#linuxcnc Logs

Sep 16 2020

12:48 AM Deejay: moin
03:20 AM Loetmichel: mornin'
03:23 AM Loetmichel: *gnah* another 200€ down the drain ordering filament for the company. Those four Ender3 printers eat that Filament like its free of charge. :)
04:33 AM solobolo2 is now known as htasta
04:56 AM JT-Cave: morning
04:58 AM XXCoder: yo
06:06 AM Tom_L: morning
06:07 AM XXCoder: yo
06:50 AM jymmmm: morning
07:06 AM jymmmm: JT what did you you find that silicone did NOT adhere to?
07:17 AM -!- #linuxcnc mode set to +v by ChanServ
07:24 AM jymmmm: Donny Darko - For whom the bell tolls (1hr mix) https://www.youtube.com/watch?v=mI0Q-4c97yI
07:25 AM XXCoder: ask not who it tolls for, it tolls for thee
07:26 AM jymmmm: I love that song, would play it on repeat once in a while, now that I found/converted/downloaded this 1hr mix, don't have to deal with that
07:36 AM jymmmm: Huh, the first time I've ever seen a rating listed as "Do not exceed 4A/mm2"
07:40 AM jymmmm: Does this sound right? 12ga to mm2 = 53.4751
07:41 AM jymmmm: 12ga = 8.2515mm diameter
07:45 AM jymmmm: The calculatr says 53, scroll down to the chart and it says 3.3, now I'm confused https://www.rapidtables.com/calc/wire/awg-to-mm.html
07:57 AM rs[m]2: AWG12 = 2mm dia
07:58 AM rs[m]2: 3.3mm²
07:59 AM rs[m]2: calculator and table on that website obviously agree on that, maybe you mistyped something. 53mm² would be HUGE
08:02 AM jymmmm: rs[m]2: select 12ga, then click n calculate next to the mm2 option
08:03 AM rs[m]2: jymm: enter 12 and press calculate next to AWG
08:04 AM jymmmm: ok, that gave me 3.3
08:05 AM rs[m]2: jymmmm: really not easy to distinguish between jymm, jymmm, jymmmm and jymmmmmmmmmmmm
08:05 AM jymmmm: but where it saw AWG, click on that, select 12, then click on the calculatte next to the mm2 dialog
08:05 AM jymmmm: rs[m]2: it's okey, they are all me
08:06 AM rs[m]2: Jymmm: enter 4 mm² and press calculate, it will show AWG nearest to 4mm²
08:06 AM jymmmm: rs yes, I did that, did you do what I explained?
08:06 AM rs[m]2: 53mm² = AWG 0
08:07 AM rs[m]2: this thing clearly has bugs
08:08 AM jymmmm: ok, just wanted to make sure, ty
08:08 AM rs[m]2: when in doubt i look here https://en.wikipedia.org/wiki/American_wire_gauge
08:09 AM jymmmm: so 3.3 * 4A = 13.2 amps for 12ga ??? That seems low
08:11 AM rs[m]2: it all depends on circumstances, e.g. material of wire, cooling, single wire or bundle, free air, moving, insulated, oxidized/clean, ...
08:12 AM rs[m]2: around here, 2,5mm² copper wire in wall is considered enough to carry 16A permanently.
08:14 AM jymmmm: This is a solar controller rated at 40A DC. Specs call for no more than 4A/mm2,
08:14 AM rs[m]2: you mean connection between controller and solar panels?
08:15 AM jymmmm: controller and battery
08:15 AM rs[m]2: is it a problem to use something like 10mm² wire?
08:16 AM rs[m]2: (AWG7 according to table)
08:16 AM rs[m]2: at 40A, even 10mm² wire will get warm
08:18 AM jymmmm: 7ga is pretty much non existant in the US, but 6ga (13.3mm) is available. I just read the specs calling for 10mm to the battery, argh
08:18 AM jymmmm: Now I have to call them and ask if the terminals will accept 6ga
08:18 AM rs[m]2: mm²
08:18 AM jymmmm: right square
08:19 AM jymmmm: right, square
08:19 AM rs[m]2: sorry ;)
08:19 AM jymmmm: no worries
08:19 AM jymmmm: on a lot of solar controllers that have/use cheap, undersized terminals sadly
08:20 AM rs[m]2: you could "thin out" a few strands at the ends to make 6gauge fit into 7gauge
08:20 AM rs[m]2: terminal
08:21 AM rs[m]2: or use 2 10gauge in parallel
08:25 AM jymmmm: two 10ga would be 8ga equiv I thought, guess I could always use 3 instead
08:27 AM jymmmm: Oh, I'll be damn, 2 10ga is 7ga, haha
08:27 AM rs[m]2: two 10ga (5.26mm²) is equiv to 7ga (10.5mm²)
08:27 AM jymmmm: https://www.wirebarn.com/Combined-Wire-Gauge-Calculator_ep_42.html
08:28 AM jymmmm: rs[m]2: awesome, tyvm
08:28 AM rs[m]2: "circular mil" is a funny unit, equivalent to the area of a circle of 1/1000 inch diameter
08:29 AM jymmmm: it's the metric that confuses me =)
08:29 AM jymmmm: then have to convert to awg
08:30 AM rs[m]2: area is area, behaves exactly the same however measured
08:30 AM jymmmm: True, but comprehesion and familiarity is another matter all it's own =)
08:30 AM rs[m]2: my favorite unit of area is the "barn", equivalent to 10^-28m²
08:31 AM jymmmm: lol
08:32 AM rs[m]2: approximately the cross section of heavy nuclei
08:33 AM jymmmm: Hmmm, 100A on a cable rated at 106A, pushing my luck?
08:35 AM rs[m]2: you have to pay attention to the connectors. cable is probably fine if it is used and loaded according to spec
08:35 AM rs[m]2: some kind of thermal imaging device is useful to check connections under load
08:36 AM jymmmm: Oh boy, thermal cameras are pricy
08:36 AM jymmmm: so for a safety margin, what, 5% 10% ?
08:38 AM rs[m]2: the cable will be fine, but check what condition it is rated for. putting cable into a conduit or burying it is different from hanging in free air
08:38 AM jymmmm: This is DC in a portable enclosure, no conduit
08:39 AM jymmmm: with active cooling (fans)
08:40 AM rs[m]2: it all depends on the cooling. surface area of the wire. more surface area -> better cooling.
08:41 AM jymmmm: I'm more prone to use parallel conductors, so more surfce area in that respect.
08:43 AM rs[m]2: unfortunately heat transfer problems like this are very hard to calculate, so usually you go look it up in some code produced at midnight with full moon on a friday the 13th in a gap year by a standards body
08:43 AM jymmmm: haha
08:48 AM jymmmm: Ok, if contonous rating is 1000W (84A), and surge is 2000W (168A) momentarily, would 26.7mm2 (3ga) be good enough?
08:49 AM jymmmm: I get 106.8A at 3ga
08:49 AM jymmmm: or is that all based on thermal disipation at that point?
08:50 AM rs[m]2: according to wikipedia, single 3ga wire @ 85Amps in free air @20°C will get 60°C hot
08:51 AM rs[m]2: smaller wire -> higher resistance -> more heating -> even more resistance
08:51 AM rs[m]2: higher temperature -> more convection and more radiation (T^4)
08:52 AM rs[m]2: somewhere an equilibrium forms (assuming everything else is "steady state"), and that should be somewhere safe
08:56 AM jymmmm: Looks like I'm looking at 1/0 instead
08:57 AM jymmmm: rs[m]2: are those wikipedia temperatures linear?
08:58 AM rs[m]2: what do you mean?
08:58 AM jymmmm: Like 42A ~= 30C ???
08:58 AM jymmmm: half the load, half the temperature
08:59 AM rs[m]2: don't think so
09:00 AM rs[m]2: what is "momentarily" in your context? a few seconds? or half an hour?
09:04 AM jymmmm: For the surge? Like a motor starting (frig/freezer), or a microwave oven
09:06 AM jymmmm: or even an air compressor starting
09:07 AM jymmmm: assuming a compressor would normally run on 1000W
09:08 AM rs[m]2: so your surge is seconds at most
09:09 AM rs[m]2: but your numbers keep increasing, an hour ago you were talking about 40A, now it's 180
09:16 AM -!- #linuxcnc mode set to +v by ChanServ
09:57 AM Tom_L: Jymmm, air compressor has start capacitors to help with that though
09:59 AM Eric__: is there any hope that wine will run 16 bit programs on the 2.8 iso version of debian?
09:59 AM Eric__: would be nice to be able to run the setup program for my drives using that
10:34 AM * jymmmm is back (this is your only warning)
10:35 AM jymmmm: rs[m]2: 40A is the max charge rate. 1000W is the cont load rate, 2000W is the surge load rate.
10:35 AM jymmmm: rs[m]2: Sorry for the confusion
10:39 AM jymmmm: rs[m]2: I MIGHT go with a 24V system instead, it really depends on what inverter I use
10:40 AM jymmmm: rs[m]2: But 12V is a bit more "universal"
10:46 AM CaptHindsight: jymmmm: fresh silicone sticks poorly to cured silicone
10:47 AM jymmmm: CaptHindsight: interesting
10:50 AM CaptHindsight: how long will the power supply Surge or run at Peak?
10:52 AM CaptHindsight: copper with especially with an insulator will heat faster than it cools
10:52 AM CaptHindsight: 2x especially for emphasis
10:53 AM jymmmm: CaptHindsight: Sssurge would be like a frig/freezer, microwave, or maybe a small compressor
10:53 AM CaptHindsight: so several minutes
10:53 AM jymmmm: The initial inrush of motor startup basically
10:54 AM CaptHindsight: ok make up your mind
10:54 AM jymmmm: I did, It's surge, not contonous
10:55 AM CaptHindsight: a microwave is run for seconds to minutes
10:55 AM jymmmm: Yes, but it's the initial startup surge
10:56 AM jymmmm: If I have a 700W microwave, the inrush might be 1200W for a few seconds
10:56 AM CaptHindsight: how long arevthe conductors in question?
10:57 AM jymmmm: Less than 2ft
11:00 AM CaptHindsight: this in a box with cooling fans that move are inside the box, through the box or move air from where to where?
11:00 AM CaptHindsight: have a pic?
11:01 AM jymmmm: No pic, 2 fans, one on each end of the box, both flowing air the same direction.
11:02 AM CaptHindsight: moves air through the box from outside through back the outside?
11:03 AM jymmmm: yes
11:05 AM CaptHindsight: are you willing to listen to math and science or do you already have some with that you are trying to use for this?
11:05 AM CaptHindsight: some wire that
11:05 AM jymmmm: CaptHindsight: I am open
11:05 AM CaptHindsight: have to ask, strange days
11:40 AM Eric__: fusion now limits you to 10 active files
11:40 AM Eric__: haven't run into that yet though
11:44 AM Tom_L: tiny assemblies
11:47 AM gloops: freecad
11:52 AM Eric__: wish I could afford solidworks
11:52 AM Tom_L: we use math and science here??
11:52 AM Tom_L: Eric__, you'd still need a cam for it
12:00 PM jymmmm: Tom_L: I always thought it was voodoo and hogwash
12:02 PM roycroft: soldworks have a cam add-on
12:02 PM roycroft: solidworks, rather
12:06 PM jymmmm: CaptHindsight: Did you have some science you wanted to share with the class by chance?
12:31 PM jymmmm: How in the heck to you get rid of stuff like this? There will be that ONE TIME when it'll be perfect for a job/project... https://i.imgur.com/4zlKetn.jpg
12:40 PM Eric__: wow, free version of fusion is really limited. Cam is very limited
12:40 PM Eric__: have to go look to see if I want to save any files
12:41 PM Eric__: end of october, I think
01:06 PM snaked: By the end of the month, it would be better thought of us a demo.
01:27 PM methods_: jymmmm and his nipple clamps
01:27 PM robotustra: got new toy
01:28 PM robotustra: center finder
01:28 PM robotustra: I'm just uber happy
01:28 PM robotustra: https://i.imgur.com/DI2l2DD.jpg
01:33 PM FinboySlick: https://1n0g9sqmedx94.wpcdn.shift8cdn.com/wp-content/uploads/2020/09/117238377_3256349451100224_5914997655199445416_n.jpg I'm mildly curious of how something like that could happen.
01:34 PM jymmmm: FinboySlick: Looks like it's in a junkyard, so maybe someone just tossed a clutch plat in the air?
01:38 PM FinboySlick: jymmm: The way things are placed, it looks more like it came from below.
01:38 PM jymmmm: FinboySlick: IDK, but the side mirrir is pointing backwards too
01:39 PM FinboySlick: Which side mirror?
01:40 PM jymmmm: I closed the link, but the drivers side, and I think there was only one side mirror attached to that blue car
01:42 PM FinboySlick: That was the door itself.
01:42 PM FinboySlick: It's an old car, look like those old V8 Fords.
01:44 PM CaptHindsight: it's usually parts of the bellhousing, flywheel as well
01:45 PM FinboySlick: CaptHindsight: One of the clutch plates caught something while spinning too fast?
01:46 PM CaptHindsight: drag racing rules require an explosion proof bellhousing or a blanket placed over it
01:51 PM CaptHindsight: the transmissions input shaft would have to of let go for the clutch disc to travel out
01:52 PM CaptHindsight: https://www.youtube.com/watch?v=knhgiA0V7ug
01:55 PM CaptHindsight: https://www.youtube.com/watch?v=1V6nBdiK118
01:57 PM CaptHindsight: that is not the side mirror, that is the reflection off the outside of the drivers door
01:58 PM CaptHindsight: fooled me at first
01:58 PM CaptHindsight: the perspective makes the door appear to be very short
01:59 PM robotustra: I didn't know that clutch can explode like this
01:59 PM CaptHindsight: usually not a problem <5k rpm
01:59 PM FinboySlick: I guess if you spin pretty much anything fast enough, it'll explode.
02:00 PM robotustra: spin something unbalanced
02:01 PM FinboySlick: Even balanced. Materials can only sustain so much tension.
02:01 PM CaptHindsight: stock flywheel at 6-9k rpm
02:01 PM robotustra: and I guess here the problem if the clutch is touching the surface unevenly
02:01 PM robotustra: or too fast
02:02 PM CaptHindsight: stock pressure plate housings might even been stamped
02:03 PM robotustra: there is an istruction how to get this efffect 1) start engine 2) disingage the clutch 3) get 6000 rpm 4) drop the clutch :)
02:03 PM robotustra: 3.5) shift the gear
02:04 PM CaptHindsight: what we used to call driving just about anywhere, then the tickets add up and you stop
02:05 PM CaptHindsight: go though too many tires
02:07 PM CaptHindsight: FinboySlick: have anymore pics of the same car?
02:11 PM CaptHindsight: https://cdn.sau.com.au/post-4657-1220271519.jpg
02:11 PM CaptHindsight: this one went down vs up https://cimg5.ibsrv.net/gimg/ls1tech.com-vbulletin/2000x1504/4539e250_2527_45a3_9722_3bbc5b9b94be_d7cadc646fe8e61be401b9a352c37406cc48552b.jpeg
02:17 PM FinboySlick: CaptHindsight: sadly, no. It's off some sort of image board.
03:04 PM Deejay: gn8
03:14 PM lostinsip: channels
03:28 PM jymmmm: CaptHindsight: Was there some math/science you were going to share by chance?
03:37 PM Tom_L: robotustra, that's what these are for: http://www.fowlerprecision.com/Products/Coax-Indicator/527100250.html
03:48 PM CaptHindsight: jymmmm: fell asleep
03:48 PM jymmmm: CaptHindsight: good morning?
03:49 PM Tom_L: it's all wizzardry
03:49 PM Tom_L: math is just so.... old school
03:49 PM jymmmm: bbiab, heading to the store
03:51 PM rs[m]2: with 20/20 hindsight you don't need math
03:52 PM CaptHindsight: jymmmm: all depends on how long you hold the surge current for
03:52 PM CaptHindsight: the longer the time the larger the wire should be
03:52 PM CaptHindsight: higher temp insulation is the other option if available
03:53 PM rs[m]2: 12V and 80A / 180A surge is wrong
03:54 PM CaptHindsight: better to have him post all his specs vs play 50 questions
03:55 PM rs[m]2: i think he is trying to build a solar charger / inverter system with 12v battery and 1kW continuous power output
04:01 PM CaptHindsight: rs[m]2: yes he is
04:39 PM jymmmm: back
04:39 PM jymmmm: and inrush surge is what I'm trying to account for, so that would be a few seconds
04:40 PM jymmmm: under 30 seconds I'd imagine
04:41 PM jymmmm: Si if contonous is 1000W, how much would I overrate the cabling for surge? 10%? 20%?
04:43 PM rs[m]2: Jymmm: a cable that is rated for whatever continuous current will take 10times that current for a second every 5 minutes
04:43 PM rs[m]2: depending on cooling ;)
04:45 PM CaptHindsight: jymmmm: it's somewhere between the wire size for 80A and 180A
04:46 PM CaptHindsight: the closer you are to the wire size for 180A the longer you will be able to run it at 180A
04:46 PM jymmmm: rs[m]2: Ok, that's perfect. I'll have active cooling for the enclosure as well.
04:46 PM CaptHindsight: yeah but that doesn't really help much without any numbers
04:47 PM CaptHindsight: 2 x 1" fans at 10LFM vs 2 8" fans that more air at 500LFM
04:48 PM CaptHindsight: and you're building one of these so just use larger cable
04:48 PM Tom_L: stranded better than solid?
04:48 PM Tom_L: like the fine strands in a welding cable...
04:49 PM CaptHindsight: is the wire flexed during its life? if yes then stranded
04:50 PM Tom_L: i didn't know if the surface area mattered
04:50 PM CaptHindsight: the cross section of the conductors will be the same
04:52 PM rs[m]2: surface area of conductor matters with high frequency (skin effect). surface area over all matters because of cooling.
05:01 PM jymmmm: Xantrex says for their 1000W/1500W, minimum 4ga and 175A fuse. Which is 83A / 4A per sqmm comes to 20.9, and 4ga is 21.1 sqmm
05:04 PM jymmmm: Tom_L: I thought solid > stranded, but not the case. stranded better in mobile applications to prevent brakage
05:04 PM jymmmm: stranded can be flexiable enough to absorb vibrations.
05:13 PM jymmmm: Ok, not too bad for 4ga... https://www.amazon.com/Battery-Welding-Flexible-Terminal-Connectors/dp/B07K4WY8XF/
05:25 PM robotustra: Tom_L, how much is that thing?
05:59 PM Tom_L: you can get cheaper ones but that's probably one of the best
06:00 PM Tom_L: https://www.amazon.com/coaxial-indicator/s?k=coaxial+indicator
06:00 PM Tom_L: that one is $250 us
06:00 PM Tom_L: 260
06:00 PM Tom_L: you can get a cheap one for $75ish
06:01 PM XXCoder: i wonder how good cheap one is
06:01 PM Tom_L: some of the guys at my kids work have cheaper ones
06:01 PM Tom_L: i got a fowler
06:02 PM Tom_L: he said they're not as heavy
06:02 PM roycroft: fowler is the "cheapest" coaxial indicator i'd consider
06:02 PM roycroft: and fowler isn't really cheap
06:02 PM roycroft: it's just not really expensive
06:04 PM Tom_L: i dunno if starret makes one but i wouldn't buy theirs anyway
06:04 PM Tom_L: it could get damaged fairly easy if you're not careful
06:05 PM Tom_L: you can hold the body still while the rest spins (slowly) if you can't figure it out
06:06 PM Tom_L: they don't come in a nice wood box anymore either
06:07 PM Tom_L: padded plastic
06:08 PM Tom_L: they work on ID and OD both
06:12 PM skunkworks: http://electronicsam.com/images/greenmachine/IMG_20200916_180712.jpg
06:16 PM Tom_L: nice one
06:20 PM Tom_L: ok skunkworks, http://tom-itx.no-ip.biz:81/~webpage/cnc/rotary/RotaryTest3.jpg
06:20 PM Tom_L: :) i cheated
06:20 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/rotary/RotaryTest2.jpg
06:21 PM skunkworks: that is the easy way to do it... I like doing it hard.
06:21 PM skunkworks: ;)
06:38 PM Tom_L: skunkworks, the finish looks better on that one
06:38 PM robotustra: Tom_L, mine is less than 100$
06:39 PM Tom_L: you can get those for that too
06:39 PM Tom_L: 70 - 75ish
06:39 PM robotustra: no, you can't
06:39 PM robotustra: 75$ usd = 100$ cad + shipment + tax
06:40 PM robotustra: it means at least 140
06:40 PM robotustra: cad
06:44 PM CaptHindsight: anybody know the status of the Rpi SD card image for 2.8?
06:45 PM CaptHindsight: is it working or available but still needs testing?
06:45 PM Tom_L: should be workable
06:45 PM Tom_L: i used jeplers and he said it handn't changed much if at all
06:45 PM CaptHindsight: there was some discussion on the ML about the menus for 2.8
06:45 PM Tom_L: idk
06:46 PM Tom_L: http://linuxcnc.org/iso/
06:46 PM Tom_L: that's the latest zip i'm aware of
06:46 PM CaptHindsight: next question is qtpyvcp on the Rpi
06:46 PM Tom_L: not by default
06:46 PM CaptHindsight: will I be blazing that trail?
06:46 PM Tom_L: i bet jt would know if that's possible
06:48 PM Tom_L: https://gnipsel.com/linuxcnc/uspace/index.html
06:48 PM Tom_L: i don't see any example there
06:49 PM Tom_L: you _could_ ask in #hazzy
06:49 PM CaptHindsight: we have a 16 or 32 core Ryzen ripper for running ARM in QEMU so building for Rpi is really fast
06:49 PM CaptHindsight: yeah already asked there
06:49 PM Tom_L: blaze on
06:50 PM Tom_L: https://www.raspberrypi.org/forums/viewtopic.php?t=257914
06:52 PM Tom_L: that might get you partway there
06:54 PM CaptHindsight: "it works but the 3d plotter has a dependency that needs to be installed from a deb"
06:55 PM Tom_L: http://www.qtpyvcp.com/install/index.html
06:55 PM jdh: I have a 3 axis table with 3 steppers & drives. I'd like some trivial way to move it via something like MDI, but without having to install linux, or mach. anyone know of any software that would do this?
06:55 PM Tom_L: i haven't done much with it
06:55 PM skunkworks: Tom_L: that was single point turned in passes - not plugned at once
06:56 PM Tom_L: jdh, grbl
06:56 PM jdh: it has a UI?
06:56 PM Tom_L: i dunno
06:56 PM Tom_L: there might be something out there for it
06:57 PM Tom_L: https://www.google.com/search?client=firefox-b-1-d&q=best+gui+for+grbl
06:57 PM jdh: taht might work. thanks
06:58 PM Tom_L: it's pretty basic
06:58 PM Tom_L: no canned cycles etc and alot of missing codes
06:59 PM jdh: I just want to manually move an object around under a camera for testing.
07:00 PM jdh: scaled movement would be cool
07:00 PM veegee: Instead of getting toe jacks, is there anything wrong with bolting on some steel to the side of the machine to act as a jack point?
07:00 PM veegee: obviously should be strong enough
07:02 PM veegee: With over sized bolts
07:03 PM CaptHindsight: Tom_L: yeah that install howto is almost correct, I'm going to do a fresh install using 2.8 and see what the real howto is
07:06 PM veegee: Same with forklift. I'm going to make a a little bracket that bolts onto the side so I can jack it up easier
07:06 PM veegee: Previous method was lifting an overweight object so the rear tilts up so I could access the oil drain
07:07 PM veegee: Shouldn't be too hard to bolt a bracket directly to the counterweight
07:08 PM veegee: made of very thick steel and very strong bolts
07:08 PM Tom_L: CaptHindsight, the pi et friends may not have enough poop for those graphics?
07:12 PM veegee: CaptHindsight and these are the only machine rollers I've found that actually work well: https://www.amazon.ca/Vestil-VHMS-15-Swivel-Machine-Capacity/dp/B0052PWAOM
07:13 PM veegee: nothing else seems to work and always flexes under load, making it very difficult to move
07:13 PM veegee: The ones I linked work so well, I bought 4 more. I have 8 now
07:13 PM veegee: Also they look like mini battle tanks
07:18 PM CaptHindsight: add some motors :)
07:22 PM veegee: Yeah that's what I was thinking too, but can't figure out a good place to add a motor and pneumatic tire for traction
07:22 PM veegee: Air bearing casters would be awesome, I should try to make one at some point
07:23 PM CaptHindsight: Tom_L: my Rpi4 runs at 2GHz
07:24 PM Tom_L: i mean, i've got 2.8 on one and tested it on my mill
07:25 PM Tom_L: 2.8 pre
07:27 PM CaptHindsight: I'll ask JT how far he got, if he even tried qtpyvcp on his Rpi
07:28 PM Tom_L: iirc he may have a front end of it on his coup
07:29 PM jdh: can you do pport step/dir with a Rpi? without additional hardware?
07:29 PM Tom_L: i dunno
07:29 PM Tom_L: not that i'm aware of
07:29 PM Tom_L: mesa has a board that it fits on and you can use spi or ethernet
07:29 PM Tom_L: i tested both the latter
07:31 PM Tom_L: jdh, the 7i90 will do spi, 7i80 ethernet
07:31 PM Tom_L: cheap boards
07:33 PM veegee: NICE I like this design: https://www.globalindustrial.com/p/material-handling/forklifts-attachments/jacks/global-industrial-forkli-jack-8800-lb-capacity
07:33 PM Tom_L: jdh, you'd have to use the gpio plug as it doesn't have a parallel port
07:33 PM veegee: time to DIY a low profile 20 ton jack
07:34 PM veegee: Also what kind of retarded forklift jack is rated for only 8,800 lbs?
07:41 PM jymmmm: veegee: you want to move, by hand, something that weights more than 9000 lbs?
07:43 PM veegee: jymmmm ideally by finger, but hand if I /really/ have to
07:43 PM jymmmm: most ppl would just use a forklift
08:21 PM CaptHindsight: jdh: you can but just not very fast
08:22 PM CaptHindsight: IIRC the Rpi4 latency jitter is 50,000 - 70,000ns with preempt_rt
08:26 PM CaptHindsight: https://www.forum.linuxcnc.org/media/kunena/attachments/27529/frederick-isolcpus.png
08:27 PM CaptHindsight: https://www.forum.linuxcnc.org/38-general-linuxcnc-questions/39663-rpi4-servo-lateceny-spikes
08:49 PM Eric__: if I want to reduce the latency of my i5 with rt prempt kernel, what kernel settings should I use?
08:49 PM Eric__: didn't we determine that isolcpus doesn't do anything?
08:50 PM Tom_L: turn off hyperthreading
08:50 PM Tom_L: etc
08:54 PM Tom_L: i'm not sure about kernel but bios ^^
08:57 PM Tom_L: https://taste.tuxfamily.org/wiki/index.php?title=Tricks_and_tools_for_PREEMPT-RT_kernel
08:57 PM Tom_L: not sure if this applies
09:04 PM zipper is now known as Guest78729
09:04 PM roycroft: holy moly
09:04 PM roycroft: the aqi is down to 31.1 by my house
09:05 PM Tom_L: better bottle some up
09:05 PM roycroft: i sure hope this is a sign that this chapter in our nightmare is coming to an end
09:06 PM roycroft: it's down to 4.1 on the coast
09:07 PM roycroft: now if we would just get some rain
09:08 PM roycroft: it looks like there's a good chance of that late tomorrow and on friday
10:02 PM skunkworks: https://www.youtube.com/watch?v=Q9zhi8Mc1fE
10:06 PM -!- #linuxcnc mode set to +v by ChanServ
10:06 PM cradek: that's so cool
10:07 PM cradek: quite the relief angles on that tool!
10:07 PM skunkworks: yes - requred...
10:07 PM skunkworks: I think it is magic
10:07 PM cradek: that's a really responsive X axis
10:08 PM skunkworks: it is iirc 50in/sec^2
10:08 PM skunkworks: small servos
10:08 PM cradek: awesome
10:09 PM skunkworks: should try 3 sided... ;)
10:11 PM skunkworks: I think we all bought bunch of them..
10:11 PM skunkworks: (servos)
10:11 PM skunkworks: or maybe it was jepler?
10:12 PM cradek: I think he did have some
10:12 PM cradek: he and I put them on a little gantry machine that he had - I forget who has it now, someone else does
10:13 PM cradek: I had some I gave to moses a long time ago for some development project
10:14 PM cradek: it's funny, I use cnc for my coil winding, but pretty much all lathe work I do now is by handheld tools
10:14 PM skunkworks: heh
10:14 PM skunkworks: like wood working?
10:14 PM cradek: heh yes but metal and tiny
10:14 PM skunkworks: I still have to send you the watches...
10:14 PM skunkworks: ugh
10:15 PM skunkworks: mine stopped again