#linuxcnc Logs
Dec 08 2019
#linuxcnc Calendar
12:26 AM elmo40: yes.
12:30 AM elmo40: maybe the motor overheats. don't know
12:30 AM elmo40: and it isn't the bed motor, it is the carriage motor.
12:30 AM XXCoder: try lower amp on controller
12:36 AM elmo40: and those cooling fits for the stepper
01:41 AM _unreal_: yawn
01:41 AM _unreal_: well I guess its time to go to bed
02:42 AM Deejay: gn8
02:42 AM Deejay: sorry, moin! :-)
03:30 AM miss0r: goodmorning.
03:30 AM miss0r: Land rover has its oil pressure back :D
03:36 AM XXCoder: thats good, its heart can keep going.
03:55 AM p0g0_: miss0r, where did the oil pressure go? some leak, a bad filter seal, what?
09:14 AM _unreal_: any one around?
10:08 AM jymmmm: Well, this is SUPER nice... https://www.youtube.com/watch?v=iIrQNZo21FM
10:18 AM SpeedEvil: https://www.youtube.com/watch?v=6HXSigszx5w - related.
02:06 PM jymmmm: Hose water meter.... https://www.amazon.com/Save-a-Drop-P0550-Water/dp/B0058EOC5M/
02:30 PM Rab: Cheaper, monochrome one in "Frequently bought together": https://www.amazon.com/Orbit-56854-Hose-Water-Meter/dp/B0187BOFD6/
02:31 PM Rab: That seems like a bargain price for a flow meter with GHT ends if it's at all accurate or reliable.
02:32 PM Rab: Seeming bargain price on carbide flush cutters, if you're cutting e.g. a lot of stainless steel wire... https://www.goldmine-elec-products.com/prodinfo.asp?number=G24417
02:44 PM miss0r: p0g0_: Actually.. It was a bit complicated. Which is why I spend a good amount of time trying to figure out where the error was.
02:45 PM miss0r: p0g0_: I had a bad(The thought at the time) roller in a tappet assembly. This roller runs on the cam moving stuff ect.. Anyway it was worn badly, and was causing issues. I determined that it must have been a sub par part and replaced all the tappet assemblys and the cam shaft.
02:47 PM Tom_itx is now known as Tom_L
02:47 PM miss0r: p0g0_: After this small repair my oil pressure had dropped from 4.1 to 3.05 bar. I was somewhat puzzled at this. At first I thought I must have dislocated a cam bearing causing excessive oil flow to escape there - that was not the case. Then I thought that one of the small metal particles from the damaged roller could've lodged intself in the regulator valve on the oil pump, causing it to never close properly - not the case either
02:49 PM miss0r: p0g0_: Turns out that I had a blockage in one of the lubrication channels for the tappet assemblys. Replacing the tappet assemblys got rid(unbeknownst to me) of said blockage.
02:50 PM miss0r: p0g0_: And then it became clear that the oilpump was worn out and now unable to keep up with demand. So I replaced the oilpump and everything is dandy :)
02:53 PM miss0r: so to celebrate getting the oil pressure back, I decided to beach the land rover earlier today: https://imgur.com/a/3AkRpbk
02:59 PM _unreal_: sigh... 3 hour tour comming up again
03:03 PM sync: miss0r: 3bar is not too bad tho
03:21 PM rmu: miss0r: you need something like this: https://www.x-ant.at/?page_id=567&lang=en
03:25 PM JT-Shop: doing some mudding miss0r
04:06 PM Deejay: gn8
04:29 PM R2E4: hi anyone here?
04:29 PM R2E4: Need some help with homing gantry.
04:44 PM R2E4: No one here?
04:45 PM Tom_L: no one with gantry help
04:46 PM Tom_L: are you running 2.8?
04:46 PM Tom_L: 2.8 simplifies it
04:47 PM Tom_L: http://linuxcnc.org/docs/2.8/html/config/ini-homing.html
04:49 PM Tom_L: Negative HOME_SEQUENCE values indicate that joints in the sequence should synchronize the final move to [JOINT_n]HOME by waiting until all joints in the sequence are ready.
04:54 PM R2E4: 2.9,
04:54 PM R2E4: I have two different type switches on the gantry. They are cause issues. How do I do via hal?
04:55 PM R2E4: problem is one switch is prox and other is mechanical and the mechanical releases at a different place then it engages.
04:56 PM R2E4: I need to do it in hal and just use XYZ instead of XXYZ
04:56 PM R2E4: with stepgen rerouting magic
04:57 PM R2E4: stepgen.03 sends to two outputs on 7i76
04:59 PM R2E4: Tom_L: you know how to do that?
05:15 PM R2E4: anyone?
05:16 PM p0g0_: miss0r, gee, that sounds like a painful journey- getting the tappet & oil pump back in spec has to be good tho.
05:25 PM burklefoo: R2E4 if you are using 2.8 or greater I would setup XXYZ and just use a single home switch (until you get matched switches)
05:27 PM R2E4: How do you tell it to just use a single home switch? If you remove the switches from the sl;ave joint it will rack because that joint wont move
05:31 PM burklefoo: you feed the single switch signal to both joint home inputs
05:34 PM R2E4: burklefoo: yeah, I'll try that, but I removed the switches from the slave joint and it starts to home. I will; see what it does, may not need to do that.
05:35 PM R2E4: that didnt woprk, it racked after it found the switch.
05:37 PM R2E4: burklefoo: I dont think sending the one switch to both home inputs wont work. It will search ok but it does one at a time after and that will rack.
05:38 PM R2E4: I need to figure pout how to send stepgen to both stepgen.0 and stepgen.1
05:48 PM burklefoo: If it works with two switches it must work with one
05:49 PM burklefoo: (that is the homing state machine cannot distinguish between 2 switches and on feed to two inputs)
05:50 PM R2E4: yes, but it will search and find the two, then backoff ond one moves, the other wont move.
05:51 PM burklefoo: Really, that sounds completely wrong (it would rack with two perfectly aligned switches if that were the case)
05:53 PM burklefoo: is should _never_ move one paired joint without the other unless the switches are not aligned
05:55 PM R2E4: The two different switches one is prox and the other is a mechanical switch where the release is not at the same place..
05:55 PM burklefoo: Thats why I suggested using only 1
05:55 PM burklefoo: (until you get matched switches)
05:56 PM R2E4: So it searches ok finds both but when it backs off the mechanical switch does not release yet causing timing issue. and gantry racks
05:56 PM R2E4: I tried that, it finds it with bopth ok, then it backs off one while the opther doesnt move and it racks.
05:57 PM burklefoo: Sounds like its not setup correctly (because one switch feed to both paired joint home inputs should not cause racking)
05:58 PM R2E4: I dint do that
05:58 PM R2E4: didnt
05:59 PM R2E4: I see what yopur sayimng, I can just jump the input on 7i76 to make both at the same time....?
06:00 PM burklefoo: you could do that (or just do it in hal)
06:05 PM R2E4: WOW, thank you.
06:05 PM R2E4: It worked....
07:32 PM _unreal_: sigh....
07:32 PM _unreal_: long ass drive
07:32 PM _unreal_: I'm back.... who didnt miss me?
07:33 PM XXCoder: I missed you
07:33 PM XXCoder: my breathing threw off my aim
07:33 PM _unreal_: duck and weave baby
07:34 PM XXCoder: you ddint seem to notice? kept walking by
07:34 PM _unreal_: I have an xwife, I'm numb to pain
09:08 PM _unreal_: hi
09:46 PM SpeedEvil is now known as Guest61040
10:04 PM ve7it: _unreal_, good news.... I found my old servo code for the stm processor... from 2015... took a while to get it compiling again
10:06 PM ve7it: _unreal_, I think it will drive your boards... it outputs pwm and dir
10:08 PM ve7it: _unreal_, it builds for the Nucleo_F411RE demo board ( which is nice... it has floating point processor )
10:10 PM ve7it: _unreal_, https://www.st.com/en/evaluation-tools/nucleo-f411re.html
10:23 PM _unreal_: hi
10:31 PM _unreal_: ve7it, still around?
10:31 PM ve7it: yup
10:31 PM _unreal_: https://drive.google.com/open?id=1dQOn5Lmg9admABkuhdn0zVQwq7D6PP-j
10:32 PM _unreal_: those are the items to be packed
10:33 PM _unreal_: 3 idential motors, one with the serial connection
10:33 PM _unreal_: for siring etc...
10:33 PM _unreal_: wiring
10:34 PM _unreal_: I'd have to show a different photo but on the metal frame the motor controller board has one DB25 header, the board below it has a db15 and a DB34
10:35 PM _unreal_: the db25 is MOTOR power and hall sensors
10:35 PM _unreal_: the db15 i know is a simple stepper controller and dc motor controller
10:36 PM _unreal_: the db34 as I understand it is mainly logic and sensory and the one quadrature, and a POT and limit switches etc.... and serial comunications and all kinds of crap
10:37 PM _unreal_: I am also including the GPS reciver. I HAVE no use for it it is just a reciver. the gps controller is mounted on the lower board.
10:38 PM _unreal_: the metal back plate is from the motor with an encoder. I'm 99% positive its compatable with these motors IF you wanted to try and swap it in to be able to mount the encoder and bolt it down.
10:38 PM ve7it: cool did you ever figure out the wiring from the motors into the driver board? any notes?
10:38 PM _unreal_: down fall though there is no shaft sticking out the back of any of those motors so
10:38 PM _unreal_: heh
10:38 PM _unreal_: into the driver board no
10:39 PM _unreal_: again i'm 99% positive its a 2 sided board only
10:39 PM _unreal_: and there are lots of traces on the bottom side
10:39 PM _unreal_: no parts on the bottom at all
10:40 PM ve7it: ok... not that many wires to work out
10:41 PM _unreal_: but then again I can only clearly identify 2 motors 3 phase traces so maybe it is a 3-4 layer pcb?
10:41 PM _unreal_: traces going from the motor controller chips to the db25
10:41 PM _unreal_: ya there are no quad encoder traces going to the motor controller board
10:42 PM _unreal_: and they are clearly orginized
10:43 PM _unreal_: I have one board I'm likely to desoder the motor controller chips
10:43 PM _unreal_: strip everything off to make it easy to pin trace
10:44 PM _unreal_: again there are only 3 logic chips, 3 motor controller chips, 18 ceramic caps, 8 resistors, a 7085 and two large electrolytic caps
10:44 PM _unreal_: thats it
10:45 PM _unreal_: if all the hall sensors are tied into the chips and all 3 chips ARE setup identical? then its just a matter of identifying everything on it
10:46 PM _unreal_: and solving how to interface the max505
10:47 PM ve7it: have you figured out what the max505 does? what pins does it drive on the output chips?
10:47 PM _unreal_: I can check
10:48 PM ve7it: http://www.micromouseonline.com/2013/02/16/quadrature-encoders-with-the-stm32f4/ if you want some light reading before bed :}
10:48 PM _unreal_: I know it drives the power control of the motor controller let me figure out what pins are the outputs that should go to the motor controller and I can tell you which its driving
10:50 PM ve7it: probably 3 outputs for the 3 driver chips.... but does it set current limit? speed? position?... might be able to tell if we know how its hooked into the driver chips
10:52 PM ve7it: that encoder article above is interesting... in the code I did, I attached interrupts and basically did 2 software encoder interfaces, but there are 2 32 bit timers in the stm32f411 that can count encoders at over 50MHz in hardware
10:53 PM ve7it: just a simple matter of programming :}
10:55 PM _unreal_: ok
10:55 PM _unreal_: just tested
10:55 PM _unreal_: so pin1 of the max505
10:55 PM _unreal_: connects to pin10 REF
10:56 PM _unreal_: directly
10:56 PM _unreal_: to one and only one of the motor controllers
10:56 PM _unreal_: pin2 connects to an other one
10:57 PM _unreal_: and pin 24 connects to the last motor controller chip
10:57 PM ItsFlipp: Any suggestions for end mills to cut a 3mm timing pulley profile in alu?
10:57 PM ItsFlipp: sadly 1/8 is ~0.1mm too big to cut the pocket :/
10:58 PM _unreal_: ItsFlipp, so go with a smaller single or dual flute
10:58 PM _unreal_: I assume this is on a rotory setup?
10:58 PM _unreal_: or are you milling a gear top down?
10:59 PM ItsFlipp: Yeah, looking to cut it on a mill (my puny lathe would probably explode if I tried to do any type of indexing)
10:59 PM ItsFlipp: so top-down
10:59 PM ve7it: max505 pins 1,2,23,24 are outputs of the 8 bit dacs
10:59 PM _unreal_: ve7it, so the DAC outputs from the max505 A B C go to the 1 2 and 3rd motor controllers
10:59 PM _unreal_: ya
11:00 PM _unreal_: driving the REF pin or pin10 of the motor controller
11:00 PM elmo40: _unreal_, what are you using the encoder for? spindle speed?
11:00 PM _unreal_: cnc drive motor
11:00 PM _unreal_: sevo
11:00 PM _unreal_: brushed/brushless
11:00 PM elmo40: coo
11:00 PM ve7it: Thanks... will see if I can figure out what ref is used for on the drivers
11:01 PM _unreal_: elmo40, I have access to luxury yacht tracking satellites now and then as they toss out old one's and buy new one's
11:01 PM _unreal_: they just throw them away
11:01 PM _unreal_: marine grade brushless motors
11:01 PM ve7it: remind me of the driver ship numbers again please
11:01 PM _unreal_: ve7it, it drives the power
11:01 PM _unreal_: I have not shipped yet
11:01 PM ve7it: chip, not ship
11:02 PM _unreal_: you didnt give me the address till what late yesterday
11:02 PM _unreal_: OH
11:02 PM _unreal_: hehehehe
11:02 PM ve7it: A3864?
11:02 PM _unreal_: maxmax505bcng and a3936sed
11:02 PM ve7it: wag
11:03 PM _unreal_: the board also has a cd4051 and a cd4099
11:04 PM _unreal_: not sure what there use is but all of there visible traces go to the short ribbon cable.
11:04 PM _unreal_: ve7it, the ref is used to drive the level of power
11:05 PM _unreal_: it uses a 0-5v ref for "power" or you could look at it like a STEP pin. and an other pin for direction
11:06 PM _unreal_: and it uses the hall sensors as reference to index its phases
11:07 PM ve7it: just looking at the data sheet, it looks like it may set the current limit on the driver
11:08 PM ve7it: https://media.digikey.com/pdf/Data%20Sheets/Allegro%20PDFs/A3936SED(-T).pdf
11:08 PM ve7it: check page 2
11:12 PM _unreal_: ve7it, page3
11:12 PM _unreal_: buffer divider current sense compairitor
11:12 PM ve7it: page 6...Current Regulation.Load current is regulated by aninternal fixed off time PWM control circuit. When theoutputs of the DMOS H-bridge are turned on, current increases in the motor winding until it reaches a value givenby:ITRIP=VREF/(10*RSENSE
11:13 PM ve7it: good to know... one more part figured out
11:14 PM _unreal_: ya
11:14 PM _unreal_: once its shipped what will you do with them ultimatly?
11:16 PM ve7it: If I dont blow up anything, it would be nice to build a 3 motor drive system that could be used with linuxcnc
11:16 PM _unreal_: I dont know if I'll be able to mail tomrow. I'm going to the machine shop that is building my bigger cnc machine they are starting monday.
11:16 PM ve7it: I have the frame for router table, but dont have any linear rails or drive screws for it... these motors my be a good fit
11:17 PM _unreal_: so tomorrow night I may desolder the motor controllers from the board I already have a few parts removed from
11:17 PM elmo40: 3-phase motor drivers? nice.
11:17 PM _unreal_: and toss those into a static bag to be part of the package
11:17 PM elmo40: 3A @ 50V.
11:17 PM elmo40: good enough for my 3D printer ;-)
11:18 PM ve7it: cool.... too bad the chips are listed as obsolete... just keep collecting those driver boards!
11:18 PM _unreal_: elmo40, bldc
11:18 PM _unreal_: ya
11:18 PM elmo40: 150W motor is decently strong.
11:18 PM elmo40: ve7it, obsolete doesn't mean anything if you already have it ;-)
11:19 PM _unreal_: ve7it, I prob could have collected about 40 ofthese board by now had I had a desire to collect them before
11:19 PM elmo40: _unreal_, what, you don't have a garage shelf FULL of randomly collected 'just in case' items?
11:19 PM _unreal_: but as I said they toss them every now and then so I WILL be collecting
11:19 PM ve7it: yup... when the price is right its time to get busy
11:19 PM _unreal_: sure just not these
11:20 PM _unreal_: ve7it, you should see the bearings I pull out of those
11:20 PM _unreal_: sat's
11:20 PM elmo40: on a side note... has anyone used a treadmill motor&driver assembly on a lathe? just wondering how to control the speed with lcnc if the controller has a manual pot on it.
11:20 PM _unreal_: https://anltup4an.cloudimg.io/height/800/x/https://kitmondo-images.s3.amazonaws.com/SiteImages/Images/Listing/2019/6/IMG_7306.JPG
11:21 PM ve7it: I would love a whole unit to put on top of the house... big old dish swinging around and pointing at the neighbors should get a few comments
11:21 PM _unreal_: upper right and left hand corners of the kind of a Y part of the frame is a HUGE 3" bearing
11:21 PM _unreal_: and the base that is the entire pivot has a 4" set of bearings
11:21 PM elmo40: nice
11:22 PM _unreal_: wouldnt last long with out the dome on it
11:22 PM _unreal_: now if you could build a clear fiberglass dome or plastic one ;)
11:22 PM elmo40: the motors look small
11:23 PM _unreal_: ve7it, https://www.riviera-cbay.com/shop/audio-visual/sea-tel-4004-marine-stabilized-satellite-tv-antenna-40/
11:23 PM _unreal_: those are photos of the very same unit I stripped these parts from
11:23 PM _unreal_: the 4004 series seatel
11:24 PM _unreal_: oh you can kind of see the 4" bearing in this https://www.riviera-cbay.com/wp-content/uploads/2019/07/IMG_0088.jpg
11:24 PM _unreal_: has a dark gray/black plate over it
11:26 PM _unreal_: I have only gotten my grubby hands on one of the big one's once. want to get my hands on the big 80" one's but they dont often toss those
11:26 PM _unreal_: BIG freaking motors on them
11:26 PM _unreal_: well I have to work in the morning
11:26 PM _unreal_: guess I should clear the shit off my bed
11:26 PM _unreal_: I did make good progress on my daughters cnc machine
11:26 PM _unreal_: this weekend
11:26 PM _unreal_: I got most of the wires run finally
11:27 PM ve7it: nice... have a good night... should be a fun project... I will have to start a web page so I can post progress
11:27 PM _unreal_: now I need to make two small pcb's to interface wires
11:27 PM _unreal_: sweet
11:28 PM _unreal_: fyi the motor controllers I'm going to strip off this board so I can assist with mapping pins are going to be loaded with conformal coating
11:28 PM _unreal_: they are Marine boards :)
11:29 PM _unreal_: but as I said before I dont know if it would be better to try and use the boards as they are? or design a custom PCB
11:30 PM _unreal_: would be nice to have a board design that allows for easy tuning/current adjustments etc..
11:32 PM _unreal_: ve7it, one of the things I dont understand and have not bothered to trace out. there are clearly 3 sets of jumpers that are hard jumped. I'm not sure what there use is
11:32 PM _unreal_: jp8 jp9 jp10
11:34 PM ve7it: once I have a board in hand, we should be able to figure it out.... for testing, I would use the driver boards as is as there are not many external connections.
11:38 PM _unreal_: the onlything I dont know if is the chips are wired differently from each other?
11:39 PM _unreal_: 2 are the same motor one is an "encoder" setup when in operation
11:39 PM _unreal_: there are 5 motors in the sat. one stepper, one one dc geared motor with a POT for reference, and 3 bldcs 2 just have hall sensors one has an encoder
11:40 PM _unreal_: the sat bases everything off of the xyz gyro and the GPS
11:42 PM ve7it: gyros could be the sensors for 2 of the motors without encoders
11:44 PM ve7it: encoders dont go to the motor driver board.... they are needed by the positioning cpus and the cpu figures out what direction and how much to turn each motor
11:44 PM ve7it: should work well for cnc purposes
11:46 PM _unreal_: the gyros go to the logic board whichis part of the package that I'm sending
11:46 PM _unreal_: gyros NOT included :)
11:49 PM _unreal_: ve7it, the gyro box has 4 sensors
11:50 PM ve7it: no problem... not needed for our use
11:50 PM _unreal_: 3 gyros yaw, pitch and tilt and there is fluid filled sensor that is like the old mercury bulbs
11:50 PM _unreal_: with multi pickups inside
11:51 PM _unreal_: the gyros sit in a metal box
11:51 PM _unreal_: that has a pitch adjustment stepper motor
11:52 PM _unreal_: ok bed time later
11:52 PM ve7it: yea... it hurts my head thinking about what they had to do to keep a 4' dish pointed at a small spot in the sky while the ship was pounding through rough seas
11:52 PM ve7it: gnite
11:52 PM _unreal_: ve7it, one thing is for sure
11:52 PM _unreal_: those sat's are VERY VERY VERY well balenced
11:52 PM _unreal_: balanced
11:53 PM _unreal_: lots of steel weighted blocks.
11:53 PM _unreal_: https://www.psicompany.com/man-prod-info/seatel/seatel-4009-satellite-antenna-without-dome.jpg
11:54 PM _unreal_: everything that is a blue colored block is steel and just dead weight for balance
11:55 PM ve7it: I would not want to make that for $15k and make it work!
11:56 PM _unreal_: 15 lol most of those sats sell new for $30k
11:56 PM _unreal_: base price
11:56 PM _unreal_: for 40" dish and up
11:57 PM _unreal_: I mean the bearings in them alone run around $500 total
11:57 PM _unreal_: ok now I mean it going to bed