Searching \ for 'Clarifying PIC control of DC-DC converter' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'Clarifying PIC control of DC-DC converter'.

Truncated match.
PICList Thread
'Clarifying PIC control of DC-DC converter'
1996\05\15@205559 by Hamish Seaton

flavicon
face
>Do you wish the PIC to be the controller or to provide some
>secondary control over and above the National parts?  If the
>PIC is the controller, would that be a cost advantage over the
>large array of switching converters available out there?  If you are
>trying to combine the switching function with other
>functions in the application, you may consider using a PIC
>with PWM to drive the switching element (i.e. logic level mosfet) and
>A/D to periodicaly test and adjust the conversion duty.  I don't have
>much experience with battery charging, so perhaps there are others
>who can suggest what parameter to regulate.
>
>                                  Mike

Thanks for the reply mike

Just to clarify, I'm trying to get away from using specilised DC-DC
converter control chips and use the PIC that is already in my circuit.  I
want the switching circuit to be similar to that used by the National Simple
Switcher Series  ie. small components- 50uH, 250uF etc operating at 20-100kHz.

I need the DC-DC converter to boost my input voltage from 12 up to 15.5V
This is used to charge the battery.  The plan is to use a 71 with an A/D and
periodically check on the charging voltage and current.

At present I plan to let the switcher run using the CLKout of the PIC as the
timebase.  I cna externally gate this signal, depending on the output
voltage level and the current drawn from the battery.  What I am unsure of
is if I will need to sample the output voltage of the switcher every
switching cycle, or if it will be sufficient to check the output voltage
every millisecond or so (as the load on the switcher will be relatively
constant).  I'm pretty sure what I'm doing is feasible, it will just be a
bit of a control & stability problem.  I will be using a logic level MOSFET
as the switching element - have read the existing letters concerning these.  Ta.
Hamish
>

1996\05\15@215320 by Scott Dattalo

face
flavicon
face
Hamish,

I have a couple of recommendations.

1) Use an external Current Mode Control Pulse Width Modulator like the x3842.
This way, cycle by cycle control is relegated to the PWM IC and not the PIC.
The duty cycle and consequently the output voltage can be controlled by varying
the voltage on the feedback/compensation of the PWM IC. In other words, you can
add or subtract to the normal feedback signal to vary the output voltage.

2) Be careful using 12 volts as your charging voltage. This will force you to
use
a flyback topology (== transformer) as opposed to a simpler boost circuit
(==inductor). The reason is that a "12 volt" lead-acid battery has a fully
charged
terminal voltage of 14.4V and a fully discharged terminal voltage of about
10.2V.
Since this range of voltages crosses the 12 volt boundary, you will be required
to
buck and boost the DC-DC converter.

3) Beware of Logic level MOSFETs. They can be controlled with 0/5V gate to
source
voltages. However, they have a fair amount of gate capacitance. Thus if you want
to have a quasi efficient power supply, you will need something with a little
umph
to drive them. (Soft edges in switching power supplies == power loss).

4) Don't forget your temperature sensor.

5) Don't count on the PIC for fail-safe operation. I am not knocking the PIC.
But,
there will be times when PIC is not active but the DC-DC converter is. (Like on
power up, and software crashes). Dangerous high voltages could develop in these
situations.



Scott

More... (looser matching)
- Last day of these posts
- In 1996 , 1997 only
- Today
- New search...