Searching \ for '[PIC] How should I power the 887?' in subject line. ()
Help us get a faster server
FAQ page: piclist.com/techref/power.htm?key=power
Search entire site for: 'How should I power the 887?'.

Exact match. Not showing close matches.
'[PIC] How should I power the 887?'
2008\06\23@131704 by

I'm using the PIC16F887 in a portable device that's going to run off
batteries.

The device will have 7 LED's on at a time, so the total power
consumption will be about 1.05 watts.
( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)

In my previous design, my power source was one of those 9 V square
batteries, which I fed into an LM7805 to give me 5 V, and I used this 5
V to power everything on the board (the uC, the logic chips, even the
LED's).

This time around, I'd like advice on:
1) What power source to use (e.g. a 9 V square battery)
2) How to get 5 volts from the power source (e.g. an LM7805)
3) Whether to power stuff like the LED's off the square battery or off
the LM7805.

I'm kind of new to this stuff in a way so I'd appreciate if you'd spell
things out instead of using abbreviations and initialisms. Also if you
have the vaguest idea I won't know what you're talking about then please
be extra elaborative!

Thanks!

On Mon, 23 Jun 2008 18:16:31 +0100, "Tomás Ó hÉilidhe" <toelavabit.com>
said:
{Quote hidden}

Do you require exactly 5 volts? In other words, are you using the A/D
converters on the PIC and need to use the supply as a reference? If the
answer is no(or maybe not, you can use an external reference or
calculate based on the internal 0.6 volt bandgap reference), then just
skip the regulator and run off three AA or AAA batteries. After all, the
ampacity of a 9 volt is tiny and you are running LEDs! AA batteries will
easily give you 5 times the run life.

Cheerful regards,

Bob

--
http://www.fastmail.fm - Does exactly what it says on the tin

The 7805 is terrible for battery usage. Its Iq (quiescent current) is
normally 2-3ma (read: very high). There's a comparison of some regulators
available at:

http://www.prc68.com/I/BTPS.shtml#FOV

But, lower q, higher \$, longer battery life.

On Mon, Jun 23, 2008 at 1:16 PM, Tomás Ó hÉilidhe <toelavabit.com> wrote:

{Quote hidden}

>
While 7805 voltage regulators are nice, they are also linear and "waste"
the extra energy as heat.    I don't know the voltage range on the 887,
but if it can run off of 4.5V instead, using 3 x LR44 batteries (AG13 same
battery) or 3 AAA batteries means you don't waste battery power in doing
the down conversion.  9V batteries are nice but also expensive.  If you
feel you have to use a 9V battery, then by all means drive your LEDs
directly off of the 9V battery and potentially in series too.  For
instance most LEDs run around a 2V voltage drop (except white that gets to
around 4V or more) so you can run 4 leds off of 9V and only need to drop
1V across a resistor that would set your current.  My main concern for you
would be what you want your battery life to be.  Sometimes running off a
3V coin cell with a charge pump to bring you up to 6V to run the micro
works best because the micro requires the least power of anything and any
conversion process just wastes energy.

Regards,

Bryon Sol

Tomás Ó hÉilidhe <toelavabit.com>
Sent by: piclist-bouncesmit.edu
06/23/2008 01:16 PM
"Microcontroller discussion list - Public." <piclistmit.edu>

To
"Microcontroller discussion list - Public." <piclistmit.edu>
cc

Subject
[PIC] How should I power the 887?

I'm using the PIC16F887 in a portable device that's going to run off
batteries.

The device will have 7 LED's on at a time, so the total power
consumption will be about 1.05 watts.
( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)

In my previous design, my power source was one of those 9 V square
batteries, which I fed into an LM7805 to give me 5 V, and I used this 5
V to power everything on the board (the uC, the logic chips, even the
LED's).

This time around, I'd like advice on:
1) What power source to use (e.g. a 9 V square battery)
2) How to get 5 volts from the power source (e.g. an LM7805)
3) Whether to power stuff like the LED's off the square battery or off
the LM7805.

I'm kind of new to this stuff in a way so I'd appreciate if you'd spell
things out instead of using abbreviations and initialisms. Also if you
have the vaguest idea I won't know what you're talking about then please
be extra elaborative!

Thanks!
Tomás Ó hÉilidhe wrote:
> I'm using the PIC16F887 in a portable device that's going to run off
> batteries.
>
> The device will have 7 LED's on at a time, so the total power
> consumption will be about 1.05 watts.
> ( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)

Are the LED's used to actualy lit something ?
Or are they just used as indicators ? 30 mA/LED
seems as way to much in that case. Use some low-power
LED and you'll be fine with a couple of mA's/LED.

And and the current for the PIC itself.

> In my previous design, my power source was one of those 9 V square
> batteries, which I fed into an LM7805 to give me 5 V, and I used this 5
> V to power everything on the board (the uC, the logic chips, even the
> LED's).

Use better regulator then the 7805, some LDO model.

Jan-Erik.

Bob Blick wrote:
> Do you require exactly 5 volts? In other words, are you using the A/D
> converters on the PIC and need to use the supply as a reference?

No I'm not using any A to D, I don't need an exact 5 volts.

>  If the
> answer is no(or maybe not, you can use an external reference or
> calculate based on the internal 0.6 volt bandgap reference),

Sorry you've lost me.

What do you mean by "external reference"?
What's the "internal .6 volt bandgap difference"?

> then just
> skip the regulator and run off three AA or AAA batteries. After all, the
> ampacity of a 9 volt is tiny and you are running LEDs! AA batteries will
> easily give you 5 times the run life.
>

Can AA batteries supply more current than those 9 V square batteries?
How do they compare when it comes to actual charge content (i.e. "amp
hours" or "watt hours").

Wikipedia tells me that AA batteries are 1.5 volts. Should I put 4 of
them in series to give me 5 V?

Jan-Erik Soderholm wrote:
> Tomás Ó hÉilidhe wrote:
>
>> I'm using the PIC16F887 in a portable device that's going to run off
>> batteries.
>>
>> The device will have 7 LED's on at a time, so the total power
>> consumption will be about 1.05 watts.
>> ( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)
>>
>
> Are the LED's used to actualy lit something ?
> Or are they just used as indicators ? 30 mA/LED
> seems as way to much in that case. Use some low-power
> LED and you'll be fine with a couple of mA's/LED.
>
> And and the current for the PIC itself.
>

They're bi-colour LED's which will be lit either green or red. I need
them nice and bright so they can be seen in daylight, so I'm thinking
I'd need *at least* 25 mA through them.

As for the current for the PIC, well I had thought this would be
negligible (nanowatt technology and all that).

On Mon, 23 Jun 2008 18:52:35 +0100, "Tomás Ó hÉilidhe" <toelavabit.com>
said:

> > answer is no(or maybe not, you can use an external reference or
> > calculate based on the internal 0.6 volt bandgap reference),
>
> Sorry you've lost me.
>
> What do you mean by "external reference"?
> What's the "internal .6 volt bandgap difference"?

If you're not using the A/D converter, it doesn't matter, ask again
later :)

> Can AA batteries supply more current than those 9 V square batteries?
> How do they compare when it comes to actual charge content (i.e. "amp
> hours" or "watt hours").

At least 5 times as much ampacity, I don't have the exact figures at
hand but the Duracell web site should have it for you.

> Wikipedia tells me that AA batteries are 1.5 volts. Should I put 4 of
> them in series to give me 5 V?

3 * 1.5 = 4.5 volts. Adequate for one-off "hobby" projects at 20MHz. It
will be close to 5 volts with fresh batteries and you'll have to see how
low you can go. A good reason to read the page on the brownout detect in
the PIC(leave it off for now).

You'll be able to do without a power switch if you learn about sleep
mode.

Cheers,
Bob

--
http://www.fastmail.fm - The way an email service should be

Bob Blick wrote:
> 3 * 1.5 = 4.5 volts. Adequate for one-off "hobby" projects at 20MHz.

I know it's a hobby project but I'm doing it with the aspiration of
actually making a marketable product, which is why I'm sticking to the
limits of spec sheets and so forth.

If I need 4 AA batteries, I'll use 4 AA batteries.

> It
> will be close to 5 volts with fresh batteries and you'll have to see how
> low you can go. A good reason to read the page on the brownout detect in
> the PIC(leave it off for now).
>

What does "brownout detect" do? Does it shut down the uC if the supply
voltage drops below a certain threshold?

>
> They're bi-colour LED's which will be lit either green or red. I need
> them nice and bright so they can be seen in daylight, so I'm thinking
> I'd need *at least* 25 mA through them.

"Modern" high-efficiency LEDs don't require nearly that much current to be bright.  However, if you find that 8 to 10mA steady current is not sufficient you can pulse higher currents at a low duty cycle.  I can't remember if your schematic had current limiting resistors or if the part number you had listed for drivers had internal current limiting.

Short answer, I doubt you will need or want 25mA.  And if you did, run a current budget and see how quickly it adds up during a game.  9V battery will be dead before the game is over.

>
> As for the current for the PIC, well I had thought this would be
> negligible (nanowatt technology and all that).

Only if you take advantage of the features (sleep, perpherial turn off, WDT wakeup, slow clock, etc).  At the back of the PIC data sheet are the AC and DC operating characteristics.  Look through the tables and you will find a series of charts giving worst case Idd for combinations of Vdd, oscillator mode and oscillator frequency.  I think something like 4.5mA for HS, 20MHz and 4.5V running full out.  While this is better than previous parts, it isn't exactly20mW is nowhere near the "nanowatt" range.  The "nanowatt" business refers to draw during sleep and perpherials turned off.

Rob

> Bob Blick wrote:
> > 3 * 1.5 = 4.5 volts. Adequate for one-off "hobby" projects at 20MHz.
>
> I know it's a hobby project but I'm doing it with the aspiration of
> actually making a marketable product, which is why I'm sticking to the
> limits of spec sheets and so forth.
>
> If I need 4 AA batteries, I'll use 4 AA batteries.
>
> > It
> > will be close to 5 volts with fresh batteries and you'll have to see how
> > low you can go. A good reason to read the page on the brownout detect in
> > the PIC(leave it off for now).
> >
>
> What does "brownout detect" do? Does it shut down the uC if the supply
> voltage drops below a certain threshold?

RTFM. ;)  Seriously, read section 14.  It does a body good.

Short and flip answer is the BOR will put the PIC into reset and hold it there until the Vdd level raises back above a certain trip point, stays there and also conforms to some risetime stuff.  Some interaction with the POR too.  And I think, please consult the data sheet to confirm, that you can read a status register after reset and determine the reason for the last reset.  Possibly useful in error recovery situations.

And one more thing, again, a data sheet check and schematic (which I don't have to reference) is if the comparator mode of the 16F887 could be used as a voltage monitor so that you can get an early warning to pending battery failure.  Save game and all that jazz.

Rob
Tomás Ó hÉilidhe wrote:
> The device will have 7 LED's on at a time, so the total power
> consumption will be about 1.05 watts.
> ( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)

Are these for illumination or are they indicators?  If just indicators, then
do they really need to be run that bright?  Most "normal" LEDs are only
rated at 20mA in the first place, and running them at 5mA will make them
quite visibile in normal office light.  A few extra cents on higher
efficiency LEDs will likely pay back on just the first set of batteries.

> In my previous design, my power source was one of those 9 V square
> batteries, which I fed into an LM7805 to give me 5 V, and I used this
> 5 V to power everything on the board (the uC, the logic chips, even
> the LED's).

That couldn't have lasted long.

> This time around, I'd like advice on:
> 1) What power source to use (e.g. a 9 V square battery)

My first reaction is to use 2 AA cells and run the LEDs from that directly.
I don't know what the micro needs to do, but plenty of PICs can run down to
2V if you drop the clock speed a bit.  If your computing requirements can
handle the PIC running at the lower clock speed, then you can run everything
directly off the two AA batteries.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
Tomás Ó hÉilidhe wrote:
> Can AA batteries supply more current than those 9 V square batteries?

Yes.

> How do they compare when it comes to actual charge content (i.e. "amp
> hours" or "watt hours").

Get a datasheet of each and do your homework.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
Tomás Ó hÉilidhe wrote:
> If I need 4 AA batteries, I'll use 4 AA batteries.

4 x 1.5V = 6V, which is more than the maximum 5.5V most PICs are rated to
run at.  If you use 4 batteries, you will have to put something between them
and the PIC.

> What does "brownout detect" do?

This is described in the "Brown-out Reset" section of the "special features
of the CPU" datasheet chapter.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

Olin Lathrop wrote:
>> How do they compare when it comes to actual charge content (i.e. "amp
>> hours" or "watt hours").
>>
>
> Get a datasheet of each and do your homework.
>

Sorry I think you've got the wrong mailing list, you want the Dickhead List.

-----Original Message-----
From: piclist-bouncesmit.edu [piclist-bouncesmit.edu] On Behalf Of
Tomás Ó hÉilidhe
Sent: Monday, June 23, 2008 4:10 PM
To: Microcontroller discussion list - Public.
Subject: Re: [PIC] How should I power the 887?

>> How do they compare when it comes to actual charge content (i.e. "amp
>> hours" or "watt hours").
>>
>
> Olin Lathrop wrote:
> Get a datasheet of each and do your homework.

>>Sorry I think you've got the wrong mailing list, you want the Dickhead
List.

Here we go again...
Little Tomas was told to do some work on his own,
AFTER he was handed so MANY pieces of the puzzle,
and he gets resentful...

This will OBVIOUSLY be Olin's fault...
Bring on the same old stuff AGAIN!
FLAME ON

*
|  __O    Thomas C. Sefranek  WA1RHPARRL.NET
(*)/ (*)  Bicycle mobile on 145.41MHz PL74.4

ARRL Instructor, Technical Specialist, VE Contact.
http://www.harvardrepeater.org

> Sorry I think you've got the wrong mailing list, you want the Dickhead List.

Oh boy, are you too lazy to google for "battery capacity" or even look
at the print on a battery? [text here deleted by self-sensorship]

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

Thomas C. Sefranek wrote:
{Quote hidden}

So, boys, it's time of my english course: lesson 1 - offences to an
ungrateful idiot!

Nic

{Quote hidden}

> Sorry I think you've got the wrong mailing list, you want the

Hey, that's enough of that

Battery datasheets are easy enough to find. You need only to
compare mAh to get started

Just Google for battery capacity (the first two words I could
think of related to batteries and their capacity)

Tomás Ó hÉilidhe wrote:

> They're bi-colour LED's which will be lit either green or red. I need
> them nice and bright so they can be seen in daylight, so I'm thinking
> I'd need *at least* 25 mA through them.

>
> As for the current for the PIC, well I had thought this would be
> negligible (nanowatt technology and all that).
>

You know where the actual values are, right ?
current "thinking" will.

Jan-Erik.
> If I need 4 AA batteries, I'll use 4 AA batteries

Fresh alkalines, what most consumers would use in a commercial
product, will be 1.65V. x 4 = 6.6V, well over spec. Better to use
3. And make the circuit run properly when skinflints like me use

On Monday, June 23, 2008 2:16 PM [GMT-3=CET],
Tomás Ó hÉilidhe  wrote:

> I'm using the PIC16F887 in a portable device that's going to run off
> batteries.
>
> The device will have 7 LED's on at a time, so the total power
> consumption will be about 1.05 watts.
> ( 5 volts * 30 milliamperes * 7 = 1 050 milliwatts)

> Thanks!

Tomas.
I count 6 not 7. Why 7?
You've said your firmware lit one colum at a time.
The colums have 6 LEDs.

Some ideas to think or try:
.- Choose the minimum frequency at which your soft can run.
.- Try to Source the hole circuit with 1.2 * 4 NiMH 2850mAh AA
.- Choose the minimum duty cycle... (it's a feature :P)
.- .... continue asking and calm down ;)

Thank you for sharing your project, which IS a very solid sign of good will.

Dennis.

Tomás Ó hÉilidhe wrote:
>>> How do they compare when it comes to actual charge content (i.e.
>>> "amp hours" or "watt hours").
>>
>> Get a datasheet of each and do your homework.
>
> Sorry I think you've got the wrong mailing list, you want the

You're way out of line, and may get booted of the list if you continue like
this.  I'm not generally in favor of that, but in this case I couldn't fault
James for doing it.

I ignored your questions for a couple of days after your last ingrateful
tirade.  When I saw a new thread from you this morning, I thought it was
time to put that behind us and answer your questions.  You snipped and
apparently ignored the other answers I wrote.

You still don't get it.  It's *your* project.  People here are quite willing
to help, but *you* have to do the work.  And yes, that often means looking
things up in datasheets.  Since you can't handle that, you need to find
another major fast.  It's been 28 years since I got my masters degree in
electrical engineering, and hardly a day goes by I don't look up something
in a datasheet or manual.  When I'm in the middle of a electrical design or
writing firmware, I often have multiple datasheets open on my desk and on my
screen.

You're going to be in for a rude surprise if you ever manage to get your EE
degree.  You'll need to be looking up stuff even more than when in school.
Things are handed to you much more in school than in the real world.  Bail
out now before you waste any more tuition.  With your laziness and attitude,
you'll never be even a mediocre engineer.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

Dennis Crawley wrote:
> I count 6 not 7. Why 7?
> You've said your firmware lit one colum at a time.
> The colums have 6 LEDs.
>

There's 6 LED's per column and there's 7 columns, so you're right in
saying that the figure is 6.

However there are 2 extra columns for the seven segment displays, and
there'll be 7 LED's on at a time for those 2 columns. Also the very
first column has an extra LED to indicate whose go it is.

If I wanted to be very specific I'd get the average current flow as:

x = current per LED = 30 mA

6(6x) + 3(7x) = 36x + 21x = 57x

Average current = 57x / 9 = 190 mA = 950 mW (at 5 volts)

I might be able to halve that figure from 950 mW to 475 mW if I play
around with different pulse widths and duty cycles.

Jan-Erik Soderholm wrote:
>> They're bi-colour LED's which will be lit either green or red. I need
>> them nice and bright so they can be seen in daylight, so I'm thinking
>> I'd need *at least* 25 mA through them.
>>
>
>

In my previous project I took a normal everyday bi-colour LED and
powered it from a PIC pin without using a current-limiting resistor. It
*still* wasn't bright enough to be seen properly in daylight. It's for
this reason that I'm thinking -- or , as you call it, "thinking" -- that
I'll need a decent current flow to get the performance I want even if I
choose some sort of "high luminosity" LED.

> Try to Source the hole circuit with 1.2 * 4 NiMH 2850mAh AA

Rechargeables is a good idea for the consumer, but that presents
challenges for the designer - charging, self-discharge (damn things
are always flat when you get the game out of the cupboard), cycle
lifetime, and especially the often dubious quality of cheap/affordable
OEM cells

> Average current = 57x / 9 = 190 mA = 950 mW (at 5 volts)

And from that you look up battery mAh ratings to work out how
long they'll last. A PP9 9V, perhaps an hour or two. AA alkalines,
perhaps 8 hours, NiCd a bit more but they're rechargeable. You
need also to look at discharge curves for the various common types.
Some are good for high drains in short times, others for consistent
moderate drains

I'd add a voltage input socket and pre-regulator so that a wide range
of household plug-packs, for example a phone charger, could be used
(so you don't have to supply one)

Olin wrote:
> Tomás Ó hÉilidhe wrote:
>
>> This time around, I'd like advice on:
>> 1) What power source to use (e.g. a 9 V square battery)
>
> My first reaction is to use 2 AA cells and run the LEDs from that
> directly.
> I don't know what the micro needs to do, but plenty of PICs can run down
> to
> 2V if you drop the clock speed a bit.  If your computing requirements can
> handle the PIC running at the lower clock speed, then you can run
> everything
> directly off the two AA batteries.

Or indeed, three AA batteries.  Nominal 4.5V makes it possible to power
blue or white LEDs, without a boost converter.  And there's more total
battery capacity, and you can run them down more.  The only problem is
that 3-way battery holders are hard to come by (at least in Australia).  I
use one in one of the kits I'm selling, but I ordered in a batch of
holders from a supplier in Taiwan - not something you'd do for a 1-off!
On the other hand, a 2-way and a 1-way in series (glued together if you
want) is quite acceptable.

I've also used 4 x AA, with a 1N400x diode to drop 0.6V - but that assumes
adequate current (put the PIC to sleep and turn off the LEDs, and you'll
be over the voltage spec - which is another reason for going the 3 x AA =
4.5V route).

David Meiklejohn
http://www.gooligum.com.au
> I'll need a decent current flow to get the performance I want even
> if I choose some sort of "high luminosity" LED.

The point about high luminosity is that you don't need high current
to get it. I've got ultrabright greens here that are uncomfortably bright
at only 5mA. Even white LEDs in torches and headlamps are using
probably only 15-20mA and you know how bright those are

Manufacturers and suppliers have selection guides

Digikey

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=524729;keywords=led

Kingbright

http://www.kingbright.com/index1.php?lang=English

Radiospares, Mouser, Farnell etc all have LED tables

>From my, admitedly limited, experience with bi-colour LEDs,
IMHO I think you have chosen the wrong part and would be
better off replacing it with two discrete LEDs

Compare these "high intensity" (300mcd @ 30mA !!!) bi-colours

http://www.ledtronics.com/ds/L200X-3C/default.asp

for price, intensity, consumption vs general purpose LEDs that are
10x, or more, the mcd at 1/10th, or better, the price

Tomás Ó hÉilidhe wrote:
>
> Olin Lathrop wrote:
>>> How do they compare when it comes to actual charge content (i.e. "amp
>>> hours" or "watt hours").
>>>
>>
>> Get a datasheet of each and do your homework.
>>
>
> Sorry I think you've got the wrong mailing list, you want the Dickhead
> List.

So Tomás, how on Earth do you imagine any of us would be able to answer
you?  Because, in the past, we've looked it up ourselves.

When I consult a data sheet, I often print it and put it in a binder for
future reference (I'm old-fashioned that way), and if you looked in my
binders, you'd find that some of the data sheets are for batteries.  Why?
Because I like to build little battery-powered devices, and I like to be
able to choose the right battery solution for the job (just like you're
trying to do with these questions), and I want to know how various
batteries perform under various load conditions.  So to learn that, I go
look it up.

It's not difficult, and there are nice graphs (discharge curves) and
tables that will give you a much better idea than someone could easily
convey to you here.

David Meiklejohn

Bob Blick wrote:
>
> You'll be able to do without a power switch if you learn about sleep
> mode.

As long as the PIC is the only thing in the circuit drawing power.  Yes,
an obvious point, and probably not relevant for this project, but
sometimes people forget when they have say a resistor divider for a
comparator reference or something that those resistors will drain power
(slowly, but it adds up over time) while the PIC is "off".  And you have
to learn to do things like tie off unused PIC inputs.  It's always a good
idea to measure the total current draw when in sleep mode, to confirm that
it's as low as it should be, or else you'll find your batteries
unexpectedly running down...

David Meiklejohn
http://www.gooligum.com.au

Jinx wrote:
> Compare these "high intensity" (300mcd @ 30mA !!!) bi-colours
>
> http://www.ledtronics.com/ds/L200X-3C/default.asp
>
> for price, intensity, consumption vs general purpose LEDs that are
> 10x, or more, the mcd at 1/10th, or better, the price
Thanks a lot for your help Jinx, these LED's look inviting. If they turn
out to be too bright I can also always use higher value resistors.

I have a matrix of LED's, 6 rows and 7 columns. The LED's are tightly
packed, so each of them will have to have a high luminous intensity so
that the user can easily see what LED's are lit and which ones aren't.

Have you any idea what minimum value of luminous intensity I should look
for for my bi-colour LED's, given that I want to be able to play the
game in normal daylight.

(My previous project board looked great in a poorly lit room but it
looked crap if the room was in any way lit)

David Meiklejohn wrote:
> So Tomás, how on Earth do you imagine any of us would be able to answer
> you?  Because, in the past, we've looked it up ourselves.

The reasons I asked about the charge capacities rather than looking at
datasheets was:
* There's more than one brand of 9 V square battery
* I don't know what brands are popular
* Different brands will have different charge capacities

Also, even if I'd looked at *all* the datasheets, it still wouldn't be
as good as having someone be kind enough to say:
"Well I've used Battery 1 in loads of different products and I've
always gotten poor run time. However I've found Battery 2 to be great, I
always get really long run times, even across the board with different
brands of battery".

Tomás Ó hÉilidhe wrote:
>
> They're bi-colour LED's which will be lit either green or red. I need
> them nice and bright so they can be seen in daylight, so I'm thinking
> I'd need *at least* 25 mA through them.

You can easily answer that by testing it.  All you need is a LED, a DMM, a
power supply (can be a battery if you want to take it outside to test in
daylight), a breadboard and a pot.  Twiddle the knob, and see what current
gives you acceptable brightness.

> As for the current for the PIC, well I had thought this would be
> negligible (nanowatt technology and all that).

Have you heard the saying about assumptions?  Better than assuming would
be to glance at the electrical characteristics section of the data
sheet...

David Meiklejohn

Jinx wrote:
>> Average current = 57x / 9 = 190 mA = 950 mW (at 5 volts)
>>
>
> And from that you look up battery mAh ratings to work out how
> long they'll last.
looking at datasheets was:
* There's more than one brand of 9 V square battery
* I don't know what brands are popular
* Different brands will have different charge capacities

Also, even if I'd looked at *all* the datasheets, it still wouldn't be
as good as having someone be kind enough to say:
"Well I've used Battery 1 in loads of different products and I've
always gotten poor run time. However I've found Battery 2 to be great, I
always get really long run times, even across the board with different
brands of battery".

And then there are these:

*/BATTERIES AMERICA/  / Rechargeable 9-Volt Nickel Metal Hydride
Batteries. // Provides proper 9-volt output . (Note: many 9V
rechargeable batteries are made with only 6 cells inside.  Our AP-LX8 is
made with 7 cells inside, to provide a higher, adequate voltage level).
/*

*/See, near the bottom of the page:

/*

http://www.batteriesamerica.com/newpage3.htm

Tomás Ó hÉilidhe wrote:
{Quote hidden}

>

David Meiklejohn wrote:
> Tomás Ó hÉilidhe wrote:
>
>> They're bi-colour LED's which will be lit either green or red. I need
>> them nice and bright so they can be seen in daylight, so I'm thinking
>> I'd need *at least* 25 mA through them.
>>
>
> You can easily answer that by testing it.  All you need is a LED, a DMM, a
> power supply (can be a battery if you want to take it outside to test in
> daylight), a breadboard and a pot.  Twiddle the knob, and see what current
> gives you acceptable brightness.

Using "normal" bi-colour LED's, I took out the current-limiting
resistors altogether and they still weren't bright enough. I'll have to
get my hands on proper LED's for the job before I can do testing.
> If they turn out to be too bright I can also always use higher value
> resistors

Given the concern about batteries, a resistor would not be the best
solution as it will waste power as heat. The better option would be
to use PWM for controlling brightness. If that makes the software
too involved, I'd suggest this -

On the i/p of each driver is an AND gate. AND (4081 for example)
ICs are ridiculously cheap. One i/p of the AND gate goes to the PIC,
the other i/p to a separate oscillator which has a settable duty cycle.
This could even be a pot for the user to adjust brightness or, if you're
ambient conditions

Jinx wrote:
>> If they turn out to be too bright I can also always use higher value
>> resistors
>>
>
> Given the concern about batteries, a resistor would not be the best
> solution as it will waste power as heat. The better option would be
> to use PWM for controlling brightness. If that makes the software
> too involved, I'd suggest this -
>

Software Involvedness is not a problem :-D

10 times out of 10, I'll opt for more complicated software instead of
spending more on hardware.

I have 9 columns to flash, and I suppose I'm a bit paranoid about being
able to flash them fast enough while at the same time having them appear
bright enough.

Should I be worried about 9 columns... or is it very workable? At what
amount of columns would it become difficult?

> Also, even if I'd looked at *all* the datasheets

It does not hurt at all to investigate but .....

My practical instinct is to look at what the end user is going to do.
They are going to buy cheap-as-chips alkalines. They will not be
interested in locating expensive and/or exotic batteries to get 10%
more playing time. I would therefore base my design on supermarket
AA alkalines

Welcome to Engineering For The Real World !!!  ;-)

Tomás Ó hÉilidhe wrote:
> Have you any idea what minimum value of luminous intensity I should look
> for for my bi-colour LED's, given that I want to be able to play the
> game in normal daylight.

I wonder would these do the trick?

http://www.farnell.com/datasheets/72810.pdf

They can give 2.5 millicandelas. Do you reckon that would be enough for
daylight?

> My practical instinct is to look at what the end user is going to do.
> They are going to buy cheap-as-chips alkalines. They will not be
> interested in locating expensive and/or exotic batteries to get 10%
> more playing time. I would therefore base my design on supermarket
> AA alkalines
>
But also make sure your device can handle the lower output voltage of AA
nicd and nimh batteries.
> > http://www.ledtronics.com/ds/L200X-3C/default.asp

> these LED's look inviting

I have a couple with similar specs and wouldn't call them spectacular.
The last product I used them for was as over-rev indicators. It helped
to have them against a black background and shielded from ambient
lighting by a raised ring on the fascia. You also have to consider view
angle too if you're going to do that, probably not a major problem in

> 10 times out of 10, I'll opt for more complicated software instead of
> spending more on hardware.

Personally, I'd cough up the extra 10c for a quick guaranteed result
instead of many many hours of coding and debugging

Tomás Ó hÉilidhe wrote:
>
> The reasons I asked *here* about the charge capacities rather than
> looking at datasheets was:
>     * There's more than one brand of 9 V square battery
>     * I don't know what brands are popular
>     * Different brands will have different charge capacities

Despite manufacturer's claims, brand differences aren't really important.
What counts is the battery type, such as alkaline, carbon, nicad, lithium,
whatever.  And packaging - 9V, AA, AAA, etc.

> Also, even if I'd looked at *all* the datasheets, it still wouldn't be
> as good as having someone be kind enough to say:
>     "Well I've used Battery 1 in loads of different products and I've
> always gotten poor run time. However I've found Battery 2 to be great, I
> always get really long run times, even across the board with different
> brands of battery".

Yes, it's perfectly valid to ask about people's experiences (that's what a
list like this is all about, really), but it didn't sound like you were
asking that.  And there was certainly no reason to start calling Olin
names.

You won't get a simple answer like "this battery is always good, that one
is bad" because "run time" isn't just about a mAh rating.  Some batteries
are more suited to a low constant current draw, others are better at
supplying a higher current in short bursts.  Some maintain a reasonably
"flat" voltage as they discharge, but then collapse quickly.  Others run
down at a more even rate.  And then there's energy density, shelf life,
and more.

But if you want a simple answer - use alkaline AAs.  But don't trust me on
that - look at a data sheet (any manufacturer), know the current draw of
your circuit and the minimum voltage it needs, and decide if they'll give

David Meiklejohn
> Using "normal" bi-colour LED's, I took out the current-limiting
> resistors altogether and they still weren't bright enough. I'll have to
> get my hands on proper LED's for the job before I can do testing.

Whilst it's desirable to give the consumer the best product you can,
I think they have to meet you half-way. It would not be reasonable
to have to make an LED be seen in bright direct sunlight for a unit
that doesn't have to be used in bright direct sunlight. You can have
some compromises that are acceptable

On Monday, June 23, 2008 7:59 PM [GMT-3=CET],
Jinx  wrote:

>> Try to Source the hole circuit with 1.2 * 4 NiMH 2850mAh AA
>
> Rechargeables is a good idea for the consumer, but that presents
> challenges for the designer - charging, self-discharge (damn things
> are always flat when you get the game out of the cupboard), cycle
> lifetime, and especially the often dubious quality of cheap/affordable
> OEM cells

You're right.
The case is that I bought a good german digital(*) camera (Maginon) with a
pack of 4 AA 2850mAh (ANSMANN). They came with an intelligent charger. I
ab-used them a lot and they still running. The charger is the photocam III.

In which meridian are you now?....
well I'm going to the 2m band now, bye.

Dennis.
ps: have you ever heard a german pronouncing digital? In my job the service
support is german from Alsbach,... digital, with, video (phideo),... an
more.

> The case is that I bought a good german digital(*) camera (Maginon)
> with a pack of 4 AA 2850mAh (ANSMANN). They came with an
> intelligent charger. I ab-used them a lot and they still running. The
> charger is the photocam III

With a high-end product you'd expect high-end performance. On the
other hand, I've got a few cordless tool battery packs that could do
with a good zapping of the individual cells. Their chargers are not what
you'd call intelligent

But you get what you pay for I guess, and that's a consideration for
Tomas or any manufacturer. What expectations does the customer
have at the price ?

> In which meridian are you now?....

Time zone ? GMT +11 (funny, Wellington has been replaced
with Auckland. Auckland is not the NZ capital. Yet)

http://www.timeanddate.com/worldclock/

2008/6/24 Jinx <joecolquittclear.net.nz>:

>>>>snipped
>
> Time zone ? GMT +11 (funny, Wellington has been replaced
> with Auckland. Auckland is not the NZ capital. Yet)
>

Is that something else that Helen is planning to change while she
hangs on with her fingertips?

RP
> Is that something else that Helen is planning to change while she
> hangs on with her fingertips?

Hehe. I did catch a snippet of something in the news the other day
about changing the capital. I'd be surprised if Helen didn't go for
Mt Albert as a tax-free principality. Might be worth checking her
previous overseas visits for a trip to Monaco

I really am curious as to why timeanddate, which I've used as a
reference for many years, has recently dropped Wellington in
favour of Auckland

But I see no PICs in sight so I'll keep any further musings to myself

Tomás Ó hÉilidhe wrote:

> The reasons I asked about the charge capacities rather than looking at
> datasheets was:
>     * There's more than one brand of 9 V square battery

Then you need to look up datasheets of more than one brand.

>     * I don't know what brands are popular

Neither does probably anyone here, at least not for your country (unless
you spell that out). Look in your local supermarket; that should give you a
hint.

>     * Different brands will have different charge capacities

Not only different brands, but different models of the same brand. Again,
you need to look up the datasheets.

That's all what Olin was telling you: the information you need for this is
out there, you just need to get it. Nobody here can do as much for you as
you reading through those datasheets can do. (And it's not that many that
you need to read, either... some twenty or so should give you a good
overview, plus a few app notes or general papers from battery
manufacturers.)

Also, not surprisingly, batteries have been discussed over and over on this
list. There is also a /lot/ of information in the archives.

Gerhard
Is anybody used MCP1702 already? Looks impressive to me, however, it could
be quite the limitation here if 7*30mA LED will be driven, as it provides up
to 250mA only - might be good though.

Tamas

On Tue, Jun 24, 2008 at 1:29 PM, Gerhard Fiedler <listsconnectionbrazil.com>
wrote:

{Quote hidden}

>
Tamas Rudnai wrote:
> Is anybody used MCP1702 already? Looks impressive to me, however, it could
> be quite the limitation here if 7*30mA LED will be driven, as it provides up
> to 250mA only - might be good though.

Yes, I've used it in a couple of devices now, one of them
battery-powered.  (<http://timothyweber.org/remote>) That's been in
daily use for over a year without a battery change.  The low quiescent
current is very nice.
--
Timothy J. Weber
http://timothyweber.org
Clara's remote, nice :-)

Tamas

On Tue, Jun 24, 2008 at 7:34 PM, Timothy Weber <twtimothyweber.org> wrote:

{Quote hidden}

> -
Thanks!  Very satisfying.

Tamas Rudnai wrote:
> Clara's remote, nice :-)
>
> Tamas
>
>
> On Tue, Jun 24, 2008 at 7:34 PM, Timothy Weber <twtimothyweber.org> wrote:
>
>> Tamas Rudnai wrote:
>>> Is anybody used MCP1702 already? Looks impressive to me, however, it
>> could
>>> be quite the limitation here if 7*30mA LED will be driven, as it provides
>> up
>>> to 250mA only - might be good though.
>> Yes, I've used it in a couple of devices now, one of them
>> battery-powered.  (<http://timothyweber.org/remote>) That's been in
>> daily use for over a year without a battery change.  The low quiescent
>> current is very nice.
>> --
>> Timothy J. Weber
>> http://timothyweber.org
>> --

Are you sure you need to run the LED's at 30mA?  That sounds like a typical
"absolute maximum" value in the datasheet.  You should consider running them
at 20mA, and for most LED's (that I use), going from 20mA down to 10mA makes
very little difference in visual brightness.  But it all depends on your
application.

TomÃ¡s Ã“ hÃ‰ilidhe-2 wrote:
{Quote hidden}

> -
>> Get a datasheet of each and do your homework.
>>
>>>Sorry I think you've got the wrong mailing list,
>
>Here we go again...

Unfortunately ...

>Little Tomas was told to do some work on his own,
>AFTER he was handed so MANY pieces of the puzzle,
>and he gets resentful...
>
>This will OBVIOUSLY be Olin's fault...
>Bring on the same old stuff AGAIN!
>FLAME ON

However a pointer to a battery datasheet or website would have been helpful
...

>Software Involvedness is not a problem :-D
...
>Should I be worried about 9 columns... or is it very workable?
>At what amount of columns would it become difficult?

Well, seeing you have a prototype working, this seems like a good time to be
experimenting with the software to ascertain some of these things.

> However a pointer to a battery datasheet or website would have been helpful

It would, but finding one takes time, and about the same amount of time
regardless of who does does the finding. So why can't the OP do that
himself?

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> a pointer to a battery datasheet or website would have been helpful

Actually, common sense suffices.  For a given battery chemistry,
capacity is approximately proportional to cell volume (or mass.)  For
a linear regulator, extra voltage doesn't buy you extra capacity.
Now compare the size of a 9V battery with the size of 3 AA cells.
Actually, compare HALF the size of the 9V battery with the 3AA cells,
since half the 9V is "wasted" in the regulator.

http://data.energizer.com/ seems to have pretty good data, including
capacity in mAH at different discharge rates (Duracell has curves
that are more difficult to interpret.)

A 9V alkaline seems to have about 250mAH at 300mA discharge, while an
AA has about 1500mAH (6x more).  Note that you still need 2x of the
AA to reach a reasonable voltage for a PIC; otherwise you need a
switchmode boost regulator that throws more variables into the equation.

I think it was Linear Technology who first pointed out in