hey the stepper driver came today!
so you can play with the lathe this weekend
hmm, mystery connectors
it has a db25 and a db15
plus the motor connectors I assume?
db25 for PC, and db15 for the switches on the machine?
I'll ignore the one I don't understand, and maybe it'll work
take the cover off and trace a few wires?
the lathe has two db25s
the machine itself has two db25s?
probably one goes to this, and one goes to the computer
seems like a lot of wires
or ... maybe something else
I think he wanted to make it interesting for me!
the lathe doesn't have a box or anything does it? wires go straight from the connector(s) to the switches and such?
there are lots of wires all taped up so you can't see them
tape can be untaped
I unwrapped some stuff, it's not as bad as I thought
a simple picture would have really helped...
crap, forgot to eat, brb
ok I'm not getting this so far
the connector I think goes to the computer has wires to 1-8, 11, 20,23,25
11 is an input, but 20,23,25 aren't
there's also a small black box (shrink tubing filled with epoxy) that I suspect holds something
that's between the switches and this connector
I hate mysterious lumps of epoxy
I'm afraid I'm going to have to cut off at least some of this shrink tubing
but there's no way to replace it without unsoldering
maybe you could ohm things out?
how big is the black box?
might just be some pullup resistors?
I don't get what they would pullup TO
you sure about those pin numbers?
1 thru 9 are outputs from the PC, 10-17 are a mix of in and out
it's one of the DBs that has the numbers on it
18 and up are ground I think
the 1-8 are one size of wire, 11-20-23-25 are bigger
I would guess 1-8 go directly to the other DB (long cable)
11-20-23-25 probably go to the black box and then the switches
the data port (usually with step/dir) is pins 2-9, not 1-8
1 is another output
yeah, but emc has never used that for step/dir
the mightydrive has an outlet marked spindle, I bet pin 1 turns it on
no sign of estop
still the pins to the mightydrive(?) are 1-8, so one is probably missing (doesn't matter)
take the cover off the mighty drive yet?
no, can't find the right size allen wrench
my organized workspace strikes again
aha, torx is close enough
aha, in the mightydrive, pin 11 is spindle control
step/dir ARE 1-8
thats really strange
especially since pin 11 is an input
pin 23 goes somewhere to the A axis driver
the A axis driver is completely different from the other 3, including its connections to the breakout board
25 is ground
20 is a mystery
[01:03:56] <cradek> http://www.microkinetics.com/optistep.htm
hey, a big clue here
heh, it wasn't designed for a parport at all
time to get out the slobbering iron
no, I bet it's wired right in this cable
the one between the two DB25s that came attached to the lathe
and the black box might be pullups, because there's +5 available on that pin 25 I think?
I don't trust this picture completely though - it shows pin 11 is a ground, but I know it's spindle control
oh, wrong 11
I have motion
what config are you using?
I started with stepper-xyza but copied it to nist-lathe
(I'll put it in cvs)
why xyza? there's no A (and no Y)
well I have four drivers...
CL is gonna be frustrating for a certain class of users
don't know anything about it...
they want one or two rungs worth of logic, emc itself doesn't do that, and CL is overkill
for example, wholepair on the other channel
he wants an external estop button
simple with CL, but I dread trying to explain it to him
I pointed him at the wiki page that ray did that goes thru exactly the setup he needs for his estop
but thats a learning curve we really don't want to force "simple machine" guys to climb
he just wants a button that signals emc to stop?
or that turns off the hardware too?
ok, just a signal to emc
I don't think a hardware path occurred to him
can't you put a button between two parport pins (the estop out and estop in)?
he has his estop hooked to pin 13 of the parport, and a drive enable hooked to pin 1
that might work
I normally think of buttons as being between vcc and input, or gnd and input
not from output to input
maybe you need a resistor too
I'm fuzzy about how it works, if you can't tell
but if he configured an output to go low when EMC is ready to go, then it could go thru the NC button and pull an input low
I'm pretty sure he already has a pullup on the input
yeah, he says "I want pin 1 (enable to drives) to go low when pin 13 (input from red button) goes high
and he said the button is normally closed, open when you hit it
so he must have a pullup (or the parport has one)
he could connect the other end of the button to an output instead of ground
this is incredibly frustrating - its one thing to tell folks how to use the software, but I don't want to, can't, and won't, teach them how to design a machine control system
thats not a job for the clueless
jmkasunich: do you know what acceleration homing uses? It sounds much higher than mdi g0 accel
it should be the axis accel
same as jogging
it ignores traj accel tho, since that applies only to coordinated motion\
try jogging and see if that sounds the same
but g0 accel is lower
maybe it's my 95% hack, but it seems like I wouldn't notice that
you could scope it (pita sticking the ddts in there)
your traj limit is the same as your axis limit?
oh, I know what it is, it's the .5 issue
forget I said anything
yep if I use g61, it sounds about the same
so it's a feature *cough cough*
* jmkasunich smiles
g61 is exact stop?
btw, there's a pinout printed on the encoder
is hooking that up the next step?
crap, I don't think I have enough inputs
the switches take 3
I think there are only 5
all pos limits, all neg limits, and all homes?
got a parport?
you could drop the homes and home on the limts
SWPadnos: I want to continue to use my laptop
or combine pos and neg limits (which is a pain, cause if you hit one, the sw doesn't know which end your're at so it can't let you motor off of the limit
actually thats a non-issue with a sherline that has cranks
crap, and I just got homing and limits to work
so leave homing alone
just hook all the limits (pos and neg) to the same parport pin
so you wind up with 5 inputs:
1) limits (all)
2) homes (all)
3) spindle enc A
4) spindle enc B
5) spindle enc Z
SWPadnos_ is now known as SWPadnos
I love those responses to bo^dick :)
he's really starting to piss me off, I should just ignore him
all the auto-tools in the world ain't gonna turn him into a circuit designer overnight
I dunno what he is thinking
heh - I've been thinking that myself
he's trying to write an eagle script to do it, I wish him luck
I think he doesn't realize it's a hard problem
* rayh wonders why he doesn't try that on an eagle list.
because he is without clue
there are some things the human eye and mind are just plain better at
assumption forced to reality is a fascinating thing to watch.
itr's a long process with him
some never complete the journey.
heh - this may be one of those cases
He's trying to lay out the board for his stepper driver, because he doesn't want to buy Geckos
one thing -> 5 things -> 100 things -> ...
PMDX is showing off a stepper drive at NAMES this weekend.
interesting. is it a design of Steve's, or his gecko "motherboard"?
rumor has it the lowest ability is 48 volt 5-6 amp.
I actually thought about going to NAMES
highest ability 80 volt 8 amp.
only 2-2.5 hours from here
but then I said fsck it
stand alone or plug in to a 4-5 axis mobo
hmm - cool
not servo though ;)
Roland was at the same gathering place NAMES is using this year.
last weekend and said parking was near impossible
I nearly cried about not being at names.
why? you miss it?
it sounds like there's a lot of neat stuff there
That was a good time for quite a few years. But hey...
I remember your jmk's year of the HAL.
A lot of stuff got worked out there.
mostly after the show tho
And the constantly long line waiting to talk to Hassan about his home made mill.
* rayh goes back to halconfig.
rayh is now known as rayh-away
rayh-away is now known as rayh
when I ask halcmd to show me pid.1.maxerror i get pid.1.maxerrorD pid.1.maxerrorI also.
I thought there was a way round this so that I got only the info for one.
it treats any string you give it as if it ended in *
so foo = foo* which matches foo, food, and foobar
I thought that a get or something worked around.
show treats the parameter as a part of the string to match, gets will only return exact matches
there is gets for signals and getp for params
does it return the whole "show" string.
is there a getpin? I don't think so
no, it returns only the value
(thats the reason for get existing, to get just the value)
okay I see that. Thanks
the -s makes no difference there?
at least, that's what I think ;)
I wonder how hard it would be to change the string match "exact match unless they specify a * "
or add a -e option for exact matches only
the * thing is just a subset of the real change, which is to use regexp
exact would be nice for some things.
I'd rather not do -e
because options apply to all commands passed to an instance of halcmd, and in theory at least, the decision between exact and wildcard is made on a command by command basis
I agree that regexp would be really nice
but more work than I want to tackle right now
I do like the current show param
or show pin
and get the whole list
or even show pin axis.0
right - I don't want to give that up
We would loose a lot without that.
if we had regexp you would do "show pin *"
or show pin axis.0.*
or even axis.*.motor-pos-cmd
that last is what I'd like
except it probably wouldn't just be *, it could be [0-3]
and regexp is probably the easiest way to do it, believe it or not
is there a libregexp?
oh that would be nice. axis.*.motor-pos-cmd
you call one function to prepare the expression, then you call another function to do the search (I think)
I'm trying to unstall linux on a little (formerly) diskless client box
got hooked up a disk and cdrom drive
got it to boot the cd, and begin the ubuntu install
was looking good, then a kernel panic
bummer. what kind of client box?
Neoware Neostation 3000
233MHz CPU, 256M ram
did you try memtest86?
the only time I've seen kernel panics during install was bad ram on my desktop machine
I think I must have misread the BIOS display of ram size
memtest says there is 29M
which is a weird amount...
it has an empty DIMM slot
either that or there's an error the BIOS is detecting, and assuming that's where RAM ends
fasking splash screen hides the bios memory test
bios reports 30208K total memory
it has a little laptop style memory module
gonna stick a 128M dimm in there and see what happens
is video using 2.5M of system RAM?
that sounds familier
(that would be 1024x768x24bits)
+ a little
then it's a 32M module
or it has 32M that's good ;)
heh, 161M now
bummer tho, with the DIMM installed, access to the PCI slot is blocked
(It has a riser board with one pci slot
(actually one PCI/ISA slot, it has both connectors)
you need a low profile dimm
or a no-profile dimm ;)
well, it detected 161M, but then the BIOS memtest failed
can you use only one module, or is the 32M soldered down?
Cyrix MediaGX CPU at 233MHz, no idea what the frontside bus is
it can be removed
try just the 128 then
the DIMM I tried is PC100
bios memtest passed, running memtest86 just for the heck of it
the shared video arch means it won't be a good RT box I don't think
yay, midi cables work to extend the steppers on the lathe
you had some laying around?
or you actually went out and spent money on them?
no, I bought them
I do have some somewhere...
but buying them is faster than finding them in the basement
actually I should shut up, I just bought the 40 to 44 pin IDE adapter so I could stick a laptop drive in this box
hmm, my inches are only half as long as I'm used to
different screw pitch? microsteppint?
I think it's off by a factor of 2, no big deal
wonder how many hours it will take to install...
I'm only doing the server install, so at least there won't be "fonts of the world" for three hours
those are the worst part aren't they
another annoying part is how it does a bit of install, then stops and asks questions, then a bit more install, and more questions, about 5-6 times
I'm gonna have to try to trashpick a bigger drive
this ones only 800M
jmkasunich: there still a bug with jerking after machine-off
I hit a limit switch and it stops, and when I hit f2 for machine on, I get a big jerk and sometimes even following error
what mode? manual (jog into limit) or mdi/auto?
g0z3, poke the switch while it's moving
the fix I did would only affect the free mode planner
in mdi, its probably coming from the tp or interpolators
I don't want to work on that, so forget I said anything
can you think of any reason why catting a 2496 byte file would print absolutely nothing?
(a log file, which I would expect to be human readable)
use 'od' or 'file' on it and find out if the contents are binary
od, file (and less) aren't available
I'm trying to read a log after getting an error during install
guess they're not installed yet
ls -l says its 2496
cat gives nothing
use "cat -v" ?
$ printf '\0\1\2\3\n' > /tmp/test; cat -v /tmp/test
turns out the cat is part of busybox
there are three files in /target/var/log with non-zero lengths
I can cat one of them the other two give nothing
no other ideas are coming to mind
cat lastlog | wc returns 0 0 2496
so the bytes are coming out of cat
0 0 means no lines and no words: so none of those characters are whitespace
and none of them are printable either
on top of that, the file that the error message said to read isn't even there
I think the odds are it's all zeros
it says to read /target/var/log/bootstrap.log
but there isn't one
it failed installing initrd-tools
I think I'll just tell it to continue and see what happens
initrd-tools sounds pretty important to getting a bootable system
yeah, its part of the base system
it won't let me go on
I tried to reinstall the base system, it bitches about the partially installed files
tried to repartition, it starts to load the partitioner, then bails
disk might be frotzed
I should get a bigger one anyway
I have lots of 3.5" disks, but I think this is the only 2.5
or not... just found a 6G one