if i update the kernel will it brake emc2? :P
naw, emc2 just won't run!
bugger, never easy to fix stuff in linux
shitload of f**kup just to get a simple touchscreen running =(
lilalinux is now known as lilalinux_away
hmm i'm having servo amp issues
one may be dead
that's not a good sign
may be a wire though
one servo is holding, the other one seems to be dead
i checked the impedence across the wires going to the motors, and one is basically open, and the other is 0 ohms
you might check the brushes
but the weird part is the one that is open is the one that is working
oh that may be a short then
the old controller worked, and so did the servos
so i doubt it's the brushes unless it went bad by sitting for a month
gotta double check your work then LOL
my work is fine
i didnt touch the wiring going to the servos
other than at the amps
but that's just 2 wires, and it's right
so your drive takes a velocity signal and an enable?
i'm going to go disconnect the wires from the amp and check the resistance
but the motors still hold if none of that is hooked up
except for enable
the z axis holds, the z axis is dead
i'm gunna go check the wires now, brb
it will only hold if the tach is wired right
I bet it's the tach, power supply, or enable/inhibit wiring
there is a drift pot on my drives that robh was telling me about that needs to be set
power supply and inh are set right
i'll double check the tach again
* jthornton wanders out of the beer cave to check on the surf and turf
* Jymmm hopes jthornton's beer is 80F (bastard not sharing with us)
Well, not so much the sharing part, as much as taunting us with it!
GonMD_ is now known as GonMD
hmm seems like the tach is wired right
MarkusBec is now known as MarkusBec_away
Hello all. Does anybody have sample hal files for use with a dual drive (X1,X2) servo setup, or know where I might look?
is there some package in the ubuntu release that will give me a working build enviroment?
seem to be missing libc headers at least
"sudo apt-get install build-essential" will install compilers and most things you need.. for compiling emc check out the wiki there are a couple others needed as well.
actually i'm trying to compile a tourchscreen driver
build-essential looks like it's just what i need. thanks
cradek: what's the input scale, or how can i figure it out?
nm, i think i can figure it out
hmm well i guess i need to figure out the gearing
danimal: the way I always did it is make sure your encoders are counting.. open halmeter and watch the encoder position.. move your machine say one inch by hand and you will get a rough idea of the input scale by the difference before and after the move..
good call, i'll try it
however i know it's extremely high
6 figures i think
4096 counts per rev of the resolver, and there's quite a bit of gear reduction involved
hmmm looks like 10:1 reduction
i set the count to 1, and made one full rev, and it came out to about 40960
now i gotta figure out the pitch
of the lead screw
204800 counts per rev
cradek: is that right?
looks right to me
i know i'm talking to myself but it helps me
yeah (by memory) I think 204800 is right
yea, i tihnk it is, at least roughly checking by my callipers
nothing's working except for the resolvers
trying to figure out what's going on
i quadruple checked my connections, it all seems ok
SWPadnos: you around?
[02:59:08] <skunkworks> http://www.electronicsam.com/images/KandT/conversion/xaxis/ballscrew.JPG
[02:59:36] <skunkworks> http://www.electronicsam.com/images/KandT/conversion/xaxis/start.JPG
[02:59:47] <skunkworks> http://www.electronicsam.com/images/KandT/conversion/xaxis/open.JPG
[02:59:56] <skunkworks> http://www.electronicsam.com/images/KandT/conversion/xaxis/stripped.JPG
skunkworks: where did oyu find the 14" tall midget?
looks odd - doesn't it?
that is my hand
oh youre a 14 tall freak, didn't know
hmm - this one now works
[03:21:07] <skunkworks> http://www.electronicsam.com/images/KandT/conversion/xaxis/stripped.JPG
must not have uploaded right
i think my amp is bad
yup, swapped amps, and now the z axis doesnt work
so it isnt my wiring
does anyone have a way to cross reference a transistor, other than google?
google is the way
dammit i cant find what i need
improve your google-fu
the part has probably been discontinued long before the interwebs
and dumbass me forgot to throw my sheets in the dryer, so now i gotta stay up for another 2 hours
hate it when that happens. you should build a robot to do your laundry
also it takes two hours do dry sheets? i think your doing it wrong.
dryer cycle is 90 minutes
i just unloaded some stuff so hopefully they'll dry faster
i need to buy a V cutting bit
i need to buy a servo drive
wonder if walmart would carry a 60 degree V bit
this is america, i want my cutting tool 5 minutes ago.
grind one yourself
my brother works at a CNC shop and my dad at another
i guess i just need a blank. and i could even go sharper then 60 degree
time for dinner, adios
danimal_garage: is the big fuse ok?
do you know what happened? do you remember having something wired wrong for a while?
yea, the fuses are good
(I have a schematic for those somewhere)
not sure, i must have done something wrong
yeah, I was wondering if you had a guess at what it was
only thing i can think of is maybe it was trying to move, but it wasnt able to
what's the symptom?
and burned somethig out
the amp does nothing
nah, they're current limited
how is the enable/inhibit line wired?
i swapped all the wires from one amp to the other, and it definitely seems to be the amp
just to ground for now
so all wiring, tach, enable, power supply all stayed the same but you swapped the amp itself?
tell me about it
any obvious stuff like burn smell? burnt traces?
broken solder joints?
i even pulled off the heat sink
not that i can see
i'd have to pull the board so i can see the back to be sure
remind me and I'll look for my schematic tomorrow. maybe we can do some basic troubleshooting.
doesnt look like there's much to it
i may pull the board and thow a meter on those big transistors
you mean just check for power there?
hey maybe skunkworks would sell you a skunkworks-special amp
no, check between the base and collector
see if it's open in both directions
and the emitter
what's a skunkworks special amp?
his homemade not-so-dumb H bridge
but you want your velocity mode if you can... forget it
on the one axis that does move, it only goes atiny bit then gets a following error
the one good thing about having a machine shop is there's no shortage of things that work as bottle openers
what tuning have you done?
not much, still trying to learn what i'm doing
i set the input and output scale
and i started the "p" at 20
i and d are at 0
you'll have to increase MIN_FERROR to get it to move very far when it's badly tuned
maybe even 0.1 or 0.5 inch
then get your halscope plots going, and start minimizing that following error
yea it's like at .01
you can zero out the steady state following error (error at constant speed) by adjusting FF1. FF1 is your main tuning knob with velocity mode
havent used halscope yet, get to learn something new
rapid speed is about 4ips, rapid velocity command is maybe 7v? so FF1 will probably be something like 4/7 = 0.6
assuming your scaling is in inches and volts
you set the output scale to 8.0, correct?
since it's 8v?
lemme see if it's on
maxoutput is 8
ok yea thats what i ment
MAX_VELOCITY = 3.33334
MAX_ACCELERATION = 20
INPUT_SCALE = 204800
OUTPUT_SCALE = -10
FERROR = .002
MIN_FERROR = .0015
MAX_OUTPUT = 8.5
oh output scale i have at -1
is the scale -10 because the gear ratio is 10:1?
for the resolver and tach
no, that's because pwm command of 1.0 is full duty cycle on the pwm = 10v for the dac output
and they are inverted I guess, so -10
well at least i got the input scale right!
I guess I have my rapids set at 200 ipm
yea, same as the original control
ah I misremembered it as 250
such a short travel it doesn't matter one bit
my brochure says 200
200 is still too fast lol
if something's in the way it sure is
however my mill seems to greep at 110
my mill is 450 but I ran it briefly at 900 to see if it would work (it did)
but that's way too fast - it seemed pretty violent
yeah it wastes no time getting into position
yea, i would think the machine would start dancing
i'd kill for 300
but 110 is pushing it with these steppers
maybe if I had a heavier floor I'd run it faster, but I don't like how I can feel the floor move
even the bandit control topped out at 100
100 is pretty typical and good for a stepper machine
yea, sometimes i wonder myself... i have nearly 10 tons on my garage slab
im not even comfortable pushing my machine that fast atm
running slow doesnt matter right now, and id rather get comfortable with my machine first
dont blame ya GonMD
my thing rapids at 30ipm =(
my mill starts stalling at 120
itll go to 60 but it starts f'in up past 45
mine used to do that too
i usually rapid mine around 60. but thats about what im comfortable with right now
ugh, i'm not looking forward to having to repair this drive
so put gasoline on it and set it on fire buy new one
go Office Space on it. a field, a baseball bat
GonMD, at 60ipm, it would take 37 seconds to get across my machine
in the x axis
big machine. lucky
not if you're in a hurry lol
what drive is it?
GE hi-ak pwm50e
being in a hurry with cnc machines isnt a good combo :p
is that a fanuc?
it is if you need to make a living with them
eric, good question, i dopnt think so
these are probably pre-fanuc
I have an old GE robot from about then
the electronics didn't look too useful
danimal_garage: what kind of work do you do with your machine?
i make bicycle parts
how about yourself?
used to work with automatic swiss Cincom lathes for 4 years, then got laid off, then did a bit of contract work over the summer and bought a CNC router, and im hoping to get into the movie prop making scene
also firearm engraving here locally. for SBR stuff mainly
thought about trying to do robotics for that
i need a good machine for engraving
I'm filing the casting marks off of a steel bottom bracket shell
so I can build it into a frame without having any casting marks on it
what kind of material?
i'm suprised it's cast
it's all investment cast nowadays
well i gotta run to get some dinner.
very little pressed steel left
my buddy uses machined parts
fillet or tig you can do that, I'm using lugs
all his BB shells are turned on a lathe from tube stock
show me some pics when it's done
i love custom frames
I just got some of the machined bbs, super light
the more homegrown, the better
yea, my friend's frames are less than 4lbs for a 29er mtb frame in steel
yea, he builds some nice stuff
he used to work for intense, so he learned all his stuff there
he's building me a scandium frame with ti seat stays
that sounds cool
bonded, I assume?
yea, it gives it a little compliance for rocly trails
only bonded where the scandium and ti meet, obviously
i'm going to make a sliding lower yoke so i dont need chain tensioners or an ecentric bottom bracket
it'll be a single speed
I was just looking at sliding dropouts, cant afford them
yea, they're not cheap
anyways, i'm off to grab a burger, adios
cradek, iirc you have a laptop with p3 7xx proc that will run emc? What is it? Friend gave me a Dell lattitude C500 with a p3 750 that is taken all apart with a broken power jack today.
lilalinux_away is now known as lilalinux
MarkusBec_away is now known as MarkusBec
how does one enable touchy when installed?
how do i enable the touchy interface in emc2-dev?
frallzor beat me to it
[12:59:21] <jthornton> http://www.linuxcnc.org/docview/devel/html//gui_touchy.html
replaxing axis with touchy just crashed it so =P
in the machine *.ini
MarkusBec is now known as MarkusBec_away
MarkusBec_away is now known as MarkusBec
lilalinux is now known as lilalinux_away
* archivist_emc puts $3 on even to win
archivist_emc you into touchy? having issues getting it up and running =)
im thikning it should be as easy as replacing "axis" with "touchy" in the machine *.ini
not tried it yet, although I do have a touch screen
gotta read the manual
with the right stuff installed
touchy is there
got some help doing that =)
but applying it to a config is harder =(
not hard at all
even I did it
still, isnt that just for making some buttons work?
no, you have to follow the instructions for touchy to work
touchy "requires" a few things to work
last touch screen on a machine I saw had a screwdriver through it.
don't work there anymore I guess...
Touchy requires you to write a file named "touchy.hal" to make these connections. Touchy executes the hal commands in this file after it has made its pins available for connection. is this what you mean?
heh - forgot to clean out the mouse nest.... http://www.electronicsam.com/images/KandT/conversion/xaxis/start.JPG
how to do that then? cant say I've done that stuff before
frallzor: I think if you have an empty file called touchy.hal it will work
where should it be placed?
in the same folder as the rest of your config files
yea, I just have an empty file called touchy.hal
done now lets see if it plays
nope same issue
placed in the machines config-folder
is that the same folder as the ini file?
are you running 2.4?
the person that helped me got it running but not with a config and same issues with a proper config
in the [EMC] section of your ini file you have MACHINE = EMC-TOUCHY
no the my-mill machine
should it be emc-touchy?
kanzure_ is now known as kanzure
I'd thought it would be as easy to use a proper config and just change from axis to touchy
it has to be MACHINE = EMC-TOUCHY
should it work then?
it doesnt =)
display = what? still axis
or something touchy related?
nothing but axis works and then its still axis
add touch to axis :)
display = touch like that or?
no I was joking
the error is about the display anyway
what's the error?
well I assumed it was called touchy so the error is it cant start it
cant execute DISPLAY program touchy
it is called touchy
then its odd
are you sure you're running the version you built?
DaViruz helped me getting it going and got an instance of it working
but not with a proper config
so it should work
ok this isn't really giving us anything to go on. could you copy and paste everything you type, and everything you get in return while trying to run your devel version, to pastebin so we can see it?
frallzor: actually it might be my fault
but you got it running so it should work? =)
do you have anything running on it now?
nothing but some editors
select my-mill now and see if it works
one error still though, for some reason it cant open /home/cnc/emc2/nc_files
never mind that now
it says so but the file in there is visible
the docs say:
$ . scripts/emc-environment
but that doesn't work, you have to start it with ./scripts/emc
else it launches the old version
if I close it now
and want it going again
yes that does work
if not, something is wrong
to launch with a specific configuration it says
$ scripts/emc configs/sim/sim.ini
yes you can do it that way too - there are several ways to start it
isnt there an easy way?
so the path should be set so you can launch it with just "emc"?
and it will start the one in the build directory?
assuming you have run emc-enviroment
well I'm off to do other stuff, hope you guys get it.
so DaViruz there isnt an easy way to do i? =)
make a "bat-file" and use as launcher?
frallzor: I sorry but I'm a dumb ass this morning
MACHINE is the name ...
my launcher command is : /home/john/emc2-dev/scripts/emc /home/emc2/configs/touchy/touchy.ini
frallzor: you can right click on the desktop and create a launcher
lilalinux_away is now known as lilalinux
frallzor: How is your build going?
frallzor: did you get touchy working?
DaViruz did a "hack" to get it going via the original menu
so you start it like normal
im just playing around now, loaded some coda via the menu but how to run it? =)
you have a touch screen?
you do that from the Auto tab I think
thats where I am
no option for run =)
I'm up in the garage atm and don't have a touchy config on this computer
is the estop reset and the machine on?
well im just playing around with no machine now
back to the manual LOL
but the "machine" should be working
you have to have a hard button for cycle start
thats no fun
[15:46:22] <jt-plasma> http://www.linuxcnc.org/docview/devel/html//gui_touchy.html#r1_1_1
guess ill have to add those when making a nice controlpanel then
or make a pyvcp one
wow it is up to 35F in here all ready
* jt-plasma wanders back to the beer cave
but all in all touchy is pretty nice =)
yes it is
hope this computer lives longer than my old laptop
died on me =P
death by dust
just need a nice little arm now with vesa 75 =)
what is vesa 75?
a standard for "attaching" monitors and such to stands, or what to call it
ok, I didn't have a clue :)
75mm bolt square?
yea, I just found 4 little rubber plugs on the back of this flat screen and they are about 75mm apart and have some threaded inserts under the plugs
for mounting to stuffs like wallmounts and such =)
something fun you can do with a touchscreen until the mill is done? :P
a touch screen would be nice, but i know the coolant would disolve the screen lol
frallzor: maybe invent interactive porn?
cradek: are you on?
danimal_garage good idea
will keep me pretty busy
i thought so
refreshing the chassis to the computer I got could keep me busy too
new coat of paint
what are you waiting on?
time to fix stuff =)
the time I got is good enough for tinkering
not serious work
I'm waiting for a clue lol
a clue as to what i'm doing
hmm I want a onscreen keyboard...
autologin would be sufficent
yea, typing passwords is for the birds
how to make ubuntu autologin?
i think you can set that up at the default login manager (in ubuntu)
i dont exactly need a password for my mill or lathe lol
you have to log off and set it there
(i don't know an another way, there might of course be one)
(and i don't use ubuntu... :P)
assistive technologies perhaps
lets see if its kicking
bye bye keyboard
how durrable are those touch screens?
this one should be pretty durable
i'd love one, but i'm afraid of my coolant eating it
some reinforce version for heavy duty use
i'm more worried about chemicals
water solulable coolant eats everything
what about one of those thin layer screen protectors?
danielfalck a plastic cover will work
a sheet if nice plastic that will stand up againgt the chemicals
the only thing i could see being a problem, is coolant buildig up and running down
so i suppose as long as you wipe it down
make a windshield wiper :P
just make sure you press fast enough to avoid it
[#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website: http://www.linuxcnc.org/,
wiki at http://wiki.linuxcnc.org/"
bored bored bored
play some guitar
or work on a car
I cad for money, anyone interested? :P
Anybody have a sample hal's for a dual drive gantry (X1,X2) using servos? I would like to see how to properly home and properly set the gantry skew.
well , i am just finishing theese test thanks to micges
[19:06:14] <frallzor> http://www.mic.com/
that nema 4 powerstation on that site, that mount for the computer
anyone seen them somewhere to be bought?
for gaintry you have to update drom brach joint axes 3
and there is sample config
perfect, thank you.
[19:08:37] <isssy> http://git.linuxcnc.org/gitweb?p=emc2.git;a=commit;h=4de3d0514cca59dcedfc0881e4a0ecf16e4c91d8
ugh i think i'm going to have to replace my servo amps
i dont know
just for fun then?
well they're NFG
has anyone built EMC on the latest version of Ubuntu?
hmm well i found some amps
i wish i knew the amperage of the servos i have
probably over 9000
9000 amps huh lol
yeah. i like to crank my machine all the way up to 11
but... but... how can that be, the dial only goes to 10!
you must be a rebel
*insert Mr T babble about hacking the code*
mainly im just bored of engraving MDF and aluminum. but its all i can do till i buy cut3d :/
i wanna cut me some video game props
i wanna cut me some cheese
cradek: are you awake yet?
you didn't smoke the other one too did you?
oh right I was going to look for the schematic
i think i did :(
do you know how many amps these motors draw?
i slipped with the meter's probe :(
saw some smoke
do you see a "catalog number"? starts with 3N2100
yes that matches my book
for that model it says peak 12.5A, 6A RMS, 90V
ah, darn, i found a nice 2 axis baldor drive for cheap but it's 5A/10A
what part smoked?
not sure, it's a little dark around the resistor i was probing
what Rxxx number?
no parts were charred
hold on i'll check
i think the probe contacted one leg of the resistor, and one leg of the next resistor as well
143 had 90v
R143,R147 are a divider from +90 to -15, used for undervoltage fault sensing
yep seems right
if you smoked one, it's faulted thinking the +90 is undervoltage
i was getting 12v on one side and 90v on the other
now what is it?
after that happened, i was only getting 16v between the 90v and bottom arm connectors in the middle of the board
hold on, i'll check
right about 7v
both amps are only getting 7v
(7v on the 90v side of the resistor)
the other side is still 12v
[20:26:39] <danimal_garage> http://cgi.ebay.com/Baldor-Motion-Dual-Axis-DC-Servo-Drive-TSD-100-05-2-U_W0QQitemZ250558402956QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3a5671d18c
Heh, they are 30 minutes away
yea, close to where i got my shizuoka
R143 hooks right to +90. why isn't it +90 now?
very good question
i'm getting 98v at the amps
but i'm loosing it somewhere
i was trying to figure that out earlier
that terminal in the middle of the board that has the 2 heavy gauge wires coming from it...90v and bottom leg..... i'm only getting 16v across those, when i used to get 98v
if i check from the 90v terminal to the common terminal coming from the power supply, it says 90v. If i check from the bottom leg terminal to the positive lead from the power supply, it only says 16v
[20:32:10] <Jymmm> http://126.96.36.199/search?q=cache:2IMZGlbeBaAJ:www.prismelectronics.net/+prism+electronics+morgan+hill+ca&cd=2&hl=en&ct=clnk&gl=us
hey my buddy has a 408 area code
CAUTION the 0 volt terminal of the bias supply _MUST_ be connected to the 0 volt terminal of the 90 volt power supply. energizing the 90 volt power supply with this connection missing will result in damage to the servo.
those resistors on top of the caps are getting warm
i dont follow
did i hook it up wrong?
I don't know
are the logic supply and servo supply grounded together? it says they need to be
not that i know of
my amps have a big sticker on them saying that thing I typed
I'd hope they fused that - sounds way too common of an issue
Jymmm: no kidding
I guess another reason to only test one axis at a time.
guess i cant read
they are fused
danimal_garage: do you have all axis connected?
fuses are out, but yes
that'd explain why i had a bad fuse earlier
danimal_garage: to be on the safe side, *I* would disconnect all but one axis, and verify/test jsut the one
yea, i'll do that
"fuses are out"??? you mean the fuses are blown?
i removed them
ah, well disconnect the control too from the unused axis too - just to play it safe
hmm well one amp is working now that i did what cradek said
cradek: one drive seems to work now. The one drive that was bad still appears to be bad.
'troubleshooting: remove the armature fuse. WITH POWER OFF check each armature terminal's resistance to both the 90 volt bus and the 0 volt bus. In each case the obmeter should read a diode drop in one direction and high impedance with the meter reversed.'
ok will do, hold on a sec
ok, both amps show 0 ohms in one direction from the 0v bus, and open loop in the other direction
ok hook 'er back up to power, and put a scope on test point PL1-C
but when i go from 90v bus to the terminal on the fuse block, it shows like 460k ohms in one direction and open loop in the other direction
is that right?
hmm, I'd think it would be symmetrical
both amps are showing the same thing
and the one is working fine both directions?
does your meter have diode measuring functionality? (i guess it's usually the 'beep' mode)
not sure if it's working in both directions, it kinda goes apeshit when i try pushing the button
it moves, then goes back to where it was
so i guess it does move in both directions
should we work on getting that one working right, or go on with the troubleshooting steps?
well i was thinking if i couldnt get the other amp working, i'd just replace them both with a matched set of something else
so i thought it made more sense to get the amp working first
does that make sense?
celeron55: i think it does have a diode mode
danimal_garage: try measuring the 0/460k ohm directions with it
though i doubt the result matters at all
well the beep mode does nothing in either direction
if you unhook the velocity command from the working axis, and jumper it to ground, you should be able to drive the motor around by turning the balance pot P1 / RA1
it should read the voltage drop on the display
the diode mode shows like .300V
to both directions?
umm... i mean both buses
just one direction
ok i'll try it cradek
if it does show like 0.300V for both buses, then it's good
danimal_garage: just to check: for the diode tests you did 8 measurements, two for each diode, and 4 ways shows 0.300 and the other 4 ways showed open?
if it doesn't... well, then it's not what the troubleshooting step said
^^ per amp
ok it shows:
fuse block terminal-0v=OL
0v-fuse block terminal=.415v
fuse block terminal-90v= .360v
90v-fuse block terminal= OL
are screws in the middle (TB2-A,B) fuse block or post?
ok the fuse block terminal is the one that goes to the heat sink then?
ok the two post (TB2) diodes seem wrong
those are D44, D45
i'll see if i can track them down
between TB2-A,B and those diodes is the current sense resistor made up of R102, 101, 100, 99, 98, 97 all in parallel
I wonder if that's messing up your meter
can you find test point PL3-D?
that is right between the two diodes
you could make the four tests from that point instead of TB2-A,B
i'm having trouble finding the diodes
they are the big ones on the heat sinks
the numbers are under the diodes, so i gotta carefully move them to see
the huge ones
44,45 are the bottom two if you're looking at the board with the logic screws to the right
ok that helps, i found them
they arent numbered from what i can tell
I'm cheating :-)
let me see if i can find that test point
at the bottom, right next to the heat sink, a 4 position thingy
yea, i think i got it, it's the only one not labeled
so what do i check from there?
the 4 diodes?
that's between the two diodes we're questioning
so do the 4 tests, to 0v and to 90v, each way
ok doing it now
when I do mine, I get 3k ohm one way and about 20k ohm the other way
(which seems weird)
all four test the same (I took off the heat sink to test them directly)
i got -.469v from d-0v, and OL when reversed
and i got 0000 from 90v-d, and .489v when reversed
0 seems pretty wrong
not sure the difference between 000 and ol
0 means shorted
OL means open
so from 90 to PL3-D is D44, the lower right one
btw, i checked d to the diodes, and the bottom one on the right was zero ohms in both directions. the rest were only shorted in one direction, and OL in the other
time for a diode?
the bottom right one sure seems wrong
maybe pull it out and test it once more out of circuit
if it's shorted, I wonder what else it took out
yea thats what i was wondering too
the things that look most susceptible are Q25, 26, 27; R103, 104, 105, 106
what are q's?
the big ones
oh wait I gave you the wrong numbers
Q28, 29, 30; R107, 108, 109, 110
q's are big resistors?
no - transistors
Q29,30 are the pair on the lower left heatsink
still looking for Q28...
yea, i was looking for those online last night
couldnt find any info on them
none of the resistors look charred at all
maybe you would get lucky and it's just that diode
parts list says only "20A diode"
i can run to the store and grab one today
I think they say 16F20 on them
yea sounds right
did you pull it out and test it?
going to right now
the internet says 16F20 is 200V 16A std recovery
.432 in one direction, and ol in the other
seems to be good, huh
the wire going to it is shorted
in both directions
that's the lower right one?
bad trans perhaps?
yeah try PL3-B to +90, meter both ways
with the diode in or out?
mag wire is insulated, isn't it?
leave the diode out
o.ooo one way, and -ol the other
that's C-E of Q26 (NPN)
the one just like it is Q29 - let's do the same test - it is from PL3-D to PL2-B
it's bad then, right?
but wouldnt it be different since that one diode is out?
c-e is supposed to be OL in both directions, right?
yes good point
I think so ...?
[22:30:04] <danimal_garage> http://www.electronics-radio.com/articles/test-methods/meters/multimeter-diode-transistor-test.php
I can't find my digital meter!
c-e is different if the base has a low R
depends on circuit
i've only been testing the dead drive, btw
oh ok thanks archivist
I'd compare the dead one to a known good one
btw archivist , found some braking resistors http://smg.photobucket.com/albums/v20/danielwilcox/?action=view¤t=DSCN0746.jpg
yea, gunna do that now Jymmm
Q26 C-E is PL2G - PL3B
Q29 C-E is PL3D - PL2B
wouldn't hurt to have them both disconnected from everything too
these pairs read the same for me, 3k one way, 18k the other way
ok i'm checking now
is the circuit online ?
I'm using an ohmmeter
I meant schematic
oh, no unfortunately
I have it on a clay tablet, er, dead tree here
oh hey I was thinking I needed a scanner - I can photograph this page
let me see if I can do it
I use a camera too for quick and dirty
amazing - it does seem readable on my phone
0l in one direction, 0 ohms in the other for both PL2G - PL3B and PL3D - PL2B
ok our meters sure read differently
they often do because the ohm driving voltage and current
[22:40:57] <cradek> http://timeguy.com/cradek-files/emc/ge-pwm50-powercircuit.jpg
sorry it's a jpg - all my stupid phone can do
* archivist_emc remove head and rotates 90 deg
that's a phone feature too, grr
my camera does that :(
danimal_garage: seems like the transistor is ok too then - wtf
fwiw I don't think we have Q25,28
they might be for the PWM99 version?
ah yes the "N0TE" says that
so that wire shouldnt be shorted in either direction, right?
the one that goes to the diode
to d on p3
wires should show 0 ohms either way
lol not what i ment
i ment from the d terminal
PL3-D is between D44,D45
I'm lost, say again what you mean
eh nm its the same on both amps
is anything different between them?
if not, we can assume it's ok and go on to the logic troubleshooting
lemme go check some more
obviously I don't understand how this circuit should test with it all assembled
(we're assuming the other one works right - not sure we know that for sure)
'it all seems the same
ok, guess that was a big waste of time...
did you put the diode back in?
heating up my soldering iron now
should i try using that pot to move the servo next?
oh to see if the working one is working?
if you want to be sure of that - otherwise we can start looking for the source of deadness on the dead one
yea let me do that real quick, just to make sure
it'll only take a minutes
it moves in both directions
but very slowly
but smooth and sounds right?
it is working right if it's smooth and controlled velocity
yea it's smooth
you could try a 1.5v battery to go a little faster if you want
just put the balance pot back where it's about stopped
yea i put it back where it was
well - I think we are going to use the drive train that is there to drive the z axis. It has about .00025 of backlash.. And we think we can get it out - (there is a washer that you grind down to take the backlash out of a bunch of split gears.)
skunkworks: neat - are you finally starting on it?
(if we can get the nut off - it is in the depths of the machine)
danimal_garage: ok let's go on looking for the source of deadness
You are worried about 0.00025" backlash?
is it all hooked up?
well should i put the heat sink back on?
are all the wires hooked up without it?
cradek: yes. I hope so.. ;)
only the one wire that goes to the heat sink is off, but i can screw it on to the sinks below it
yeah just do that so you can get to everything
not like it's going to get warm :-/
the z axis goes through a 'ball spline shaft' > right angle spir gears > 3 sets of split helical gears.
the 3 shafts are x ball screw - z drive train and b axis. all comes down to that spot.
That's a lovely bit of kit
yeah it sure is in good shape
ok fired up the laptop so i dont have to walk as far
cradek: 90v bus on or off?
/inhibit tied to gnd so it should enable
That must have cost a _fortune_ when it was new.
check your +90 +12 -12 for the heck of it
check for voltage at PL4-B
i can hear the one working servo
is the working one safe to drift for a while? (you could inhibit it by taking off that jumper)
Oh noes... The drive of DOOM!!!! Eeeeeeek
yea its fine
no that should be logic level...
what ground are you using?
ok 12v on the good drive, nothing on the non working one
how about PL4-A
Sounds like it is broken to me. :-)
12v on the good drive
nothing on either drive
ok PL1-B=0 means it's not in overcurrent trip
I spent most of yesterday evening trying to figure out my PWM-voltage convertor block. Eventually I realised that the main relay was open (as might be expected when the controller is open on the bench) so there was 5v but not 12v or 24v....
how about PL1-A
andypugh: arg! we've all done that kind of thing.
12v on the bad, 0 on the good
ok that means it's in 90v overvoltage trip
I am disturbed that cradek knows this much detail about a drive.
andypugh: No, you're just disturbed, period.
to clear the 90v overvoltage trip, remove then reapply input power (that's not gonna cut it, is it)
i can try it again
but it's been on and off several times
that's just in the instructions for how to test it
remove power wiat 30s?
for the caps to discharge
That phrase always sounds odd to my ears. We would say "you are just disturbed, full stop". Mentioning periods brings to mind something completely else.
you say your 90v bus is 98?
it shouldn't overvoltage trip until 130
she's right around 98v
have a scope handy?
I found that only one channel on my scope works. That's going to make adjusting my encoder tricky.
dan1mal: run to the store and buy one
ok use your meter, probably have to set it for AC, test for voltage at PL1C to ground
Anybody know if the 25MHz Owon scopes are any good?
no 25MHz scope is any good
andypugh: SWPadnos says no way
ok that's supposed to be the triangle wave, at least we know something is there
hard to guess if it's right
Are they £180 better than no scope at all?
mind you , he's a premadonna too, but I BARELY got him to say "ok" to a tektronics
what voltage at PL1-E? (back on DC)
I quite like the fact they are thin and won't eat a whole cupboard.
andypugh: 100MHz and the VERY VERY minimum if you ever work on digital circuits
cradek, iirc you had a laptop running emc? Someone gave me a latitued c500 with pIII750 all in pieces
-1.7 compared to the 0v signal ground??
ok that's bogus
what about PL1-F
All I want to look at is step signals and encoder outputs.
well now it says 0
skinnYPuP: yeah it's an old gateway thing I use at cnc workshop sometimes
dan1mal: I believe 0 a lot more than -1.7
andypugh: and you never want to use it again? Never want to play with microcontrollers?
ok PL1F is 100v but PL1E is 0v?
cradek, cool , may spend a lil time to try this one out then.
I certainly hope never to play with microcontrollers :-)]
one sec, calculating some stuff
sorry its 3v now
my meter is really slow
PL1E is 3v?
pl1e is 3v and f is 100
pl1e on the good drive is about 7v
f is 100
ok seems like pl1e should be 5.3v if my math is right
good drive is 7.5v
do you see BD7?
what is it?
a zener diode
it's 4 parts above PL4F
starting at PL4 going left: big cap, big resistor, weird looking zener BD7
i think i see it
pretty sure i got it
test for DC right across that - red lead to the end with the stripe
cant see the number without removing it
check on the good one too
the working one is 6.7v
Hmm, 500MHz HP digitising scope currently at £6.05. Only drawback is that it doesn't work. And that it is vast.
ok perfect, I expected 6.7
this biases the transistor that tells it when it's overvoltage
see if R139 looks burnt
cradek: shouldn't he check before both zeners too?
it's the big one right below BD7
andypugh, a wavetek for £31 in somerset poor description though
sorry, what do you mean before?
let's turn it off and let the +90 bleed down, then check the resistance of R139
cradek: my bad, nm I thought you had him check across it.
doesnt apear burnt
* Jymmm is only half paying attention
i did check across it
That wavetek is only 20Mhz though.
what did you check?
across the legs of the diode
positive probe on the cathode
negative on the anode
Though my current 1-broken-channel, doesn't focus Gould is 15Mhz so it would be ab upgrade
oh right, I caught up now
I did mean for you to test across it - 6.7 is right
andypugh, I got a freebee gould, donator may have another
did you let the 90v bleed down?
I can probably do what I need to do with Halscope, if I run the base thread rather faster than normal.
andypugh, model number of yours?
yea it's bled down
(The OS could well stand for Out of Skip, because that is where I found it.