►
From YouTube: 1" Universal Axis Accuracy
Description
-----------------
What you see here at Open Source Ecology is an ambitious open source project for the common good. Join our development team:
http://opensourceecology.org/wiki/OSE_Developers
We run One Day Extreme Build workshops in many locations. Please put yourself on our map if you'd like us to host a workshop near you:
https://microfactory.opensourceecology.org/request-a-workshop/
Take a minute to subscribe to our email newsletter (updates, workshops, etc): http://bit.ly/1LtcM44
A
All
right
so
recordings
so
on
the
torch
table
right
now.
We've
got
definitely
some
issues
regarding
stiffness
of
the
motion,
so
it's
there
is
some
good
stiffness.
So
we've
got
two
axes
on
the
x
two
axes
on
the
y.
A
So
the
thing
is
that
in
principle,
if
you've
got
brass,
bushings
like
we
do
have
they
have
a
coefficient
of
friction
of
about
0.1
or
so.
In
other
words,
if
you've
got
like
10
pound
load,
friction
you're,
gonna
get
out
of
that
is,
is
about
one
pound
in
our
system.
The
y
axes
are
carrying
about
80
pounds.
A
If
you
count
up
the
weight
of
the
rods
actually
show
some
of
these
numbers
here.
If
you
look
at
just
weight
of
shaft
weight
of
the
whole
system,
I'm
getting
about
80
pounds
total
that
the
the
y
axis
are
carrying
so
each
bushing
by
itself
slides
smoothly.
Naturally,
as
it
should,
it's
all
impregnated
coefficient
friction
of
about
0.1,
but
what
we're
getting
into
is
definite
stiffness
and
we
can
trace
that
back
to
some
very
particular
things.
A
B
A
A
If
it's
point
one,
it
means
that
you
can
push
it
only
with
like
one
tenth
of
its
weight,
so
you
can
use
that
basic
kind
of
this
basic
principles.
Thinking
here,
you've
got
smooth
rods.
You've
got
steel
rods
that
are
factory
made,
they're
actually
called
rolled
steel,
which
is
a
little
bit
smoother
than
standard
steel
like
standard
steel
is
called,
it's
called
hot
roll
typically,
but
we've
got
smooth
rods.
A
We've
got
bushings
that
are
designed
specifically
for
low
friction,
they're
oil
impregnated
bushings
that
are
one
inch
and
then
you,
google,
what's
the
coefficient
of
friction
of
bronze
on
steel
and
you'll,
get
values
between
like
0.08
to
up
to
as
much
as
0.3,
0.3
being
like
the
worst
case,
where
it's
not
lubricated,
but
between
0.8
and
0.3.
A
So
if
you
have
lubrication
like
these,
these
bushings
do
have
some
lubrication
in
them
and
then
we
can
also
oil
it
up
with
a
little
diesel
or
other
lubricants,
but
we're
expecting
like
if
we
look
at
what's
going
on
we're,
expecting
good
motion
like
the
other
side
of
the
y.
That's
it's
getting
us
decent
motion.
Now
we
do
have
a
lot
of
a
lot
of
force
in
the
motors
themselves.
What
do
they
have?
A
They
were
looking
at
those
numbers
and
they've
got
up
to
68
pounds
per
motor
and
therefore
we
were
actually
getting
more
than
a
belt
could
hold,
because
if
you
look
at
some
of
the
numbers
for
the
gt2
belts
that
we
were
using
before
their
tooth
skipping
force
is
what
you
look
at
and
that
gets
to
be.
I
was
looking
at
some
of
those
numbers
and
it
was
like
it
was
definitely
lower
than
68.
It
was
like,
maybe
up
to
30.
Actually
I
have
some
of
those
numbers
here.
A
Look
at
this
thing:
let's
see
where.
A
A
It's
a
little
bit
of
a
different,
slightly
different
profile
of
the
way.
The
belt
looks
it's
three
millimeter
pitch.
So
instead
of
two
millimeter
spacing
between
the
teeth,
it's
three
htd
is
height
height,
torque
drive,
that's
what
it
stands
for
now.
Gt
is
actually
deemed
to
be
a
little
better,
but
I
could
not
find
any
place
that
had
gt
3
belt,
like
I
just
couldn't
find
it
like
online.
The
place
that
sells
this,
which
is
gates
belts
like
they
had
like
12-week,
lead
time
and
stuff.
A
I
don't
know
why
I
couldn't
find
it,
but
I
could
find
some
hdd,
which
is
pretty
good.
It's
actually
stronger,
stronger,
it's
not
as
accurate,
but
here
we're
talking
about
accuracy.
Requirements
for
ourselves
are
like
400,
microns
or
like
a
little
under
a
millimeter.
That's
that's
fine
for
what
the
hdd
can
do.
Anyway.
If
you
look
at
these
numbers
here
tooth
jump
torque
for
two
millimeter
gt2
now
they
they
show
five
millimeter
belt.
We
we
had
six
but
look
at
those
very
low
numbers.
The
two
millimeter
gt
gt2,
that
is
4.8
millimeter.
A
On
the
left
hand,
side
we've
got
six,
but
look
at
that.
It's
only
like
eight
pounds
before
you
start
jumping
teeth
on
a
20,
groove
pulley,
so
the
size
of
the
pulley
will
determine
that.
If
it's
a
bigger
pool,
you've
got
more
teeth
catching
so
you'll
have
a
different
number.
We
happen
to
have
a
20
groove,
fully
those
half-inch
little
pulleys
that
we
have
and
we're
using
the
six
millimeter
belt,
so
you're
expecting
to
actually
start
jumping
at
eight
pounds.
A
A
A
Diameter
is
half
inch,
so
that
means
the
actual
radius
of
drive
is
a
quarter
inch,
which
means
you
take
this
eight
inch
pounds.
Eight
pounds
is
what
you
have
at
an
inch,
but
at
quarter
inches
you've
got
pounds,
in
other
words,
if
the
motor
is
like
68
inch
pounds
we're
tapping
about
half
of
its
power
half
of
its
power
before
you
start
skipping
teeth,
because
the
torque
is
the
idea
that
inch
pounds
means
that
you've
got
so
much
force
at
an
inch.
A
But
if
you
go
inside
like
if
it's
a
rotating
pulley
like
the
closer
you
go
to
the
shaft,
the
more
force
you
have,
the
harder
it
is
to
stop
it
like.
If
you
have
a
big
lever
on
a
pulley
like
that,
it's
very
easy
to
stop
it
like
if
it's
10,
inches
out
and
you're
trying
to
hold
that
pulley
it'll,
be
like
one
tenth
of
the
four
pounds.
You
only
need,
like
4.4
pounds,
to
hold
that
pulley
at
a
far
radius,
but
very
close,
and
you
get
high
force.
A
So
in
our
case,
because
the
pulleys
are
so
tiny,
like
quarter
inch
radius,
you
get
about
32
pounds.
According
to
this
table
here
for
two
millimeter
gt2
belt,
which
is
still
half
of
what
the
motors
can
put
out
and
the
numbers
for
the
motor
the
inch
pounds
of
the
motors
we
have
are
they're
like
68
pounds,
well
68
pounds
at
that
quarter.
Inch.
A
A
A
A
Or
more,
in
other
words,
these
belts
should
be
finally
should
not
be
skipping
with
our
motors,
and
I
don't
think
they
are,
and
we
have
observed
that
already
what
happened
was
when
the
axis
was
stuck.
I
was
looking
at
the
pulley
and
the
belt
was
not
skipping.
It
was
actually
the
motor
was
skipping
because
it
couldn't
have
it
didn't,
have
enough
force,
so
it
tried
to
go
and
then
it
kind
of
like
snapped
back
so
we're
good.
We're
cool
on
that.
So
belt
is
not
the
issue
here.
So
friction
is
the
issue.
A
What
do
we
do
about
it?
The
friction,
so
you
start
with
like
first
principles
of
friction.
These
coefficients
of
friction
so
so
bushing
is
a
line
on
the
shaft.
That's
cool,
like
that
very
low
friction.
You
can
move
them
very
easily
with
your
finger.
If
you
put
down
a
bunch
of
your
weight
on
it,
then
still
try
to
move
it.
It's
still
rather
smooth.
A
You
have
no
problem
moving,
if
you,
even
if
you
like,
hold
it
down
really
hard,
it's
a
little
more
than
like
ball
bearings,
because
ball
bearings
would
roll
even
smoother
their
coefficient.
I
mean,
what's
the
coefficient,
the
friction
of
linear
bearings.
A
A
Yes,
so
it's
10x,
it's
10x
better!
It's
super
low
at
that
point
now
we're
not
using
ball
bearings
for
a
couple
of
reasons,
one
at
high
loads,
the
ball
bearings.
Unless
you
have
chrome
rods
which
are
more
expensive,
you
can
do
it
with
chrome
rods,
but
the
bulb
ball
bearings
tend
to
eat
up
your
rods
like,
even
if
you
have
stainless
or
whatever
you
use.
A
I've
seen
that
on
other
printers,
we're
using
the
rods
and
and
the
rods
would
just
get
eaten
up
by
the
balls,
and
so
the
two
disadvantages
were
the
noise
and
the
that
your
rods
get
destroyed.
They
are
lower
friction,
but
only
if
you've
got
them
aligned.
Well,
if
you
have
a
lot
of
tension
in
it,
the
balls
actually
like
drive
into
the
rods
that
we
have
so
we
went
away
from
them.
A
The
plastic
is
really
good
for
multiple
reasons:
one
we
can
print
them,
two
they're
very
quiet
and
they're
adequate
they're
good
enough
so,
but
they
do
have
like
for
this
brass,
it's
higher.
It's
higher
than
ball
bearings,
so
one
solution
would
be
ball
bearings,
but
still
that
doesn't
solve
your
issues
if
you're,
if
your
axes
are
not
if
the
prints
are
crappy,
but
let's
look
at
what
a
crappy
print
means.
A
Crappy
print
means,
you're,
gonna,
get
discrepancies,
say
you're,
printing,
with
a
point
four
millimeter
nozzle
I
mean
the
best
accuracy
you
can
get
is
maybe
like
10
of
that.
So
if
it's
like
40
400
microns,
which
is
0.4
millimeter,
it's
like
40
microns.
A
A
A
A
little
burp
of
irregularity,
your
bushings
are
not
aligned
and
you
get
for
a
lot
of
force.
So
the
only
way
your
bushings
are
sliding
smoothly
is,
if
they're
aligned
with
the
with
the
shafts
but
any
force
on
them,
so
say
you
can
find
them
now
in
your
carriage
and
your
carriage
is
not
perfect
print.
You
talk,
you
have
to
talk
about
precision,
machining,
meaning
like
1
000
or
like
10
micron.
For
you
to
not
see
these
kinds
of
effects,
because
the
fit
around
the
bushing
of
the
bushing
around
the
rod
is
quite
precise.
A
How
precise
is
it
you
would
look
at
the
bushing
that
we
use,
and
you
would
say
it's
interior
tolerance
and
that
determines
the
amount
of
inaccuracy
in
the
print
like
say,
you've
got
a
print
imperfection,
that's
as
big
as
that
tolerance,
and
at
that
point
you
start
jamming.
So
what
is
that
tolerance?
In
other
words,
that
little
burp
I
drew
there?
A
What
is
the
maximum
size
that
we
can
sustain
without
stuff
getting
tight?
So
we
go
to
say
the
mcmaster
car?
Let's
look
at
the
specs
of
those
all
impregnated
bushings.
A
A
Let's
look
here
so
let's
do
shaft
diameter
one
inch.
A
One
inch
it's
one
and
a
quarter
here,
one
and
a
quarter
this
one
here.
A
I
mean
their
their
load
capacity,
they're
huge.
It's
like
you
can
hold
like
3
000
pounds
on
these
things
at
60
rpm,
it's
huge,
but
let's
see
so
look
at
their
imperfection
id
tolerance.
A
A
C
A
C
A
So
you
have
to
get
it
tight
enough,
but
not
over
tight,
like
what
I
can
tell
you
is
the
the
case
that
we
were
doing
in
a
shop
there
with
a
torch
table.
It
is
over
tight.
It's
I
mean
the
two
carriages
are
bound
in
such
there's
that
metal
plate,
but
whenever
we
we
have
two
imperfect
holders
and
that
plate
like
fixes
it.
You
know
it's
it's
what
I
showed
here
any
time
you
would
fix
your
bearing,
and
this
is
a
talking
about
the
y-axis
that
applies
to
any
of
our.
D
A
A
But
let's
well,
we
don't
have
them
right
here,
but
we
can
make
make
provide
solutions
with
these
already.
Yes,
so
longer
one
you
can
do
one
for
instead
of
two
yeah.
That
could
be
a
a
good
idea
like
how
much
do
they
cost
like
the
question
would
be
cost
or
actually
do
they
have
them
so
length
the
max
length
so
we're
using
the
one
and
a
half
a
one
and
a
half
length
there?
A
C
A
A
Because
that's
how
the
industry
works
when
it's
a
non!
This
is
not
a
common
part.
So
once
you
go
to
non-common
parts,
which
is
actually
as
a
side
note,
an
advantage
of
open
source
construction
set
construction
set
approaches.
Is
it
doesn't
matter
if
you
make
a
short
one
or
a
big
one,
you
still
have
the
same
tooling
for
each,
but
that's
just
how
industry
works.
So
here,
okay,
it's
not
you
know.
A
So
no,
let's,
let's
see
what
we
can
do
with
the
existing
ones
that
we
do
have
so
for
one
well
here
in
this
carriage,
yes,
that
that
can
be
an
issue
now,
if
you
have
two
carriages,
that's
a
bigger
issue,
because
now
you're
binding,
like
we
bound
the
two
carriages
together,
if
they're
not
now
you're
forcing
the
bushings
to
be
not
aligned,
you
get
more
jamming
and
that's
exactly
what
we're
seeing.
So
what
are
the
solutions
here?
So
basically,
it's
like.
A
Let's
reduce
the
number
of
bearings
for
one
like,
for
example,
each
each
carriage
if
we
still
use
two
carriages
just
put
two
barrel
like
just
redo,
the
prints
put
one
bearing
one
on
top
one
on
the
bottom.
We
actually
have
beside
where
we
were
working
it.
We
actually
were
taking
out
bushings
and
we
saw
that
it
moved
easier.
Yes
yeah,
but
it's
a
natural
thing.
A
A
A
B
A
C
A
A
So
what
about
so?
Can
we
do
this?
How
content
are
we
that
this
is
pretty
robust?
Like
do
we
wanna,
because
the
thing
to
do
right
now
is
now
that
we're
retrofitting
into
now
like
we're
working
with
what
we
have.
We
should
probably
spend
the
time
to.
Okay,
maybe
make
this
dedicated
one
one
bushing
piece,
but
we
still
have
to
connect
them
together
in
an
effective
way.
A
A
It
could
be
if
we
find
this
doesn't
work
when
we
follow
a
good
procedure.
I
mean
right
now
we're
running
into
some
difficulties,
so
maybe
in
the
future
we
say:
oh
man.
This
is
just
too
hard
for
for
people
we
migrate
to
something
else,
but
in
the
other
systems
you
also
have
its
own
challenges
too.
So,
in
any
case,
you
have
to
know
what
you're
doing
so
to
me,
just
conceptually
speaking,
I
mean
I've
done
the
rails
and
all
that
done
it.
I
actually
don't
think
it's
easier
than
this.
A
A
A
Yeah
springs
are
springs
and
adjusters
are
the
way,
but
that's
a
lot
of
mechanical
parts
here,
we're
trying
to
say.
Oh,
we
have
these
rods
and
the
end
pieces
are
at
a
very
well
defined
distance.
Therefore
we're
already
exactly
parallel
in
the
middle.
The
rods
can
flex
a
little
bit,
so
you
actually
have
have
tolerance
with
in
the
middle.
It's
much
it's
very
easy
to
move
in
the
middle
area.
B
A
You
know
yeah,
no,
it's
you
have
to
look
at
the
the
cold
rolled
steel
specifications,
but
that's
where
you
have
to
go.
A
A
So
what
to
do
here
so
say
we
solve
the
the
the
y
axis
say
you
build
the
two
independent
carriages
with
two
bushings.
So
already
there
we're
we're
having
the
number
of
bushings.
That's
a
good
start.
Now
what
about
on
the
x-axis?
A
The
way
we
have
because
the
torch
head
is
so
heavy.
You
really
do
need
unless
we
strip
everything
out
of
it
and
maybe
just
do
the
very
simple
prototype
without
the
gas,
solenoids
and
igniter.
That
thing
is
going
to
be
pretty
heavy,
so
you
do
need,
like
both
sided
support.
Otherwise
I
mean
it's
really
hard
to,
but.
A
A
So
I
mean
it's
pretty
heavy.
I
mean
when
you
lift
it.
It
may
not
feel
so
with
the
springs,
because
it's
easy
to
move
it
up
and
down,
but
the
the
actual
weight
of
it.
A
A
You
could
print
custom.
Bearings
is
true.
If
you
print
your
own
delrin,
that's
a
printable
material,
so
I
would
keep
keep
it
at.
I
would
still
keep
the
two
axes
on
that
on
the
axe
unless
we're
just
gonna
strip
that
completely
remove
all
the
parts
and
we
could
go
at
that
point.
You
can
literally
go
to
like
the
eight
millimeter
rods,
just
even
with
a
just
put
the
torch
head.
Just
you've
got
basically
the
nozzle
thing
and
hoses.
C
A
Yeah
one
thing
that
would
definitely
work
is
like
say
the
smaller
rods,
but
the
big
motor,
like
with
a
small
motor
that
gets
questionable
unless
you
have
well
balanced,
out,
springs
and
stuff
like
that,
so
yeah,
but
probably
in
order
would
be.
I
don't
know
like
if
we
do
this
simple
holder
with
two
two
bushings
just
use
metal
plate
in
between
it,
but
whenever
we
do
that,
make
it
such
that
that
can
take
that
assembly
and
make
sure
it's
absolutely
easy
to
slide
like
pound
or
less
it's
just
really
smooth.
A
Keep
it
loose
the
thing
about
the
auto
the
x
parallel
thing
where
the
rods
can
go
in
and
out.
I
think
that
is
still
important,
because
if
the
it
is
quite
important
because
there
is
no
guarantee
of
parallel
on
the
on
the
y-axis,
but
those
are
tight
anyway,
we
loosened
them
before
they
were
super
tight,
and
that
was
what
was
loosening
things
down.
A
A
A
You
should
add
that
add
that
get
rid
of
the
bearings,
if
I
mean
do,
we
want
to
just
try
to
keep
use,
use
the
same
parts
and
then
just
keep
removing
bearings
now
the
bearing
is
going
to
just
free
float
inside
too.
C
C
The
measuring
tape
fits
very
well
yeah.
It
has
a
nice
diameter,
okay,.
B
B
B
Couldn't
you
just
even
like
you
did
just
some
little
nubs,
nothing
hot
yeah,
yes
with
the
like
soldering
iron,
so
that
the
bearing
doesn't
yeah
I'm
actually
coming
out.
You
might
have
to
take.
A
D
A
A
That
would
be
useful,
but
but
definitely
the
y
parallel
the
y
paralleling
mechanism.
We
should
definitely
put
it
in.
What's.
C
A
A
D
C
A
Yeah
they
lock
up
well
for
that.
So
you
have
the
two
two
half
pieces,
the
holder
pieces,
and
then
they
have
to
be
connected
with
a
stiff
thing
like
a
like
quarter,
inch
steel,
it's
kind
of
like
what
we
did
before
that
worked,
I
mean
we,
we
had
our
big
five
by
ten
table
moving
before
it
wasn't
super
stable,
because
the
axis
started
wobbling
just
sagging,
but
that
all
moved,
and
that
was
on
nema
17
motors.
A
So
we
know
this
thing
can
work
so
stiff
connection
between
two
pieces
that
you
still
have
some
play
like
you
will
tighten
down
hard,
but
when
you
tighten
it
down
hard
at
that
point,
you
make
sure
that
you
haven't
locked
anything,
that's
the
key
there.
So
it's
a
very
stiff
structure.
You
can
now
connect
things
to
it
without
the
worry
of
this
part
now
binding
up
when
you
connect
other
things
to
it.
That's
the
thing
so
two
half
pieces
with
metal
in
between
would
be
a
good
solution.
A
Yeah,
let's
refresh
on
that.
A
Well,
we
got
to
solve
it,
so
you
know
we're
finding
out.
This
is
a
huge
issue
like
I
didn't
really
expect
that
that
that
kind
of
tightness
would
arise
here,
but
it's
it
is
quite
sensitive
to
that
I
mean,
and
we
looked
at
the
number
there.
It's
like
one
thousandth
play
in
that.
Bearing
I
mean
yeah.
If
anything
is
off,
it's
going
to
start
binding.
A
Yeah-
and
I
mean
that
would
be
okay
if
they
were
oversized
and
you
then
fix
them
together
in
a
stiff
way
right
now,
our
connections
are
not
particularly
stiff,
like
we
have
those
thin
little
bars,
so
things
can
still
move
around.
So
I
think
if
we
do
a
fixed
connection,
so
that
would
mean
two
plates
one
one
on
each
side,
so
do
a
sandwich
with
the
so
effectively
do
like
an
elongated
carriage
piece,
metal,
plastic,
composite
yeah,
no
view
that
would
do
it
yeah.
A
So,
let's
design
that
thing
oh
get
on
that
we've
got
steel,
I
mean.
We've
got
quarter
inch,
steel
that
we
can
use
for
that,
so
we
can
still
print
it.
On
the
universals,
I
mean
the
pieces
we
can
as
long
as
we
oversize
them,
they're
gonna
work,
but,
like
you
can
think
of
it.
That,
however,
you
printed
it's
all
wonky
and
stuff
like
that,
but
when
you
hold
it
it
just
like
holds
it
with
just
the
right
amount
of
play.
That's
what
we're
looking
for.
A
So
when
you're
now
screwing
them
down
tight,
we
fixed
them
in
a
position
that
still
allows
full
motion.
We
gotta
test
that,
like
so
obviously,
the
two
individual
characters
are
gonna
slide
easily.
Once
you
fix
them,
it's
gonna
be
time
for
trouble.
But
at
that
point
that's
when
we
have
to
do
all
the
work
before
we
go
any
further.
It's
like
either
heat
gun
or
shift
it
around
a
little
bit,
ream
it
out
whatever.
A
But
that
has
to
move
perfectly
still
and
even
like
you
can
think
about
the
two
top
bearings
since
the
gravity
pulls
on
them.
First,
you
can
think
about
the
two
top
bearings
being
like
the
ones
that
provide
all
that
precision,
because
you
can't
do
this.
You
release
all
degrees
of
freedom
outside
of
like
twisting
like
that,
so
the
bottom
ones
you
we
can
think
about
the
bottom
bearings
being
looser.
The
top
ones
like
super
nice
and
tight.
The
bottom
ones
are
just
like
helping
us
move
along,
but
even
those
two
at
the
top.
A
Once
we
have
the
the
x
in
between
them,
it
would
in
principle,
be
sufficient
to
have
like
two
and
and
the
bottom
rod
just
like
free
floating,
even
right,
because
the
point
is
you've
got
one
two
and
then
three
four
on
the
other
side,
four
is
all
you
need
for
yeah,
so
maybe
even
just
even
just
make
the
bottoms
free-floating
and
get
rid
of
those
two
points
of
lockup.
A
C
A
D
A
A
A
A
When
we
put
the
yeah,
I
think
I
think
it
would
work.
The
only
issue
is
there
like.
If
you
look
at
the
auto
paralleling
mechanism
that
has
to
be
connected,
like
the
two
rods
like
on
one
side,
they're
connecting
to
your
long
carriage.
C
A
Yeah
I
mean
you
still
haven't.
We
still
haven't
addressed
the
parallel
issue
like
yeah,
you
can
have
the
two
on
each
side,
but
if
they're
like
going
out,
I
can't
have
that.
That's
total
binding
we're
talking
about
okay,
look
at
this.
If
we
have
1000
accuracy
on
a
frame,
then
we
don't
have
to
worry
about
it,
but
we
don't
right.
A
That's
as
tight
as
the
bearings
are.
A
A
But
if
you
have
two
bushings,
you
can't
really
test
it
somewhat.
I
mean
you
it'll
move
slow,
move
smooth,
but
the
bottom
could
wobble
back
and
forth
a
little
bit,
but
maybe
the
term
there
is
just
using
the
three
bushings
instead
of
four.
A
C
C
A
More
slop,
but
the
slop
we're
looking
for
is
1
16
inch.
Accuracy
on
the
torch
table
is
more
than
what
we
need.
That's
that's
what
we're
shooting
for.
So
we
can
be.
We
kind
of
like
1
16
slot
somewhere.