#linuxcnc Logs
Sep 01 2022
#linuxcnc Calendar
12:55 AM Deejay: moin
03:18 AM randy: morning
04:41 AM Tom_L: morning
04:51 AM JT-Cave: morning
05:05 AM mrec: I think I'm getting further with hobbing, does anyone have experience with that? I wonder should I go for a single pass or do multiple passes for a small m0.3 gear?
05:06 AM jpa-: it will depend so much on material, cutter and rigidity of the setup that probably best to just try out
05:07 AM mrec: for what I've learned so far the hobber is way better than a simple gear cutter
05:07 AM mrec: if the gear cutter is worn out you can just dump it .. but the hobber can be sharpened/maintained at least
09:51 AM chris_99: Hi, i just designed a metal plate for a linear actuator XY stage to sit on, just wondering if anyone might have any suggestions for the thickness of metal to use (thinking SS304), the linear actuator XY stage weighs around 5kg, and the plate is 450x450mm, i was originally thinking 2mm, but not sure if that's thick enough now
09:56 AM roguish[m]: chris_99: how is it held? what is it holding? loads, loads, loads..... what, where, how much, ??? fairly simple math. equally important to strength, is stiffness. how much will it deflect? does it matter?
09:58 AM CloudEvil: 3mm is 5* as stiff as 2mm, for example.
09:59 AM chris_99: anfractuosity.com/files/out.dxf - i created two slots, that match up to the ridges of the X rail, and will attach bolts in several places. Ah, i didn't realise that CloudEvil, cheers
10:01 AM roguish[m]: chris_99: post an image if you can. a .dfx is a text file.....has to be opened by a cad program....
10:06 AM chris_99: sure, https://www.anfractuosity.com/files/out.jpg
10:07 AM Rab: chris_99, the actuator will be supported by the plate, and the plate will be supported at the four corners?
10:07 AM chris_99: yeah that's right
10:07 AM chris_99: with feet on each corner
10:08 AM Rab: Seems pretty springy. Is there a specific reason you're using 304 stainless? Corrosion resistance?
10:08 AM CloudEvil: Stainless is no better than steel in this app
10:08 AM CloudEvil: thick steel would be better than thinner stainless.
10:08 AM jdh: or a chunk of 6061
10:09 AM CloudEvil: And, indeed, moderately thicker aluminium
10:09 AM chris_99: i was thinking mainly for not rusting, re 304
10:09 AM jdh: are the slots for 1 rail or two parallel rails?
10:10 AM Rab: chris_99, can you describe your application?
10:10 AM CloudEvil: Steel is about 5* more rigid than stainless steel. But, once you look at increased section, you need to get to (4th root)5 to be as stiff with aluminium, which is 1.5* the thickness of an equivalently stiff steel section
10:10 AM CloudEvil: Wait - is it 3, or 4 the factor.
10:11 AM * CloudEvil is not awake.
10:11 AM CloudEvil: It's 3, recalculate above respectively.
10:11 AM CloudEvil: (stiffness of a beam with thickness
10:11 AM chris_99: for one X rail, that has two slots in it (hope that makse sense) i can post a photo of the actual linear actuator thing, it's to make a scanning optical microscope
10:12 AM chris_99: https://www.anfractuosity.com/files/DSC09304.jpg
10:13 AM Rab: chris_99, not my area of expertise, but it sounds like you would need as little deflection/vibration as possible. Is this a DIY effort, i.e. are you scrounging materials or are you limited to catalog items?
10:14 AM chris_99: yeah diy, although getting the plate laser cut, and will bolt everything together then
10:15 AM Rab: chris_99, I guess there are ridges on the bottom of the actuator extrusion?
10:15 AM chris_99: yeah
10:16 AM Rab: And one axis is completely supported by the other one, i.e. floating in the air?
10:16 AM chris_99: yeah
10:16 AM chris_99: which probably isn't great i guess
10:18 AM Rab: Not ideal, and the floating axis will probably sag and change attitude across its travel. But I assume there isn't much load.
10:20 AM roguish[m]: CloudEvil:regular steel and stainless steel are the same stiffness for a given geometry. Modulus of Elasticity is approx. 30,000,000 psi for both, give or take a very small percentage for alloying elements. Aluminum modulus is 10000000, titanium is 17000000
10:20 AM CloudEvil: roguish[m]: yes, but per price, you'll get much more with steel, I mean.
10:20 AM CloudEvil: (if thickness is acceptable to vary)
10:20 AM chris_99: just looking at some rough prices, 6mm mild steel is approx £43 , SS304, £112, do you folks think 6mm might be overkill?
10:20 AM roguish[m]: basic material science
10:20 AM Rab: My preference would be to find some type of cast metal base, either aluminum or iron. But I think if you build a frame under your plate, you'll have a much more rigid assembly than trying to size the plate to be self-supporting.
10:21 AM Rab: Even just a couple of reenforcing ribs from angle or box.
10:24 AM roguish[m]: CloudEvil: yes, price is a totally different matter......
10:24 AM roycroft: you could build a frame and fill it with epoxy granite
10:24 AM roycroft: with some reinforcement ribs as well
10:24 AM roguish[m]: chris_99: if you only need to cut the plate, not weld it, check out 303. has some sulfer to aid machinability
10:25 AM roguish[m]: What roycroft said.....
10:25 AM roguish[m]: or start with a granite slab.
10:26 AM roycroft: yes, axminster should have surface plates pretty cheap
10:26 AM chris_99: alas 303 wasn't an option at the laser cutter. That sounds interesting re.slab
10:26 AM roguish[m]: chris_99: most importantly is .... what are your performance requirements and expectations, and budget...
10:27 AM roycroft: https://www.axminstertools.com/axminster-granite-surface-plate-105640
10:28 AM roycroft: it looks like they only carry small ones, and are out of stock
10:28 AM roycroft: but there are other vendors for that in the uk, i'm sure
10:28 AM chris_99: not much expectations tbh, this is a silly project to try to image a floppy disk optically via a microscope, among other things, the cheaper the better
10:28 AM roguish[m]: very high 90's today. 100+ for the weekend. 107 forcast for Monday
10:29 AM roycroft: 34 here today
10:29 AM roycroft: and then we'll be back in the upper 20s for a few days
10:30 AM Rab: Another issue is that cutting those long slots across the plate will remove most of the perpendicular rigidity of the plate, and the weight will be centered right at the point where it's weakest. If it were me, I would cut a shim to go between the slide and the plate.
10:33 AM chris_99: yeah that's a good point, re rigidity
11:42 AM Tom_L: !
11:42 AM Tom_L: the weekend has arrived!
11:46 AM roycroft: not yet
11:46 AM roycroft: not even nearly
11:46 AM roguish[m]: couple daze early
11:46 AM roycroft: i am almost finished with the email server migration, though
11:46 AM roguish[m]: roycroft: how are you with subnets and stuff?
11:47 AM roycroft: then i get to build the membership database for the country fair from a backup, and make hundreds of updates without the front end software
11:47 AM roycroft: what about subnets and stuff?
11:49 AM roguish[m]: I have a Netgate SG-1100. I want to set up a network so I can separate all the stuff I don't want to access the outside world or be accessed, and all the stuff that's ok to do so.
11:49 AM roguish[m]: therefore subnets. I think.
11:50 AM roycroft: there's actualy a simple solution to that
11:50 AM roycroft: that may work for you
11:50 AM roguish[m]: all ears...
11:50 AM roycroft: if you want all the hosts to be able to reach each other internally, then you can keep them all in one network
11:51 AM roycroft: you'll need to set up static ips for the ones you don't want to reach the outside
11:51 AM roycroft: so let's say you're using 192.168.1.0/24 for your internal network
11:51 AM roycroft: your gateway address will probably be 192.168.1.1 - the address on your firewall/router
11:52 AM roguish[m]: my Asus router has a limited number of blocks, 16 ips, I have more than that, that I wish to block
11:52 AM roycroft: so set up the dhcp pool to use 192.168.1.128-192.168.1.253, and use dhcp for your hosts that you wish to have access to the outside
11:52 AM roycroft: then for the ones you want to keep isolated, set them up with static ips in the range 192.168.1.2-192.168.1.127
11:53 AM roguish[m]: Most everything has static ips.
11:53 AM roycroft: but when you set them up, set the gateway address to 192.168.1.254, not 192.168.1.1
11:53 AM roguish[m]: quite a speadsheet to track it all.
11:53 AM roycroft: and make sure there is no device listening on 192.168.1.254
11:53 AM roycroft: use the same netmask (255.255.255.0 in this case) on all the hosts
11:54 AM roycroft: those with the correct gateway address will be able to reach the outside world
11:54 AM roycroft: those with the bogus one will not
12:01 PM roguish[m]: interesting strategy.
12:01 PM roguish[m]: i'll consider...
12:01 PM roycroft: it works well
12:01 PM roycroft: and a nice benefit is that if you ever need to connect one of the machines to the internet, for, say a software update, you just fix the gateway, do what you need to do, and then change it back to the bogus one
12:01 PM roycroft: if you don't want to do it that way you'll need either a smart switch that does vlanning, or you'll need a switch for each subnet, and perhaps another router
12:03 PM roycroft: i'm not familiar with that device
12:03 PM roycroft: if you have any specfic questions about subnetting i can answer them
12:04 PM roycroft: if you have general questions about subnet strategies, i can probably answer them
12:04 PM roycroft: if you have questions about configuring that device i probably cannot answer them
12:04 PM roycroft: but honestly, i've been using the bogus gateway strategy for many many years, and it works brilliantly
12:04 PM roycroft: and it's simple and device-independent
01:23 PM roycroft: oh, the insanity
01:23 PM roycroft: i'm still restoring mail folders for our customers to the new mail server
01:24 PM roycroft: meanwhile, my boss has sent me five feature requests for the new mail system, and wants me to implement them right now
01:24 PM roycroft: like before i finish getting the files restored
01:24 PM roycroft: i should note that the new mail server already has all the features of the old one, and more
01:24 PM roycroft: these are brand new requests
01:31 PM jdh: it's good to have goals
01:38 PM roycroft: sanity is more important than goals, though
01:39 PM roycroft: and i've been working on this project for a year and a half
01:39 PM roycroft: why, in all that time, were these new feature requests not expressed?
01:40 PM roycroft: oh well, i've added them to my list of things to work on when i have the time
01:51 PM RoboticsNut1189[: Can I ask why we moved to a new discord server?
01:55 PM CaptainHindsight: Robotics Nut (@_discord_570343600026550287:jauriarts.org) tes
01:55 PM CaptainHindsight: yes even
01:56 PM roycroft: i think it's because this is the year of moving
01:56 PM CaptainHindsight: the admin did not respond to any messages
01:56 PM roycroft: most youtube makers seem to have moved this year, or are in the process of moving
01:57 PM CaptainHindsight: so we had to start a new server in order to bridge from Discord and Matrix to IRC
01:57 PM roycroft: even matt cremona, who just bought a new house a couple years ago and moved there, is now moving out of his warehouse into a new building he's constructing on his property
02:02 PM mrec: does anyone know is it possible to mix the parallel port pin configuration with a network configuration? (eg. using stepgen+parport, but getting encoder feedback via ethernet)?
02:03 PM CaptainHindsight: Robotics Nut (@_discord_570343600026550287:jauriarts.org) we can't even get a hold of him now, he never responds
02:03 PM mrec: I have a De2-115 development board lying around which has an ethernet connector
02:04 PM mrec: the parallel port is fully occupied, so I wonder if I can just extend the configuration with some ethernet configuration
02:04 PM CaptHindsight[m]: mrec: yes you can mix and match
02:05 PM CaptHindsight[m]: I often have LPT and PCIe with LCNC going at the same time
02:06 PM CaptHindsight[m]: mrec: encoder feedback might be an issue if the motors and encoders are not in the same FPGA
02:07 PM CaptHindsight[m]: mrec: Ethernet is on a servo thread and the LPT port is on a base thread
02:07 PM mrec: hmm.. I could take the limit switches off the parport
02:08 PM CaptHindsight[m]: mrec: can you put the steppers and encoders on the same Ethernet FPGA and then leave LPT for non servo/closed loop devices?
02:09 PM mrec: I will have a look
02:09 PM CaptHindsight[m]: mrec: Encoder reading on the LPT port is limited by the base thread speed, FPGA's can't count to the MHz vs 10's of KHz on the LPT port
02:10 PM CaptHindsight[m]: sorry can't/can
02:15 PM mrec: that's just fine
02:15 PM mrec: I only need it for gear hobbing it will run on low rpm
02:16 PM mrec: I did threading on the lathe with the spindle encoder attached tot he parallel port
02:16 PM mrec: that worked just fine
02:37 PM CaptHindsight[m]: I'm getting a: caught signal 8 - dumping core when I spin the encoder too fast on the LPT port
02:38 PM CaptHindsight[m]: new simple config, encoder connected as a spindle encoder
02:38 PM CaptHindsight[m]: and only one stepper motor
02:38 PM CaptHindsight[m]: it can happen before the machine is even out of e-stop
02:42 PM mrec: I'm using 2.9pre
02:43 PM mrec: I'm not sure how fast I ran the encoder very sure <1000rpm
02:44 PM mrec: I even forgot the settings (how many PPR) it's quite some time ago already
02:46 PM mrec: https://ams.com/documents/20143/36005/AS5047P_DS000324_3-00.pdf
02:46 PM mrec: this one has quite a few settings on page 19
02:47 PM mrec: oh on the lathe I'm only using A/I no B
02:47 PM CaptHindsight[m]: 1000 count per rev encoder is 4000 counts per rev and you need to sample at at least 2x, so 8000 reads per rev minimum,
02:48 PM mrec: I don't need so many for threading or hobbing I'm sure
02:50 PM mrec: the thread was pretty good back then, I had other problems back then ... bldc motor controller too weak and the belt was slipping, also rigidity on the crossslide
02:50 PM mrec: I have a solid toolpost and 1.2kw motor installed, I still need to replace the belt
03:02 PM perry_j1987: hows it goin
03:05 PM JT-Shop: hmm the Automation Direct P1-01 PLC is made in florida
03:05 PM JT-Shop: I take that back it's the power supply that's made in florida
03:10 PM CaptHindsight[m]: JT-Shop: it looks decently made vs the *duinos, even though it uses a similar processor
03:14 PM CaptHindsight[m]: JT-Shop: which PLC are they out of stock of?
03:18 PM CaptHindsight[m]: oh wow, only the Productivity3000 and ProductivityOpen Controllers are in stock
03:21 PM JT-Shop: click is mostly out of stock
03:59 PM perry_j1987: man that was janky setup heh but it worked
04:00 PM perry_j1987: had to fly cut a part on the lathe using strap clamps stacked...
04:00 PM perry_j1987: to reach the spindle centerline haha
04:13 PM JT-Shop: just blew up the power supply
04:17 PM Tom_L: no wonder they are in short supply!
04:18 PM JT-Shop: ordered the wrong one
04:19 PM Tom_L: woops
04:21 PM JT-Shop: $50 mistook
04:22 PM Tom_L: good you're not at the bottom of the chain
04:36 PM JT-Shop: might have damaged the plc...
04:37 PM Tom_L: that could be a more serious problem
04:43 PM JT-Shop: hmm even though it seems to power up with just the usb I think it still needs 24vdc
04:46 PM Tom_L: these for the copy lathes?
04:53 PM JT-Shop: this is for the double chucker, I hope the click will be back in stock by the time I do the final copy lathe or nearly so
04:54 PM JT-Shop: between me and my buddy at Revere we can come up with the click plc for the copy lathe
05:00 PM rene-dev56 is now known as rene-dev5
05:04 PM XXCoder: https://youtu.be/CmOxuwojt_c
05:04 PM XXCoder: pretty cool
05:05 PM XXCoder: hes making isulation for giant steel door
05:05 PM XXCoder: door was at 120f inside side wow
05:16 PM XXCoder: it seems to be very effective apparently
05:18 PM CloudEvil: Shading also a good plan
05:18 PM XXCoder: yeah though dont know what his lot conditions is like. do he own it or is it rental
05:19 PM perry_j1987: that was odd
05:20 PM perry_j1987: just noticed the servos acting a bit odd while jogging the z axis it started sputtering like the servo was trying to find itself for a split second
05:26 PM JT-Shop: odd I had to put a usb hub between the windblows 7 pc and the P1AM or the sketch would fail to upload
05:27 PM Tom_L: drawing too much current?
05:27 PM Tom_L: hub probably has more power
05:29 PM JT-Shop: some kind of issue with windblows 7 I think
05:29 PM JT-Shop: and the MKR usb connection
05:29 PM Tom_L: most ports only put out .5A
05:30 PM JT-Shop: I don't think it's a power issue but a communications issue
06:13 PM Birdman3131[m]1: Should Linuxcnc boot off a usb hdd ok?
06:15 PM Birdman3131[m]1: Not positive this computer is gonna work anyways as old as it is but if i can boot from a usb hdd i won't need to track down a pata drive.
06:16 PM CloudEvil: If the motherboard supports booting off USB, it should.
06:33 PM perry_j1987: https://www.fluidlinesystems.net/Bijur-D3607-V5W-Gear-Pump-OBSOLETE-p/d3607.htm and https://www.fluidlinesystems.net/Bijur-5-Gallon-Reservoir-D3050-p/d3050.htm i got these two units sitting here wow are they really worth that much?
06:39 PM CloudEvil: If you can get someone to pay that, yes.
06:46 PM perry_j1987: you buyin? :P
06:48 PM CloudEvil: I will pay $15, shipped.
06:49 PM CloudEvil: (for the pump)
06:49 PM perry_j1987: lol
06:50 PM perry_j1987: wow thats very generous $15 dollars a day for 288.23 days :)
07:12 PM CaptHindsight[m]: if I spin the spindle encoder quickly by hand I get: rtapi_app: caught signal 8 - dumping core
07:12 PM CaptHindsight[m]: https://pastebin.pl/view/0880f7a8
07:12 PM CaptHindsight[m]: even if the machine is still in e-stop
07:12 PM CaptHindsight[m]: simple config using stepconf
07:13 PM CaptHindsight[m]: https://pastebin.pl/view/3d30cd86
07:14 PM CaptHindsight[m]: LINUXCNC - 2.9.0~pre0
07:15 PM CaptHindsight[m]: all default setting other than manual edit of LPT base address
07:18 PM CaptHindsight[m]: and changed
07:18 PM CaptHindsight[m]: setp encoder.0.position-scale 4000.000000
07:18 PM CaptHindsight[m]: default is like 100 vs 4000
07:20 PM CaptHindsight[m]: the only mention of a similar error: https://forum.linuxcnc.org/24-hal-components/35892-hal-twopasses-and-ethercat-lcec?start=10
07:20 PM CaptHindsight[m]: "has happened to me many times, due to duplicating signals in HAL with the same name"
07:39 PM perry_j1987: well im officially peeved
07:39 PM perry_j1987: just cannot stop this qctp from spinning
07:40 PM perry_j1987: need to ditch it and finish building the gang tool setup for this new lathe
08:14 PM CaptHindsight[m]: why does stepconf now default to BASE_PERIOD = 100000 no matter what you enter for you maximum base period jitter?
08:15 PM XXCoder: sounds like a bug
08:30 PM CaptHindsight[m]: I think another bug in stepconf is causing my signal 8 error as well
08:31 PM Tom_L: jitter bug
08:37 PM solarwind: I'm so frustrated. I used to be able to play the guitar. Now every chord just comes out muted
08:38 PM CaptHindsight[m]: Removed the option of manually inserting the parallel port Why since this was too handy to have in stepconf?
08:38 PM solarwind: Especially F, being so close to the nut
08:42 PM CaptHindsight[m]: WTF https://github.com/LinuxCNC/linuxcnc/commit/06a34258f8603b46614f2f47186b02a6e65edebc
08:45 PM Tom_L: :)
08:45 PM Tom_L: no reason given?
08:47 PM CaptHindsight[m]: solarwind: could be worse, pianos have way more keys than guitars have strings and each one has a different tone
08:47 PM CaptHindsight[m]: Tom_L: not that i can find, was this too confusing?
08:47 PM Tom_L: well, i can't really say since i've never used it
08:47 PM Tom_L: does seem a bit odd though
08:48 PM CaptHindsight[m]: handy feature to test if you have the right base address
08:48 PM Tom_L: you haven't noticed it in 2 yrs though
08:48 PM CaptHindsight[m]: now if stepconf doesn't automagically detect your LPT card you have to manually edit the config and try loading LCNC
08:49 PM CaptHindsight[m]: yeah i was using an older version of stepconf
08:49 PM CaptHindsight[m]: now in 2.9 it's a fun new feature
08:50 PM Tom_L: you use the parport quite a bit?
08:50 PM CaptHindsight[m]: now they are fixing things that aren't broken
08:51 PM Tom_L: i'd question it if i were you
08:51 PM Tom_L: don't think it was hurting anything being there
08:51 PM CaptHindsight[m]: easy enough to back out
08:51 PM CaptHindsight[m]: but why?
08:51 PM Tom_L: i'd ask him
08:52 PM CaptHindsight[m]: where?
08:52 PM CaptHindsight[m]: they don't hand out anywhere
08:52 PM CaptHindsight[m]: hand/hang
08:52 PM Tom_L: sign in and leave a comment
08:53 PM Tom_L: it will generate an email
08:53 PM Tom_L: it's ok in 2.8?
08:55 PM CaptHindsight[m]: weird stuff going on with LCNC
09:02 PM Tom_L: some
09:04 PM Tom_L: i don't do github so i'm not that familiar with how it works
09:04 PM Tom_L: what are the blue lines indicating?
09:04 PM Tom_L: i know the red/green
09:06 PM CaptHindsight[m]: I just asked C Morley
09:06 PM Tom_L: strange it would get dumped with no explanation
09:07 PM CaptHindsight[m]: the BASE_PERIOD that stepconf generates is based on what you enter for max feed rates and accel
09:08 PM CaptHindsight[m]: not just base thread latency jitter
09:09 PM Tom_L: does that justify the edit?
09:23 PM CaptHindsight[m]: I'm just used to testing RTAI and you just leave everything maxed out
09:25 PM CaptHindsight[m]: but if your system can run a 10-20uS base thread BUT you never plan on rapids or accel fast enough to need a base thread that fast you can lower the base thread to what you have set for max feed rates and accel
09:26 PM CaptHindsight[m]: BUT if you change your max feeds and accel later then you have to also lower the base thread timing
09:26 PM CaptHindsight[m]: or you'll get RT errors
09:27 PM CaptHindsight[m]: I'm testing out a bunch of PCIe LPT cards as well
09:27 PM CaptHindsight[m]: why I am discovering these issues
09:28 PM CaptHindsight[m]: will also try to fix the CH382 not working with the HAL LPT driver
09:28 PM CaptHindsight[m]: the CH382 cards are <$10ea
09:28 PM CaptHindsight[m]: skunkworks and I are on a LPT card mission!
09:30 PM * roycroft points capthindsight and skunkworks towards the nearest pc museum
09:30 PM CaptHindsight[m]: LPT cards work on the latest PC hardware
09:31 PM CaptHindsight[m]: I wish that PC's had GPIO ports. LPT cards are the next best thing
09:32 PM CaptHindsight[m]: AMD did put some GPIO pins on some embedded chipsets but not enough pins to run a 3-axis mill with steppers
09:36 PM CaptHindsight[m]: I think it was only 6 pins
09:40 PM Tom_L: too bad scsi couldn't be used
09:41 PM Tom_L: i dunno how many IO it had but sure had a bundle of wires
09:45 PM Tom_L: how many signals do you get on LPT? 12?
09:46 PM CaptHindsight[m]: http://linuxcnc.org/docs/2.4/html/hal_parallel_port.html#fig:Parport-block-diag
09:47 PM CaptHindsight[m]: up to 17
09:47 PM Tom_L: usin it for all it's worth
09:48 PM CaptHindsight[m]: every new Ryzen we test gets under 20uS for base jitter
09:48 PM CaptHindsight[m]: closer to 10uS
09:49 PM Tom_L: how many cores can you get on those?
09:49 PM CaptHindsight[m]: IF there was a low cost and available FPGA with a hard PCIe express port it would be nice
09:49 PM Tom_L: the high end i7 i saw 8+4
09:49 PM CaptHindsight[m]: Tom_L: the newest are yeah something like that ^^
09:50 PM CaptHindsight[m]: new ones were just released this past Tuesday
09:50 PM Tom_L: i don't keep up anymore
09:50 PM Tom_L: 16
09:50 PM CaptHindsight[m]: https://www.tomshardware.com/news/amd-zen-4-ryzen-7000-release-date-specifications-pricing-benchmarks-all-we-know-specs 16 core/32 threads
09:51 PM Tom_L: just about to post that
09:51 PM Tom_L: ..
09:51 PM CaptHindsight[m]: up to 5.7 GHz
09:52 PM Tom_L: i bet that eats some watts
09:54 PM Tom_L: of course you can't get them..
09:54 PM CaptHindsight[m]: TDP 230W for th max cores and core speed
09:55 PM CaptHindsight[m]: https://wccftech.com/amd-ryzen-7000-zen-4-cpus-run-hot-ryzen-9-7950x-up-to-95c-230w-ryzen-5-7600x-up-to-90c-at-120w-rumor/
09:55 PM Tom_L: it takes ~2hrs per build on this silly iso
09:56 PM Tom_L: it will probably take linux a while to catch up with them
09:58 PM CaptHindsight[m]: soon I will make a LCNC img for an ARM board like the https://pine64.com/product/quartz64-model-b-4gb-single-board-computer/
09:58 PM CaptHindsight[m]: down the road we can get the PRU's to software step as well
09:59 PM CaptHindsight[m]: but it can work with Mesa over Ethernet and they are only $60ea with 4GB RAM and in stock
10:48 PM skunkworks: CaptHindsight[m]: I ran your config on buster with no issues. I don't have an index - but it counts on a and b..
10:51 PM skunkworks: (just built master.
11:00 PM CaptHindsight[m]: skunkworks: thanks
11:01 PM skunkworks: could you post the extra files?
11:01 PM CaptHindsight[m]: it counts for me but if I spin it fast enough it throws the rtapi_app: caught signal 8 - dumping core
11:01 PM CaptHindsight[m]: sure, it's just the defaults that stepconf creates
11:01 PM skunkworks: nothing really in them?
11:02 PM skunkworks: does this have spindle speed indicator in the xml?
11:02 PM CaptHindsight[m]: yes
11:03 PM skunkworks: ok - link those somewhere (jsut to be sure)
11:03 PM * CaptHindsight[m] posted a file: (1KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/zFJgtxuWxbOCNpFIemBOPrXU/custompanel.xml >
11:04 PM * CaptHindsight[m] posted a file: (2KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/wifjzWVERKnIgQFHtNixIbiZ/linuxcnc.var >
11:04 PM * CaptHindsight[m] posted a file: (1KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/IAUvJrNEYTrmdFBuOkVvLyiP/pyvcp_options.hal >
11:04 PM * CaptHindsight[m] posted a file: (4KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/qsZWlXwtSbNLwAWLCuOTZAsM/my-mill.hal >
11:04 PM * CaptHindsight[m] posted a file: (3KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/AtfaWyFzTeVGiZEfYAenCTje/my-mill.ini >
11:06 PM CaptHindsight[m]: it's pretty much a default config for a 3-axis mill with spindle encoder connected to LPT pin 10 and 11
11:06 PM CaptHindsight[m]: looks like I found some new bug
11:07 PM CaptHindsight[m]: I haven't tried this with different LPT controllers yet
11:09 PM skunkworks: CaptHindsight[m]: custom.hal?
11:09 PM CaptHindsight[m]: LCNC 2.9pre on Bullseye, nope
11:09 PM CaptHindsight[m]: no custom hal
11:10 PM CaptHindsight[m]: keeping it simple to try and fine why the signal 8 error
11:10 PM CaptHindsight[m]: fine/find
11:10 PM skunkworks: your simple.ini and hal files point to
11:11 PM skunkworks: HALFILE = my-mill.hal
11:11 PM skunkworks: HALFILE = custom.hal
11:11 PM skunkworks: POSTGUI_HALFILE = postgui_call_list.hal
11:12 PM CaptHindsight[m]: # Include your custom_postgui HAL commands here
11:12 PM CaptHindsight[m]: # This file will not be overwritten when you run stepconf again
11:12 PM CaptHindsight[m]: heh for both HALFILE = custom.hal
11:12 PM CaptHindsight[m]: POSTGUI_HALFILE = postgui_call_list.hal
11:13 PM * CaptHindsight[m] posted a file: (0KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/wzMzFaOhorKthVuRLghWIpcr/postgui_call_list.hal >
11:13 PM CaptHindsight[m]: oops
11:13 PM skunkworks: lol
11:13 PM CaptHindsight[m]: what a mess it made
11:13 PM * CaptHindsight[m] posted a file: (1KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/VAVTEMzizFcRCqvzcDdtNFKF/pyvcp_options.hal >
11:14 PM CaptHindsight[m]: what the hell is stepconf doing?
11:16 PM skunkworks: lol - I have only used it about 2 times
11:16 PM skunkworks: need custom_postgui.hal also
11:17 PM * CaptHindsight[m] posted a file: (0KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/DHooFnAYeBJyqPipcGJCKZuJ/custom_postgui.hal >
11:17 PM CaptHindsight[m]: i just wanted to test a simple config since I got the error
11:17 PM CaptHindsight[m]: also empty file
11:17 PM skunkworks: ok
11:17 PM CaptHindsight[m]: I have to make one myself :)
11:18 PM skunkworks: hey - it boot
11:18 PM skunkworks: heh - I get that error!
11:18 PM CaptHindsight[m]: ah ok good
11:20 PM CaptHindsight[m]: the only message on the forums about this error is if you have duplicate entries in HAL
11:20 PM CaptHindsight[m]: might be the case with how stepconf created all those other files
11:21 PM skunkworks: I cannot get it to do it in 2.8.2
11:22 PM CaptHindsight[m]: ah ok
11:22 PM CaptHindsight[m]: was going to try 2.8 later
11:23 PM skunkworks: happens pretty darn quick in 2.9
11:23 PM CaptHindsight[m]: yeah and I need 2.9 for QTPYVCP and all Python 3
11:23 PM CaptHindsight[m]: so more fun
11:27 PM skunkworks: weird - I am running 2.9 on the little emco at the shop.. but that is running mesa hardware..
11:28 PM skunkworks: hmm - now try to remember how to do a git bisect..
11:28 PM skunkworks: maybe another day
11:28 PM skunkworks: CaptHindsight[m]: it isn't just you ;)
11:29 PM CaptHindsight[m]: 😢
11:30 PM CaptHindsight[m]: I'll try a new config later with the encoder not as a spindle but for an axis
11:32 PM CaptHindsight[m]: I plug things in all the time just to test
11:32 PM CaptHindsight[m]: and I use stepconf without issue
11:33 PM CaptHindsight[m]: this PC has 3 LPT ports and a 6i25
11:34 PM CaptHindsight[m]: have room for 2 more 6i25's
11:34 PM skunkworks: lol
11:35 PM CaptHindsight[m]: skunkworks: thanks for saving me some time
11:55 PM miss0r: morning. I'm machining up two heat sinks for my ridiculous dual pentium pro dream PC :) I am just wondering: Is it possible to do some sort of surface treatment to the copper to have it not fade over time? What is this process called?
11:55 PM CaptHindsight[m]: miss0r: paint?
11:56 PM miss0r: Maybe, But I fear that would hurt the heat transfer to air property
11:56 PM CaptHindsight[m]: very thin coat
11:56 PM XXCoder: thin nickel coat would work I think
11:56 PM miss0r: Is that realy the only way?
11:57 PM miss0r: XXCoder: Maybe yes. But I must admit I like the copper look :)
11:57 PM CaptHindsight[m]: or plating yeah
11:57 PM XXCoder: https://itigic.com/why-is-nickel-used-in-heatsinks/
11:57 PM miss0r: It just seems to me that I remember owning a heatsink back in the day that was copper coloured
11:57 PM XXCoder: actually alum can be used too. its second best compared to copper
11:57 PM CaptHindsight[m]: probably had a thin coat of clear on it
11:57 PM XXCoder: (well under silver but more money)
11:58 PM CaptHindsight[m]: aluminum plating of copper is no cheap either
11:58 PM CaptHindsight[m]: we have some pennies we aluminum plated and then anodized
11:58 PM XXCoder: that site says nickel hardly affects perforance
11:58 PM miss0r: I wonder how long it will stay shiny in a dry indoor environment without fading untreated
11:58 PM CaptHindsight[m]: a very thin plate or coat of anything
11:59 PM CaptHindsight[m]: the bulk is copper and that is what spreads the heat into the air
11:59 PM miss0r: hmm. Yeah. I guess plating is the way to go. I will see if I can find someone who does that
11:59 PM XXCoder: wonder what coating capt menioned
11:59 PM miss0r: paint