App note: Long Delay Generation Using the AVR Microcontroller

in app notes by DP | 4 comments

long_delay_gen_AVR

Here’s an app note from ATMEL presenting a solution for microcontroller long delays.

The solution presented here shows how the AVR AT90 series microcontrollers (AT90S2313, AT90S4414 and AT90S8515) generate and handle long delays. On-chip timers are used without any software intervention, thus allowing the core to be in a low-power mode during the delay. Since the timers are clocked by the system clock, there is no need of any additional components.
Due to the very long timing capability, this implementation combines high system performance with long delay generation. For example, an AVR Microcontroller running at
20 Mips can generate delays as long as half an hour.

This entry was posted in app notes and tagged , , , .

Comments

  1. Tom says:

    “thus allowing the core to be in a low-power mode during the delay”

    I don’t know what they’re smoking at atmel but having a high-frequency clock running does not count as low-power mode in my book. And the mcus they are talking about are ridiculously power-hungry — their “idle supply current” (at 3 V) is about the same as the active mode current of an ARM at twice the clock speed!

    • Tiersten says:

      It isn’t too surprising that the chips have bad specs compared to the current generation uCs as these are very old designs which were first put into production over 10 years ago. The 3 chips listed have been discontinued.

      AT90S2313 was replaced by the ATtiny2313

      AT90S4414 was replaced by the AT90S8515 which was replaced by the ATmega8515

      • Tom says:

        Looks like the “idle supply current” of the ATtiny2313A at 20 MHz and 5 V (*) is about 2.4 mA which is still about the same as the active mode current consumption of a Cortex M0+. Except that the ARM is happy to run below 2 V, so we’re still looking at twice the power consumption.

        (*) It really makes me sad that it is 2013 and people are still using microcontrollers that need 5 V to reach their full clock rate. Appearently even the obsolete AT90S2313 could run on 3 V.

  2. Sleepwalker3 says:

    Tom said – “I don’t know what they’re smoking at Atmel but having a high-frequency clock running does not count as low-power mode in my book. And the mcus they are talking about are ridiculously power-hungry ”

    Obviously you were referring to them ‘Smoking’ their chips and not their vegie patch of tomato plants. One wonders why they would be so “ridiculously [power] hungry” – probably something to do with smoking the chips (from all that excessive current ;)

    The data sheet appears to be from January 2004, so it’s not really surprising it’s a bit behind the times, but the principles might still hold some merit in some applications perhaps (though probably better ways to do it now).

    Plenty of 5V gear out there that people need to interface to, me included. Some chips have come out with decent setups with separate I/O supplies, so you can have your core doing the LV and happily have your I/O doing 5V, but unfortunately it’s more the exception.

    …Of course, in a few years time the teens will be telling us that we should get off that crazy ancient 3V stuff and get with the latest Quantum Slipstream 0.3mV 100THz, 1fA 500GFlop chip with double kitchen sink that can be powered from the vibrations of a snoring dust mite :)
    – Gotta get some early samples of those!

    :D

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.