Homebrew_PCBs

Hi everyone, it's me again. I'm starting to work on a homebrew CNC
machine project. My question is about the motors. I know I need a
stepper motor, but what else? Steps, torque, voltage, etc.

I had an idea for the Z axis. Most of the use of this machine will be
drilling holes, so I thought of a spring mechanism. That is, a
solenoid pushes a spring, which in turn pushes the arm holding the
bit, which may be cheaper than a Z motor and also will let the bit do
the drilling on its own, not forced by a motor. Is this a good idea?

hjf

--- In Homebrew_PCBs@yahoogroups.com, "hjf2k2" <drgenio@h...> wrote:

> Hi everyone, it's me again. I'm starting to work on a homebrew CNC
> machine project. My question is about the motors. I know I need a
> stepper motor, but what else? Steps, torque, voltage, etc.
>
> I had an idea for the Z axis. Most of the use of this machine will be
> drilling holes, so I thought of a spring mechanism. That is, a
> solenoid pushes a spring, which in turn pushes the arm holding the
> bit, which may be cheaper than a Z motor and also will let the bit do
> the drilling on its own, not forced by a motor. Is this a good idea?
>
> hjf

Bad idea,

You will have no control over the force - with a spring, you would
simply get a single feedrate for everything. Believe me, a .010"
diameter drill requires a much different feedrate than a .125"
diameter drill.

Stick with a stepper drive Z axis. As for what size steppers, that
depends on the overall system design. Here are few examples of why
you have to consider the system as a whole...

Ball screws are 95% efficient
Lead screws are about 40% efficient
A smaller motor can move a bigger load with ball screws.

Linear bearings have low frictional and slip-stick losses
Metal on metal ways have higher frictional losses and higher slip-
stick.
A smaller motor can be used to move a bigger load when the way system
is using linear bearings.

There is no quick and easy answer on motor sizing,

Chris

> Bad idea,
>
> You will have no control over the force - with a spring, you would
> simply get a single feedrate for everything. Believe me, a .010"
> diameter drill requires a much different feedrate than a .125"
> diameter drill.

Why? If I compress the spring a little more, I'll have more strength.
Or so I think. So I can put a screw to adjust the spring tension for
the different bits (I use .75 and 1mm bits anyway...)

> There is no quick and easy answer on motor sizing,

Well, I'll be making my machine with whatever I have handy. I can't
seem to find a supplier of ball screws for my country (I'm sure there
are, but anyway the price will be too high), so I'd have to build it
with a regular threaded and nuts. I don't mind the efficiency either,
I don't care if it's a little oversized. I ask this as I have
absolutely no idea about stepper motors.

hjf

--- In Homebrew_PCBs@yahoogroups.com, "hjf2k2" <drgenio@h...> wrote:

> > Bad idea,
> >
> > You will have no control over the force - with a spring, you

would

> > simply get a single feedrate for everything. Believe me,

a .010"

> > diameter drill requires a much different feedrate than a .125"
> > diameter drill.
>
> Why? If I compress the spring a little more, I'll have more

strength.

> Or so I think. So I can put a screw to adjust the spring tension

for

> the different bits (I use .75 and 1mm bits anyway...)
>
> > There is no quick and easy answer on motor sizing,
> Well, I'll be making my machine with whatever I have handy. I can't
> seem to find a supplier of ball screws for my country (I'm sure

there

> are, but anyway the price will be too high), so I'd have to build

it

> with a regular threaded and nuts. I don't mind the efficiency

either,

> I don't care if it's a little oversized. I ask this as I have
> absolutely no idea about stepper motors.
>
> hjf

While similar methods have been used on drilling machines WAY BACK
in time, they too suffered from the lack of control. You would not
have any way of knowing if it's too fast till you break the tools.
With a controlled axis, you simply tell it how fast to feed in
inches per minute. For small drills, the difference between 1 IPM
and 2 IPM will be the difference between a broken tool and a tool
that will drill all the holes.

For a stepper size reference, my CNC router uses 160 oz-in motors -
direct drive to ball screws running dry. The way system is very
basic, UHMW plastic against aluminum. It developes around 50 pounds
of cutting force - very approximate value.

These 160 oz-in motors are running at 44 volts and 1.5 amps through
a unipolar chopper driver. I can get rapid speeds up to 150 IPM but
only with some pretty slow accell/decell ramps.

That should give you some parameters to start with.

> You would not have any way of knowing if it's too fast till you

break the tools.

> With a controlled axis, you simply tell it how fast to feed in
> inches per minute. For small drills, the difference between 1 IPM
> and 2 IPM will be the difference between a broken tool and a tool
> that will drill all the holes.

Well, yea, guessing I get the parameters of the board and the bits,
which I won't, because the board, I don't know the material, and the
bits I buy are sold by the unit, attached to a piece of paper with
tape so they don't get lost. That should give you an idea of how we do
things down here :(

> For a stepper size reference, my CNC router uses 160 oz-in motors -
> direct drive to ball screws running dry. The way system is very
> basic, UHMW plastic against aluminum. It developes around 50 pounds
> of cutting force - very approximate value.

Thanks, I'll figure out how to convert all that to metric and make an
order :D

> That should give you some parameters to start with.

Thanks a lot. This is a first-time project anyway so I'm allowed to
make some mistakes :D

hjf

lcdpublishing wrote:

>While similar methods have been used on drilling machines WAY BACK
>in time, they too suffered from the lack of control. You would not
>have any way of knowing if it's too fast till you break the tools.
>With a controlled axis, you simply tell it how fast to feed in
>inches per minute. For small drills, the difference between 1 IPM
>
>

Spring/solenoid can actually work quite well, PWM the solenoid against
the spring and you have control, some things are done this way. But
it's more work to reengineer this new system than to just copy the
stepper system from the other axes.

If room isn't an issue, there is also a very simple way for stepper
control for a drill..

H H




D SM


Looking down, H and H are two hinges on a triangular piece of wood.
Come out about a foot, and D is the drill with SM being the motor. At a
foot out, the arc of the circle is very close to a line. Get the bit
aligned so it's tangent where it's drilling, and the error is probably
less than a couple percent for .2" of drilling through a board. Use two
standard $1 door hinges, mount half to the bottom and slightly cock the
other half against it when attaching the triangle and you will have near
zero play. If you come out 3 or 4 feet, this is almost exactly linear
for short depths while being very easy to construct, I've used it for
simple manual guidance on several occasions. Since it's so simple, I've
also looked at ways to correct it to truly linear, but haven't found one
yet that was viable over a rail system. Stepper mounted to either base
or triangle screwing through a nut on the other side gives it easy control.

Arc of the bit tip can easily be corrected for with the other axis to
make it fully linear in X, Y, and Z too. Tip would be precisely
located, the tool would just have a small change in angle to the work
surface at different depths. Not really needed for just drilling of course.

This is the minimum system I would use, not really any harder to make
than a spring/servo, 95% or more of the correctness of a full axis for a
short depth, yet easier to make and align than a full railed axis.

Alan

--- In Homebrew_PCBs@yahoogroups.com, Alan King <alan@n...> wrote:

> If room isn't an issue, there is also a very simple way for

stepper

> control for a drill..
>
> H H
>
>
>
>
> D SM
>
>
> Looking down, H and H are two hinges on a triangular piece of

wood.

> Come out about a foot, and D is the drill with SM being the

motor. At a

> foot out, the arc of the circle is very close to a line. Get the

bit

> aligned so it's tangent where it's drilling, and the error is

probably

> less than a couple percent for .2" of drilling through a board.

Use two

> standard $1 door hinges, mount half to the bottom and slightly

cock the

> other half against it when attaching the triangle and you will

have near

> zero play. If you come out 3 or 4 feet, this is almost exactly

linear

> for short depths while being very easy to construct, I've used it

for

> simple manual guidance on several occasions. Since it's so

simple, I've

> also looked at ways to correct it to truly linear, but haven't

found one

> yet that was viable over a rail system.

Perhaps the thing to do is NOT convert it to a linear system but
instead work within the constraints of the rotary coordinates. You
may be interested in this page:
http://techref.massmind.org/techref/idea/lccam.htm which is a redux
of
http://freeandeasy.sourceforge.net/ Thier java simulation at
http://freeandeasy.sourceforge.net/rotsim.php is really worth
looking at.

The links describe a method of moving the table using two rotary
joints. I had not considered just using a pair of hinges, but I
wonder why that wouldn't be perfectly acceptable...

James Newton wrote:

>Perhaps the thing to do is NOT convert it to a linear system but
>instead work within the constraints of the rotary coordinates. You
>may be interested in this page:
>http://techref.massmind.org/techref/idea/lccam.htm which is a redux
>
>

Well that works fine enough for the table part, and really not that
hard to convert to X,Y from the polar. But even for that, after looking
at it the rail system I made wasn't much harder to do and had the more
standard X,Y..

Can't just leave it rotary for Z though, at least not for a drill if
you don't want to snap bits. But it's close enough as is as long as the
circle is large enough and you only use a small arc out of the circle.
Easiest way to make it more linear is to move the hinge point further
away, it just gets very unwieldy very fast.

Actually there is a simple way to make it linear now that it hit me, a
scissor jack type of arrangment that looks like an accordion. Similar
to this scissor gate:

http://www.concordelevator.com/Gallery/Infinity%20Fixtures/gate-scissor.jpg

While linear these aren't very dimensionally stable esp when extended,
but with a little thought they might be made so.

Alan