#linuxcnc Logs

Apr 26 2019

#linuxcnc Calendar

12:33 AM pink_vampire: how are those bearings?
12:37 AM pink_vampire: https://www.ebay.com/itm/302853475603
02:48 AM Loetmichel: Jymmm: because no coil dattached
02:48 AM Loetmichel: -d
02:48 AM Loetmichel: its the maximum it can measure
03:00 AM Loetmichel: mornin'
03:19 AM Loetmichel: btw: got that sheet alu finished yesterday... only 2 hours overtime... looks like that will even work... (Dual all in one thin client with 24" monitor in front) MAAAN that will be a fat box (and i hope those cooling hexagons are sufficient for 2 times 30W)... -> http://www.cyrom.org/palbum/main.php?g2_itemId=17358&g2_imageViewsIndex=1
04:18 AM XXCoder: pink_vampire: dunno but cheap enough to buy and test i guess?
04:33 AM pink_vampire: XXCoder: ??
04:33 AM XXCoder: the bearing you asked about
04:34 AM pink_vampire: oh i see
04:34 AM pink_vampire: yeah
04:34 AM XXCoder: 4 cm in size?
04:34 AM XXCoder: pretty large lol
04:34 AM XXCoder: I have 2 bearings from work, they was 50 bucks and 200 bucks lol
04:34 AM pink_vampire: also find this one https://www.fastenal.com/products/details/0474380
04:35 AM XXCoder: those bearings were part of spindle, and shot.
04:35 AM pink_vampire: this is just for the gear in the milling head
04:35 AM XXCoder: can also make pretty big fidget spinner lol
04:36 AM XXCoder: i may make novelity giant spinner with my big bearing
04:36 AM XXCoder: it may not work for spindle, but it spins well enough
04:36 AM pink_vampire: now the bearing make very grinding noise
04:36 AM pink_vampire: and i move it by hand
04:37 AM XXCoder: sounds like shot i guess
04:38 AM pink_vampire: yeah
04:50 AM pink_vampire: i need to find a capacitor for the milling head
04:50 AM pink_vampire: no idea what is the value
04:50 AM XXCoder: have a old obe?
04:51 AM pink_vampire: http://vi.vipr.ebaydesc.com/ws/eBayISAPI.dll?ViewItemDescV4&item=390618796315&t=1368219078000&tid=10&seller=emco_hobby_shoppe&excSoj=1&excTrk=1&lsite=0&ittenable=false&domain=ebay.com&descgauge=1&cspheader=1&oneClk=1&secureDesc=0
04:52 AM jthornton: morning
04:52 AM XXCoder: hey
04:53 AM pink_vampire: hi jthornton
04:53 AM XXCoder: 20 uF +- 10%, 330v
04:56 AM pink_vampire: huge capacitor
04:56 AM XXCoder: https://www.repairclinic.com/Shop-For-Parts/b129c12/Bosch-Capacitor-Parts maybe
04:57 AM XXCoder: no
04:59 AM XXCoder: found it. https://www.ebay.co.uk/itm/CBB60-Wired-Capacitor-Start-Run-Motor-Generator-Compressor-/233187696309
04:59 AM XXCoder: they sell various uF range at 450ac
04:59 AM XXCoder: *v
04:59 AM XXCoder: though I cant recall if higher volts is acceptable as long as uF is same or not.
05:00 AM pink_vampire: no idea
05:07 AM XXCoder: apparently yes though not too far
05:07 AM XXCoder: 330v to 450v probably too far
05:07 AM XXCoder: https://www.ebay.co.uk/itm/ALUMINUM-ELECTROLYTIC-CAPACITOR-36-43UF-330V-20-QC-NWK-PN-PSU3630/352409847389?epid=16006703102&hash=item520d43865d:g:rW8AAOSwuOxZ15NF
05:08 AM pink_vampire: so the 450V will be ok?
05:08 AM XXCoder: problem is i dont know what it means by "20%:
05:08 AM XXCoder: ah! Capacitance Tolerance : 20%
05:08 AM XXCoder: i guess error range of 20%
05:09 AM XXCoder: i dont know if type of cap it is matters. it says electrolytic
05:09 AM pink_vampire: https://en.wikipedia.org/wiki/Motor_capacitor
05:10 AM XXCoder: so what its made from matters. it says electrolytic cant do constant cycle
05:11 AM XXCoder: spindle its probably that
05:12 AM Loetmichel: XXCoder: electrolyte CAn do constant cycle
05:13 AM XXCoder: really? wiki seems to say no
05:13 AM Loetmichel: just stay within the power ratings
05:13 AM Loetmichel: it doesent like to get cooked
05:13 AM Loetmichel: or did you mean electrolytics cant do AC?
05:13 AM XXCoder: no just constantly being used
05:14 AM XXCoder: maybe you can find the same part pink linked a picture to
05:14 AM pink_vampire: it is from 1972
05:14 AM Loetmichel: you can use electrolytics constantly. they just tent to start boiling so you should derate the power you send thru it so it stays cool
05:15 AM Loetmichel: an electrolytic at 80°C will live weeks, the same cap at 45°C will live half a century
05:15 AM XXCoder: beefy heat sinks ;)
05:15 AM Loetmichel: its actually so that old electrolytes that sat for decades without use tend to be damaged
05:16 AM pink_vampire: Loetmichel: this is the "original" http://www.maxton.com/ebay/emco/maximat7_milldrillcap_0420.jpg
05:16 AM pink_vampire: but I didn't got it with the machine
05:16 AM Loetmichel: you need to intriduce them to voltage quite slowly, the internal oxide layer that insulate them disappears without use
05:16 AM Loetmichel: pink_vampire: classic motor starter cap
05:17 AM pink_vampire: but it is rated for 330V
05:18 AM pink_vampire: do you think can use a capacitor with higher voltage rating? like 450V or 500V?
05:18 AM XXCoder: pink my later link is 330v 40uf same as yours, but 20% varancy rather than 10% of yours
05:19 AM XXCoder: which means you may have to buy a few and find one thats close to your orginial
05:21 AM Tom_L: morning
05:21 AM XXCoder: hey
05:45 AM Loetmichel: pink_vampire: please use a similar designated motor starter cap
05:45 AM Loetmichel: the design of those is quite different from a standard cap
05:46 AM Loetmichel: "self healing" and stuff
05:52 AM XXCoder: self healing ehj
06:03 AM sync: sure, the metal foil explodes away around the isolation fault
06:04 AM sync: so you loose a tiny bit of capacitance but you keep running
06:04 AM XXCoder: capactor heal thyself
06:53 AM Loetmichel: *hrrrrmph* just "repaired" an aluminium-swarf damaged VFD... or so i thought... replaced the blown FET, reassembled it, plugged spindle and mains in: booting... looks normal... pressed start button... *ffffFIIIIIp* half a turn and off again with "dl" in display. something more damaged. THAT will take a while now. Or i'll have to buy a new HY01D523B VFD. Luckily i can still read the parameters...
06:53 AM Loetmichel: never written them down the first time.
09:58 AM Loetmichel: *gnhihi* "bad codec" https://www.isnichwahr.de/r276986-schlechter-codec.html
10:24 AM Rab: Loetmichel, how was the VFD damaged by swarf?
10:27 AM roycroft: i don't know about loetmichel's situation, but many of the inexpensive chinese vfds have cooling fans that suck air over the electronics without a filter
10:27 AM roycroft: a little bit of swarf flying in the wrong direction could easily get sucked in and short something out
10:28 AM Rab: I figured something like that.
10:28 AM roycroft: i use those kinds of vfds, but i put filters on the air inlet
10:28 AM roycroft: cheap insurance
10:28 AM Rab: Or bridging the terminals, maybe.
10:28 AM roycroft: yes
10:29 AM roycroft: i filter the air flowing through the vfd on my router table, not because i'm concerned about shorts, but because the sawdust could coat the output transistors and cause them to overheat
10:30 AM Rab: Or absorb moisture if your shop's not well climate-controlled.
10:30 AM roycroft: that too
10:31 AM roycroft: i do think it's funny that some of the folks who poo-poo the huanyang vfds use the lack of air filtration as their excuse for hating on them
10:32 AM roycroft: because adding a filter is very cheap and very easy to do
10:32 AM roycroft: there are legitimate reasons for dislike huanyang, but lack of filtration is not one of them, imo
10:33 AM Rab: I think it's a good argument for putting all the electronics in a cabinet with good filtered ventilation, but of course there are plenty of applications for a standalone VFD where that's not reasonable.
10:33 AM roycroft: that said, my new belt grinder is still under construction, but i'm using it to finish building it
10:34 AM roycroft: and right now the vfd is just sitting on the bench next to the unit
10:34 AM roycroft: i should probably build a filtered enclosure for it very soon - i'm really taking a gamble using it like it is
10:36 AM roycroft: it will be ready for final cleanup and paint this weekend, though, so i'm almost done using it as a bootstrapping machine
10:36 AM roycroft: but i'll probably build a stand for it before i mount the vfd, and the stand will be welded steel
10:36 AM roycroft: so more grinding
10:36 AM jesseg: is huanyang the outfit that makes all those cheap VFD's that are on ebay?
10:36 AM roycroft: they are the "good" cheap ones that are on ebay
10:36 AM jesseg: ahh OK.
10:37 AM roycroft: there are huanyang clones all over ebay that are utter crap
10:37 AM jesseg: I think they have a firmware bug relating to controlling DC cap overvoltage during braking
10:37 AM jesseg: maybe that's the kind I had
10:42 AM roycroft: my memory is failing me on this, but something in the back of my mind tells me that there have been braking issues with the huanyang clones, but not with the genuine ones (other than user error)
10:43 AM roycroft: the biggest braking issue with genuine huanyang is not with the units themselves, but with vendors who are either unethical or ignorant
10:43 AM roycroft: the external braking resistor circuitry is not present in all huanyangs
10:43 AM jesseg: well, I was instant reversing with the mechanical reverser switch as a test to see if it could handle it. You might call that user error, but as an electronics design engineer who knows they *should* be able to handle it, to me, it was a valid test to evaluate the product.
10:44 AM roycroft: because it costs an extra 37 cents or so to add it
10:44 AM roycroft: but almost all the ebay vendors advertise the ones they sell as being capable of handling an external resistor, and most of the ones they sell are missing that circuitry
10:45 AM jesseg: but even without the resistor, the firmware *should* just let the motor freewheel the instant the DC cap bank voltage reaches the limit, thus preventing destruction.
10:45 AM roycroft: probably
10:46 AM roycroft: although iirc the manual discusses not setting the braking parameters too aggressively
10:46 AM roycroft: and warns of possible damage to the vfd if you do
10:46 AM jesseg: In free wheel mode, the motor can never produce a higher voltage than the power it runs on. Not unless you have a HUGE fly weight on your chuck and you quickly change your gearhead ratio to speed the motor drastically above full speed
10:47 AM roycroft: if i'm correct on that then the argument could be made that they are not designed to instant reverse like that
10:47 AM jesseg: well that's evidence of stupid firmware then, if the manual warns
10:47 AM roycroft: and that's not a bug, but a design decision to keep the cost down
10:47 AM jesseg: it's a few lines of code that once it's done, it's done. The hardware is already in there. There is no additional per unit cost.
10:47 AM roycroft: i've always felt a little uneasy about huanyang vfds
10:48 AM jesseg: They already monitor cap bank voltage and individual leg currents.
10:48 AM jesseg: You can go into the menus and read all those parameters.
10:48 AM roycroft: but at $150 for a vfd vs. $500+ for a high quality one, i grit my teeth and put up with the uneasy feeling
10:49 AM roycroft: that said, i'm already considering watching for a good used high quality vfd for my belt grinder
10:49 AM roycroft: and swapping out the huanyang as soon as i find one for a good price
10:50 AM roycroft: that grinder is so heavy-duty the vfd is pretty much the weak link
10:51 AM jesseg: In fact, the firmware did sometimes go into a freewheel state and throw an error code when I was doing instant reverses, so I know the code is already in there.. it's just buggy :D But like you said, for $150, I just got a bigger one and I try to not instant reverse, although it seems to handle use of the mechanical reverser switch if I just let it stop before starting the other way.
10:51 AM roycroft: i've been getting overspec huanyang vfds
10:51 AM roycroft: i.e with a 2hp load i use a 3hp vfd
10:51 AM jesseg: I did autopsy on the blown up unit, and was actually impressed with the PCB layout and stuff. They had opto isolated gate drivers and everything.
10:52 AM jesseg: yeah, my mill has a 3HP motor so when I blew up my 3HP vfd I found I could get the "7.5hp" unit for same cost LOL
10:52 AM jesseg: or was it 7.5KW?
10:52 AM roycroft: probably 7.5kw
10:52 AM jesseg: I don't remember but it was bigger both physically and parametrically
10:53 AM roycroft: one thing i've never sorted out is whether the load spec is with 3 phase input power or both 3 phase and single phase
10:53 AM roycroft: i don't have 3 phase power in my shop, so i have to feed single phase to the vfd
10:53 AM jesseg: yeah that has to cut the output power
10:53 AM roycroft: and i've read some discussion that they need to be oversize of being fed single phase power
10:54 AM roycroft: s/of/if/
10:54 AM jesseg: because they have a nice cap bank but it obviously sags between peaks, so the output voltage obviously has to sag too, and so does the power
10:54 AM roycroft: that makes sense
10:54 AM jesseg: and I'm sure they advertize it for optimal conditions then double it for ad copy :P
10:55 AM roycroft: i wasn't sure if that was the case, but the cost difference between a 1.5kw and 2.2kw was just a few dollars
10:55 AM roycroft: so to me it made sense to get bigger ones as an extra safety factor
10:56 AM roycroft: i have to say, btw, that i'm really impressed with my new grinder, and i'm really glad i put a 3hp motor on it
10:57 AM roycroft: i was grinding some 1.5" cold roll steel bar yesterday and the thing did not dog down at all
10:57 AM roycroft: i was able to push the bar until it started turning blue and the motor still did not slow down
10:58 AM roycroft: i just hope the vfd doesn't blow up soon
11:01 AM Connor: Someone help me out.. my maths isn't working right today. Old machine.. looks like maybe steps where set to 4064. Lead screw pitch is 1/2-20
11:01 AM Connor: I think the old controller was set up metric.
11:02 AM Connor: 4064 / 5.08 = 800 which makes sense in some way..
11:02 AM Connor: I do know that when machine is commanded to move 1 inch, it moves around .4985"
11:03 AM gregcnc: just do the math for the new setup?
11:03 AM Connor: I don't know what the controller is going step wise.. I.E. Full, half, 1/4 etc..
11:04 AM Connor: So, I'm trying to figure that out..
11:04 AM gregcnc: screw pitch, pulley ratio?
11:04 AM Connor: 20tpi, no pulley
11:05 AM Connor: assuming 1.8degree steppers
11:09 AM Loetmichel: Rab: user error. coworker used to blow swarf off the CNC router while in use like this... -> https://www.youtube.com/watch?v=m7eO5cu-fXg
11:10 AM Loetmichel: btw: VFD works again, had to change TWO output FETs
11:10 AM Loetmichel: and now has kitchen exhaust filters on all air vents ;)
11:11 AM Connor: Hmm. wonder if it's 1.25mm pitch.
11:12 AM Connor: 20.32 (1.25mm) * 200 = 4096 which is what the old controller had set.
11:12 AM Connor: errr. 4064
11:13 AM gregcnc: that looks like the case
11:13 AM gregcnc: 200 steps /rev, full steps
11:13 AM Connor: new controller setup up with 4064 per inch, and it only moved .4985
11:14 AM jdh: 2033
11:15 AM gregcnc: was it accurate before?
11:16 AM Connor: No idea. it came loaded with mach3 pre-configured.
11:16 AM gregcnc: yeah well
11:16 AM jdh: what is it
11:16 AM Connor: Little taig cnc lathe.
11:16 AM Connor: http://www.microproto.com/taigturn_2000dsls.htm
11:17 AM Connor: I didn't see any where in mach3 config for 1/2 step mode.. and it had 4064.
11:17 AM Jymmm: Loetmichel: gotcha
11:17 AM gregcnc: closed loop?
11:18 AM Connor: going to double the count and try it again and see what the measurement is.
11:18 AM Connor: Not to the controller, only to the stepper drivers.
11:19 AM gregcnc: this will depend on the encoder resolution, not the stepping
11:19 AM gregcnc: look up the drive manual
11:19 AM Connor: It's proprietary.
11:19 AM gregcnc: ask the mfg about the screws, there is no need ot be guessing here
11:20 AM gregcnc: proprietary as in they wiped the chinese label off or they built their own?
11:20 AM Connor: Built on their own.
11:20 AM Connor: Board was designed in 2004 or 2005
11:25 AM gregcnc: it seems like these are 20tpi
11:26 AM Connor: based on?
11:26 AM gregcnc: that's what taig produced
11:30 AM gregcnc: based on their resolution claims they were half stepping
11:30 AM gregcnc: did you check x or Z?
11:30 AM Connor: That's what we're working with.
11:30 AM Connor: both are set @ 4064
11:31 AM jdh: 33
11:31 AM jdh: erk
11:32 AM Connor: I just can't get over the fact that it's so close to 1/2" set at 4064 which makes me think the screws were changed out to 1.25mm pitch
11:35 AM gregcnc: don't let assumptions hold you back, turn the lead screw 10 times and measure actual travel
11:35 AM gregcnc: 12.5 vs 12.7 should be quite clear
11:38 AM Connor: Can linuxcnc single point thread with index only now? I know that I couldn't do power tapping way back with just index..
11:38 AM Connor: never understood how mach3 could but linuxcnc couldn't
11:39 AM gregcnc: just depends on how accurate you want the threads
11:39 AM pcw_home: LinuxCNC could (almost) always thread with index only
11:40 AM Connor: could (almost) ??
11:40 AM pcw_home: well a long long time ago it could not
11:40 AM pcw_home: (2.2?)
11:40 AM Connor: okay.
11:40 AM Connor: Power tapping different?
11:40 AM pcw_home: needs encoder interpolated position
11:41 AM gregcnc: rigid tapping is different
11:41 AM Connor: ok.
11:41 AM Connor: Maybe that's what it was then..
11:41 AM pcw_home: rigid tapping needs a full encoder
11:41 AM pcw_home: true of any controller
11:42 AM pcw_home: well I guess an servo/stepper spindle could do without
11:42 AM Connor: pcw_home: Had heck of a time with this old controller. had to pysically swap pins 2/3 5/6 and 16/17 because the controller was expecting step and dir flipped.. and needed spindle on a non stepgen pin.
11:43 AM pcw_home: I think Mach3 can only thread with index
11:43 AM pcw_home: what controller? I could have made a config if its a new one
11:44 AM Connor: and the way pncconf displays the pins/ports for the g540 is hard to follow when trying to match it up to parallel port pinout.
11:44 AM Connor: The controller in this thing.. http://www.microproto.com/taigturn_2000dsls.htm it's custom to this company..
11:44 AM Connor: I just ended up making a db25 to db25 adapter.. I need to pull off extra wires anyway for the limit/home switches.
11:45 AM pcw_home: Yeah it would be nice if pncconf should show the PP pins on PP compatible FPGA cards
11:45 AM pcw_home: would show
11:46 AM pcw_home: the info is available from the pinout file or mesaflash --readhmid
11:46 AM Connor: I ended up hooking it up to a different machine with a pp to make sure I had the pin mapping correct.. then used the .pin file to map it out for the 5i25
11:52 AM gregcnc: it appears the manual for these machines included steps to correct for lead screw error
11:53 AM Connor: link?
11:53 AM gregcnc: I'm guessing the preivous owner made mistakes as it shows 8000steps/inch to be default for the mill
11:53 AM gregcnc: https://www.soigeneris.com/Document/Taig/MicroMill_DSLS_Manual.pdf
11:54 AM Connor: it's possible.. that's for the mill not the lathe.
11:54 AM gregcnc: so have you measured the screw yet?
11:54 AM Connor: I'm going to test it with mach3 and see what it does, the try 8000 and 8128
11:54 AM Connor: I'm not at the machine.
11:55 AM Connor: at home right now doing research for when I get back in front of it.
11:56 AM gregcnc: but if you were short on travel with 4064, it's certainly not going to move 1" at 8000
11:57 AM Connor: right.. which is what was confusing the crap out of me.
11:57 AM gregcnc: that's when you start measuring
11:58 AM Connor: Other possibility is, it steps on low vs high on the stepper driver.. I saw one of those (not direction) inverted. But that made no sense to me either.
12:05 PM gregcnc: 4064 makes sense for 2.5mm pitch screws
12:05 PM Connor: at half distance?
12:05 PM Connor: double number for 1 inch..
12:06 PM gregcnc: no, but 4064 isn't right anyway it seems
12:07 PM gregcnc: 4604 steps /inch is 400step/rev with 2.5mm pitch
12:11 PM gregcnc: Taig currently versions with 2.5mm ballscrew as an option. ...with the microproto control
12:12 PM Connor: So, it's possible that either a) it was configured for ball screw and it's not.. or b) it's half stepping and need to double the counts.
12:14 PM gregcnc: 4064 would be for 2.5mm pitch, and 8000 for 20tpi, if it's something other than that who knows how it got that way.
12:15 PM gregcnc: half step in both cases
12:22 PM Connor: .9995" at 8128
12:23 PM Connor: within the tolerance of measuring equipment we're using..
01:00 PM gregcnc: connor on X or Z?
01:00 PM Connor: Z
01:04 PM gregcnc: hmm. so is it 1.25mm pitch or 800 step/rev
01:06 PM Connor: we tested both.
01:08 PM gregcnc: both??
01:08 PM Connor: Yea. Both X and Z.. have the same pitch
01:09 PM gregcnc: in diameter mode X on a lathe should move half of commanded
01:11 PM Connor: We did it using jog. I'm not sure if it was in diameter mode or not.
01:27 PM gloops: The worlds largest 3D printed rocket engine set to power new UK launcher spearheading Britain's revitalised space ambitions http://ow.ly/kX0F50schTo
01:31 PM jym: gloops: SMH
01:33 PM gloops: nothing can go wrong lol
01:34 PM jym: gloops: Nothing can go right either it seems =(
01:37 PM gloops: we could be on the moon before Greece
01:48 PM jym: gloops: been there done that, got the tshirrt
01:48 PM roycroft: we've been on the moon, before anyone else
01:49 PM sysop is now known as Guest80004
01:49 PM gloops: yes, fair amount of scepticism over that though roycroft
02:03 PM net|: https://www.thingiverse.com/thing:3590144 for canadians with old doors
03:05 PM XXCoder: roycroft: yeah its so easy to use velco around intake and use fiber mat on it
03:35 PM SpeedEvil is now known as Guest59942
04:01 PM BitEvil is now known as SpeedEvil
05:20 PM Wolfmetalfab is now known as Wolf__
05:31 PM andypugh: Quiet night?
05:32 PM XXCoder: looks like
05:32 PM andypugh: Still trying to make RTAI work
05:34 PM andypugh: But I do now have Stretch running on 3.16.52-RTAI or 4.9.80
05:34 PM andypugh: I just can’t get RTAI to work on either.
06:00 PM skunkworks: andypugh: did you see
06:00 PM skunkworks: http://www.electronicsam.com/images/KandT/testing/8300/Screenshot_2019-04-26_17-53-42.png
06:01 PM andypugh: Good numbers. The machine I am struggling with is nearly that good with preempt-rt.
06:01 PM skunkworks: That is what this is too
06:01 PM skunkworks: amazine
06:01 PM skunkworks: amazing
06:01 PM andypugh: But I rather feel we need an RTAI option.
06:02 PM skunkworks: right - it would be nice
06:02 PM andypugh: But we need a LiveCD that works more.
06:03 PM andypugh: In the spirit of “it’s easier to ask forgiveness than permission” I think I will change the download link this weekend.
06:04 PM skunkworks: I played with it and it seems like it will do 25us base thread. that would be 40khz..
06:04 PM skunkworks: (this thing has a printer port too)
06:13 PM jthornton: I think the download link should be like before and go to a page that lists all the downloads
06:14 PM jthornton: it would be so much easier for folks supporting older systems