Mail Archives: geda-user/2011/12/21/14:56:11
On Dec 21, 2011, at 12:32 PM, John Griessen wrote:
> On 12/21/2011 01:09 PM, DJ Delorie wrote:
>> I'd assumed rotor inertia would be sufficient, but I suppose "that
>> depends" as usual.
>
> I'm sure the inertia is plenty to keep it going. The smoothing can be good if the motor
> inductive response of the bare motor makes big volt spikes between
> times it is driven by transistors.
Typically you handle those spikes with diodes, often the body diodes of the MOSFET switches.
> A series wound brush DC motor with some added
> capacitance at the motor, downstream of the FET switches will be less noisy,
> less likely to have a big volt spike if run hard then cut off.
>
> On 12/21/2011 01:16 PM, John Doty wrote:
> > The problem with added capacitance is that it increases the dissipation of the driver, since it then has to charge/discharge the capacitors when it switches. With enough capacitance to smooth the waveform, this will be a big effect.
>
> I'm suggesting capacitance enough to make the PWM square wave disappear, not round its edges.
> and it's at the motor. When the FETs cut on they just have to top up the voltage the cap stored since
> last time.
But if you turn the FET fully on, it will charge the cap up to the supply voltage, defeating the PWM.
Usually with PWM you use inductance (possibly intrinsic to the thing you're driving) to smooth the current, and let the voltage swing rail to rail, with larger excursions prevented by diodes.
John Doty Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com
- Raw text -