Alan B. Pearce wrote:
>> Most chips will probably survive that, but it's still
>> something bad that shouldn't be allowed to happen.  I
>> would fix the buck regulator design.
>> There is no excuse for it doing that.
>
> There is probably nothing you can do (see my previous post).  The
> problem arises because of the time it takes the feedback loop to
> settle, which is a significant number of microseconds, and long
> enough to destroy components which do not have a absolute max voltage
> rating that can take the overshoot.

There is probably nothing you can do if you are using that particular
regulator chip.  However, this is not an inherent characteristic of a buck
regulator, even if there are apparently a number of bad designs out there.

Lately I've been using 10F204 for switching power supply controllers.  They
are cheaper, very effective, and it is easily possible to design a switching
regulator that does not have startup overshoot.

In my opinion, digital control of switching power supplies is the preferred
way to go, and the 10F PICs with built in comparator, absolute voltage
reference, small size, and low cost, are a great way to do it.  The last
time I designed in an analog switching regulator chip was over a year ago.

One of my current projects is powered either from ethernet or an optional
wall wart.  It has two buck regulators.  The first takes anywhere from 20 to
60 volts and makes 5.5 volts out.  The second takes the 5.5V and makes 3.3V.
Both are controlled by 10F204, although the two use rather different
algorithms.  The first runs in discontinuous mode and the firmware modifies
the pulse with in addition to the time between pulses to optimize
efficiency.  The second runs in continuous mode and uses a very simple
algorithm.  The first supply hovers around 90% efficiency for most of its
range.  I haven't measured the second except with the finger test to see
that it wasn't getting very warm.  Neither of these supplies have startup
overshoots.  The algorithms just don't work that way.

I hear there is going to be an article on a more high end digitally
controlled switching power supply in the January Circuit Cellar.  It does a
PFC (power factor correction) front end, then a buck regulator to make 12V
DC, all controlled from a single 30F2010.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist