lordpil's blog

wtf just do it

Posted by renesis at 15:14 | permalink | 0 comments

the steppers just kind of desync and bounce back and its against the momentum of the machine
usually machine just gives up and stepper gets stuck
most the time it starts up against when the drive pulses stop then start again
k work laters

Posted by renesis at 00:38 | permalink | 0 comments

its kind of normal for machines to perform a bit rough until they kind of wear into themselves
this isnt like a $50K CNC the mating surfaces need time to kinda rub themselves aligned
yeah then its going to be able tuning the machine, if its possible
it might be able like, just how tight the thing is bolted together in certain spots, heh
so yeah i would suggest setting acceleration way low
and then trying to see what kinds of max speeds you can get
then do like 66% of your reliable max
and then bring the accel time down
and do alot of direction changes with the machine in different places
try to get it to lockup
then yeah same thing, find a reliable accel speed, then set it to 66% or so of that
if the machine is open loop step, no encoders, which is likely, then the machines cant ever lock up and skipped stepped need to be almost non existant

Posted by renesis at 00:33 | permalink | 0 comments

if you can adjust backlash or gibs, you want them tight so there is no slack
but loose enough you dont get binding when you try and go fast
obviously there is compromise between repeatability and speed
however machine too lose is completely fail
as it prob wont go fast and will have less repeatability do to backlash or the table having a bit of play in the axes
lose loose
one the machine seems kind of set up
assuming you can even do that shit
its about motor tuning
setting up max speed should be like, him racing the machine around trying to find the speed it can go fastest pretty reliably
then doing maybe like 66% of that speed for rapids
he needs to try and get them machine moving at the ends of the ways
and try and change direction on different parts of the envelope
in mach i think its all in 'motor tuning' dialog
but set accel reeeally low at first
so you can hear it spool up over maybe 2 or 5 seconds
and try and find max speed values
also the machine needs to break in

Posted by renesis at 00:28 | permalink | 0 comments

tv hurts people
some illegal stuff halps peoples
i think the point is the monitor wasnt 55"?
i think hes bragging how his phone lcd is small
or something
i didnt get that either
set the machine up if there is any way to

Posted by renesis at 00:23 | permalink | 0 comments

hahaha wat
man aliens seeing the shuttle is embarassing
i hope they run into f22 first
i think sr71 are gone
if you call engineers standing in front of congress trying to explain why they let shuttles go up with sensors throwing out of spec data
'yeah we dont know why they do that so we just call them anomolies'
the shuttle was a compromised system to begin with
oh structure of that long sentence is awful there should be something saying like about results

Posted by renesis at 00:14 | permalink | 0 comments

besides the time thing, engineers and techs always talking about how they were taken out after work a lot
because factories trying to make a good impression and youre there helping out
haha also wtf plane trip i think it is like 100 hours or something
rab: oh yeah when were you going to colorado?
oh, hi

Posted by renesis at 00:09 | permalink | 0 comments

rab: hi
rab: omg they are like for sure going to punt me to korea they are asking korea factory about dates =\
'k we need that guy when the guitars are hanging in the paint booth'
which sounds epic
i kinda am!
but like, every time i seen people go on these trips
2 weeks becomes 6 weeks ot something =\
and like, i have a cat and stuff and i friend girl has to take care of them while im gone
so i dunno that seems kinda sucky if it goes forever
the place and the work are awesome, looking forward

Posted by renesis at 00:04 | permalink | 0 comments

Top | Add to Technorati Favorites

© 2007 lordpil.   XHTML 1.0! CSS! Site design by GNAA  Blog Engine by pbx | MULTI2 | ian hanschen | lolwat