShH> Loetmichel: and should her stock be empty? THEN WHAT?
[04:43:34] <archivist> this pc is fighting...is is stubbornly refusing to boot from cd
[04:44:19] <Loetmichel> GuShH: she is a woman. Her stock is NEVER empty ;-)
[04:44:34] <Loetmichel> archivist: bad drive?
[04:45:03] <kengu> archivist: i have tried to boot from cds and from usb sticks and then finally just used networkboot that works like a charm
[04:46:37] <Loetmichel> *sneeeze* [grabs a paper tissue] *TOOOOOOT*... hmm, i should bring the paper bin out... the first soft tissues a falling out (full) ;-)
[05:23:00] <archivist> bios boot order was important
http://ubuntuforums.org/showthread.php?t=764497
[08:24:12] <Tom_itx> archivist no usb to boot from?
[08:24:31] <Tom_itx> i guess if it has floppy, probably not
[08:24:44] <archivist> its running linuxcnc now
[08:25:38] <Tom_itx> horray!
[08:25:49] <Tom_itx> my sserial isn't showing up yet
[08:26:21] <Tom_itx> my next step is to make sure the config line is ok
[08:27:05] <archivist> I have also been out in the garage playing with a dividing head
[08:27:20] <Tom_itx> wish i had one or a rotary head
[08:30:01] <archivist> I search broken on fleabay and found 221319084790 bit rusty and a few parts missing but not siezed
[08:30:04] <atom1> CONFIG="firmware=hm2/7i43-4/SVST2_4_7I47SS.BIT num_encoders=2 num_pwmgens=0 num_stepgens=4 sserial_port_0=3x"
[08:30:12] <Tom_itx> that's what i've been using
[08:30:27] <Tom_itx> i get no errors but it doesn't show up that i can tell
[09:25:16] <Tom_itx> where can i find out about sserial modes? the doc says there are 9 modes but doesn't list them. somewhere else in the docs it mentions only 2 modes have been implemented
[09:26:42] <pcw_home> Modes depend entirely on the remote (If you are not sure, select 0 as 0 is always valid)
[09:27:50] <Tom_itx> i posted my config line above and it loads with no errors but i don't see any sserial in dmesg or halcmd show pins
[09:28:59] <pcw_home> usually that means that no device was found
[09:29:25] <Tom_itx> i'm not sure why sserial has =3x' behind it other than your recomendation
[09:30:00] <pcw_home> thats probably invalid
[09:30:09] <Tom_itx> i'm wondering that myself now
[09:30:27] <pcw_home> you need to look at the supported modes in the manual
[09:30:37] <Tom_itx> the device manual?
[09:30:47] <Tom_itx> i was looking at that
[09:30:57] <Tom_itx> 0 hardware 1 software?
[09:31:02] <pcw_home> I think master is better at reporting sserial errors
[09:31:42] <pcw_home> master should report a invalid mode error
[09:31:54] <Tom_itx> how do i run it as master?
[09:32:11] <pcw_home> Linuxcnc master
[09:32:26] <Tom_itx> oh from a terminal
[09:32:47] <pcw_home> no just the latest (master) version of linuxcnc
[09:33:01] <Tom_itx> ok
[09:33:49] <pcw_home> in any case i think the 7I71 only has one mode so 3 is definately wrong
[09:33:59] <Tom_itx> k, i'll try 0
[09:34:35] <Tom_itx> that may have been set for the 7i84
[09:34:40] <pcw_home> (see the software process data modes section of the manual) it may have 2 modes not sure
[09:35:53] <pcw_home> mode cannot really be illustrated in the man page because they can mean entirely different things for different remotes
[09:35:55] <pcw_home> (LinuxCNC master lists the modes and mode names)
[09:36:05] <Tom_itx> yeah i see
[09:36:28] <Tom_itx> do the card pdf's describe it for each card?
[09:36:41] <pcw_home> yes,
[09:37:06] <pcw_home> but of course new modes may be added by firmware updates
[09:37:14] <Tom_itx> yeah the 84 has mode0..2
[09:37:18] <Tom_itx> yep
[09:37:37] <Tom_itx> mode 0 looks pretty safe
[09:37:45] <pcw_home> so it nice the the latest LinuxCNC reports the available modes
[09:38:04] <pcw_home> mode 0 always works
[09:38:17] <Tom_itx> i will want an encoder on the 84 though
[09:38:27] <Tom_itx> i'll start with 0 until it works
[09:38:36] <pcw_home> hardware modes are different (set with option jumpers typically)
[09:38:57] <Tom_itx> looks like on the 7i84 i may need mode2 for encoder
[09:39:52] <Tom_itx> if you run a chain of these off one tx rx pair how do you determine the chain order?
[09:40:02] <pcw_home> since the 7I71 is an digital output only device it doesn't have many options
[09:40:11] <Tom_itx> yeah i can see that
[09:40:49] <pcw_home> currently they are RS-422 only (one device per RX/TX pair)
[09:40:57] <Tom_itx> ok
[09:41:36] <pcw_home> this is for a couple of reasons (mainly that it scales better)
[09:42:11] <pcw_home> (32 channels is not significantly slower than 2)
[09:43:14] <Tom_itx> well at least the bit file seems to be working. the other functions remap the pins like the file indicates...
[09:43:36] <Tom_itx> i posted the before and after dmesg files
[09:44:23] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/cnc/configs/sherline/bitfiles/
[09:44:36] <pcw_home> I do think the pinout reporting is broken a bit (it really should list pins as sserial even if a remote is not found)
[09:44:52] <Tom_itx> it would be helpful
[09:45:12] <pcw_home> for debugging certainly
[09:45:20] <Tom_itx> just shows IOPort right now
[09:45:22] <pcw_home> did you try mode 0?
[09:45:32] <Tom_itx> gettin ready to... multitasking here
[09:46:35] <pcw_home> what version of linuxcnc? (I think sserial is not very good until maybe 2.5.2)
[09:48:29] <Tom_itx> 2.5.3
[09:48:53] <Tom_itx> still shows IOPort on IO4 & IO20 where the TX RX pins are
[09:49:11] <Tom_itx> in dmes
[09:49:12] <Tom_itx> g
[10:00:44] <Tom_itx> i'm fairly sure the card is wired in right
[10:00:44] <Tom_itx> 7i71 RXA -> 7I47 TX0
[10:00:44] <Tom_itx> 7i71 RXB -> 7I47 /TX0
[10:00:44] <Tom_itx> 7i71 TXA -> 7I47 RX0
[10:00:44] <Tom_itx> 7i71 TXB -> 7I47 /RX0
[10:03:47] <Tom_itx> 7i71 has mode0 and mode1 -> ouputs + readback field voltage
[10:12:12] <pcw_home> Do you have a red LED (and no green) illuminated on the 7I71?
[10:15:51] <Tom_itx> the one by the power supply was replaced.. it's blue but illuminated as well as the red
[10:16:29] <pcw_home> Yeah we tested blue LEDS but found them too annoying
[10:16:42] <Tom_itx> me too, i tried some once
[10:17:27] <Tom_itx> i've got field and vin wired together at 24v
[10:17:33] <pcw_home> We use them for the HV on the 8I20 since its meant to catch your attention (could be 400V)
[10:21:33] <pcw_home> what are your cat5 colors vs 7I47 pins?
[10:23:13] <pcw_home> and do you have the 5V cable power supplied to the 7I71
[10:23:14] <Tom_itx> just a sec
[10:23:28] <Tom_itx> 5v thru the cat5e yes
[10:23:34] <Tom_itx> double checked that
[10:24:52] <Tom_itx> blue blue/white are GND
[10:25:07] <Tom_itx> brown brown/white are +5v
[10:25:48] <Tom_itx> green is TXB, green/white is TXA
[10:26:16] <Tom_itx> orange is RXB, orange/white is RXA
[10:26:37] <Tom_itx> wires per previous post ^^
[10:28:25] <pcw_home> Yeah but Im pretty sure RXA is /RX and RXB is RX etc
[10:28:41] <Tom_itx> well i couldn't find that anywhere
[10:28:59] <Tom_itx> doesn't RXA need to go to TX?
[10:29:05] <Tom_itx> or is it RX to RX?
[10:29:30] <Tom_itx> i'm getting the physical pins for you
[10:29:54] <pcw_home> TX to RX and RX to TX
[10:30:33] <pcw_home> 7I71 manual shows physical pins on 7I71
[10:31:22] <pcw_home> probably the polarity is swapped
[10:31:54] <Tom_itx> blue blue/white are GND
[10:31:55] <Tom_itx> brown brown/white are +5v
[10:31:55] <Tom_itx> green/white is TXA -> 7i47 P4 Pin1 IO4
[10:31:55] <Tom_itx> green is TXB -> 7i47 P4 Pin2
[10:31:55] <Tom_itx> orange/white is RXA -> 7i47 P4 Pin19 IO20
[10:31:55] <Tom_itx> orange is RXB -> 7i47 P4 Pin20
[10:32:22] <Tom_itx> i can swap em and try again
[10:32:26] <Tom_itx> the A & B
[10:32:33] <Tom_itx> but not the TX RX pairs
[10:33:04] <Tom_itx> all that isn't that clear on this board arrangement
[10:33:13] <Tom_itx> which i understand is a bit odd
[10:37:32] <Tom_itx> ok, swapped wires on pins 1 & 2 and on 19 & 20
[10:38:26] <Tom_itx> that was it
[10:38:36] <pcw_home> Ahh
[10:39:13] <atom1> [ 5749.826557] hm2/hm2_7i43.0: IO Pin 004 (P4-09): Smart Serial Interface #0, pin RxData0 (Input)
[10:39:15] <atom1> [ 5749.826664] hm2/hm2_7i43.0: IO Pin 020 (P4-41): Smart Serial Interface #0, pin TxData0 (Output)
[10:39:16] <pcw_home> did you get a blinky green light (thats a good sign)
[10:39:39] <Tom_itx> no blinky lights yet
[10:40:12] <pcw_home> 7I71 should have a blinkly green light if the interface and threads started
[10:40:47] <atom1> i had closed lcnc, yes they're blinking with it loaded
[10:41:22] <Tom_itx> so RXA is - and RXB is +
[10:41:48] <Tom_itx> and apparently all the help you did on the bit file works too
[10:43:01] <Tom_itx> thanks
[10:43:34] <pcw_home> if you want the latest SSLBP you can just download any fpga zipfile and update your source
[10:44:52] <Tom_itx> source on the board or in linuxcnc?
[10:45:21] <Tom_itx> the 7i43 is a bit older... ver 41 iirc
[10:46:00] <Tom_itx> now to figure out pin access...
[10:46:34] <pcw_home> SSLBP is the firmware (processor code) that runs in the FPGA but you will not see any
[10:46:35] <pcw_home> difference with code > 35
[10:46:37] <pcw_home> (last change was just patching a bug that only happens with bad remote firmware)
[10:46:38] <pcw_home> important for us so that we can debug bad firmware without SSLBP hanging
[10:47:06] <Tom_itx> you updated the card firmware you said
[10:47:08] <pcw_home> all FPGA firmware is the same (currently v 43)
[10:47:23] <Tom_itx> unless i have problems i'll probably leave it for now
[10:47:31] <Tom_itx> may be good to know how to do it though
[10:47:53] <Tom_itx> does hal configuration show the sserial pins?
[10:48:09] <pcw_home> EPP/SPI/PCI/Ethernet/USB everything at the Hostmot2 level and below is the same
[10:48:37] <pcw_home> It shows the remote I/O pins
[10:48:41] <Tom_itx> ahh yes there they are
[10:50:02] <Tom_itx> now, i can split the vfield and vin and keep vin at 24v and set vfield to 5v can't i?
[10:50:14] <Tom_itx> wanna put a couple leds on the outputs and test
[10:51:11] <Tom_itx> ^^ with W1
[11:23:39] <pcw_home> Yes the latest firmware allows I think 4V minimum vfield
[11:24:11] <pcw_home> I think the 7I71 already has output LEDs
[11:25:59] <Tom_itx> the docs are a bit confusing on that. one place says 5v the other says 8v
[11:26:14] <Tom_itx> yeah it does have leds
[11:27:37] <Tom_itx> the section under power supply indicates 5-28v but the power connectors pinout shows field power 8-32v
[11:28:17] <Tom_itx> in the 7i71 pdf
[11:44:34] <Jymmm> Eh, just split the difference =)
[11:44:48] <pcw_home> Yes not everything has been updated to reflect the latest firmware changes
[11:45:24] <Jymmm> Um, how can firmware define voltage limits?
[11:48:01] <pcw_home> via operational limits
[11:49:54] <pcw_home> I also need to update the specifications to show that output current is limited at 5V
[11:49:55] <pcw_home> (which is why we originally would not allow < 8V)
[11:56:12] <Tom_itx> limited to what?
[11:56:20] <Tom_itx> 20ma or so?
[12:02:18] <pcw_home> 200 mA or so
[12:03:22] <pcw_home> nominal rating is 350 mA per output but also 1.4 A/driver chip
[12:04:35] <pcw_home> at 5V its more like 200 mA /output and 800 mA/driver chip
[12:04:46] <Tom_itx> plenty to drive xmas tree light relays
[12:06:02] <pcw_home> also you normally dont need clamp diodes
[12:06:15] <Tom_itx> i think i read that somewhere
[12:06:48] <pcw_home> (only if you cycle so fast the the inductive energy is significant)
[12:07:09] <Jymmm> ULN2803
[12:07:15] <Tom_itx> i'm not doing the lights.. just happened to think of that
[12:08:09] <pcw_home> we really like the NCV7608s
[12:09:09] <pcw_home> Overcurrent and overtemperature protection, 5 to 32V range (safe to about 40V)
[12:09:22] <Tom_itx> who makes those?
[12:09:37] <pcw_home> On-Semi
[12:09:52] <Tom_itx> i've used some of their stuff before
[12:10:11] <pcw_home> They have some newer push pull ones as well
[12:10:35] <Tom_itx> i used to be on their mail list.. not sure what happened
[12:11:02] <pcw_home> so you can have software programmable high-side/low-side or push-pull with the same driver
[12:11:15] <Tom_itx> nice
[12:11:21] <pcw_home> They are made for automotive use
[12:13:17] <IchGuckLive> Hi all B)
[12:13:35] <pcw_home> other nice thing is the overcurrent and overtemp status are available from the SPI interface
[12:16:04] <IchGuckLive> i woudt like to thank pcw for his fabouless electronics all lasmas are right up and running as fast and good as they shoudt
[12:16:54] <pcw_home> Good to hear (sometimes all I seem to hear about are problems)
[12:27:02] <Tom_itx> having never done any of this before, walking me thru modding the bit file and setting up the sserial was relatively easy with your assistance
[12:28:35] <Tom_itx> i did take notes... was able to rebuild the bit file again
[12:33:00] <cpresser> did anyone in here do a retrofit on a colchester lathe. i am looking forward to get a 'tornado 300' and rebuild it
[12:34:06] <archivist> dont remember seeing one mentioned in here
[12:34:08] <cpresser> basically i a looking for hints, tips, tricks, possible problems, ...
[12:39:23] <archivist> I just grepped the logs on this box, your question is the only tornado hit
[12:40:21] <archivist> so the next person will want to read your blog and progress :)
[12:52:24] <cpresser> archivist: i dont have the machine yet. transportation is an issue :/
[12:52:54] <cpresser> but its quite cheap and in working condition (mechanically). the control broke down. perfect for a retrofit
[12:53:18] <Tom_itx> how big it it?
[12:53:47] <cpresser> 2.7x1.7x2meter, 3.5tons
[12:55:01] <Tom_itx> yeah, not something you could put in your trunk
[12:59:04] <IchGuckLive> cpresser: thats the size im dealing now for plasmas big sheet
[13:21:28] <IchGuckLive> im off BYE
[14:23:59] <_DJ_> re
[16:42:15] <_DJ_> gn8
[16:43:51] <kengu> evening
[17:03:46] * Jymmm hears crickets....
[17:04:03] <Jymmm> (and see's dead peope)
[17:15:48] <K0D3R> I'm trying to make a bootable usb stick with linuxcnc
[17:16:17] <K0D3R> I just used dd and copied the iso to the usb device but it doesn't works,
[17:16:40] <K0D3R> anyone tried this?
[17:18:27] <CaptHindsight> http://askubuntu.com/questions/116942/any-way-to-manually-make-a-bootable-usb-from-iso
[17:18:59] <kengu> i have tried. sometimes it did work and .. but.. yeah. I usually just use netboot nowadays
[17:19:54] <K0D3R> CaptHindsight, that's what I tried
[17:20:27] <K0D3R> usually it works
[17:21:05] <CaptHindsight> you must have done something different or have a different format on the usb stick
[17:21:46] <cpresser> K0D3R: custom ubuntu-variants usually dont work for usb-boot.
[17:21:55] <CaptHindsight> unless you have a goofy BIOS issue that play now you see it now you don't with drives
[17:22:16] <cpresser> K0D3R: if you convert them (eg with unetbootin), you get a messed up boot prompt which only offers the regular install
[17:22:22] <somenewguy> can you boot that machine from a usb stick normally?
[17:22:27] <somenewguy> cause that is often a nightmare for me
[17:23:22] <andypugh> K0D3R: Just use the boot disk creator in Ubuntu
[17:24:27] <Tom_itx> unetbootin
[17:25:54] <K0D3R> It's an old machine but it can boot from usb, I have done this before many times
[17:26:27] <andypugh> Why use unetbootin when Ubuntu has a built-in utility (that has always worked for me)
[17:29:56] <K0D3R> as I don't use ubuntu I will try unetbootin, thanks
[17:30:47] <CaptHindsight> http://www.datamation.com/open-source/the-burning-bridges-of-ubuntu.html
[17:44:04] <K0D3R> unetbootin worked... though it printed hundreds of error messages...
[17:49:58] <andypugh> working generally doesn't print hundreds of error messages
[18:26:45] <NickParker|2> andypugh: So I finally got the Xilinx sdk installed on my pc. Remind me where I find the files I need to modify?
[18:27:25] <andypugh> Download the latest ZIP from Mesa, it is all in there.
[18:27:47] <NickParker> Do I not need anything special to make table mode work?
[18:28:29] <andypugh> I think all you need to do is allocate some more pins to your stepgens in the vhdl file
[18:29:31] <andypugh> You will need to give the pins the right names, I think that all the names appear in a different file.
[18:40:44] <NickParker> alright I *believe* I've changed PIN_7I76x2_34.vhd to what I want. I even grouped the pins so i don't need so many criss crossed jumpers. what do I open next to compile a bit file with these pins?
[18:41:55] <andypugh> No idea :-) I have only ever used the LinucCNC "in-house' version which has a script
[18:42:08] <Tom_itx> i should post the session pcw did with me
[18:42:16] <Tom_itx> it explains the steps
[18:42:23] <Tom_itx> i haven't cleaned it up yet though
[18:42:25] <NickParker> that would be appreciated
[18:42:33] <andypugh> I guess you edited something like a PIN_7I76......vhf file?
[18:42:34] <Tom_itx> it's just a log from here
[18:42:55] <Tom_itx> is it epp?
[18:42:59] <pcw_home> You probably _dont_ want to edit PIN_7I76x2_34.vhd but rather a copy of it
[18:43:06] <Tom_itx> mine may be different than yours
[18:43:16] <NickParker> andypugh: yes
[18:43:29] <andypugh> NickParker: Listen to pcw_home
[18:43:36] <andypugh> Ignore me
[18:44:08] <NickParker> Alright. pcw_home why should I edit a copy? This is on my PC not my main machine, and I just unpacked the mesa zip.
[18:44:35] <pcw_home> because a pin file that works can be used as a reference
[18:45:22] <NickParker> well I still have the original in the archive so no worries. Do you have any pointers on what I do next?
[18:46:21] <pcw_home> rename the new pin file to your desired name
[18:46:39] <Tom_itx> this is a log capture of the steps pcw had me take:
http://tom-itx.dyndns.org:81/~webpage/cnc/configs/sherline/bitfiles/Launching_webpack.txt
[18:47:16] <Tom_itx> i plan to clean it up a bit but haven't had a chance
[18:48:27] <Tom_itx> you may need to open a different project file depending on what the bitfile is
[18:49:03] <pcw_home> start xilinx webpack, , select open project from the menu
[18:49:04] <pcw_home> and open fivei25.xise in the directory where you unzipped the hm2 source
[18:50:03] <pcw_home> in the bottom of the source pane you should have a library tab
[18:50:18] <NickParker> I don't have anything named webpack, will SDK or platform studio work? SDK opens .xise files at least
[18:50:38] <Tom_itx> ise is what launches it
[18:50:49] <Tom_itx> you need to set an environment var first
[18:51:05] <Tom_itx> windows or linux?
[18:51:07] <NickParker> windows
[18:51:13] <Tom_itx> nevermind then
[18:51:21] <NickParker> I've got fivei25.xise open in ISE project navigator.
[18:51:34] <NickParker> found the library tab
[18:51:46] <NickParker> i have one library called work
[18:51:54] <Tom_itx> open it
[18:52:15] <NickParker> ok, lots of files.
[18:52:35] <NickParker> adding new pin file
[18:53:00] <pcw_home> in that pane right click and select add source (and add your new pin file)
[18:53:23] <pcw_home> then go back to the design pane
[18:53:52] <NickParker> ok
[18:55:10] <pcw_home> in the design pane double click on the TopPCIHostMot2 file
[18:56:16] <NickParker> ok
[18:56:49] <NickParker> --use work.updated_pins.all?
[18:57:13] <pcw_home> Yes without the --
[18:57:42] <pcw_home> and make sure all other pin files are commented out
[18:57:50] <NickParker> yup.
[18:58:06] <pcw_home> and make sure all cards are commented out except i25x9
[18:58:34] <NickParker> so leave "use work.i25_x9card.all" uncommented?
[18:59:10] <pcw_home> yes
[18:59:12] <pcw_home> (make sure the @CARD@ and @PIN@ entries are commented out)
[18:59:32] <NickParker> yep got them.
[19:00:41] <pcw_home> now make sure the top level file (TopPCIHostMot2) is selected
[19:00:42] <pcw_home> and in the process pane click on "generate bitfile"
[19:01:21] <pcw_home> (a submenu item of "implement design")
[19:02:33] <NickParker> under implement design i see translate, map, and place & route
[19:03:56] <pcw_home> maybe its under the next submenu (configure device) not sure
[19:08:44] <Tom_itx> i also had to click on "generate programming file" in order to get a bit file
[19:09:10] <Tom_itx> from the design pane
[19:11:05] <NickParker|2> just tried that and got "Process "Synthesize -XST" failed
[19:11:35] <pcw_home> So the you need to look at the first error
[19:11:40] <pcw_home> then
[19:13:08] <NickParker> Says std_logic_vector is not defined for 3 lines, but those 3 lines are in the middle of probably 15 others that use the same variable
[19:14:36] <pcw_home> that normally indicates that you are missing a pin file or card file
[19:16:46] <NickParker> i have both my updated_pins.vhd and unchanged i25_x9card.vhd in work library. should I be saying "us work.updated_pins.vhd" instead of .all?
[19:17:00] <NickParker> I went with all because it seemed like what all the others did
[19:19:58] <pcw_home> Did you change the updated_pins package name?
[19:20:20] <NickParker> nope, i'll go do that.
[19:20:24] <pcw_home> (in your new .vhd file)
[19:20:32] <pcw_home> at top and bottom
[19:20:43] <NickParker> yep i had noticed that and thought nothing of it
[19:22:04] <pcw_home> file names can be anything... only package names are important
[19:23:47] <pcw_home> well the xise (project file) remembers the file names
[19:26:39] <NickParker> alright I can make .bit files now.
[19:27:37] <NickParker> one more question, how do I add pins for a pwmgen? Can I just find the lines used to assign pwmgen pins in another file and copy them over, or do I need to change this ModuleID stuff as well?
[19:27:38] <andypugh> Don't understimate how special that makes you.
[19:27:47] <Tom_itx> :D
[19:27:54] <Tom_itx> is it an elite group?
[19:28:05] <andypugh> Yeah, there are 5 of you
[19:28:14] <Tom_itx> that's it?
[19:28:27] <pcw_home> You need to add the PWMGen to the module ID section
[19:28:38] <andypugh> I think it is single figures.
[19:28:43] <Tom_itx> i'm gonna document both mine and this session and post it somewhere
[19:28:46] <pcw_home> (if theres not one there already)
[19:30:12] <NickParker> exactly that, "PWMGen"? I've got "PWMNumRegs" "PWMTag" "PWMMPBitMask" and "PWMValAddr&PadT"
[19:30:43] <Tom_itx> pcw_home, i forgot how many entries were supposed to be in the table?
[19:31:07] <Tom_itx> i kept it the same on mine but forgot how many there were
[19:31:19] <andypugh> NickParker: Look at another file that already contains pwmgens, and plagiarise.
[19:31:47] <NickParker> yep i just found one. either they're a bit rare or I was getting unlucky. will do
[19:32:25] <andypugh> I think most 5i25 configs don't use PWM, they use serial D to A, so you may need to cast the net wider to find a pwmgen example.
[19:33:06] <NickParker> You know, I should probably just use serial D to A. mesa sells them I assume?
[19:33:21] <pcw_home> tag, version, clock tag, number, base address, numregs, strides. MPBitMask
[19:33:23] <Tom_itx> my 7i43 has pwmgen
[19:33:30] <andypugh> 5i20 / 5i22 / 5i23 configs will have them at least half the time, bit rather few 5i25 ones do
[19:34:12] <pcw_home> Some do (7I78 or 7I85S PWM/DIR versions)
[19:34:46] <andypugh> NickParker: You could just buy a 7i77 :-)
[19:35:08] <Tom_itx> pcw_home, is the format of the .vhd files consistent throughout?
[19:36:01] <andypugh> Why do you need 4-pin stepgens _and_ pwm anyway? You are Mesa's most eccentric user!
[19:36:08] <NickParker> andypugh: You must remember I'm a poor student.
[19:36:22] <NickParker> andypugh: I have a very eccentric mill.
[19:36:40] <pcw_home> Tom_itx: the pinout files? yes
[19:37:35] <pcw_home> K wallace may be interested in the table mode stepgen, He asked about it a couple months ago
[19:38:01] <NickParker> we have identical mills, except my spindle is more eccentric
[19:38:08] <pcw_home> (the firmware has had it since day one)
[19:39:02] <andypugh> Talking mills, I spent today trying to figure out how to fit my (new, expensive, preloaded, special order) ballnut in the mill. The problem is that the catalogue drawing shows a 1" cylindrical ball nut, with a thread to screw in to the housing.
[19:40:25] <andypugh> The actual part looks exactly the same. Plus a large external ball-transfer tube which means that the nut actually can't screw in to anything.
[19:42:38] <andypugh> I ended up trying to clamp the nut inside a rather slender bit of steel:
http://imagebin.org/279742
[19:42:55] <andypugh> Those screws are M4.
[19:43:34] <NickParker> That seems like it would add any backlash you had eliminated with your new nut
[19:44:08] <andypugh> It's not as stiff as I would like, certainly
[19:44:51] <andypugh> Though I don't see how it adds backlash?
[19:45:16] <andypugh> The location is still on the thread.
[19:46:04] <NickParker> I figured the clamp would probably be able to move along the axis of the screw slightly.
[19:47:37] <andypugh> The spigot on the brown part fits into the table, and is bolted up. The nut fits in the bore of the part, which is threaded to suit the nut on the far side.
[19:48:13] <andypugh> The previous part was non-preloaded M12, this is preloaded M16
[19:48:26] <NickParker> Oh I hadn't noticed the threads. Nevermind me
[19:48:50] <andypugh> The previous part was a cylindrical nut that just screwed in and was locked by a nut.
[19:49:19] <andypugh> I didn't model the threads in the brown part.
[19:50:30] <andypugh> I would be incapable of designing that without 3D CAD, how did they do it with paper
[19:51:07] <Tom_itx> if you're not using all 3 pins of an encoder function can the remaining pins be remapped to another function or would that present an error?
[19:51:17] <Tom_itx> in the bitfile
[19:52:51] <Tom_itx> for example if i'm using an MPG on A & B, can i remap the Index channel to another function?
[19:58:50] <andypugh> Tom_itx: 7i73 features 2-pin encoders, but I am not sure that is directly comparable
[19:59:41] <Tom_itx> i just don't know if the bitfile will error if i drop the index and reuse it
[20:00:06] <Tom_itx> it would sure be handy if i could
[20:07:08] <Jymmm> Tom_itx: Why couldn't you?
[20:07:34] <Tom_itx> i don't know if it would screw up the internal encoder function firmware
[20:18:28] <Jymmm> Tom_itx: would it hurt to just test it out?
[20:19:09] <Tom_itx> that's the _other_ way to find out :)
[20:19:50] <Tom_itx> probably won't tonight
[20:20:00] <Tom_itx> been updating all my pin documentation
[20:34:25] <NickParker> Why has pwmgen got a direction output?
[20:34:48] <NickParker> Or rather, can I ignore it?
[21:00:18] <pcw_home> Tom_itx: you can normally just use the the pins you need
[21:00:20] <pcw_home> (so a encoder counter with just an 'A' pin is fine for example)
[21:02:38] <pcw_home> Unused module input pins are tied to logic 0
[21:02:40] <pcw_home> (like the 'B' and index pins of an encoder counter that only has a 'A' pin)
[21:09:30] <Tom_itx> pcw_home yes but can i reassign them in the bit file?
[21:09:37] <Tom_itx> i wanted to map the sserial to a couple of them
[21:10:31] <Tom_itx> the MPG will not have an index channel so it could easily be used
[21:29:03] <pcw_home> Any pin can be pretty much anything
[21:30:18] <Tom_itx> i wasn't sure if it would interfere with the internal functions not having all the pins assigned to it
[21:30:45] <pcw_home> thats what I was saying, you can make a stepgen with just one pin if you like
[21:30:47] <pcw_home> (This is done on some configs like the G540 as a charge pump signal generator)
[21:31:03] <Tom_itx> ok
[21:31:24] <Tom_itx> that'll help quite a bit
[21:31:52] <Tom_itx> i may give it a try here in a bit
[21:31:54] <pcw_home> some things dont make sense of course but are still possible (a index only quadrature counter for example)
[21:32:08] <Tom_itx> well sure
[21:33:33] <Tom_itx> i'm not using B or index on the spindle and I on the MPG so there's 3 i could use
[21:34:06] <Tom_itx> i may not mess with the spindle since i could some day want index on it
[21:35:09] <pcw_home> In general most sub-setting of pins is possible
[21:36:19] <pcw_home> PWMgens normally have PWM/DIR/ENA but you can use just one pin (or maybe no pins at all) if you like
[21:36:51] <pcw_home> a no output stepgen could still be used as a timer (reading the feedback register)
[21:37:27] <andypugh> Goodnight chaps
[21:37:53] <Tom_itx> makes for well use of pins when necessary
[21:39:03] <pcw_home> yeah, you can make a configuration with just the things you need
[22:37:43] <sabotender> moo
[22:38:02] <sabotender> so then, i have Debian installed *finally*
[22:38:27] <sabotender> what do I need to download to work with cnc crap? I want to look at the software
[22:50:56] <Tom_itx> best way to 'look' at the software is download the live ubuntu cd
[22:52:18] <Tom_itx> http://www.linuxcnc.org/index.php/english/download
[22:54:25] <sabotender> Tom_itx: huh? debian is already set up on my computer
[22:54:43] <Tom_itx> you need RTAI
[22:55:17] <Tom_itx> i haven't built lcnc from scratch but it can be done
[22:56:17] <Tom_itx> http://buildbot.linuxcnc.org/
[22:56:23] <Tom_itx> may be a place to start
[22:57:54] <roycroft> linuxcnc has a lot of dependencies
[22:58:16] <roycroft> if you're just wanting to explore the software right now, the livecd can be put on a thumb drive and you can boot off the thumb drive and run it from there
[22:58:36] <roycroft> or you can install some virtualization software on your linux machine and run the live cd from the virtualization manager
[22:59:02] <roycroft> i would recommend either of those before trying to install it on your existing machine
[22:59:18] <roycroft> you need to test the hardware to see if it performs well enough for linuxcnc anyway
[22:59:52] <roycroft> booting off a thumb drive and running the hardware tests from there would be a good way of doing that without disturbing your existing linux installation