CAP sizing

[ltj]

I have a 3.3V module that draws 60mA when idle and jumps to 1A about every
second. I had a 10uf tantalum on the output of the LM1086 regulator but
that doesn't seem to be enough. How do I figure what caps to put on the IN
and OUT of the regulator?
thx

 

[John Fields]

On Fri, 12 Dec 2003 10:32:10 -0500, "ltj" <ltj@yahoo.com> wrote:

I have a 3.3V module that draws 60mA when idle and jumps to 1A about every
second. I had a 10uf tantalum on the output of the LM1086 regulator but
that doesn't seem to be enough. How do I figure what caps to put on the IN
and OUT of the regulator?
---

You need to know what the input voltage to the LM1086 is when you put a
60 mA load on it and what happens to the input voltage when you change
the load to 1A. You also need to know how often (with a little more
precision than "about") the load changes from 60mA to 1A and how long
the 1A load lasts before it goes back to 60mA. You also need know how
the input to the regulator is being generated. That is, do you have a
transformer connected to the mains driving a full wave rectifier, or
what?

--
John Fields

 

[ltj]

"John Fields" <jfields@austininstruments.com> wrote in message
news:98gktv4rbldk3je4esnqfqh131eb9t0pbf@4ax.com...

On Fri, 12 Dec 2003 10:32:10 -0500, "ltj" <ltj@yahoo.com> wrote:

I have a 3.3V module that draws 60mA when idle and jumps to 1A about
every
second. I had a 10uf tantalum on the output of the LM1086 regulator but
that doesn't seem to be enough. How do I figure what caps to put on the
IN
and OUT of the regulator?
---
You need to know what the input voltage to the LM1086 is when you put a
60 mA load on it and what happens to the input voltage when you change
the load to 1A. You also need to know how often (with a little more
precision than "about") the load changes from 60mA to 1A and how long
the 1A load lasts before it goes back to 60mA. You also need know how
the input to the regulator is being generated. That is, do you have a
transformer connected to the mains driving a full wave rectifier, or
what?

--
John Fields

The 1A loading occurs every 2.2 seconds (or longer). Best I can tell it is
actually a series of ten pulses, 3mS each, lasting a total of .3 seconds.

I need to run this from a battery pack. Initial attempts with 4xAA alkaline
dropped to near zero volts when pushed to 1 amp. I know alkalines are the
worst for high draw circuits so I'll try a NiMH pack when I get home. I'm
not concerned with battery life since this is a portable test instrument
that will only be used occasionally for an hour at a time. I'd like to have
caps provide this current but will go to C or D cells ($$$ in NiMH) if that
is the only solution.

tom

 

[Nick Hull]

In article <3fda4f98$0$30456$a04e5680@nnrp.fuse.net>,
"ltj" <ltj@yahoo.com> wrote:

"John Fields" <jfields@austininstruments.com> wrote in message
news:98gktv4rbldk3je4esnqfqh131eb9t0pbf@4ax.com...
On Fri, 12 Dec 2003 10:32:10 -0500, "ltj" <ltj@yahoo.com> wrote:

I have a 3.3V module that draws 60mA when idle and jumps to 1A about
every
second. I had a 10uf tantalum on the output of the LM1086 regulator but

The 1A loading occurs every 2.2 seconds (or longer). Best I can tell it is
actually a series of ten pulses, 3mS each, lasting a total of .3 seconds.

I need to run this from a battery pack. Initial attempts with 4xAA alkaline
dropped to near zero volts when pushed to 1 amp. I know alkalines are the
worst for high draw circuits so I'll try a NiMH pack when I get home. I'm
not concerned with battery life since this is a portable test instrument
that will only be used occasionally for an hour at a time. I'd like to have
caps provide this current but will go to C or D cells ($$$ in NiMH) if that
is the only solution.

How about putting 2 Alk "D" cells to power the 3.3v module directly and
use a very simple charger circuit to keep the D cells charged? I did
something similiar with an electric clock that would only work on a
really fresh cell; I kept the cell charged so it worked for years
without discharging.

If you want to go the cap route, I use a similiar circuit to run
electric fencers. They take a big current surge every second. I
usually use a big (about 2000 mf) cap to provide the surge. A quick
test to see if it is big enough is to disconnect the power source (in
my case a junk 12 v auto battery) and see how many pulses I get out of
the stored energy in the cap. If I can get about 3 pulses then I figure
the cap is big enough.

--
free men own guns - slaves don't
www.geocities.com/CapitolHill/5357/

 

[ltj]

NiMH did not help. I scoped a voltage drop down to 3.0V (not 0V) when the
surge occurs.
tom



"ltj" <ltj@yahoo.com> wrote in message
news:3fda4f98$0$30456$a04e5680@nnrp.fuse.net...

"John Fields" <jfields@austininstruments.com> wrote in message
You need to know what the input voltage to the LM1086 is when you put a
60 mA load on it and what happens to the input voltage when you change
the load to 1A. You also need to know how often (with a little more
precision than "about") the load changes from 60mA to 1A and how long
the 1A load lasts before it goes back to 60mA. You also need know how
the input to the regulator is being generated. That is, do you have a
transformer connected to the mains driving a full wave rectifier, or
what?

--
John Fields


The 1A loading occurs every 2.2 seconds (or longer). Best I can tell it
is
actually a series of ten pulses, 3mS each, lasting a total of .3 seconds.

I need to run this from a battery pack. Initial attempts with 4xAA
alkaline
dropped to near zero volts when pushed to 1 amp. I know alkalines are the
worst for high draw circuits so I'll try a NiMH pack when I get home. I'm
not concerned with battery life since this is a portable test instrument
that will only be used occasionally for an hour at a time. I'd like to
have
caps provide this current but will go to C or D cells ($$$ in NiMH) if
that
is the only solution.

tom