Re: [CAD_CAM_EDM_DRO] Re: FS: New 4 axis microstepping driver $45.00 USD

On Monday 22 March 2004 10:41 pm, Mariss Freimanis wrote:

> I think regulating the power supply does not address the bigger
> problem of returned energy and in fact, will make it much worse.
>
> Those of you who have scopes may want to try a little experiment that
> could destroy your drive.
>
> 1) Place the scope probe across your DC power supply. Set the input
> coupling to DC, 10V / div vertical scale, 100 milliseconds / div on
> the timebase. Trigger set to AUTO.
>
> 2) Spin up a NEMA-23 or NEMA-34 motor to 2,000 to 3,000 RPM.
>
> 3) Stall the motor while watching the scope.
>
> You will see a significant voltage "bump" of at least 15 to 20 VDC on
> top of your power supply voltage. This means a 35V supply will bump
> up to 50 or 55 VDC for about 100 to 200 milliseconds.
>
> What you are seeing is the mechanical energy stored in the motor
> being returned as a regenerated current. If the supply is
> unregulated, the filter cap will absorb it and give a smaller "bump".
>
> With a regulator in place, this current has no place to go. A zener
> clamp won't help either; a 33V, 5W zener (1N5364B) has a series
> resistance of 10 Ohms. 2 amps of returned current turns this 33V
> zener into a 53V zener. This kills L297/298 everytime; they go into
> destructive secondary breakdown. I'm not as familiar what the Allegro
> part would do.
>
> What's needed is an active circuit voltage clamp to make it
> unconditionally safe.

That's why, whenever I build something that has a regulator in there, I
always put a couple of diodes in place, mostly to protect the regulator.
One goes across the output, but so that it's reverse-biased. The other goes
across the regulator, likewise normally reverse-biased, cathode connected
to the filter cap and anode connected to the output. In a scenario like the
one you describe this diode goes into conduction and the excess energy gets
dumped into the filter cap. The output voltage never goes any higher than
one diode drop above the regulated output voltage...