Searching \ for '16F84 Voltage Regulation' in subject line. ()
Help us get a faster server
FAQ page: techref.massmind.org/techref/power.htm?key=voltage
Search entire site for: '16F84 Voltage Regulation'.

Truncated match.
'16F84 Voltage Regulation'
1999\03\27@183450 by

I've been experimenting with 16F84 PIC for a short time.  As a first
project I'm going to make a flashing "anti-theft" LED for my car.  This
red LED will flash once every second to ward off would-be thieves.  The
car battery is +12V.  The 16F84 operates on +5V.  In the interest of
efficiency, what is the best way to regulate the car battery down to +5V
to power the PIC?  My initial thought was to use a LM7805 +5V
regulator.  Is this device efficient?  It seems that this device waists
a lot of energy through heat?

>I've been experimenting with 16F84 PIC for a short time.  As a first
>project I'm going to make a flashing "anti-theft" LED for my car.
This
>red LED will flash once every second to ward off would-be thieves.
The
>car battery is +12V.  The 16F84 operates on +5V.  In the interest of
>efficiency, what is the best way to regulate the car battery down to
+5V
>to power the PIC?  My initial thought was to use a LM7805 +5V
>regulator.  Is this device efficient?  It seems that this device
waists
>a lot of energy through heat?

The best answer would be a self-flashing led, which is available at
12V.
They draw amazingly little power.

Well, Dave i dont you have to worry to much about that, the 7805 Generates 4 ¡C/W (junction-case) and has a maximum output of 1A, the PIC can drive LED«s up to 25mA ( normal for a 5mm LED is aprox 10mA ) per i/o pin, and you only need one and the PIC itself only use something like 26uA. so i wouldnt worry much about the 7805 draining your battery. I say it would be just fine for this purpose. a simple resistor should do just fine to ( with some noise filtering of course ) since the PIC16F84 works in such a wide voltage area 2.0 - 6.0 Volts thus if you choose your resistor visely and made shure the battery is in "normal" condition, aprox 12.5 - 13.5 volts DC when engine is not running, then choose the resistor so you get something like 4.5 - 5 volts to the PIC. then the battery can go down to 9 V and it would work even if your car wont start.

But still i would recommend the 7805 voltage regulator. more reliable.

Greets
Robert

{Original Message removed}
After doing some home work I found that an average vehicle's "key-off" battery l
oad is around 10ma-20ma.  This includes the engine control's keep alive memory,
remote keyless system, alarm system etc...  To keep power consumtion to a minimu
m I'll put the PIC to sleep in between flashes.  Also, I'll PWM the LED.  Does a
nyone have any better ideas?  Thanks.
Kevin.

Robert Mellgren wrote:

> Well, Dave i dont you have to worry to much about that, the 7805 Generates 4 0
C/W (junction-case) and has a maximum output of 1A, the PIC can drive LED4s up t
o 25mA ( normal for a 5mm LED is aprox 10mA ) per i/o pin, and you only need one
and the PIC itself only use something like 26uA. so i wouldnt worry much about
the 7805 draining your battery. I say it would be just fine for this purpose. a
simple resistor should do just fine to ( with some noise filtering of course ) s
ince the PIC16F84 works in such a wide voltage area 2.0 - 6.0 Volts thus if you
choose your resistor visely and made shure the battery is in "normal" condition,
aprox 12.5 - 13.5 volts DC when engine is not running, then choose the resistor
so you get something like 4.5 - 5 volts to the PIC. then the battery can go dow
n to 9 V and it would work even if your car wont start.
>
> But still i would recommend the 7805 voltage regulator. more reliable.
>
> Greets
> Robert
>
> {Original Message removed}
I would recommend the LM2940 from National. (http://www.national.com). This part is
a low drop out (LDO) regulator that won't use up tons of power in the form
of heat like the 7805. They are slightly more expensive (7805 = \$.30 LM2940
= \$1.50 US). I use them for mobile robotic designs and they work great. They
are also designed specifically for automotive use, additionally they are
truly "idiot proof" if you install them backwards they simply won't work (No
harm to the device). The LM2941 is also cool as it can be adjusted for 3 to
30 volts and has an enable pin so you can selectively turn on different
parts of your supply (these are also LDO). I must concur with my fellow
PICLISTERS that the flashing LED can be a single part solution from a nearby
Radio Shack. It would be neat to have the led rate change every now and then
to create the illussion of a more advanced alarm system.

Bill Ruehl

enpassant wrote:

> I've been experimenting with 16F84 PIC for a short time.  As a first
> project I'm going to make a flashing "anti-theft" LED for my car.  This
> red LED will flash once every second to ward off would-be thieves.  The
> car battery is +12V.  The 16F84 operates on +5V.  In the interest of
> efficiency, what is the best way to regulate the car battery down to +5V
> to power the PIC?  My initial thought was to use a LM7805 +5V
> regulator.  Is this device efficient?  It seems that this device waists
> a lot of energy through heat?

William Ruehl wrote:

> I would recommend the LM2940 from National. (http://www.national.com).  This
> part is a low drop out (LDO) regulator that won't use up tons of power
> in the form of heat like the 7805.

I'm sorry, but you guys are getting entirely confused as to what LDO
regulators are.  There is as they say, "no such thing as a free lunch"
and translated to regulators, this means that LDO regulators draw *more*
current than an "ordinary" one (10mA nom vs. 3mA nom for a 78L12).

I refer to the "quiescent" or idle current.  At full load of 1A, the
LM2940 uses "only" 30mA (nominal) unless of course you run it at low
drop-out (less than 3V differential) when it uses *more*!

This current penalty may be fine for robotics regulating 5V from a 6V
battery, but for an automotive application, the quiescent current of the
LM2940 represents a possible doubling of the battery load even *before*
you switch on the LED.

You are in fact better off using a plain old 78L05 for the job.  If
you *really* want to save current, you could use a low power regulator.
These are *not* low-dropout.

Can I tell you a trick?  Put the LED in *series* with the 12V to the
regulator, and have the PIC just switch a 270 ohm load resistor.  The
LED will switch between the minimum 3mA of the regulator quiescent drain
which you cannot avoid anyway (except by using a true LP regulator) and
an 18mA "on" current.

Forget PWM unless you propose to use an inductor (and commutation
diode).  If you think awhile on the theory, you'll realise it does
nothing.
--
Cheers,
Paul B.

enpassant wrote:
>
> I've been experimenting with 16F84 PIC for a short time.  As a first
> project I'm going to make a flashing "anti-theft" LED for my car.

Have a look at my web site for a suitable power supply arrangement on
the circuit shown.

http://www.picnpoke.com/projects/ignition.html

I think the power saved by making the PIC sleep between flashes is
probably insignificant compared to the overall current drawn when the
LED is on. I also wouldn't worry too much about your PIC circuit
flattening the car battery. I think it would take a long time to do so.
If you look at a lot of late model cars, I'm sure there is a larger
'trickle' current drawn by the myriad of electronic systems anyway.

If you want a simpler idea, just use a flashing LED.

I'll bet that your brain is ticking over though, and is itching to write
that PIC program. Why not add in an alarm to your flashing LED as well.

Here is a project that may give you some ideas.

http://www.picnpoke.com/projects/alarm.html

--
Best regards

Tony

PicNPoke - Multimedia 16F84 Beginners PIC Tools.
*** FREE PCB OFFER ***

http://www.picnpoke.com
Email picnpokecdi.com.au

While we are discussing the 78 series regulators and there heat output. I
would like to know if it would be a bad thing to put a few 1N4007 diodes in
series with the input to a 78 regulator to drop the voltage it is regulating
from. Since the heat produced is related to the difference between the input
and the output isn't it. The reason I would be tempted to do this is simply
to make the thing run cooler and maybe save some heat sinking. I have heard
that 1N4007 and the like are slowing switching and maybe this would cause a
problem. Is this the done thing ?

Steve

{Original Message removed}
Steve Ridley wrote:

> While we are discussing the 78 series regulators and there heat
> output.

Well, that depends.  The original discussion was in regard to
minimising overall current drain to conserve power.  Heat dissipation
was the *last* aspect of concern, which is what the message you quoted
was pointing out, rather subtly.

> I would like to know if it would be a bad thing to put a few 1N4007
> diodes in series with the input to a 78 regulator to drop the voltage
> it is regulating from.  Since the heat produced is related to the
> difference between the input and the output isn't it.

That is as possibility, but the *big* problem is that by doing so, you
are adding to the drop-out voltage and drop-out will occur in an
automotive application during engine starting.  The argument for a LDO
regulator is in fact to make the whole system more reliable in the face
of engine starting with a weak battery etc.

> The reason I would be tempted to do this is simply to make the thing
> run cooler and maybe save some heat sinking.

"The thing"?  The whole assembly including the diodes will dissipate
exactly the same power.  Trying to be parsimonious about the heatsink
for the regulator (such as trying to do without one) is a *bad* approach
to design.  You are much better off to think positively - in the first
case, find the largest possible heatsink, usually the (cast) aluminium
case, and bolt the regulator directly to this.

Second ploy, fit a larger regulator if necessary, or the third
approach, cascade a PNP power transistor to the regulator, mounted on
the same heatsink.

> I have heard that 1N4007 and the like are slowing switching and maybe
> this would cause a problem. Is this the done thing ?

They are indeed slow switching.  Since this application does not
involve switching at all, let alone HF commutation, you can safely
forget that.
--
Cheers,
Paul B.

Hi,

On 29 Mar 99 at 0:53, Paul B. Webster VK2BZC wrote:
> William Ruehl wrote:
> > I would recommend the LM2940 from National. (http://www.national.com).  This
[snip]
>
>   I'm sorry, but you guys are getting entirely confused as to what LDO
> regulators are.  There is as they say, "no such thing as a free lunch"

In this case there is a free lunch, you can get low-power low-dropout
regulators such as LP2950 with  quiescent current in the microamps

[snip]
>   I refer to the "quiescent" or idle current.  At full load of 1A, the
> LM2940 uses "only" 30mA (nominal) unless of course you run it at low
> drop-out (less than 3V differential) when it uses *more*!

Yes but if I've used a 78L05 and a LM2940 both at 1A and both at
their minimal input voltage (that still gives regulated 5V) which one
would get hotter?  If we ran both with 12V input I don't think you'd
notice the extra 30mA compared to the 7 Watt that has to go
somewhere.

Whatever method you choose for voltage regulation don't forget to
provide protection against voltage transients and rfi.  Automotive
systems should be able to handle voltages of up to 200V.  There's a
good application note from Intel on this:
AP-125: Designing Microcontroller Systems for Electrically Noisy
Environments - Get it at their web site.
A LED flasher isn't a very critical system but you'd still like it to
keep on flashing past the first day...

Hope this helps,
Guy

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...