Bode plotting for error amplifiers

Sometimes you get presented with a very tricky bit of power electronics to control.

Where the phase delay of the power stage is significant - but also -

the power stage gain is very high as you progress up to full power.

For this you need to understand every part of the loop and what your error amplifier is really contributing.

We plotted one of the trickier EA solutions above - note you can't get this response with a "type 3" controller - more like a type 5.

( copy and paste image into IRfanview for a good look )

Note the relatively high gain at DC, the linearly falling gain ( per decade ), to 700Hz or so,

a wee bit of phase lift around 7 Hz,

and a lot of phase lift from a few 100 Hz to 20kHz or so,

a wee bump in EA gain at 20kHz, with everything rolling off above that.

another view of the same circuit driven from the + input:

Just the ticket to give a total loop that is stable and fast responding - for the hardware we were given.

Remember - once you have the bode plot for your power stage - and the bode plot for for EA or controller - you can sum the two and see if your overall system is stable ( at least 45 deg phase margin and 6dB of gain margin ) if not you can then tweak the EA on paper before you make the prototype go !

for more of the good stuff, go to, pwrtrnx.com