Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
Power Supply / Regulator Question
Started by ●January 17, 2007
Reply by ●January 17, 20072007-01-17
Jeff Shapiro wrote:
> Hi,
>
> I have an RCM2200 design that I suspect had problems in some
> occasions because the LM7805 (TO-220 with heat sink) was getting very
> hot. Depending on the enclosure the customer used and the ambient
> temperature I believe it led to some Rabbit module failures. It
> seemed like we got more failures during the warmer months especially
> from places in the country that are warmer (i.e. Texas vs Michigan).
> In a later revision, a "quick fix" we did was to add about 6 diodes
> in series between the 12V input and the 5V regulator to cut some of
> the voltage prior to entering the regulator.
>
> The input to the board is about 12.5V and the output from the
> regulator is 5V at about 150mA driving the RCM2200 and some external
> parts.
>
> I'm starting a new design using a RCM3700 and will need a regulated
> 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> at about 1A and be able to supply 12V at atleast 750mA and 5V at
> maybe 250mA. The "easy" (and least expensive) way would be to just
> use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> I'm worried about the heat issues I had with my other product.
>
> Any recommendations for something that won't add too much cost to
> the total project?
>
(12-5)*0.15=1.15 watts
At .25a it is almost 1.9 watts
If the input is unregulated, the heat will be even worse. Most 12
adapters start at almost 17v unloaded. Plan a larger heatsink than
necessary. I like to keep them hardly noticably warm in free air.
Use one of the National Semi Simple Switcher supplies, I use the LM2975
on several designs. They work well. I usually get about 75% efficiency.
I have done switching supplies with 90% efficient. The ZW schematic on
the proto boards includes a sample 5v circuit. The data sheet has tables
to pick the correct inductor size. Usually it is not that critical, the
best value will maximize efficiency. The added cost is minimal; usually
less than the labor an parts cost of mounting the reg and installing a
heat sink (all parts can be automatically assembled.)
Be sure to add extra caps before and after a switching supply. I put as
large a cap as I can use before the switcher. Also have Polyfuse PTC
fuse and a cap on the input. Beside being a fuse, it also adds an
addition R/C filter on the input.
For 12V, you would probably want a LDO (Low Drop out) regulator. Most
78xx regulators need 2.5v more in than out to work well. The LDO type
usually only need about 0.6v.
------
| Scott G. Henion| s...@shdesigns.org |
| Consultant | Stone Mountain, GA |
| SHDesigns http://www.shdesigns.org |
------
Rabbit libs: http://www.shdesigns.org/rabbit/
today's fortune
Hhhmmmmmmmm
waterbeds for cows
eleet
Culus: why would a cow need a waterbed?
cas: To be comfy warm
> Hi,
>
> I have an RCM2200 design that I suspect had problems in some
> occasions because the LM7805 (TO-220 with heat sink) was getting very
> hot. Depending on the enclosure the customer used and the ambient
> temperature I believe it led to some Rabbit module failures. It
> seemed like we got more failures during the warmer months especially
> from places in the country that are warmer (i.e. Texas vs Michigan).
> In a later revision, a "quick fix" we did was to add about 6 diodes
> in series between the 12V input and the 5V regulator to cut some of
> the voltage prior to entering the regulator.
>
> The input to the board is about 12.5V and the output from the
> regulator is 5V at about 150mA driving the RCM2200 and some external
> parts.
>
> I'm starting a new design using a RCM3700 and will need a regulated
> 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> at about 1A and be able to supply 12V at atleast 750mA and 5V at
> maybe 250mA. The "easy" (and least expensive) way would be to just
> use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> I'm worried about the heat issues I had with my other product.
>
> Any recommendations for something that won't add too much cost to
> the total project?
>
(12-5)*0.15=1.15 watts
At .25a it is almost 1.9 watts
If the input is unregulated, the heat will be even worse. Most 12
adapters start at almost 17v unloaded. Plan a larger heatsink than
necessary. I like to keep them hardly noticably warm in free air.
Use one of the National Semi Simple Switcher supplies, I use the LM2975
on several designs. They work well. I usually get about 75% efficiency.
I have done switching supplies with 90% efficient. The ZW schematic on
the proto boards includes a sample 5v circuit. The data sheet has tables
to pick the correct inductor size. Usually it is not that critical, the
best value will maximize efficiency. The added cost is minimal; usually
less than the labor an parts cost of mounting the reg and installing a
heat sink (all parts can be automatically assembled.)
Be sure to add extra caps before and after a switching supply. I put as
large a cap as I can use before the switcher. Also have Polyfuse PTC
fuse and a cap on the input. Beside being a fuse, it also adds an
addition R/C filter on the input.
For 12V, you would probably want a LDO (Low Drop out) regulator. Most
78xx regulators need 2.5v more in than out to work well. The LDO type
usually only need about 0.6v.
------
| Scott G. Henion| s...@shdesigns.org |
| Consultant | Stone Mountain, GA |
| SHDesigns http://www.shdesigns.org |
------
Rabbit libs: http://www.shdesigns.org/rabbit/
today's fortune
Hhhmmmmmmmm
waterbeds for cows
eleet
Culus: why would a cow need a waterbed?
cas: To be comfy warm
Reply by ●January 17, 20072007-01-17
Use switch mode regulators.
You can get a 5V switchmode that is a pin for pin replacement for the 7805. We have used them in the past to get ourselves out of trouble. I'm not in the office at the moment so can't tell you the details of them.
Regards,
Nathan
________________________________
From: r... on behalf of Jeff Shapiro
Sent: Thu 1/18/2007 11:41 AM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
You can get a 5V switchmode that is a pin for pin replacement for the 7805. We have used them in the past to get ourselves out of trouble. I'm not in the office at the moment so can't tell you the details of them.
Regards,
Nathan
________________________________
From: r... on behalf of Jeff Shapiro
Sent: Thu 1/18/2007 11:41 AM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
Reply by ●January 17, 20072007-01-17
Pop a little switcher on there...take a look at the National LM267X parts.
Don't those diodes get awfully hot? ;-)
_____
From: r... [mailto:r...] On
Behalf Of Jeff Shapiro
Sent: Wednesday, January 17, 2007 4:41 PM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
Don't those diodes get awfully hot? ;-)
_____
From: r... [mailto:r...] On
Behalf Of Jeff Shapiro
Sent: Wednesday, January 17, 2007 4:41 PM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
Reply by ●January 17, 20072007-01-17
A quick and cheap fix would be a power resistor.
http://www.ckp-railways.talktalk.net/pcbcad28.htm
Dave Moore wrote:
Pop a little switcher on there...take a look at the National LM267X parts.
Don't those diodes get awfully hot? ;-)
---------------------------------
From: r... [mailto:r...] On Behalf Of Jeff Shapiro
Sent: Wednesday, January 17, 2007 4:41 PM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
---------------------------------
New Yahoo! Mail is the ultimate force in competitive emailing. Find out more at the Yahoo! Mail Championships. Plus: play games and win prizes.
http://www.ckp-railways.talktalk.net/pcbcad28.htm
Dave Moore wrote:
Pop a little switcher on there...take a look at the National LM267X parts.
Don't those diodes get awfully hot? ;-)
---------------------------------
From: r... [mailto:r...] On Behalf Of Jeff Shapiro
Sent: Wednesday, January 17, 2007 4:41 PM
To: r...
Subject: [rabbit-semi] Power Supply / Regulator Question
Hi,
I have an RCM2200 design that I suspect had problems in some
occasions because the LM7805 (TO-220 with heat sink) was getting very
hot. Depending on the enclosure the customer used and the ambient
temperature I believe it led to some Rabbit module failures. It
seemed like we got more failures during the warmer months especially
from places in the country that are warmer (i.e. Texas vs Michigan).
In a later revision, a "quick fix" we did was to add about 6 diodes
in series between the 12V input and the 5V regulator to cut some of
the voltage prior to entering the regulator.
The input to the board is about 12.5V and the output from the
regulator is 5V at about 150mA driving the RCM2200 and some external
parts.
I'm starting a new design using a RCM3700 and will need a regulated
12V and 5V. I'd like to use an unregulated "wall wort" supply rated
at about 1A and be able to supply 12V at atleast 750mA and 5V at
maybe 250mA. The "easy" (and least expensive) way would be to just
use a 7812 to get my 12V and then run that into a 7805 to get 5V but
I'm worried about the heat issues I had with my other product.
Any recommendations for something that won't add too much cost to
the total project?
Thanks!
Jeff
---------------------------------
New Yahoo! Mail is the ultimate force in competitive emailing. Find out more at the Yahoo! Mail Championships. Plus: play games and win prizes.
Reply by ●January 17, 20072007-01-17
Yup, I'm with the majority here too.. National Simple Switcher is the way
to
go.
I've used the 2575's but am more fond of the 5007 lately for smaller loads.
They're not what I would consider "cheap" though, pushing $8 in parts is a
bit steep IMHO.
I like it/them because they are truly 'simple', the 5007 especially is teeny
tiny,
and quite flexible.. Not to mention readily available from your favorite
distributor.
National even has a feature on their web site that will visually show you a
calculate expectancy of your heat dissipation (very cool).
here's a pic of 1 of 2 of my on board power supplies using the 5007:
http://ssminnow.maesoft.net/reflow/IMG_1021.jpg
Smaller than a dime, and no heat sink needed...
Dan...
On 1/17/07, Scott Henion wrote:
>
> Jeff Shapiro wrote:
> > Hi,
> >
> > I have an RCM2200 design that I suspect had problems in some
> > occasions because the LM7805 (TO-220 with heat sink) was getting very
> > hot. Depending on the enclosure the customer used and the ambient
> > temperature I believe it led to some Rabbit module failures. It
> > seemed like we got more failures during the warmer months especially
> > from places in the country that are warmer (i.e. Texas vs Michigan).
> > In a later revision, a "quick fix" we did was to add about 6 diodes
> > in series between the 12V input and the 5V regulator to cut some of
> > the voltage prior to entering the regulator.
> >
> > The input to the board is about 12.5V and the output from the
> > regulator is 5V at about 150mA driving the RCM2200 and some external
> > parts.
> >
> > I'm starting a new design using a RCM3700 and will need a regulated
> > 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > maybe 250mA. The "easy" (and least expensive) way would be to just
> > use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> > I'm worried about the heat issues I had with my other product.
> >
> > Any recommendations for something that won't add too much cost to
> > the total project?
> >
> (12-5)*0.15=1.15 watts
> At .25a it is almost 1.9 watts
>
> If the input is unregulated, the heat will be even worse. Most 12
> adapters start at almost 17v unloaded. Plan a larger heatsink than
> necessary. I like to keep them hardly noticably warm in free air.
>
> Use one of the National Semi Simple Switcher supplies, I use the LM2975
> on several designs. They work well. I usually get about 75% efficiency.
> I have done switching supplies with 90% efficient. The ZW schematic on
> the proto boards includes a sample 5v circuit. The data sheet has tables
> to pick the correct inductor size. Usually it is not that critical, the
> best value will maximize efficiency. The added cost is minimal; usually
> less than the labor an parts cost of mounting the reg and installing a
> heat sink (all parts can be automatically assembled.)
>
> Be sure to add extra caps before and after a switching supply. I put as
> large a cap as I can use before the switcher. Also have Polyfuse PTC
> fuse and a cap on the input. Beside being a fuse, it also adds an
> addition R/C filter on the input.
>
> For 12V, you would probably want a LDO (Low Drop out) regulator. Most
> 78xx regulators need 2.5v more in than out to work well. The LDO type
> usually only need about 0.6v.
>
> ------
> | Scott G. Henion| s...@shdesigns.org |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
> today's fortune
> Hhhmmmmmmmm
> waterbeds for cows
> eleet
> Culus: why would a cow need a waterbed?
> cas: To be comfy warm
>
>
>
go.
I've used the 2575's but am more fond of the 5007 lately for smaller loads.
They're not what I would consider "cheap" though, pushing $8 in parts is a
bit steep IMHO.
I like it/them because they are truly 'simple', the 5007 especially is teeny
tiny,
and quite flexible.. Not to mention readily available from your favorite
distributor.
National even has a feature on their web site that will visually show you a
calculate expectancy of your heat dissipation (very cool).
here's a pic of 1 of 2 of my on board power supplies using the 5007:
http://ssminnow.maesoft.net/reflow/IMG_1021.jpg
Smaller than a dime, and no heat sink needed...
Dan...
On 1/17/07, Scott Henion wrote:
>
> Jeff Shapiro wrote:
> > Hi,
> >
> > I have an RCM2200 design that I suspect had problems in some
> > occasions because the LM7805 (TO-220 with heat sink) was getting very
> > hot. Depending on the enclosure the customer used and the ambient
> > temperature I believe it led to some Rabbit module failures. It
> > seemed like we got more failures during the warmer months especially
> > from places in the country that are warmer (i.e. Texas vs Michigan).
> > In a later revision, a "quick fix" we did was to add about 6 diodes
> > in series between the 12V input and the 5V regulator to cut some of
> > the voltage prior to entering the regulator.
> >
> > The input to the board is about 12.5V and the output from the
> > regulator is 5V at about 150mA driving the RCM2200 and some external
> > parts.
> >
> > I'm starting a new design using a RCM3700 and will need a regulated
> > 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > maybe 250mA. The "easy" (and least expensive) way would be to just
> > use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> > I'm worried about the heat issues I had with my other product.
> >
> > Any recommendations for something that won't add too much cost to
> > the total project?
> >
> (12-5)*0.15=1.15 watts
> At .25a it is almost 1.9 watts
>
> If the input is unregulated, the heat will be even worse. Most 12
> adapters start at almost 17v unloaded. Plan a larger heatsink than
> necessary. I like to keep them hardly noticably warm in free air.
>
> Use one of the National Semi Simple Switcher supplies, I use the LM2975
> on several designs. They work well. I usually get about 75% efficiency.
> I have done switching supplies with 90% efficient. The ZW schematic on
> the proto boards includes a sample 5v circuit. The data sheet has tables
> to pick the correct inductor size. Usually it is not that critical, the
> best value will maximize efficiency. The added cost is minimal; usually
> less than the labor an parts cost of mounting the reg and installing a
> heat sink (all parts can be automatically assembled.)
>
> Be sure to add extra caps before and after a switching supply. I put as
> large a cap as I can use before the switcher. Also have Polyfuse PTC
> fuse and a cap on the input. Beside being a fuse, it also adds an
> addition R/C filter on the input.
>
> For 12V, you would probably want a LDO (Low Drop out) regulator. Most
> 78xx regulators need 2.5v more in than out to work well. The LDO type
> usually only need about 0.6v.
>
> ------
> | Scott G. Henion| s...@shdesigns.org |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
> today's fortune
> Hhhmmmmmmmm
> waterbeds for cows
> eleet
> Culus: why would a cow need a waterbed?
> cas: To be comfy warm
>
>
>
Reply by ●January 18, 20072007-01-18
I agree with Scott.
Another option you can consider is the LNK500 family. Most agents also
stock the magnetics required. This solution will allow you to drop the
Walmart adaptor. I do however agree that the simpler solution using a
simple switcher with input protection is the way to go. Maybe the
addition of a transzorb and a reverse polarity protection diode would be
worth considering. An interesting thought - your design is only as
robust as your PSU is. ST also have a part the L5973D, they cost less
than a $ and can supply up to 2.5 amps.
Alexis
Scott Henion wrote:
>
> Jeff Shapiro wrote:
> > Hi,
> >
> > I have an RCM2200 design that I suspect had problems in some
> > occasions because the LM7805 (TO-220 with heat sink) was getting very
> > hot. Depending on the enclosure the customer used and the ambient
> > temperature I believe it led to some Rabbit module failures. It
> > seemed like we got more failures during the warmer months especially
> > from places in the country that are warmer (i.e. Texas vs Michigan).
> > In a later revision, a "quick fix" we did was to add about 6 diodes
> > in series between the 12V input and the 5V regulator to cut some of
> > the voltage prior to entering the regulator.
> >
> > The input to the board is about 12.5V and the output from the
> > regulator is 5V at about 150mA driving the RCM2200 and some external
> > parts.
> >
> > I'm starting a new design using a RCM3700 and will need a regulated
> > 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > maybe 250mA. The "easy" (and least expensive) way would be to just
> > use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> > I'm worried about the heat issues I had with my other product.
> >
> > Any recommendations for something that won't add too much cost to
> > the total project?
> >
> (12-5)*0.15=1.15 watts
> At .25a it is almost 1.9 watts
>
> If the input is unregulated, the heat will be even worse. Most 12
> adapters start at almost 17v unloaded. Plan a larger heatsink than
> necessary. I like to keep them hardly noticably warm in free air.
>
> Use one of the National Semi Simple Switcher supplies, I use the LM2975
> on several designs. They work well. I usually get about 75% efficiency.
> I have done switching supplies with 90% efficient. The ZW schematic on
> the proto boards includes a sample 5v circuit. The data sheet has tables
> to pick the correct inductor size. Usually it is not that critical, the
> best value will maximize efficiency. The added cost is minimal; usually
> less than the labor an parts cost of mounting the reg and installing a
> heat sink (all parts can be automatically assembled.)
>
> Be sure to add extra caps before and after a switching supply. I put as
> large a cap as I can use before the switcher. Also have Polyfuse PTC
> fuse and a cap on the input. Beside being a fuse, it also adds an
> addition R/C filter on the input.
>
> For 12V, you would probably want a LDO (Low Drop out) regulator. Most
> 78xx regulators need 2.5v more in than out to work well. The LDO type
> usually only need about 0.6v.
>
> ------
> | Scott G. Henion| s...@shdesigns.org
> |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
>
> today's fortune
> Hhhmmmmmmmm
> waterbeds for cows
> eleet
> Culus: why would a cow need a waterbed?
> cas: To be comfy warm
>
>
Another option you can consider is the LNK500 family. Most agents also
stock the magnetics required. This solution will allow you to drop the
Walmart adaptor. I do however agree that the simpler solution using a
simple switcher with input protection is the way to go. Maybe the
addition of a transzorb and a reverse polarity protection diode would be
worth considering. An interesting thought - your design is only as
robust as your PSU is. ST also have a part the L5973D, they cost less
than a $ and can supply up to 2.5 amps.
Alexis
Scott Henion wrote:
>
> Jeff Shapiro wrote:
> > Hi,
> >
> > I have an RCM2200 design that I suspect had problems in some
> > occasions because the LM7805 (TO-220 with heat sink) was getting very
> > hot. Depending on the enclosure the customer used and the ambient
> > temperature I believe it led to some Rabbit module failures. It
> > seemed like we got more failures during the warmer months especially
> > from places in the country that are warmer (i.e. Texas vs Michigan).
> > In a later revision, a "quick fix" we did was to add about 6 diodes
> > in series between the 12V input and the 5V regulator to cut some of
> > the voltage prior to entering the regulator.
> >
> > The input to the board is about 12.5V and the output from the
> > regulator is 5V at about 150mA driving the RCM2200 and some external
> > parts.
> >
> > I'm starting a new design using a RCM3700 and will need a regulated
> > 12V and 5V. I'd like to use an unregulated "wall wort" supply rated
> > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > maybe 250mA. The "easy" (and least expensive) way would be to just
> > use a 7812 to get my 12V and then run that into a 7805 to get 5V but
> > I'm worried about the heat issues I had with my other product.
> >
> > Any recommendations for something that won't add too much cost to
> > the total project?
> >
> (12-5)*0.15=1.15 watts
> At .25a it is almost 1.9 watts
>
> If the input is unregulated, the heat will be even worse. Most 12
> adapters start at almost 17v unloaded. Plan a larger heatsink than
> necessary. I like to keep them hardly noticably warm in free air.
>
> Use one of the National Semi Simple Switcher supplies, I use the LM2975
> on several designs. They work well. I usually get about 75% efficiency.
> I have done switching supplies with 90% efficient. The ZW schematic on
> the proto boards includes a sample 5v circuit. The data sheet has tables
> to pick the correct inductor size. Usually it is not that critical, the
> best value will maximize efficiency. The added cost is minimal; usually
> less than the labor an parts cost of mounting the reg and installing a
> heat sink (all parts can be automatically assembled.)
>
> Be sure to add extra caps before and after a switching supply. I put as
> large a cap as I can use before the switcher. Also have Polyfuse PTC
> fuse and a cap on the input. Beside being a fuse, it also adds an
> addition R/C filter on the input.
>
> For 12V, you would probably want a LDO (Low Drop out) regulator. Most
> 78xx regulators need 2.5v more in than out to work well. The LDO type
> usually only need about 0.6v.
>
> ------
> | Scott G. Henion| s...@shdesigns.org
> |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
>
> today's fortune
> Hhhmmmmmmmm
> waterbeds for cows
> eleet
> Culus: why would a cow need a waterbed?
> cas: To be comfy warm
>
>
Reply by ●January 18, 20072007-01-18
Dear Dan
I looked at the photo you posted, it certainly looks promosing for
the size front except for one thing. Is that big electrolytic
necessary ?
I was thinking of trying Switching regulators after years of using
conventional regulators (ie chuck away any surplus volts as heat).
The capacitors are often a critical factors I have found in the long
life reliability of circuits. In particular I refer to electrolytic
Caps which have a short life expectancy than most other devices. To
this end I have always tried to avoid them where possible. Tantalums
were my original perfered option, although more expensive but now we
have much larger capacity ceramics available which are usually
cheaper and are'nt polarised.
Can your LM5007 design be made to work with just ceramic or tant
caps ? Can you post schematic of the regulator section ?
I have found I can buy the LM5007 from Farnell (NewarkInOne for those
in the US) as one-off for 1.57 (about $3).
--- In r..., "Dan Allen"
wrote:
>
> Yup, I'm with the majority here too.. National Simple Switcher is
the way to
> go.
>
> I've used the 2575's but am more fond of the 5007 lately for
smaller loads.
> They're not what I would consider "cheap" though, pushing $8 in
parts is a
> bit steep IMHO.
> I like it/them because they are truly 'simple', the 5007 especially
is teeny
> tiny,
> and quite flexible.. Not to mention readily available from your
favorite
> distributor.
> National even has a feature on their web site that will visually
show you a
> calculate expectancy of your heat dissipation (very cool).
>
> here's a pic of 1 of 2 of my on board power supplies using the 5007:
> http://ssminnow.maesoft.net/reflow/IMG_1021.jpg
> Smaller than a dime, and no heat sink needed...
> Dan...
>
> On 1/17/07, Scott Henion wrote:
> >
> > Jeff Shapiro wrote:
> > > Hi,
> > >
> > > I have an RCM2200 design that I suspect had problems in some
> > > occasions because the LM7805 (TO-220 with heat sink) was
getting very
> > > hot. Depending on the enclosure the customer used and the
ambient
> > > temperature I believe it led to some Rabbit module failures. It
> > > seemed like we got more failures during the warmer months
especially
> > > from places in the country that are warmer (i.e. Texas vs
Michigan).
> > > In a later revision, a "quick fix" we did was to add about 6
diodes
> > > in series between the 12V input and the 5V regulator to cut
some of
> > > the voltage prior to entering the regulator.
> > >
> > > The input to the board is about 12.5V and the output from the
> > > regulator is 5V at about 150mA driving the RCM2200 and some
external
> > > parts.
> > >
> > > I'm starting a new design using a RCM3700 and will need a
regulated
> > > 12V and 5V. I'd like to use an unregulated "wall wort" supply
rated
> > > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > > maybe 250mA. The "easy" (and least expensive) way would be to
just
> > > use a 7812 to get my 12V and then run that into a 7805 to get
5V but
> > > I'm worried about the heat issues I had with my other product.
> > >
> > > Any recommendations for something that won't add too much cost
to
> > > the total project?
> > >
> > (12-5)*0.15=1.15 watts
> > At .25a it is almost 1.9 watts
> >
> > If the input is unregulated, the heat will be even worse. Most 12
> > adapters start at almost 17v unloaded. Plan a larger heatsink than
> > necessary. I like to keep them hardly noticably warm in free air.
> >
> > Use one of the National Semi Simple Switcher supplies, I use the
LM2975
> > on several designs. They work well. I usually get about 75%
efficiency.
> > I have done switching supplies with 90% efficient. The ZW
schematic on
> > the proto boards includes a sample 5v circuit. The data sheet has
tables
> > to pick the correct inductor size. Usually it is not that
critical, the
> > best value will maximize efficiency. The added cost is minimal;
usually
> > less than the labor an parts cost of mounting the reg and
installing a
> > heat sink (all parts can be automatically assembled.)
> >
> > Be sure to add extra caps before and after a switching supply. I
put as
> > large a cap as I can use before the switcher. Also have Polyfuse
PTC
> > fuse and a cap on the input. Beside being a fuse, it also adds an
> > addition R/C filter on the input.
> >
> > For 12V, you would probably want a LDO (Low Drop out) regulator.
Most
> > 78xx regulators need 2.5v more in than out to work well. The LDO
type
> > usually only need about 0.6v.
> >
> > ------
> > | Scott G. Henion| shenion@... |
> > | Consultant | Stone Mountain, GA |
> > | SHDesigns http://www.shdesigns.org |
> > ------
> > Rabbit libs: http://www.shdesigns.org/rabbit/
> > today's fortune
> > Hhhmmmmmmmm
> > waterbeds for cows
> > eleet
> > Culus: why would a cow need a waterbed?
> > cas: To be comfy warm
> >
> >
>
I looked at the photo you posted, it certainly looks promosing for
the size front except for one thing. Is that big electrolytic
necessary ?
I was thinking of trying Switching regulators after years of using
conventional regulators (ie chuck away any surplus volts as heat).
The capacitors are often a critical factors I have found in the long
life reliability of circuits. In particular I refer to electrolytic
Caps which have a short life expectancy than most other devices. To
this end I have always tried to avoid them where possible. Tantalums
were my original perfered option, although more expensive but now we
have much larger capacity ceramics available which are usually
cheaper and are'nt polarised.
Can your LM5007 design be made to work with just ceramic or tant
caps ? Can you post schematic of the regulator section ?
I have found I can buy the LM5007 from Farnell (NewarkInOne for those
in the US) as one-off for 1.57 (about $3).
--- In r..., "Dan Allen"
wrote:
>
> Yup, I'm with the majority here too.. National Simple Switcher is
the way to
> go.
>
> I've used the 2575's but am more fond of the 5007 lately for
smaller loads.
> They're not what I would consider "cheap" though, pushing $8 in
parts is a
> bit steep IMHO.
> I like it/them because they are truly 'simple', the 5007 especially
is teeny
> tiny,
> and quite flexible.. Not to mention readily available from your
favorite
> distributor.
> National even has a feature on their web site that will visually
show you a
> calculate expectancy of your heat dissipation (very cool).
>
> here's a pic of 1 of 2 of my on board power supplies using the 5007:
> http://ssminnow.maesoft.net/reflow/IMG_1021.jpg
> Smaller than a dime, and no heat sink needed...
> Dan...
>
> On 1/17/07, Scott Henion wrote:
> >
> > Jeff Shapiro wrote:
> > > Hi,
> > >
> > > I have an RCM2200 design that I suspect had problems in some
> > > occasions because the LM7805 (TO-220 with heat sink) was
getting very
> > > hot. Depending on the enclosure the customer used and the
ambient
> > > temperature I believe it led to some Rabbit module failures. It
> > > seemed like we got more failures during the warmer months
especially
> > > from places in the country that are warmer (i.e. Texas vs
Michigan).
> > > In a later revision, a "quick fix" we did was to add about 6
diodes
> > > in series between the 12V input and the 5V regulator to cut
some of
> > > the voltage prior to entering the regulator.
> > >
> > > The input to the board is about 12.5V and the output from the
> > > regulator is 5V at about 150mA driving the RCM2200 and some
external
> > > parts.
> > >
> > > I'm starting a new design using a RCM3700 and will need a
regulated
> > > 12V and 5V. I'd like to use an unregulated "wall wort" supply
rated
> > > at about 1A and be able to supply 12V at atleast 750mA and 5V at
> > > maybe 250mA. The "easy" (and least expensive) way would be to
just
> > > use a 7812 to get my 12V and then run that into a 7805 to get
5V but
> > > I'm worried about the heat issues I had with my other product.
> > >
> > > Any recommendations for something that won't add too much cost
to
> > > the total project?
> > >
> > (12-5)*0.15=1.15 watts
> > At .25a it is almost 1.9 watts
> >
> > If the input is unregulated, the heat will be even worse. Most 12
> > adapters start at almost 17v unloaded. Plan a larger heatsink than
> > necessary. I like to keep them hardly noticably warm in free air.
> >
> > Use one of the National Semi Simple Switcher supplies, I use the
LM2975
> > on several designs. They work well. I usually get about 75%
efficiency.
> > I have done switching supplies with 90% efficient. The ZW
schematic on
> > the proto boards includes a sample 5v circuit. The data sheet has
tables
> > to pick the correct inductor size. Usually it is not that
critical, the
> > best value will maximize efficiency. The added cost is minimal;
usually
> > less than the labor an parts cost of mounting the reg and
installing a
> > heat sink (all parts can be automatically assembled.)
> >
> > Be sure to add extra caps before and after a switching supply. I
put as
> > large a cap as I can use before the switcher. Also have Polyfuse
PTC
> > fuse and a cap on the input. Beside being a fuse, it also adds an
> > addition R/C filter on the input.
> >
> > For 12V, you would probably want a LDO (Low Drop out) regulator.
Most
> > 78xx regulators need 2.5v more in than out to work well. The LDO
type
> > usually only need about 0.6v.
> >
> > ------
> > | Scott G. Henion| shenion@... |
> > | Consultant | Stone Mountain, GA |
> > | SHDesigns http://www.shdesigns.org |
> > ------
> > Rabbit libs: http://www.shdesigns.org/rabbit/
> > today's fortune
> > Hhhmmmmmmmm
> > waterbeds for cows
> > eleet
> > Culus: why would a cow need a waterbed?
> > cas: To be comfy warm
> >
> >
>
Reply by ●January 18, 20072007-01-18
neil2452 wrote:
> Dear Dan
>
> I looked at the photo you posted, it certainly looks promosing for
> the size front except for one thing. Is that big electrolytic
> necessary ?
>
Lage eletrolytics help reduce ripple.
> I was thinking of trying Switching regulators after years of using
> conventional regulators (ie chuck away any surplus volts as heat).
>
I have been doing switching regs for many years. Really the only way to
go. I do prefer linear if the board has analog.
> The capacitors are often a critical factors I have found in the long
> life reliability of circuits. In particular I refer to electrolytic
> Caps which have a short life expectancy than most other devices. To
> this end I have always tried to avoid them where possible.
I have never had any fail. As long as you do not run them hard they are
fine. Spec'ing industrial temp caps help if they are in a warm area.
PC's have had them fail as they are running 100amp supplies with a huge
ripple current. The caps degrade due to the heat caused by continuous
A/C current. The better MB's are now using industrial temp caps.
> Tantalums
> were my original perfered option, although more expensive but now we
> have much larger capacity ceramics available which are usually
> cheaper and are'nt polarised.
>
Tantalums have a habit of shorting out and catching fire. They do not
like voltage spikes. They are only good after a regulator. Back when I
was repairing car electronics, it was probably one of the most common
failures I saw. Was hard to repair boards with holes burned in them ;)
Then again cars often have 150V spikes on the 12V. I could not believe
the designers that put them on the input side of the power supplies.
Using wall-wart supplies will not have so huge a spike, but they still
have spikes/noise on them. The PTC/cap filter I use works well.
> Can your LM5007 design be made to work with just ceramic or tant
> caps ? Can you post schematic of the regulator section ?
>
Probably. I use electrolytics on the input side. For low ripple, I use
electrolytic's paralleled with a ceramic.
> I have found I can buy the LM5007 from Farnell (NewarkInOne for those
> in the US) as one-off for 1.57 (about $3).
>
The LM5007 look good. Nice to see National came out with MOSFET
versions. That eliminates the inefficiency of the Darlington versions
they had.
------
| Scott G. Henion| s...@shdesigns.org |
| Consultant | Stone Mountain, GA |
| SHDesigns http://www.shdesigns.org |
------
Rabbit libs: http://www.shdesigns.org/rabbit/
today's fortune
I only know what I read in the papers.
-- Will Rogers
> Dear Dan
>
> I looked at the photo you posted, it certainly looks promosing for
> the size front except for one thing. Is that big electrolytic
> necessary ?
>
Lage eletrolytics help reduce ripple.
> I was thinking of trying Switching regulators after years of using
> conventional regulators (ie chuck away any surplus volts as heat).
>
I have been doing switching regs for many years. Really the only way to
go. I do prefer linear if the board has analog.
> The capacitors are often a critical factors I have found in the long
> life reliability of circuits. In particular I refer to electrolytic
> Caps which have a short life expectancy than most other devices. To
> this end I have always tried to avoid them where possible.
I have never had any fail. As long as you do not run them hard they are
fine. Spec'ing industrial temp caps help if they are in a warm area.
PC's have had them fail as they are running 100amp supplies with a huge
ripple current. The caps degrade due to the heat caused by continuous
A/C current. The better MB's are now using industrial temp caps.
> Tantalums
> were my original perfered option, although more expensive but now we
> have much larger capacity ceramics available which are usually
> cheaper and are'nt polarised.
>
Tantalums have a habit of shorting out and catching fire. They do not
like voltage spikes. They are only good after a regulator. Back when I
was repairing car electronics, it was probably one of the most common
failures I saw. Was hard to repair boards with holes burned in them ;)
Then again cars often have 150V spikes on the 12V. I could not believe
the designers that put them on the input side of the power supplies.
Using wall-wart supplies will not have so huge a spike, but they still
have spikes/noise on them. The PTC/cap filter I use works well.
> Can your LM5007 design be made to work with just ceramic or tant
> caps ? Can you post schematic of the regulator section ?
>
Probably. I use electrolytics on the input side. For low ripple, I use
electrolytic's paralleled with a ceramic.
> I have found I can buy the LM5007 from Farnell (NewarkInOne for those
> in the US) as one-off for 1.57 (about $3).
>
The LM5007 look good. Nice to see National came out with MOSFET
versions. That eliminates the inefficiency of the Darlington versions
they had.
------
| Scott G. Henion| s...@shdesigns.org |
| Consultant | Stone Mountain, GA |
| SHDesigns http://www.shdesigns.org |
------
Rabbit libs: http://www.shdesigns.org/rabbit/
today's fortune
I only know what I read in the papers.
-- Will Rogers
Reply by ●January 18, 20072007-01-18
Yea I have seen Tants go bang and emit acrid plooms of smoke, usually
because I got the polarity wrong, very dramatic. You get similar
results with elec caps too.
A lot of the designs I do are for permantantly running installations.
I recently repair some cicuits I installed 12 years ago by replacing
the Elec Caps. In hind sight I am surprised they lasted so well.
If you look at the data sheets for Aluminium electrolytic caps you
should see an operational life figure quoted in hours. Figures vary
from device to device but figures of 2000-8000 hours is quite
typical. If you look around you can find some in the 30,000h range
but these tend to be much bigger value caps (greater than 1000uf). I
don't think the elec cap fails on the 8001 hour but I guess the
performace starts to fall off. If its had an easy life it should
last a lot longer, ie not too hot and a low ripple current.
Its not to say you should never use elec caps but consider what the
consequneces may be if it fails after 8,000 hours. When your circuit
does'nt work anymore after 5yrs which part are going check first ?
Will it be the CPU with an MTBF of 100,000hours or that Elec Cap with
a stated life of 8000 hours ?
--- In r..., Scott Henion wrote:
>
> neil2452 wrote:
> > Dear Dan
> >
> > I looked at the photo you posted, it certainly looks promosing
for
> > the size front except for one thing. Is that big electrolytic
> > necessary ?
> >
> Lage eletrolytics help reduce ripple.
> > I was thinking of trying Switching regulators after years of
using
> > conventional regulators (ie chuck away any surplus volts as
heat).
> >
> I have been doing switching regs for many years. Really the only
way to
> go. I do prefer linear if the board has analog.
> > The capacitors are often a critical factors I have found in the
long
> > life reliability of circuits. In particular I refer to
electrolytic
> > Caps which have a short life expectancy than most other devices.
To
> > this end I have always tried to avoid them where possible.
> I have never had any fail. As long as you do not run them hard they
are
> fine. Spec'ing industrial temp caps help if they are in a warm area.
>
> PC's have had them fail as they are running 100amp supplies with a
huge
> ripple current. The caps degrade due to the heat caused by
continuous
> A/C current. The better MB's are now using industrial temp caps.
>
> > Tantalums
> > were my original perfered option, although more expensive but now
we
> > have much larger capacity ceramics available which are usually
> > cheaper and are'nt polarised.
> >
> Tantalums have a habit of shorting out and catching fire. They do
not
> like voltage spikes. They are only good after a regulator. Back
when I
> was repairing car electronics, it was probably one of the most
common
> failures I saw. Was hard to repair boards with holes burned in
them ;)
> Then again cars often have 150V spikes on the 12V. I could not
believe
> the designers that put them on the input side of the power supplies.
>
> Using wall-wart supplies will not have so huge a spike, but they
still
> have spikes/noise on them. The PTC/cap filter I use works well.
> > Can your LM5007 design be made to work with just ceramic or tant
> > caps ? Can you post schematic of the regulator section ?
> >
> Probably. I use electrolytics on the input side. For low ripple, I
use
> electrolytic's paralleled with a ceramic.
> > I have found I can buy the LM5007 from Farnell (NewarkInOne for
those
> > in the US) as one-off for 1.57 (about $3).
> >
> The LM5007 look good. Nice to see National came out with MOSFET
> versions. That eliminates the inefficiency of the Darlington
versions
> they had.
>
> ------
> | Scott G. Henion| shenion@... |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
> today's fortune
> I only know what I read in the papers.
> -- Will Rogers
>
because I got the polarity wrong, very dramatic. You get similar
results with elec caps too.
A lot of the designs I do are for permantantly running installations.
I recently repair some cicuits I installed 12 years ago by replacing
the Elec Caps. In hind sight I am surprised they lasted so well.
If you look at the data sheets for Aluminium electrolytic caps you
should see an operational life figure quoted in hours. Figures vary
from device to device but figures of 2000-8000 hours is quite
typical. If you look around you can find some in the 30,000h range
but these tend to be much bigger value caps (greater than 1000uf). I
don't think the elec cap fails on the 8001 hour but I guess the
performace starts to fall off. If its had an easy life it should
last a lot longer, ie not too hot and a low ripple current.
Its not to say you should never use elec caps but consider what the
consequneces may be if it fails after 8,000 hours. When your circuit
does'nt work anymore after 5yrs which part are going check first ?
Will it be the CPU with an MTBF of 100,000hours or that Elec Cap with
a stated life of 8000 hours ?
--- In r..., Scott Henion wrote:
>
> neil2452 wrote:
> > Dear Dan
> >
> > I looked at the photo you posted, it certainly looks promosing
for
> > the size front except for one thing. Is that big electrolytic
> > necessary ?
> >
> Lage eletrolytics help reduce ripple.
> > I was thinking of trying Switching regulators after years of
using
> > conventional regulators (ie chuck away any surplus volts as
heat).
> >
> I have been doing switching regs for many years. Really the only
way to
> go. I do prefer linear if the board has analog.
> > The capacitors are often a critical factors I have found in the
long
> > life reliability of circuits. In particular I refer to
electrolytic
> > Caps which have a short life expectancy than most other devices.
To
> > this end I have always tried to avoid them where possible.
> I have never had any fail. As long as you do not run them hard they
are
> fine. Spec'ing industrial temp caps help if they are in a warm area.
>
> PC's have had them fail as they are running 100amp supplies with a
huge
> ripple current. The caps degrade due to the heat caused by
continuous
> A/C current. The better MB's are now using industrial temp caps.
>
> > Tantalums
> > were my original perfered option, although more expensive but now
we
> > have much larger capacity ceramics available which are usually
> > cheaper and are'nt polarised.
> >
> Tantalums have a habit of shorting out and catching fire. They do
not
> like voltage spikes. They are only good after a regulator. Back
when I
> was repairing car electronics, it was probably one of the most
common
> failures I saw. Was hard to repair boards with holes burned in
them ;)
> Then again cars often have 150V spikes on the 12V. I could not
believe
> the designers that put them on the input side of the power supplies.
>
> Using wall-wart supplies will not have so huge a spike, but they
still
> have spikes/noise on them. The PTC/cap filter I use works well.
> > Can your LM5007 design be made to work with just ceramic or tant
> > caps ? Can you post schematic of the regulator section ?
> >
> Probably. I use electrolytics on the input side. For low ripple, I
use
> electrolytic's paralleled with a ceramic.
> > I have found I can buy the LM5007 from Farnell (NewarkInOne for
those
> > in the US) as one-off for 1.57 (about $3).
> >
> The LM5007 look good. Nice to see National came out with MOSFET
> versions. That eliminates the inefficiency of the Darlington
versions
> they had.
>
> ------
> | Scott G. Henion| shenion@... |
> | Consultant | Stone Mountain, GA |
> | SHDesigns http://www.shdesigns.org |
> ------
> Rabbit libs: http://www.shdesigns.org/rabbit/
> today's fortune
> I only know what I read in the papers.
> -- Will Rogers
>
