RE: [CAD_CAM_EDM_DRO] Re:low cost stepper driver

> This is what I was wondering. I appreciate that you get a back-EMF spike

when

> it's disconnected, but it's *DISCONNECTED* - it's not the driver

transistor

> switching off so it shouldn't reach the driver. But intermittent

connections (or a

> spark that jumps the gap, as Tony suggested) is possibly the reason.
>
> I note that some good quality drivers have tranzorbs which can presumably
> absorb more energy than the internal diodes (and also don't dump it into

the

> power supply).

You get back-EMF when the coil stops being energized, unplugging the motor
is just one way of doing that.

Think of the coil as a spring; you push down on it (apply power), the spring
resists somewhat (inductance), when you let it go it'll spring back at you
(back EMF) at high speed (high voltage).

The classic circuit is a relay activated by a transistor. Your
microprocessor or parallel port provides a small signal that switches the
transistor on, and current flows through it to the coil.

When you turn the transistor off, the coil is still charged. That power
needs to go somewhere.

As said the magnetic field promptly collapses and generates a current in the
wire. This has two properties that interest us; one is the polarity is
reversed, and the other is the voltage is much higher than we used to charge
it.

Transistors don't like power going backwards across them (reverse bias) so
that can cause damage, but it's the high voltage that causes more hassles.
The high voltage can cause arcing in the transistor itself, causing damage
that may not show until months later. Worse it can escape, travelling down
the path to your uC or port. The high voltage can damage or disrupt that.
Optocouplers have a gap the high voltage can't jump, that's how they save
your parallel port.

The diode placed across the relay coil the only allows power to flow in one
direction, so when the back EMF is generated the diode absorbs it.

Years ago a local firm had up some alarm boards made for me, nice boards but
they didn't work properly. Yep, they left the diode off the relay, and
that'd cause the uC to glitch.

For unipolar steppers (6 or 8 wire) their driver boards are (at a basic
level) no more complicated than driving a relay. You have 4 coils, and you
turn them on and off. Across each coil you place a diode for protection.
Indeed you can make a driver with nothing more than 4 transistors, not
recommended though.

Bipolar steppers get more complicated as you need to reverse the voltage
across the coils, so now the diode trick doesn't work. Back EMF causing a
problem with reversed polarity isn't that much of a problem as the drivers
expect that. H-bridge circuits do have diodes, but across the transistors
rather than the coil to deal with that. If you read up on those you can see
how the protection works.

So long as the motor remains connected the driver can deal with the back
EMF, however if you disconnect it then you may create the spike when the
driver isn't ready for it, and that can damage it.

Transorbs are like diodes that allow current in both directions, but they
clamp the high voltage spikes. They cost money, and are only useful in very
few circumstances (the oops! ones) so are left off the board. Well, the
boards we're talking about anyway.

Don't forget the other half of wire damage is short circuits, but that's
another story.

Some drivers are smart enough to monitor the current, notice when something
goes wrong and protect themselves.

Tony