SLEEP command question

Archiver

In a message dated 7/22/2004 4:43:01 PM Eastern Daylight Time,
wschini03@y... writes:


> To anyone who might know,
> I am using the BS2pe for datalogging operations and I want to
> set up a sample rate. I am thinking of using the SLEEP command to
> cause the BS2pe and other circuitry to pause and go to low power mode
>

The acutal duration of the sleep command is 2.3 secs for 1, 2.3 seconds for
2,
4.6 secs for 3 and so on. The actual duration is the multiple 0f 2.3 that is
closest to the duration you specify. 30 sec duration would acutally be 32.2
secs, which is the closest multiple of 2.3 to 30. If you want to calculate
the actual duration, divide you intended duration, say 60 secs, by 2.3, then
round it up to the next integer. 60/2.3 = 26.086, so round it up to 27. 27 *
2.3 = 62.1
secs, which would be the actual duration.

Sid


[noparse][[/noparse]Non-text portions of this message have been removed]

Archiver

To anyone who might know,
I am using the BS2pe for datalogging operations and I want to
set up a sample rate. I am thinking of using the SLEEP command to
cause the BS2pe and other circuitry to pause and go to low power mode
while waiting for the next sample time. The Stamp Manual is quite
confusing in what the timing is for the Period of the SLEEP command.
Is the period 1 second or 2.304 seconds? Their example states that
the command "SLEEP 10" is 11.52 seconds long (5 x 2.304s). This does
not make sense. Should'nt "SLEEP 10" give you either 10 seconds or
23.04 seconds. Also, the maximum time for the command should either
be 65,535 seconds (18.204 hours) or 65,535 x 2.304s = 150,992.64
seconds (41.9424 hours). They give the example of "SLEEP 10" as if
the period is 2.304 seconds, but the maximum SLEEP time as if the
period is 1 second. Please help me with this if you can. They are
very vague in the explanation for the timing of the SLEEP command.
Thank you,
Bill Schini, KC9CFO

Archiver

Hi Bill,

Quick answer to your question, SLEEP XXX will give XXX seconds of
sleep, but probably a second or two longer than XXX.

Long answer: The SLEEP command is governed under the Stamp hood by
an RC timer that is built into the Ubicom SX48 micro. It is somewhat
temperature dependent, and it's period is different from one Stamp to
the next.

On the original PIC-based BASIC Stamp the period was around 2.3
seconds. (one digit precision--that 2.304 figure in the manual is
totally bogus precision!) On the SX-based BASIC Stamps (2sx, 2e, 2p
and 2pe), the period is shorter, closer to 2 seconds. Often around
1.9 seconds in my experience.

When you command SLEEP 10, the result will have a granularity of
about 2 seconds (depending on the individual stamp and on the
temperature). The interval will be close to but a little longer than
10 seconds.

You might think that longer intervals might become wildly inaccurate,
as the Stamp cannot really guess what the period of its own RC
watchdog timer is. But no, the Stamp _can_ and in fact _does_ know
the period of its own watchdog and calibrates longer intervals
against its ceramic resonator. So intervals of minutes will turn out
at +/- only a few seconds.

If you have an RTC in your data logger, you can use that in
conjunction with SLEEP to make things happen with great regularity.
If your RTC has an interrupt output, you might also look into the
POLLWAIT feature of the BS2pe.

The same internal watchdog timer is behind the timing of the NAP and
POLLWAIT commands, and its timing can be adjusted down to about 15
milliseconds. I have an essay on the BS2pe and about SLEEP posted at
these URLs:
<http://www.emesystems.com/BS2pe.htm>
<http://www.emesystems.com/BS2power.htm>

-- best regards
Tracy Allen
electronically monitored ecosystems
http://www.emesystems.com
mailto:tracy@e...



>To anyone who might know,
> I am using the BS2pe for datalogging operations and I want to
>set up a sample rate. I am thinking of using the SLEEP command to
>cause the BS2pe and other circuitry to pause and go to low power mode
>while waiting for the next sample time. The Stamp Manual is quite
>confusing in what the timing is for the Period of the SLEEP command.
>Is the period 1 second or 2.304 seconds? Their example states that
>the command "SLEEP 10" is 11.52 seconds long (5 x 2.304s). This does
>not make sense. Should'nt "SLEEP 10" give you either 10 seconds or
>23.04 seconds. Also, the maximum time for the command should either
>be 65,535 seconds (18.204 hours) or 65,535 x 2.304s = 150,992.64
>seconds (41.9424 hours). They give the example of "SLEEP 10" as if
>the period is 2.304 seconds, but the maximum SLEEP time as if the
>period is 1 second. Please help me with this if you can. They are
>very vague in the explanation for the timing of the SLEEP command.
>Thank you,
>Bill Schini, KC9CFO
>