Sign in

username:

password:



Not a member?

Search piclist



Search tips

Subscribe to piclist



piclist by Keywords

12F675 | 16F628 | 16F84 | 16f877 | 16F877A | 16F88 | 18F458 | ADC | AVR | Bootloader | CAN | CCS | CRC | EAGLE | EEPROM | ICD | ICSP | IDE | JDM | LED | Macros | Microchip | MPLAB | PCB-CAD | PIC10F | Pic12f675 | PIC16F84 | PIC16F84A | PIC16F877 | PIC18 | PIC18F452 | PicBasic | PICC | PICSTART | PWM | RS-485 | RS232 | SMT | SPI | UART | USART | USB | Wireless | Wisp628 | Xilinx

Ads

Discussion Groups

Discussion Groups | Piclist | Re: Re: Real Time Clock ( was e: wakeup ? (long)

A discussion group for the PICMicro microcontroller. Also called the Microchip PIC, this list is dedicated to the use and abuse of this fine, simple, microcontroller. Close to topic posts are welcome, ie. general electronics.

wakeup ? - Dave Mucha - Mar 16 11:54:00 2004

Hi all,

What is the preferred method of saving batteries ?

I was thinking of sleep for 30 seconds, wake to check to see if the
power is on a pin, re-sleep if no power, run if the switch is closed
and there is 5v on that pin.

Or, is there a better way to wake up the PIC when in sleep mode ?

Dave




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )


Re: wakeup ? - Wilson - Engenharia Contemp - Mar 16 12:48:00 2004

Why don´t sleep at an interrupt? I say, every time occur an interrupt,
automatically MCU wake from sleep mode.

Regards, Wilson Antonieti Engenharia de Desenvolvimento Tel.: (11) 4223-5117 Fax.:
(11) 4223-5103 Visite nosso site:
www.contemp.com.br PRECISÃO AO SEU ALCANCE!!!
----- Original Message -----
From: "Dave Mucha" <>
To: <>
Sent: Tuesday, March 16, 2004 1:54 PM
Subject: [piclist] wakeup ? > Hi all,
>
> What is the preferred method of saving batteries ?
>
> I was thinking of sleep for 30 seconds, wake to check to see if the
> power is on a pin, re-sleep if no power, run if the switch is closed
> and there is 5v on that pin.
>
> Or, is there a better way to wake up the PIC when in sleep mode ?
>
> Dave >
>
> to unsubscribe, go to http://www.yahoogroups.com and follow the
instructions
> Yahoo! Groups Links




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? - upand_at_them - Mar 16 16:31:00 2004

Like Wilson said you can use interrupts. I'm not sure if all PIC
have it, but my experience with interrupts on the AVR is that the
power-on interrupt can signal if it was a power-on or a reset.

Refresh my memory...What are you trying to do? And how can you even
do anything if there is no power at Vdd?

Mike --- In , "Dave Mucha" <davemucha@j...> wrote:
> Hi all,
>
> What is the preferred method of saving batteries ?
>
> I was thinking of sleep for 30 seconds, wake to check to see if the
> power is on a pin, re-sleep if no power, run if the switch is
closed
> and there is 5v on that pin.
>
> Or, is there a better way to wake up the PIC when in sleep mode ?
>
> Dave




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - Dave Mucha - Mar 17 13:47:00 2004

--- In , "upand_at_them" <upand_at_them@y...>
wrote:
> Like Wilson said you can use interrupts. I'm not sure if all PIC
> have it, but my experience with interrupts on the AVR is that the
> power-on interrupt can signal if it was a power-on or a reset.
>
> Refresh my memory...What are you trying to do? And how can you
even
> do anything if there is no power at Vdd?
>
> Mike
I guess I'm thinking ahead of myself…. Sorry for not being as clear
as I could be.

My brother runs a greenhouse and wanted me to make him a portable
controller(s) that can be used for different plants, different types
of plants and different watering routines.

The variables are frequency of watering, duration of watering and
monitor soil moisture and temperature.

The idea is to make a half dozen units that would all have the same
basic type of program, and then offer field settings of watering
times and such.

The board would have a real time clock and serial mammary chip and
probably the 16F876 or similar.

The sprinkler connection has 24VAC so it would be no problem adding a
power supply on the board.

I am hoping to be able to program the PIC and then store user setting
in serial memory. The options are the 31k of scratchpad ram on the
clock or a section of memory in the serial memory chip.

The clock will have a button battery so there will be some battery on-
board, but not enough to do much of anything but back up the clock.

When one plugs the unit in to the power at the sprinkler, the PIC
would wake up, look at the parameters stored in the memory and then
use those as settings for how to operate in that location.

Periodically, someone would go to the unit, connect a PALM, download
the data of soil moisture and temperature, then take that back to the
office and someone there would figure out what changes to make.

Then someone would go back to the unit, again with a PALM, connect to
the board and upload new operation settings of water frequency and
duration.

The original concept was using a 9v or some AA batteries, but that
seemed like I would have to make the unit twice as large as 4 AA
batteries are as large as the board I've started to lay out.

I'm on the fence about batteries and not using batteries. I figure
that if one of these would sit on the shelf for a season, 3 months or
more, the batteries would be dead. That was the reason I want to
understand more about sleep and wake-up conditions.

If using batteries, sleep could be ended with the application of the
external power, or I could routinely wake-up, check for an external
setting then either go back to sleep, or run. I'm not familiar with
interrupts that would wake up the PIC.

Another question that I have not resolved (please change the subject
line if you respond to this question) is one of writing to memory.
512k serial memory chips can be page written, but I do not have a
page of memory to write unless I store data on a secondary chip and
then write the whole lot. But, that means bunches of smaller writes
to that secondary memory chip which means a shorter life for that.
Please offer alternatives if I missed something.

Also, I have not thought about storing data after a power failure,
resuming from the last write. I don't mind losing a page of data
or whatever, but if I'm storing 2 weeks data and lose power after a
week for one hour, I want to retain the first week and just have a
gap where there was no data stored.

Hope the bigger picture makes more sense.

Dave





(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Re: wakeup ? (long) - sergio masci - Mar 17 13:57:00 2004



> The board would have a real time clock and serial mammary chip and
> probably the 16F876 or similar.

That's one mother of a chip ;-)




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - Dave Mucha - Mar 17 15:32:00 2004

--- In , "sergio masci" <smypl@x> wrote: > > The board would have a real time clock and serial mammary chip
and
> > probably the 16F876 or similar.
>
> That's one mother of a chip ;-)

yeah, I'm nursing this project along......



(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Re: wakeup ? (long) - jrem - Mar 17 16:03:00 2004

just a couple of thoughts . . .

a) you really don't need an external clock, if you use a crystal you
can make a PIC to clock functions that are within a minute-per-week

b) why not just write the data to the PIC data eeprom, one write a
minute would let the eeprom live for ninteen years or so.

c) I would consider powering it off of the available 24vdc, and use a
9vdc battery for a backup if the 24 goes down. A couple of diodes does
this nicely. Or maybe even put a lithium ion rechargeable in there and
build a charging curcuit for it.

d) Interrupts are not that hard to do. You can use the interrupt
routine to do your clock functions, and to wake up the mu from sleep
mode.

Regards, John. --- Dave Mucha <> wrote:
> --- In , "upand_at_them" <upand_at_them@y...>
> wrote:
> > Like Wilson said you can use interrupts. I'm not sure if all PIC
> > have it, but my experience with interrupts on the AVR is that the
> > power-on interrupt can signal if it was a power-on or a reset.
> >
> > Refresh my memory...What are you trying to do? And how can you
> even
> > do anything if there is no power at Vdd?
> >
> > Mike
> > I guess I'm thinking ahead of myself…. Sorry for not being as clear
> as I could be.
>
> My brother runs a greenhouse and wanted me to make him a portable
> controller(s) that can be used for different plants, different types
> of plants and different watering routines.
>
> The variables are frequency of watering, duration of watering and
> monitor soil moisture and temperature.
>
> The idea is to make a half dozen units that would all have the same
> basic type of program, and then offer field settings of watering
> times and such.
>
> The board would have a real time clock and serial mammary chip and
> probably the 16F876 or similar.
>
> The sprinkler connection has 24VAC so it would be no problem adding a
>
> power supply on the board.
>
> I am hoping to be able to program the PIC and then store user setting
>
> in serial memory. The options are the 31k of scratchpad ram on the
> clock or a section of memory in the serial memory chip.
>
> The clock will have a button battery so there will be some battery
> on-
> board, but not enough to do much of anything but back up the clock.
>
> When one plugs the unit in to the power at the sprinkler, the PIC
> would wake up, look at the parameters stored in the memory and then
> use those as settings for how to operate in that location.
>
> Periodically, someone would go to the unit, connect a PALM, download
> the data of soil moisture and temperature, then take that back to the
>
> office and someone there would figure out what changes to make.
>
> Then someone would go back to the unit, again with a PALM, connect to
>
> the board and upload new operation settings of water frequency and
> duration.
>
> The original concept was using a 9v or some AA batteries, but that
> seemed like I would have to make the unit twice as large as 4 AA
> batteries are as large as the board I've started to lay out.
>
> I'm on the fence about batteries and not using batteries. I figure
> that if one of these would sit on the shelf for a season, 3 months or
>
> more, the batteries would be dead. That was the reason I want to
> understand more about sleep and wake-up conditions.
>
> If using batteries, sleep could be ended with the application of the
> external power, or I could routinely wake-up, check for an external
> setting then either go back to sleep, or run. I'm not familiar with
> interrupts that would wake up the PIC.
>
> Another question that I have not resolved (please change the subject
> line if you respond to this question) is one of writing to memory.
> 512k serial memory chips can be page written, but I do not have a
> page of memory to write unless I store data on a secondary chip and
> then write the whole lot. But, that means bunches of smaller writes
> to that secondary memory chip which means a shorter life for that.
> Please offer alternatives if I missed something.
>
> Also, I have not thought about storing data after a power failure,
> resuming from the last write. I don't mind losing a page of data
> or whatever, but if I'm storing 2 weeks data and lose power after a
> week for one hour, I want to retain the first week and just have a
> gap where there was no data stored.
>
> Hope the bigger picture makes more sense.
>
> Dave __________________________________
">http://mail.yahoo.com


(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - peterhawken - Mar 17 17:04:00 2004

I would agree with most of these thoughts but I would have selected a
different type of battery. I have just completed a project that
includes a high performance battery pack. Lithium ion was my first
choice until I studied them closely. Lithium suffers from a loss of
capacity over time that is independant of the number of charges, so
that after one year, it is reasonable to expect the battery to be as
little as 60% of it's new capacity. The rate of decay is also linked
closely to temperature. In a high temerature, the batteries can lose
a huge amount of their capacity in a matter of weeks or months. In
my case, the battery pack was likely to live inside a car for
significant parts of its life and therefore was likely to suffer a
reduced life as a result. Typical temperatures in a greehouse are
likely to severely limit the lide of lithium ion batteries. Our
choice was for NiMh which, although not the same energy density,
appears to offer up to about 500 full charges (or the equivalent in
partial charges). There is very little decay over time and they are
comfortable in most temeratures. Bearing in mind that this
application is unlikely to discharge the battery to any extent, I
think that one can consider NiMh suitable for the life of the
product - maybe not the 19 years the eprom offers, but certainly a
good proportion of that time. The battery pack we used was a 3500mAH
8.4V unit (7 x 1.2V C size cells). This pack would drive a camcorder
sized video recorder and screen for 8 hours. When the recorder and
screen are turned off, the battery will operate the PIC and
associated circuitry for 4 weeks without putting any device into
sleep. I am sure that a PP3 NiMh battery will offer 24 - 48 hours
life from fully charged and therefore is almost certainly sufficient
to cope with intermittant power failures.

Charging is a little problematic, but with a little care, the
charging algorythms can operate within the remainder of the PIC
programme. I can offer some suggestions for charging if you want to
explore that route.

Finally, John suggests avoiding the real time clock - I would agree,
but remember that the PIC will accurately count elapsed time but is
not so good at real time, so if an accurate time and date stamp is
required, a separate clock might be the way to go. Personaly, I
always look for ways to avoid using external clocks and employ
various methods to record time such as timed loops that wait for user
input or counting fixed length routines as well as using the internal
timers and counters.

Good luck in your project. It sounds interesting and I for one would
like to hear of your progress as you develop it.

Peter

--- In , jrem <jrem123@y...> wrote:
> just a couple of thoughts . . .
>
> a) you really don't need an external clock, if you use a crystal you
> can make a PIC to clock functions that are within a minute-per-week
>
> b) why not just write the data to the PIC data eeprom, one write a
> minute would let the eeprom live for ninteen years or so.
>
> c) I would consider powering it off of the available 24vdc, and use
a
> 9vdc battery for a backup if the 24 goes down. A couple of diodes
does
> this nicely. Or maybe even put a lithium ion rechargeable in there
and
> build a charging curcuit for it.
>
> d) Interrupts are not that hard to do. You can use the interrupt
> routine to do your clock functions, and to wake up the mu from sleep
> mode.
>
> Regards, John. > --- Dave Mucha <davemucha@j...> wrote:
> > --- In , "upand_at_them"
<upand_at_them@y...>
> > wrote:
> > > Like Wilson said you can use interrupts. I'm not sure if all
PIC
> > > have it, but my experience with interrupts on the AVR is that
the
> > > power-on interrupt can signal if it was a power-on or a reset.
> > >
> > > Refresh my memory...What are you trying to do? And how can you
> > even
> > > do anything if there is no power at Vdd?
> > >
> > > Mike
> > >
> >
> >
> > I guess I'm thinking ahead of myself…. Sorry for not being as
clear
> > as I could be.
> >
> > My brother runs a greenhouse and wanted me to make him a portable
> > controller(s) that can be used for different plants, different
types
> > of plants and different watering routines.
> >
> > The variables are frequency of watering, duration of watering and
> > monitor soil moisture and temperature.
> >
> > The idea is to make a half dozen units that would all have the
same
> > basic type of program, and then offer field settings of watering
> > times and such.
> >
> > The board would have a real time clock and serial mammary chip
and
> > probably the 16F876 or similar.
> >
> > The sprinkler connection has 24VAC so it would be no problem
adding a
> >
> > power supply on the board.
> >
> > I am hoping to be able to program the PIC and then store user
setting
> >
> > in serial memory. The options are the 31k of scratchpad ram on
the
> > clock or a section of memory in the serial memory chip.
> >
> > The clock will have a button battery so there will be some battery
> > on-
> > board, but not enough to do much of anything but back up the
clock.
> >
> > When one plugs the unit in to the power at the sprinkler, the PIC
> > would wake up, look at the parameters stored in the memory and
then
> > use those as settings for how to operate in that location.
> >
> > Periodically, someone would go to the unit, connect a PALM,
download
> > the data of soil moisture and temperature, then take that back to
the
> >
> > office and someone there would figure out what changes to make.
> >
> > Then someone would go back to the unit, again with a PALM,
connect to
> >
> > the board and upload new operation settings of water frequency
and
> > duration.
> >
> > The original concept was using a 9v or some AA batteries, but
that
> > seemed like I would have to make the unit twice as large as 4 AA
> > batteries are as large as the board I've started to lay out.
> >
> > I'm on the fence about batteries and not using batteries. I
figure
> > that if one of these would sit on the shelf for a season, 3
months or
> >
> > more, the batteries would be dead. That was the reason I want to
> > understand more about sleep and wake-up conditions.
> >
> > If using batteries, sleep could be ended with the application of
the
> > external power, or I could routinely wake-up, check for an
external
> > setting then either go back to sleep, or run. I'm not familiar
with
> > interrupts that would wake up the PIC.
> >
> > Another question that I have not resolved (please change the
subject
> > line if you respond to this question) is one of writing to
memory.
> > 512k serial memory chips can be page written, but I do not have a
> > page of memory to write unless I store data on a secondary chip
and
> > then write the whole lot. But, that means bunches of smaller
writes
> > to that secondary memory chip which means a shorter life for
that.
> > Please offer alternatives if I missed something.
> >
> > Also, I have not thought about storing data after a power
failure,
> > resuming from the last write. I don't mind losing a page of
data
> > or whatever, but if I'm storing 2 weeks data and lose power after
a
> > week for one hour, I want to retain the first week and just have
a
> > gap where there was no data stored.
> >
> > Hope the bigger picture makes more sense.
> >
> > Dave
> >
> >
> >
> >
> >
> >
> >
> >
> > __________________________________
> ">http://mail.yahoo.com



(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - upand_at_them - Mar 17 18:37:00 2004

Sounds cool. Here's what I'm thinking. And correct me if I have
your idea wrong...

I'm thinking you probably don't need the RTC; you probably don't need
to know what temperature and soil conditions at a specific date and
time as much as you want to see trends over reasonably accurate
periods. If that's so, then I would use the watchdog timer - and
with the scaler and a couple lines of code you could get it to
operate at almost any period.

So when it sleeps it does nothing and when it is awake and the proper
period has been reached (using the clock frequency in the
calculations to get your approximate period - I think this will be
more than accurate enough)...then you would switch power on to a
serial EEPROM (no sense powering it while sleeping), write a few
bytes of measurements, control whatever else you have to control
(water, etc.) and then go back to sleep.

Your circuit would check for changed parameters when they're
uploaded, no? There's no reason to code for it in your interrupt
routine if it only changes when someone uploads data to it (and it'll
have to be awake for that anyway).

I was also thinking about using an IR communication instead of a
physical connection to download/upload data.

Mike






(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - Dave Mucha - Mar 18 10:27:00 2004

--- In , "upand_at_them" <upand_at_them@y...>
wrote:
> Sounds cool. Here's what I'm thinking. And correct me if I have
> your idea wrong...
>
> I'm thinking you probably don't need the RTC; you probably don't
need
> to know what temperature and soil conditions at a specific date and
> time as much as you want to see trends over reasonably accurate
> periods. If that's so, then I would use the watchdog timer - and
> with the scaler and a couple lines of code you could get it to
> operate at almost any period.
>
> So when it sleeps it does nothing and when it is awake and the
proper
> period has been reached (using the clock frequency in the
> calculations to get your approximate period - I think this will be
> more than accurate enough)...then you would switch power on to a
> serial EEPROM (no sense powering it while sleeping), write a few
> bytes of measurements, control whatever else you have to control
> (water, etc.) and then go back to sleep.
>
> Your circuit would check for changed parameters when they're
> uploaded, no? There's no reason to code for it in your interrupt
> routine if it only changes when someone uploads data to it (and
it'll
> have to be awake for that anyway).
>
> I was also thinking about using an IR communication instead of a
> physical connection to download/upload data.
>
> Mike

Really neat ideas.

You are correct that exact time is not a high priority, so knowing it
was in the morning is probably good enough.

If the PIC can keep withing about a minute a week, then 3 months
would be within about 15 minutes and that is pleanty accurate.

Dave






(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: wakeup ? (long) - upand_at_them - Mar 18 16:58:00 2004

> If the PIC can keep withing about a minute a week, then 3 months
> would be within about 15 minutes and that is pleanty accurate.

I haven't tested the accuracy of a uC based clock, but you could
always have a RTC on the hand circuit and upload the time whenever an
upload takes place if you were trying to minimize cost.

Mike






(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Real Time Clock ( was e: wakeup ? (long) - Dave Mucha - Mar 19 12:09:00 2004

--- In , "upand_at_them" <upand_at_them@y...>
wrote:
> > If the PIC can keep withing about a minute a week, then 3 months
> > would be within about 15 minutes and that is pleanty accurate.
>
> I haven't tested the accuracy of a uC based clock, but you could
> always have a RTC on the hand circuit and upload the time whenever
an
> upload takes place if you were trying to minimize cost.
>
> Mike The idea of the RTC is interesting in that it adds scratch pad RAM.
I have an RTC on order so I'll at least test it in the circuit.

I have not determined if the back-up battery holds that with power
off so it may not be as valuable as it could be.

I think I will see just how accurate the timing is on a PIC with a
crystal.

Another question I have is about the Guard Ring on the DS1302 RTC
chip.

The spec sheet calls for a Guard Ring around the crystal pins
connected to ground. I am looking at an 8 pin SOIC chip and am
using a PCB etching machine so my isolation between traces is limited
to a minimum of about 10 mill. With a 10 mil trace, and two 10 ten
mill isolation spaces, there is little left over when there is only
50 mil between pads.

Can anybody offer any experience with the differences between using
and not using a Guard Ring on one of these clocks ?

Or, maybe offer a list/site/url where I can get someone with personal
experience ?

Dave
(Funny how the days waiting for the parts generates so many
questions....)




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - upand_at_them - Mar 19 14:19:00 2004

> The idea of the RTC is interesting in that it adds scratch pad RAM.

True. But you won't have room in the PIC's data EEPROM?

> I think I will see just how accurate the timing is on a PIC with a
> crystal.

And I think I'll try an experiment with a PIC and the RC oscillator
option and post the results.

I don't know anything about the guard ring, but I had a DS1302
sitting in a breadboard for a week at it was still accurate to the
second.

Mike --- In , "Dave Mucha" <davemucha@j...> wrote:
> The idea of the RTC is interesting in that it adds scratch pad
RAM.
> I have an RTC on order so I'll at least test it in the circuit.
>
> I have not determined if the back-up battery holds that with power
> off so it may not be as valuable as it could be.
>
> I think I will see just how accurate the timing is on a PIC with a
> crystal.
>
> Another question I have is about the Guard Ring on the DS1302 RTC
> chip.
>
> The spec sheet calls for a Guard Ring around the crystal pins
> connected to ground. I am looking at an 8 pin SOIC chip and am
> using a PCB etching machine so my isolation between traces is
limited
> to a minimum of about 10 mill. With a 10 mil trace, and two 10
ten
> mill isolation spaces, there is little left over when there is only
> 50 mil between pads.
>
> Can anybody offer any experience with the differences between using
> and not using a Guard Ring on one of these clocks ?
>
> Or, maybe offer a list/site/url where I can get someone with
personal
> experience ?
>
> Dave
> (Funny how the days waiting for the parts generates so many
> questions....)




(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - Dave Mucha - Mar 19 14:33:00 2004

--- In , "upand_at_them" <upand_at_them@y...>
wrote:
> > The idea of the RTC is interesting in that it adds scratch pad
RAM.
>
> True. But you won't have room in the PIC's data EEPROM? The question here is about what happens to the variables when the
power is lost. If there is no battery to the PIC, I thought
everything except what you programmed was lost and that nothing was
stored ?

Never had to try to do that before so I just don't know. Dave






(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Re: Real Time Clock ( was e: wakeup ? (long) - jrem - Mar 19 18:32:00 2004

two things . . .

a) the data memory is EEPROM, that's eraseable electronic programmable
read only memory, it keeps data when no power is on the chip

b) I took a pic of my 16F628 running a stop watch routine, the LCD is
next to a Casio dive watch (the watch shows minutes/seconds/hundreths,
the LCD is hours/minutes/seconds). I started them both last night, and
I was probably within a quarter second on each device. 23:46 later
they are about a second apart. I had 00:00 and 23:46 movies but they
didn't turn out due to lighting. The 16F628 is running at 4mhz doing
an interrupt routine for the clock.

The jpg is here: http://jrem.webhop.net/dscn0617.jpg Regards, John. --- Dave Mucha <> wrote:
> --- In , "upand_at_them" <upand_at_them@y...>
> wrote:
> > > The idea of the RTC is interesting in that it adds scratch pad
> RAM.
> >
> > True. But you won't have room in the PIC's data EEPROM? > The question here is about what happens to the variables when the
> power is lost. If there is no battery to the PIC, I thought
> everything except what you programmed was lost and that nothing was
> stored ?
>
> Never had to try to do that before so I just don't know. > Dave
__________________________________
">http://mail.yahoo.com


(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - upand_at_them - Mar 19 18:55:00 2004

There are three types of memory in a PIC:

1) Program memory. This is where the program resides and it,
obviously, remains untouched when the power is lost.

2) RAM. This is what the program uses for variables, etc. When
power is lost so are the contents.

3) EEPROM data memory. This is like RAM, but when power is lost the
contents remain. This is where people put calibration factors,
etc. It can be programmed internally, through the program, or
externally, through the device programmer. Not all PICs have this.

And, btw, the RAM in the DS1302 is SRAM, so once power is
lost...that's it.

Mike --- In , "Dave Mucha" <davemucha@j...> wrote:
> The question here is about what happens to the variables when the
> power is lost. If there is no battery to the PIC, I thought
> everything except what you programmed was lost and that nothing was
> stored ?
>
> Never had to try to do that before so I just don't know. > Dave



(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - Scott Lee - Mar 19 18:55:00 2004

--- In , jrem <jrem123@y...> wrote:
> b) I took a pic of my 16F628 running a stop watch routine, the LCD
is
> next to a Casio dive watch (the watch shows
minutes/seconds/hundreths,
> the LCD is hours/minutes/seconds). I started them both last night,
and
> I was probably within a quarter second on each device. 23:46 later
> they are about a second apart. I had 00:00 and 23:46 movies but
they
> didn't turn out due to lighting. The 16F628 is running at 4mhz
doing
> an interrupt routine for the clock.

Coded properly, there is no reason why a PIC clocked from a crystal
won't be as accurate a time keeper as the cystal is.

Thus if you have a 100ppm part then you should see less than 1 second
drift in ~2.78 hours. If you have a 25ppm part it would no more than
1 second in ~11.11 hours. Note that these are worst case scenarios.
This testing seems to confirm this. (And would anyone expect
different?)





(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - Phil - Mar 19 19:11:00 2004

--- In , "upand_at_them" <upand_at_them@y...>
wrote:
...
>
> And I think I'll try an experiment with a PIC and the RC oscillator
> option and post the results.
>
> I don't know anything about the guard ring, but I had a DS1302
> sitting in a breadboard for a week at it was still accurate to the
> second.
>
> Mike

the problem with RC (and intosc) is that RC drifts with the
temperature. even if you found a combination that appeared to be
accurate for a given test, change the temp and your frequency will
drift.



(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )

Re: Real Time Clock ( was e: wakeup ? (long) - Chad Russel - Mar 20 22:53:00 2004

I have not looked at the specs for the part, but the ring guard is used
for very high impedance inputs to reduce surface leakage current on the
PWB. Very high impeadance means long battery life, but also very
little drive requirements, which means any leakage may kill the
oscillator, or if it has AGC will increase the battery current.

Chad

<<< Snip >>>
>
> Another question I have is about the Guard Ring on the DS1302 RTC
> chip.
>
> The spec sheet calls for a Guard Ring around the crystal pins
> connected to ground. I am looking at an 8 pin SOIC chip and am
> using a PCB etching machine so my isolation between traces is limited
>
> to a minimum of about 10 mill. With a 10 mil trace, and two 10 ten
>
> mill isolation spaces, there is little left over when there is only
> 50 mil between pads.
>
> Can anybody offer any experience with the differences between using
> and not using a Guard Ring on one of these clocks ?
>
> Or, maybe offer a list/site/url where I can get someone with personal
>
> experience ?
>
> Dave
> (Funny how the days waiting for the parts generates so many
> questions....) =====
My software has no bugs. Only undocumented features.

__________________________________





(You need to be a member of piclist -- send a blank email to piclist-subscribe@yahoogroups.com )