#robotics Logs

Dec 26 2014

01:48 rue_shop4 aha, I found the problem
01:48 rue_shop4 its a .48mm shoulder between two parts of the feeder
02:10 rue_shop4 another bad thing happened, I lost my .42mm bit
02:10 rue_shop4 when I switched to the 1mm
02:10 rue_shop4 I'm sure I put it down somwhere safe
02:11 rue_shop4 I was trying to work out how to mark the size on it
02:11 rue_shop4 oh then I took it over and lceaned it
02:11 rue_shop4 ah, there is where I lost it, its still in the vise
02:11 rue_shop4 :)
02:19 rue_shop4 its quite possable that the other one was less than .48mm and thus didn't have the problem as pronounced
03:01 rue_shop4 yup, that was it, every jam since I built it, a .48mm shoulder
03:33 darthmachinist so how do i send information onto something that rotates?
03:33 darthmachinist data
03:33 darthmachinist I see something about rotary inductors
03:37 darthmachinist an led?
03:39 deshipu led is probably the cheapest
03:39 deshipu there is also radio
03:39 darthmachinist is there no easy way to use radio or microwaves?
03:39 deshipu there are radio modules
03:39 deshipu quite easy to use
03:40 darthmachinist something that can hanlde spinning. I've used radio for really simple stuff, never for transmitting data. a low amount of electronics would be important, since the thing has to spin.
03:41 darthmachinist leds would work though
03:41 deshipu I would use an IR led and a tv remote receiver
03:41 darthmachinist that's a great idea
03:41 deshipu how fast does it spin?
03:42 darthmachinist fast, like, 7k rpm
03:43 darthmachinist maybe this isn't the best idea
03:43 deshipu uh
03:43 deshipu your microcontroller will fall apart?
03:43 darthmachinist I'd imagine I could take steps to prevent that from happening...
03:44 deshipu there is also the problem of aiming the led at the receiver
03:44 darthmachinist balance it, encase it in something. experiment where it won't blow up in my face.
03:44 darthmachinist I can put it at the end
03:44 darthmachinist or have multiple recievers
03:44 deshipu may I ask what are you trying to do?
03:45 darthmachinist erm, you'll think it's dumb
03:45 deshipu I already do
03:45 darthmachinist lol
03:51 darthmachinist I was wondering if it was possible to use a gyroscope as an encoder
03:52 Jak_o_Shadows What's your opposition to radio? Seems the easiest to get running anyway
03:52 darthmachinist sinse I already needed to sense torque
03:52 Jak_o_Shadows unless you're going straight to a finished product
03:52 darthmachinist not knowing what's necessary, like I said, i have to reduce the electronics on the rotor as much as possible, cause they'd likely fly apart.
03:53 Tom_itx keep the mass to the center
03:53 Jak_o_Shadows eh, have you looked at figuring out the acceleration on the chips? I know 7k rpm, but what acceleration does that make?
03:53 darthmachinist that's easy to do
03:53 darthmachinist hmm
03:53 darthmachinist the datasheet said 10,000 gs for 0.1 ms
03:53 darthmachinist lol
03:54 deshipu darthmachinist: wait, you are using a rotating mass gyroscope?
03:54 deshipu darthmachinist: why not mems?
03:54 darthmachinist mems
03:54 darthmachinist of course
03:54 Jak_o_Shadows dude, a = v^2 / r (v = tangential velocity, r = radius from centre).
03:54 deshipu darthmachinist: so what's rotating?
03:54 rue_shop4 and it jams up again
03:54 deshipu darthmachinist: also, just put a magnet on it and use a hal sensor :)
03:57 darthmachinist I was thinking the shaftt to a small gas engine, just as a sensor for me so I can see what's going on. mems looked really nice.
03:58 deshipu you would need to put the muc and battery on it, then remember to recharge the battery, etc.
03:58 darthmachinist sliprings would work for power
03:59 deshipu not at 7k rpm
03:59 darthmachinist no need to worry about noise
03:59 darthmachinist they make them for alternators
03:59 deshipu and they heat
03:59 deshipu and have to be replaced because of wear
03:59 darthmachinist I'm not that worried about wear
04:01 darthmachinist brushed dc motors run at ,like 12k rpm
04:02 deshipu yeah, and are tricky to make
04:03 deshipu I mean, it seems to be a lot of trouble just to get the position with very poor accuracy
04:05 darthmachinist shaft speed is different from position
04:05 deshipu speed is position over time
04:06 darthmachinist things like the speed when combustion occurs versus having 100 points measured by an encoder.
04:06 darthmachinist and, just, balance
04:07 deshipu the engine has quite some inertia, I doubt it will change speed rapidly
04:08 darthmachinist what about the inertia of a connecting rod? There's lots of little moving parts, and that kind of data could be very valuable, especially when combined with a torque sensor, which needs to communicate somehow anyway.
04:10 Jak_o_Shadows dude, what are you even doing?
04:10 darthmachinist I dunno yet
04:11 deshipu if you want exact position, put a sticker with some pattern on it, and read that with a bunch of photo sensors
04:12 deshipu you can even make that analog
04:12 deshipu a stciker with a gradient
04:12 deshipu sticker
04:13 darthmachinist an analog optical encoder.
04:16 darthmachinist or I could just use a little mems device
04:19 deshipu the plus side of an optical encoder is that the rotating part is simple, light, cheap and doesn't require power
04:20 deshipu also study and reliable
04:21 darthmachinist okay, how do I get an analog optical encoder?
04:22 darthmachinist i was looking forward to my accelerometers.
04:22 deshipu darthmachinist: you take a LED, a photoresistor, and a sticker with a gradient
04:22 deshipu darthmachinist: you put the sticker on your shaft, and point the led and the photoresistor at it
04:23 darthmachinist that works
04:23 deshipu darthmachinist: you power the led, and measure the resistance of the photoresistor
04:24 darthmachinist could I get more accuracy by having multiple led/sensor pairs?
04:24 deshipu sure, why not
04:24 deshipu if you want to get fancy, you can even take a sensor from an optical mouse :)
04:24 darthmachinist ew, no.
04:25 darthmachinist oh, an optical mouse
04:25 darthmachinist hmm
04:25 darthmachinist that seems the way to do it.
04:26 darthmachinist I could use ir
04:28 darthmachinist I could torque sense by using multiple encoders, if I can get the position accurate enough.
04:30 darthmachinist would line cmos sensors give me that cheaply? and if it's so easy to do, why are they so expensive?
04:31 deshipu accuracy is expensive
04:32 deshipu actually it's not about accuracy, it's about repeatability
04:32 darthmachinist yeah, there's lots of noise/vibration, but it tends to cycle with everything else
04:34 darthmachinist now how to make the sticker
04:34 darthmachinist or, like, a translucent piece of plastic
04:36 deshipu print it?
04:36 darthmachinist and depend on it reflecting?
04:37 deshipu you don't need that much of a difference in the refelcted light
04:38 deshipu the photoresistors are quite sensitive
04:38 darthmachinist photoresistors/sensors are that accurate?
04:38 deshipu of course you have to shield it from outside light
04:39 darthmachinist of course
04:41 darthmachinist better than having a bunch of junk on the shaft, low mass is very important there
04:43 darthmachinist reflective it is.
04:43 darthmachinist accuracy in printing doesn't really matter, just spin it once.
04:43 darthmachinist and record the values
04:45 darthmachinist okay, i got it
06:20 giddles hello
06:22 Jak_o_Shadows hi
06:23 giddles hi im not a adruino printing experd or 3d print psycho i got an irobot and have some trouble with it. does anyone own also a irobot roomba 880? :)
13:49 Veverak anybody experience
13:50 Veverak with reprpa's heated bed not heating evenly?
13:50 Veverak I mean, three corners: 85C, fourth corner and center: 95C
13:51 rue_house there are a lot of cases of heaters goingwrong
13:51 rue_house there are other things that are factors
13:52 Veverak hmm
13:52 Veverak morel ike, what can I do about it? :/
13:52 rue_house heat = power in / power dissipation rate
13:52 rue_house increase the power going into that area or change the power dissipation rate
13:53 Veverak what?
13:53 Veverak are we talking about reprap's mk2 bed?
13:53 Veverak :)
13:54 rue_house insulate the cold areas or add extra power to go into the cold spots
13:54 Veverak that got one power input and that's all? how am I supposed to change the power distribution?
13:54 rue_house I think you might have a bad heater tho
13:54 rue_house they are made from a circuit board, they put current thru the traces to get them to heat up
13:55 Veverak yeah I know that
13:55 rue_house if the traces start to degrade then they wont give off the same heat as other areas
13:55 rue_house how long you had it?
13:56 Veverak from February?
13:56 rue_house and about what temp you been running it at?
13:56 Veverak 120C
13:57 Veverak well, 120C was written in config on RAMPS/ARduino - like board
13:57 rue_house are any parts of it black?
13:57 Veverak hmm, it looked fine last time I did something with it
13:57 Veverak but worth checking
14:12 Veverak nah
14:12 Veverak there are scratches on the other side
14:12 Veverak quite a deep ones
14:13 Veverak rue_house: I suppose that explains everything
14:13 Veverak now question is
14:13 Veverak can I add some heat buffer ontop of the bead that would distribute heat more evenly?
14:13 Veverak and the main question, should I even care?
14:27 Veverak well well well
14:29 Veverak hmm
14:29 Veverak e3d-online got a quite nice price
14:29 Veverak hmm
14:30 Veverak (on heated bed that is)
14:37 rue_more copper plate
14:38 rue_more the red board with the traces has deep scrathces?
14:39 Veverak yeah
14:39 Veverak unfortunetly on the side with traces
14:39 Veverak and I suppose there are ... shorts? or whta's the proper term? :D
14:39 Veverak I can clearly see the copper
14:42 rue_more usually cuts will cause opens, not shorts
14:43 Veverak anyway it doesn't look like I will be able to "repair" it
14:44 rue_more your printing abs?
14:45 Tom_itx so
14:46 Tom_itx what we workin on today??
14:46 Veverak rue_more: yep
14:46 Veverak Tom_itx: debuging printer... again
14:46 rue_more yea, I have to fix the head on mine
14:47 Veverak yeah, have to buy temp sensor
14:47 rue_more it WAS working great too
14:47 Veverak but I need to decide if I have to buy heated bed or not
14:49 Veverak hmm
14:49 Veverak that one on e3d looks pretty cheap, but dunno if that's ok
14:49 Veverak http://e3d-online.com/PCB-Heated-Bed-MK2B-12v-and-24v
14:52 verak got 25A power sou
14:55 rue_more yea
14:55 rue_more 4 of them
14:56 GuShH rue_more: https://www.youtube.com/watch?v=DXAzJE9Zs8s
14:56 GuShH I don't know whether to admire this guy or not
14:57 GuShH the rapid reverse with the electric driven leadscrew is nice though
14:58 rue_more lots of $ int hat
14:59 GuShH and time
14:59 GuShH this is one of those "can I do it" things for sure
14:59 GuShH otherwise you would just buy a used machine
14:59 rue_more I'd say that he has about $2000 in things finished purchased
14:59 GuShH at least
14:59 GuShH a good chuck and bearings, etc. are not cheap, including the motor....
15:00 GuShH pulleys, etc. he also has what looks like a gearbox from a vehicle
15:00 GuShH but sometimes you already have all the crap around and it just clicks in your head
15:00 GuShH you never have EVERYTHING though for something like this...
15:00 Veverak well
15:00 verak got one entire forum about people building this th
15:00 GuShH I need a reverse tumbler for my china lathe... asap heh
15:00 GuShH Veverak: oh?
15:01 GuShH I just like watching videos of homemade lathes and mills for inspiration
15:01 GuShH lathes are more common it seems
15:01 GuShH than mills
15:01 GuShH the horizontal "engine" mill is a nice idea
15:01 Veverak well
15:01 GuShH using an engine block for the headstock
15:01 Veverak more like CNC though
15:01 ShH eyer
15:01 GuShH so what, kids buying china shit and adding steppers and drivers
15:01 GuShH not interested
15:01 Veverak what?
15:02 Veverak I said building things
15:02 GuShH I don't care for cheap conversions
15:02 Veverak which means that they are proper working machines
15:02 Veverak GuShH: don't know the people, stfu :P
15:02 GuShH why don't you stfu and learn to express yourself properly? pretty please.
15:03 GuShH now why youtube thinks a "homemade gun silencer" is related to a homemade metal lathe, I don't know.
15:03 verak didn't said that they are children making cheap mach
15:03 GuShH I did, because that's what most of homemade CNC is.
15:03 GuShH just cheap shit with more shit added on top
15:04 GuShH they spend more on ballscrews than the machine is worth
15:04 Hyratel http://kerbaldevteam.tumblr.com/post/106251499939/get-ksp-for-40-off-on-steam
15:04 GuShH if you start with crap you get crap, no matter how long you polish a turd it still remains a turd.
15:04 Hyratel a very shiny turd
15:04 GuShH but a turd nontheless
15:05 Veverak well yeah, so when somebody is talking about some specific group and you point out shitty opinion based on knowledge about average population
15:05 GuShH (also why would you want a shiny turd, must be some asian fixation)
15:05 Veverak you got nice pretty STFU in your face and I don't feel sorry for that
15:05 verak doesn't say that it's not true, most of the machines you find are sh
15:05 GuShH Veverak: I feel sorry about your English, but that's another story.
15:05 GuShH The specific group which remains unknown...
15:05 Veverak GuShH: yeah, I know :)
15:05 Tom_itx GuShH you polishing turds again?
15:05 GuShH Tom_itx: you know it!
15:05 Veverak well yeah, not english speaking forum, is tehre a point?
15:06 Tom_itx you know what?
15:06 Veverak :)
15:06 Veverak :D
15:06 Tom_itx it's still a turd when you're done
15:06 GuShH yes it is! but a lovely one at that.
15:06 GuShH (didn't I just say that)
15:06 Tom_itx dunno
15:06 GuShH I did :(
15:06 Tom_itx wasn't paying attention
15:06 GuShH haha
15:06 Tom_itx especially to you
15:06 Tom_itx :D
15:06 GuShH aww.
15:06 GuShH the analogy to starting a cnc off of a cheap machine is that it still is a cheap machine.
15:07 Tom_itx no
15:07 GuShH and that's how most diy cncs go
15:07 Tom_itx but you can pile more shit on it and make it very nice
15:07 GuShH Tom_itx: cheap as in low quality.
15:07 GuShH to almost no added benefit, since they never work on the most important aspects of automation
15:08 GuShH motion control is just one little part of it
15:08 Veverak Tom_itx: yeah, but to mate shit nice you practically replace entire machine? :)
15:08 GuShH Tom_itx: for instance your conversion is not from a harbor freight micro/mini lathe, which would constitute a shit conversion.
15:08 GuShH or whatever, they have mills also.
15:09 GuShH Veverak: I would rather convert a big old machine instead, but big means $$$ in parts.
15:09 ShH is waiting for a come-along he ordered so he can move some p
15:10 Veverak GuShH: yeah, seen one really big
15:10 GuShH If you want to make life simpler, make sure you can pick up every part on your own... otherwise things get harder and harder for a single person :(
15:10 Veverak but these are machines when to "move them" starts to be problem
15:10 Veverak :)
15:11 GuShH Veverak: some of these require an overhead crane just to move a dividing head... forget about moving the whole thing on your own, even with pipes... there are certain limits to my stupidity, that is one limit.
15:11 Veverak lol
15:11 Veverak well yeah :D
15:11 GuShH Tom_itx goes "good to know..."
15:11 verak wanted to build CNC
15:12 Veverak then I figured out what it takes to make proper one and gave up
15:12 Veverak for now... :)
15:12 GuShH don't give up :/
15:12 m_itx says not
15:12 Veverak well
15:13 Veverak I prefer more creating and designing things to be made on CNC than taking care of the machine itself
15:13 GuShH lol
15:13 Veverak and I know that my first CNC = more repairing than actually making something
15:13 GuShH babysitting!
15:14 GuShH you kinda work more than with a manual machine
15:14 GuShH and are prone to more scrap / errors in fact, until everything is calibrated... then something goes wrong and you are back to square 1
15:14 GuShH derp
15:14 Veverak yeah
15:14 Veverak repeat until pro
15:15 verak got that know with Re
15:15 Veverak *until you are pro
15:15 Veverak *that now
15:16 Veverak fell like more debugging something than actually printing lately :/
15:16 Veverak but I've got the feeling that if I would remove the urge to make it better and JUST use it, it would print more :D :D:
15:17 GuShH hah
15:17 GuShH have two printers, one to print with, one to tinker with
15:20 Tom_itx you gotta tinker with em all the time just to keep em printing
15:20 Tom_itx ask rue_more
15:22 Veverak yeah
15:22 Veverak I suppose you have to
15:22 Veverak but you can minimalize it to maintenaice and fast debugging :)
15:22 Tom_itx not if they were built right to begin with
15:28 GuShH Tom_itx: oh no you didn't! are you saying all repraps are hardware store projects?
15:28 ShH g
15:28 Tom_itx not at all
15:28 ShH hides rue's broken ho
15:28 GuShH nothing to see here
15:28 GuShH most of the mainstream designs are harware store printers though
15:29 GuShH but it has moved onto full blown commercial these days
15:29 GuShH for the moneis.
15:29 GuShH I need to forge a spanner...
15:30 GuShH rephrase: I have to try to forge / fab a spanner (for a lock nut)
15:30 GuShH bbl
15:30 Veverak Tom_itx: yeah, but to build something right is not really easy :)
15:31 Tom_itx nothing worthwhile is
15:32 Veverak true
15:33 rue_more .... you know you live in a small town when the shopping mall is CLOSED ON BOXING DAY
15:33 rue_more the whole thing.
15:33 rue_more :/
15:34 Veverak fuu
15:34 verak just remembered the loong list of things he wants to improve on his reprap,
15:35 Tom_itx short trip ehh rue_more?
15:35 rue_more i have to get mine working again
15:35 rue_more yea
15:36 rue_more there is a nearby store that is open, apparently they have so many customers that the malls parking lot is almost full
15:36 Veverak rue_more: yeah
15:36 rue_more but the mall itself is closed
15:36 Tom_itx hah
15:36 rue_more the hardware store, I can understand.
15:36 rue_more THE MALL?
15:37 Veverak thermocouple, dual hotend, replace all drivers, now repair heated bed, add optoendstop, make the plate 200x400.... damn it
15:37 rue_more I got a proper flex cable off my carrige last night
15:37 rue_more I'm gonna turn the shop upside down and see what I can find for stainless to try to make a new hotend
15:38 verak just fell in live with e3ds new inven
15:38 Veverak Vulcano
15:38 Tom_itx i got 2 nice size bars here
15:43 wolfmanjm really? I just got a Volcano and am VERY disappointed
15:43 Veverak not as shiny as on pictures?
15:44 Veverak wolfmanjm: tell me more! :D
15:44 wolfmanjm what version dod you get? the 1.75
15:44 Veverak no no no
15:44 Veverak I didn't buyed it yet :D
15:44 wolfmanjm well I got it (the 3mm), and have been tryign to get it to worj with the 1.0mm nozzle for the last two days
15:45 wolfmanjm I basically have to slow my prints down to 40mm/sec to get a reasobable flow rate
15:45 wolfmanjm seems it is over hyped
15:45 wolfmanjm I usually print at 120mm/sec with a 0.5mm Jhead
15:46 wolfmanjm but the flow rate is too high, at 40mm/sec it can barely keep the PLA molten
15:46 wolfmanjm which really makes this whole "double the speed" thing moot
15:46 wolfmanjm yea it is half the layers, but then it is half the print speed
15:49 wolfmanjm Veverak: https://flic.kr/p/qxrsP8
15:49 wolfmanjm pretty ugly
15:50 wolfmanjm this was better https://flic.kr/p/qg27wc going at 1mm width, but the edges are bad
15:50 wolfmanjm the infill at least is relatively solid and consistent
15:50 Veverak mybe play with temperature?
15:50 wolfmanjm yea I did that ;)
15:50 Veverak hmm
15:50 wolfmanjm did not make much difference
15:50 Veverak dunno
15:51 wolfmanjm went as high as 220°C which is pretty hot for PLA
15:51 Veverak yeah
15:51 Veverak well, I print only ABS so no exp. with that :/
15:51 wolfmanjm I mean you can get it to work, but the slow down in speed negates the advantage of big thick beads
15:51 wolfmanjm ahh ABS maybe a whole other experience ;)
15:52 wolfmanjm I'm told (and have no clue if it is true) that ABS will work much better as it melts quicker and cools faster
15:52 Veverak you hot complete set?
15:53 wolfmanjm Yea I bought all the pieces as I did not have an e3d V6 to upgrade
15:53 wolfmanjm you can get the bits you need to make a complete hotend without buying an e3d and a volcano
15:54 Veverak yeah, I've got v5
15:54 wolfmanjm works out to £47.84 total
15:54 wolfmanjm not sure if you can upgrade a V5
15:55 wolfmanjm personally I have genuine jheads, I have never had any problem with them ever, but I have wasted a lot of money on other hotends that never worked
15:55 wolfmanjm I even made a hotend with solid copper heater block :)
16:00 wolfmanjm The main problem with large nozzles is even with the large meltzone, the filament moves through so fast it cannot melt it all so it comes out too cool, then clogs up entirely until it can melt again, so on a long thick bead you see it drying up and stop extruding then start again later. and the bead that does get layed down is too cold to adhere to the previous layer.
16:01 Veverak well
16:01 wolfmanjm I thought volcano would solve that, but apparently not. I did not really expect to be able to print at 100mm/sec, but I though maybe at least 60mm/sec would work, dropping to 40mm/sec is not a compromise I think is worth making
16:01 Veverak I ragequitted on prusa nozzle
16:01 Veverak and now live happy with v5
16:02 wolfmanjm yea all metal hotends have a spotty track record at best :)
16:02 Veverak well never heard bad word on v5 :
16:02 wolfmanjm I have a franken Jhed /magma/aluhotend that mostly works
16:02 Veverak D:
16:02 wolfmanjm but it is not all metal as it has a ptfe sleave.
16:03 Veverak v5? nope
16:03 Veverak v5 is full metal
16:03 wolfmanjm I meant my franken hotend
16:03 Veverak oh, yeah
16:03 Veverak anyway
16:03 wolfmanjm I have heard alot of people complain about the v5 with PLA jamming
16:03 wolfmanjm the v6 is better
16:03 Veverak I still suppose that volcano would be better for me than actual setup
16:03 Veverak wolfmanjm: well, not for ABS :D
16:04 wolfmanjm yea try it let me know how it works with ABS :)
16:04 Veverak I was able to print over 100mm/s :)
16:04 Veverak wolfmanjm: point is that I still don't know if I want to buy Vulcano
16:06 wolfmanjm well do you want to use a large nozzle?
16:06 wolfmanjm like 1mm ?
16:06 Veverak well, I've got 0.8mm
16:06 Veverak prefer 0.4
16:06 Veverak would like to have 0.25 working :D
16:06 wolfmanjm I think 0.8 works better
16:07 Veverak got it because of timberfill
16:07 wolfmanjm yea so then you probably don;t need a volcana if the 0.8 works for you
16:07 verak would like volcano with
16:07 wolfmanjm I don;t think they do that ;_
16:07 Veverak Volcano got much bigger area in which the material is heated
16:07 Veverak and that's what I like :D
16:07 Veverak wolfmanjm: yeha :/
16:07 Veverak that's what keeping me from buying it
16:09 wolfmanjm the new hexagon from reprap discount has 0.4 with a larger meltzone, that works quite well, although I have only printed calibratoin cubes with it so far
16:09 wolfmanjm but you need to get the new one, not sure if it is in the store yet they made it for lulzbot
16:09 Veverak I see
16:09 Veverak well, there is another point
16:10 Veverak I am happy with actual thingie :D
16:10 Veverak only it's natural for me to search for faster prints
16:10 Veverak another thing is that I consider switching from 3mm to 1.75mm
16:11 wolfmanjm yea I am stuck with 3mm got a tom of 3mm pla, and I'd have to switch out all my extruders and hotends, that'll be expensive
16:11 Veverak :D
16:11 verak got one nozzle and one extr
16:12 Veverak while extruder is working awesome
16:12 Veverak it's quite big :/
16:12 Veverak (not dual extrusion friendly)
16:19 rue_shop4 I suspect that changing the peice to stainless is part of my problem
16:19 rue_shop4 I think this head has qutie a sharp temperature transition point
16:20 giddles hi
16:20 giddles vaccum cleaning robotics knowlegend here?
16:20 giddles i got an easy problem that ruins my day
16:23 rue_shop4 aha
16:23 rue_shop4 yes, when the glass it hot, the pla will stick to it
16:24 rue_shop4 giddles, do tell
16:24 giddles k i cant set the clock
16:24 giddles he starts new and --:--
16:24 giddles its an miele scout rx1 thats similar to the irobot 880
16:24 wolfmanjm rue_shop4: only if the glass is very very clean ;)
16:24 lfmanjm has b=never got PLA to stick to bare glass reli
16:25 rue_shop4 it stuck pretty easy
16:25 rue_shop4 if came off when it cooled
16:25 rue_shop4 but thats not helping an extruder design
16:25 wolfmanjm yea that is a trick I have not learned
16:25 wolfmanjm extruder or hotend design?
16:25 rue_shop4 hot air gun on glass for a min, rub with pla, pla sticks to glass
16:25 wolfmanjm I have some experience in hotend design :)
16:26 rue_shop4 mine worked, then it stopped
16:26 rue_shop4 seen mine?
16:26 wolfmanjm nope
16:26 LoRez diluted elmer's glue, spread out on the glass
16:26 rue_shop4 let me show ya
16:26 LoRez either that or hairspray
16:26 rue_shop4 no, the glass in in my extuder, I need it to NOT stick
16:27 wolfmanjm LoRez: tried the PVA glue no luck for me... blue tape wiped with IPA works everytime though, except I can;t then get the print off :)
16:27 LoRez oh, I mean on your bed.
16:27 rue_shop4 http://ruemohr.org/~ircjunk/projects/reprap/p1040477.jpg
16:27 rue_shop4 wolfmanjm, ^
16:27 wolfmanjm heh glass heat break, that is interesting
16:28 rue_shop4 quartz
16:28 rue_shop4 about 1/4 the thermal conductivity of stainless
16:28 rue_shop4 BUT after an inncodent I'v having jamming problems
16:28 wolfmanjm nice that dispenses with those ugly heatsinks
16:28 LoRez how're you pushing through that successfully?
16:29 rue_shop4 if the glass and plastic get just hot enough up enough, the plastic fuses to the glass and everythnig stops
16:29 wolfmanjm tried running a fan over the glass while extruding?
16:29 rue_shop4 it was working ok, I thought I might do better if that landing bolt was stainless,
16:29 rue_shop4 yes
16:30 rue_shop4 the pla will stick to stainless too, so how come the normal heads dont jam up?
16:31 rue_shop4 where is the transtion point..
16:31 wolfmanjm rue_shop4: most good heads have a ptfe lining
16:31 rue_shop4 anyone ever cut a head in half?
16:31 wolfmanjm in the heatbreak
16:31 rue_shop4 ah
16:31 rue_shop4 so, even if it melts it cant stick
16:31 wolfmanjm most all metal hotends have trouble with PLA jamming
16:32 wolfmanjm ABS is fine
16:32 rue_shop4 ptfe vs teflon, one creeps?
16:32 rue_shop4 oooh rrrly!?
16:32 wolfmanjm I think they use ptfe
16:32 rue_shop4 hmmm
16:32 rue_shop4 hmmmmmmmmmmmm
16:32 rue_shop4 I have no lining
16:32 wolfmanjm I think the new e3d V6 has mostly fixed the PLA jamming issue, not sure if polishing the inside of the SS tube did it or the extreme cooling
16:33 rue_shop4 so my head was doing amazingly well
16:33 rue_shop4 then there is the hot glue gun canundram
16:33 rue_shop4 they just have a rubber boot off the back of the hot end
16:36 wolfmanjm rue_shop4: you mean like this... https://flic.kr/p/q8XY4j
16:37 rue_shop4 the whole thing, loaded with plastic, cut in half like that
16:37 wolfmanjm not that I have seen
16:37 rue_shop4 so you can see where the transition piont is
16:37 wolfmanjm the transition point should be just above the meltzone I believe
16:38 wolfmanjm which is what the heatbreak is for
16:39 wolfmanjm https://flic.kr/p/gGt9QX <--- My franken hotend, home made copper heater, a 0.75mm nozzle from a metal magma, and the SS ptfe lined tube from a aluhotend and a magma heatsink :)
16:40 rue_shop4 knowing that all the heads are lined is new to me
16:42 rue_shop4 ... why not just have the tube?
16:42 rue_shop4 what if it was bare
16:42 rue_shop4 maybe it would collapse over
16:47 wolfmanjm not all the hotends use ptfe, but al the ones that work well do :) like jheads
16:47 wolfmanjm e3d v6 is not lined, but I don;t know what they fixed to get it to work
16:48 rue_shop4 my ptfe tube wont fit in my quartz tube
16:48 wolfmanjm I found that if yo don't aggressively cool the fins on an all metal it will jam with PLA pretty fast
16:49 rue_shop4 aka quartz
16:49 rue_shop4 4x less cooling
16:50 rue_shop4 I'm redesigning...
16:54 wolfmanjm I think the pressure can also get quite high, so quartz may shatter :)
16:54 wolfmanjm which is why they use a SS tube with a ptfe linign rather than just ptfe
16:57 rue_shop4 the pressure shouldn't
16:57 rue_shop4 mine did break, but I suspect it was caused by a bad seating arrangment
18:16 rue_more lasted for 2 or 3 weeks
18:16 rue_more was dialed in perfect too
18:17 Tom_itx down for maintenance?
18:17 rue_more it still wont work again
18:23 rue_more glue guns have quite a cavity of molten material in them
18:26 Tom_itx did you cut one apart?
18:27 rue_more no I just took it apart
18:27 rue_more from the dollar store
18:27 rue_more they have a large heater section, and they use a rubber pipe on the back going into it
18:27 deshipu this way you can squirt quite a lot of it in one go
18:28 deshipu and if you don't, it gets solid again so no waste
18:32 Tom_itx so get a hydraulic cylinder and put a .003" hole in the end of it. fill it with pla granules and wrap a big heater around it. wait for the whole thing to heat up and use it like a cauking gun
18:32 Tom_itx no more spools of plastic
18:32 rue_more I'm tempted
18:33 Tom_itx or use your auger idea to direct print
18:35 rue_more if nothing else, I'm annoye3d to be set this far back, after that much time of getting it working
18:36 Tom_itx seems that's how reprap goes
18:36 rue_more you heard this story before?
18:36 Tom_itx don't know first hand though
18:36 Tom_itx seems they gotta do alot of continual tweaking
18:37 Triffid_Hunter rue_more, wolfmanjm: fwiw the arcol v4 doesn't have any PTFE in the thermal break and it never jams for me
18:37 rue_more what the barrel?
18:37 rue_more stainless?
18:37 Triffid_Hunter rue_more: the thermal break is about 14mm above the nozzle, just at the top of the heater block. Also the SS is extremely thin on arcol v4, some 50µm or so I'm told
18:38 Triffid_Hunter so it can break if you're not careful.. I've only broken one so far, it got slammed into something at full speed so entirely my own fault
18:38 rue_more well my quartz was working fine
18:38 Tom_itx does it have fins on the SS?
18:38 Tom_itx or just a thinwall tube
18:39 rue_more http://d2n4wb9orp1vta.cloudfront.net/resources/images/cdn/cms/0711ptDirect1.jpg
18:39 rue_more 1000psi eh?
19:12 Tom_itx Triffid_Hunter, what thread is on that SS Piece?
19:12 Triffid_Hunter Tom_itx: M6 I believe
19:12 Triffid_Hunter Tom_itx: no wait, M10
19:12 Tom_itx and the hole diameter?
19:12 Tom_itx 3mm?
19:12 Tom_itx that looks fairly easy to make
19:12 Triffid_Hunter Tom_itx: bore is 3.17mm
19:13 Triffid_Hunter Tom_itx: thing is, SS work-hardens so you need good tooling to cut it accurately
19:13 Tom_itx yeah
19:14 Tom_itx you just gotta be careful cutting it
19:14 Tom_itx somewhat
19:15 Tom_itx the thin wall must be the key to heat transfer loss?
19:19 Tom_itx so the force of the grub screw against the PTFE keeps the SS part tight in the aluminum heatsink it appears
19:20 Tom_itx since it says not to force tighten the SS in the aluminum heatsink
19:22 Tom_itx interesting idea
19:26 Triffid_Hunter Tom_itx: yep and yep, works really well
19:41 GuShH wolfmanjm: alu is m6
19:41 GuShH m10 may be the budda
19:41 GuShH e3d has been m6, not sure on the latest
19:41 GuShH (don't care)
19:41 GuShH that sure is a franken-thing
19:42 GuShH but as long as it works
20:15 wolfmanjm yea it didn't work too well... went back to my trusty Jheads
21:12 rue_shop4 this desk is filling up with every varriation possable of my head
21:14 Tom_itx all you need is one good one
21:15 rue_shop4 I figure there is 1 parapimter tht is wrong
21:16 Tom_itx that one trif linked looked interesting
21:17 rue_shop4 knowing that most heads have to use a lining to do pla really helps
21:17 rue_shop4 I been running what 2, 3 weeks? without any lining
21:17 rue_shop4 and almost no cooling
21:19 rue_shop4 one of the things I did was to sit the quarts lower into the bolt it sits in,
21:19 rue_shop4 knowing now that the plastic can stick to the glass, I'v raised it up
21:19 rue_shop4 there is almost no overlap
21:20 rue_shop4 I'v also removed a .48mm shoulder that the plastic hit on the way down
21:21 rue_shop4 if any of the last 7 itterations hadn't failed so badly, I'd say this is the one that'
21:21 rue_shop4 s bound to work
21:27 rue_shop4 ok, here we go again, revsision 8
21:29 rue_shop4 crap, I think the temp sensor died
21:33 rue_shop4 damnit
21:39 Tom_itx maybe you should get a good hot end
21:39 Tom_itx stop fussing around
21:41 rue_shop4 you know, I would put money into it if I were trying to make money
21:41 rue_shop4 this is soemthing I do to bide my time, its doing a perfectly fine job
21:42 rue_shop4 when I'm not at work I'm not making money, and so I dont go to length to spend money
21:42 rue_shop4 if I wanted I could spend $5000 easy on a printer
21:43 rue_shop4 and the end of the day, I'd just be $4700 shorter
21:43 rue_shop4 my mortguage would just as much like the $4700
21:44 Tom_itx i suppose
21:44 Tom_itx just seem to be getting frustrated
21:47 rue_shop4 yea, its part of the thrill when it all gets working again
21:47 e_shop4 dumps 4 laser printer fusors on the shop f
21:48 rue_shop4 lets see, NTC
21:48 rue_shop4 resistance goes down as temp goes up
21:48 rue_shop4 these are all cold from being outside
21:48 rue_shop4 so I'm looking for more than 100k
21:49 rue_shop4 hopefully one of these is what I want, its cold out
21:49 rue_shop4 :)
21:49 rue_shop4 Tom_itx, you have to admit, if making it work weren't a challange, the victory wouldn't be so great
21:51 Tom_itx yeah i suppose
21:54 e_shop4 follows two little orange wires into the fusors
22:00 rue_shop4 nope, about 2Mr
22:00 rue_shop4 next!
22:02 rue_shop4 nope, about 300kR
22:02 rue_shop4 next!
22:05 rue_shop4 drat, 3 of these are the same
22:06 Jak_o_Shadows right. so my height of tree function is SCREWED.
22:06 Jak_o_Shadows it works for balancing though
22:07 rue_shop4 height of tree?
22:08 Jak_o_Shadows Learning AVL Trees
22:09 rue_shop4 AVL?
22:09 Jak_o_Shadows uh, a complete binary tree
22:10 rue_shop4 just binary trees?
22:10 rue_shop4 whats avl stand for?
22:11 Jak_o_Shadows Binary tree's arn't automatically complete, they have blank gaps and stuff. So they can be quite tall. So they have O(n) stuff. AVL Trees have O(logn) stuff
22:11 Jak_o_Shadows It is named after the inventors
22:11 rue_shop4 how does a tree get generated with a gap?
22:12 rue_shop4 you mean as far as balancing?
22:12 rue_shop4 how could a tree hight fucntion screw up?
22:13 rue_shop4 and more importantly why the hell did the most obscure temperature sensor in the world get used on the reprap
22:13 rue_shop4 why the hell didn't they just use a PT100 like everything else does
22:18 Jak_o_Shadows well, because it was wrong.
22:19 Jak_o_Shadows yeah, balacning
22:36 rue_more how do you even make a memory structure for a binary tree thats got a gap in it
22:36 rue_more ya know, I DO have a bunch of K type thermocouples
22:37 rue_more AND the really expensive chips that read them
22:38 rue_more ugh, they are using maxim chip now
22:38 Jak_o_Shadows eh, you just do linked nodes.
22:38 rue_more and completely miss links!?
22:39 rue_more AD595?
22:39 Jak_o_Shadows I feel I am communicating badly.
22:40 Jak_o_Shadows My understanding of a complete binary tree is that looking at the visualization of it, going from the left for each "row" of keys, there is an item there until the last row.
22:40 rue_shop4 yup I have 3 AD595
22:41 Triffid_Hunter rue_shop4: we need a temperature sensor where the derivative of measurement is still usably high at 200°C
22:41 Triffid_Hunter 100k thermistors do well in that regard, with a 1k-4k7 pullup
22:42 Triffid_Hunter we skipped thermocouples because the converter chips are expensive, and thermistors are available to 300°C
22:43 Triffid_Hunter I use the honeywell axial ones fwiw, they have sturdy leads whereas the leads on most of the other types are hair-thin rubbish
22:43 rue_shop4 haha, I just found out my ".42mm" nozzel is .38mm
22:43 Triffid_Hunter rue_shop4: doesn't matter too much, nozzle diameter cancels out of most of the math
22:43 rue_shop4 hah
22:44 rue_shop4 and everyone scorns my 1mm
22:44 rue_shop4 ok I have 9 of the AD595, so I think I'm gonna switch to thermocouples
22:44 rue_shop4 then I can adjust other marlin things
22:44 Triffid_Hunter rue_shop4: basically the only thing it affects is sensible width/height, width should be larger than nozzle and height should be less than 80% of nozzle
22:44 rue_shop4 that or end up with a hopelessly broken machine
22:45 Triffid_Hunter other than that, volume in the top = volume out the bottom regardless of nozzle diam ;)
22:45 rue_shop4 heght of the layer?
22:45 Triffid_Hunter yep
22:46 Triffid_Hunter so with your 1mm nozzle you could use 0.8 layer height and 1.2 width and it should come out looking wonderful
22:46 rue_shop4 that explains a few things I was discovering
22:46 Triffid_Hunter I have an 0.35 nozzle, currently printing at 0.25 height 0.6 width
22:46 rue_shop4 how is the trace width adjusted? just indirectly?
22:47 Triffid_Hunter rue_shop4: yeah it's a slicer setting. when you increase it, it pushes more plastic and spaces the lines further apart
22:47 Triffid_Hunter all the math is volumetric, so 0.25 height * 0.6 width * 10mm length = some volume, then it calculates volume / ((filament diameter / 2)^2 * pi) and that's the E word in the gcode
22:47 rue_shop4 the multiplier?
22:48 rue_shop4 hmmm
22:48 Triffid_Hunter I don't use the multiplier, easier to send M92 Ennn mid-print to adjust things, but you don't have to recalibrate just to change width, the slicer already knows the volume of the bead being extruded and the cross-sectional area of the filament, the rest is basic math
22:49 Triffid_Hunter in fact I don't even use my real filament diameter, instead I put 1.128379 as my filament diameter which is a magic number such that ((d/2)^2*pi) = 1.0 and all the E words in my gcode are in mm^3 (volume) rather than mm (length)
22:50 Triffid_Hunter then I tell my firmware how many steps per mm^3 (volume) and everything works :)
22:53 rue_more hmm
22:55 rue_more http://reprap.org/mediawiki/images/0/0a/Thermocouple_1.0_schematic.png
22:56 rue_more thats it?
22:56 rue_more huh
22:56 rue_more nowonder they are such expensive chips
22:58 rue_more oh these only do K
22:58 rue_more nowonder its so simple
23:03 Triffid_Hunter rue_more: yeah all the hard stuff is in the chip
23:04 Triffid_Hunter rue_more: also, the chip should be kept at the same temperature as the TC's connector so the cold-junction compensation works properly
23:04 Jak_o_Shadows thermocouples are a bit of a pain in the ass really.
23:04 Triffid_Hunter yep, thermistors are a lot simpler :)
23:05 rue_more unless you dont ahve one
23:05 rue_more casue someone used one thats ideal, but obscure as hell in cunsumer equip
23:05 rue_more :)
23:06 rue_more laser printers use 300k
23:06 rue_more or 1.2M
23:07 rue_more industry uses pt100
23:08 Triffid_Hunter rue_more: http://au.rs-online.com/web/p/thermistors/2508333140/ <-- these are the ones I use
23:08 rue_more I think I just bought 20
23:09 rue_more 10 of the small glass ones,
23:09 rue_more and 1 thats encapsulated
23:09 rue_more I will see them in ~22 days
23:09 Triffid_Hunter hm, the epcos ones? need some serious strain relief on those, the wires are literally hair-thin
23:09 rue_more yea, thats what I have now from a laser printer
23:09 rue_more but ya know
23:10 Triffid_Hunter I broke stacks of them from even minor movement where the wire exits the glass until I got jack of it and switched to the honeywell ones
23:10 rue_more I think its acutally a rewire I did thats screwing me over right now
23:10 rue_more anyone ever tried using a diode?
23:10 rue_more I dont know if the junction would be ok with the prolonged heat
23:11 Triffid_Hunter rue_more: no they're not, diodes are only good to 150° and even then they have a drastically reduced lifetime
23:11 Triffid_Hunter might be able to use one for the bed heater
23:11 rue_more http://reprap.org/mediawiki/images/0/0a/Thermocouple_1.0_schematic.png
23:13 Triffid_Hunter what chip is that? AD597 didn't have that many pins from memory
23:14 Triffid_Hunter unless it's the tin can package which I never paid any attention to because they're hilariously quaint
23:14 wolfmanjm FWIW I was trying 1.2width and 0.6 ht with my 1.0mm nozzle and it came outawfull, switched to 1mm width and 0.6mm ht and it looks a lot better
23:14 Triffid_Hunter wolfmanjm: was there a flat around the nozzle? the plateau needs to be bigger than the width you want
23:14 wolfmanjm guess i could try 1.2/0.8
23:15 Triffid_Hunter 1.2/0.6 should work fine
23:15 wolfmanjm yea there is supposed to be a flat
23:15 Triffid_Hunter if it has a nipple rather than a flat, then anything != nozzle diameter is gonna look bad
23:16 wolfmanjm https://flic.kr/p/qxrsP8 vs https://flic.kr/p/qg27wc
23:16 wolfmanjm it is the volcano they say they have a big flat
23:16 wolfmanjm yea it has a huge flat
23:17 wolfmanjm it may be a flow issue 1 reduced it enough to be able to keep up
23:17 Triffid_Hunter ah yeah, you've been finding it won't go fast
23:17 wolfmanjm may try 0.8/1.1
23:17 Triffid_Hunter then again, 1.2*0.6 is less than 1*0.8
23:18 wolfmanjm no I was doing 1.0/0.6
23:18 wolfmanjm haven;t tried 0.8 ht
23:19 rue_shop4 aha
23:19 wolfmanjm oh yea 1/0.8 is more
23:19 rue_shop4 the resistor is ok
23:19 rue_shop4 its not a 100k, so I thought it was dead
23:19 rue_shop4 but I changed a wire in the middle, wrong
23:19 wolfmanjm seems my flow rate is limited to around 1.66
23:20 wolfmanjm I mean 0.6
23:20 Triffid_Hunter wolfmanjm: 0.6 what? cm/s?
23:20 wolfmanjm 1 * 0.6
23:20 wolfmanjm w * h
23:21 rue_shop4 ah, hooking up the sensor helps too
23:21 Triffid_Hunter ah, that's cross sectional area of the extruded bead, not flowrate.. flowrate is that multiplied by feedrate ;)
23:21 wolfmanjm so my flow rate is around 24 mm^2/s
23:21 wolfmanjm yea
23:22 Triffid_Hunter WxHxL = mm^3 not mm^2 :P
23:22 rue_shop4 Triffid_Hunter, thats the AD595
23:22 rue_shop4 there is a smaller i2c chip from maxim
23:22 Triffid_Hunter rue_shop4: ah ok, AD597 is a bit cheaper, has less pins, comes in soic8
23:22 Triffid_Hunter rue_shop4: the max31855? it's actually SPI
23:22 Triffid_Hunter if you see max6675 around, it's been replaced by max31855
23:23 wolfmanjm oh ok so I need to calcualte the volume
23:23 rue_shop4 ah
23:23 Triffid_Hunter wolfmanjm: then divide by filament cross-sectional area ((d/2)^2*pi) to find extruder mm/s if that's what you're after
23:23 rue_shop4 I have 9 of the ad595 from industrail scavange
23:23 Triffid_Hunter rue_shop4: use 'em then :)
23:24 wolfmanjm nah I already did that to find the speed the stepper needed to go, the stepper was idling at around 2 revs/sec
23:24 rue_shop4 I just got the ntc wroking again, I really dont want to recompile marlin
23:24 Triffid_Hunter rue_shop4: should get yourself a smoothieboard then, just edit a textfile to configure, no need to compile anything ;)
23:24 rue_shop4 hah
23:25 rue_shop4 everyone is eager for me to spend more money on my printer
23:25 wolfmanjm he is saving money by buildoing his own hotend, I doubt he will fork out for a smoothie ;)
23:25 rue_shop4 I'm at $300 w/2kg of PLA
23:25 wolfmanjm but $50 for a jhead is a good inveetment ;)
23:25 rue_shop4 I have a lathe
23:26 rue_shop4 if I dont use it I should be beaten senseless
23:26 wolfmanjm just get a genuine one from hotends.com
23:26 wolfmanjm making hotends on lathes is tricky
23:26 rue_shop4 maybe for th next machine
23:26 rue_shop4 why?
23:26 rue_shop4 I do the final hole in the drillpress
23:26 Triffid_Hunter rue_shop4: well cut yourself an arcol-v4 clone or similar with it.. aluminium combined heaterblock+nozzle, SS barrel with thermal break cut into the side, aluminium heatsink on the top
23:26 wolfmanjm ask reifsnyderb
23:27 rue_shop4 I like the nozzel seperate, I keep changing them
23:27 rue_shop4 I have 4 of them
23:27 rue_shop4 1mm, .5mm .35?mm and ".42mm"
23:27 wolfmanjm I bought some brass stock to try to make a nozzle :)
23:27 Triffid_Hunter why brass rather than alu?
23:28 rue_shop4 hell no, make them from brass bolts
23:28 rue_shop4 brass machines easier, not stickey
23:28 wolfmanjm becuase everyone makes the nozzle from brass, no idea why
23:28 rue_shop4 there do seem to be alu nozzels out there
23:28 wolfmanjm I suspect alu nozzle will wear out quicker
23:28 rue_shop4 gack, I think I have a cold
23:29 rue_shop4 :) all that high friction plastic
23:29 rue_shop4 ok, here goes the farm
23:29 wolfmanjm well it appears making a good horend is more voodoo than science
23:30 rue_shop4 I bet that $350000 medical analizer I have on the deck has some good temp sensors in it
23:30 wolfmanjm I now have 6 different types of hotend, the only one that works with out fail is the jhead
23:31 wolfmanjm although I haven;t really tested the hexagon much
23:31 rue_shop4 well it didnt' instantly jam
23:32 rue_shop4 lets try a print...
23:33 wolfmanjm if it looks like this https://flic.kr/p/fC9EcJ then you did a good job ;)
23:33 Triffid_Hunter wolfmanjm: mine are alu, haven't worn out at all and I've run heaps of plastic, few dozen kilograms at least
23:34 rue_shop4 and...
23:34 rue_shop4 were...
23:34 rue_shop4 j...
23:34 wolfmanjm that was on my large delta, before I "upgraded" to gt2 belts and it has not worked since :(
23:34 Tom_itx are there dimensions for that arcol v4 anywhere?
23:34 wolfmanjm IMO if you are going to make one make a jhead.. it is open source
23:35 Tom_itx i did
23:36 rue_shop4 jammed
23:36 rue_shop4 damnit
23:36 Tom_itx jammit
23:36 wolfmanjm here is an idea, rub vegetableoil on the PLA before it enters the hotend,
23:37 wolfmanjm I think they said canola oil is good
23:37 rue_shop4 now, I might be able to clear it if I cool it down, set it to way too hot and try to extrude again