Shop OBEX P1 Docs P2 Docs Learn Events
How to cut down power consumption of the prop for battery- operation? — Parallax Forums

How to cut down power consumption of the prop for battery- operation?

Christof Eb.Christof Eb. Posts: 1,285
edited 2010-07-10 12:16 in Propeller 1
Hi all,
up to this·week I have seen PropBasic as a way to improve computing speed drastically over SPIN - and avoid assembler.
I am now thinking in the opposite direction: Cut down clock speed and still have enough performance.

The aim is to use batteries. There should be some kind of idle mode with very low power consumption. From time to time, the system should wake up and do something (15Minutes a day @ 1MHz?).

What is the best way to do this?
How long can a set of batteries 3V, 2500mAh last?
What is the minimum supply voltage needed?
What is the minimum quarz frequency? Are there ways to use 32768Hz?
Can I just switch the pll somehow (??) to *16, if speed is needed?

In the datasheet I see RC-slow with 20kHz and power consumtion of 8 cogs 2*10^-5 A at this frequency. Can I really calculate 2.5Ah/(2e-5 / 24(h/day)) = 5200days for continuously working at 20kHz with 8 active cogs? (-Of course pin currents have to be added, ok, and the battery itself will not work·as long.)

If there are ideas, I would be thankful!
Christof



·

Comments

  • BeanBean Posts: 8,129
    edited 2010-07-09 11:21
    Christof,
    When waiting for a WAITCNT, WAITPEQ or WAITPNE the cog is in low-power mode.

    On the propeller you can switch clock sources on the fly. So you could run from the 20KHz until you need to do some fast work, then switch to a 5MHz crystal clock, then switch back again.

    See the CLKSET command in spin and pasm. In PropBasic you'll use an in-line pasm command \CLKSET varname

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    Use BASIC on the Propeller with the speed of assembly language.
    PropBASIC thread http://forums.parallax.com/showthread.php?p=867134

    March 2010 Nuts and Volts article·http://www.parallax.com/Portals/0/Downloads/docs/cols/nv/prop/col/nvp5.pdf
    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    There are two rules in life:
    · 1) Never divulge all information
    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    If you choose not to decide, you still have made a choice. [noparse][[/noparse]RUSH - Freewill]
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2010-07-09 17:23
    Minimum supply is 2.7V. A 3V supply is already pushing it.
    A buck-boost supply may be useful for you.

    I do some battery operated stuff, and the ability to shut down some of the hardware on whatever board you are using can be useful.
    When on battery, I run the prop on internal 20KHz and bump it up to 40MHz to do operations then drop it back down. Also, shut down as many as cogs as possible when in slow mode and restart them once everything is up to speed. Depending on what is running in the other cogs, when you slow it down it might loose proper timing and freeze up.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!

    Some of my objects:
    MCP3X0X ADC Driver - Programmable Schmitt inputs, frequency reading, and more!
    Simple Propeller-based Database - Making life easier and more readable for all your EEPROM storage needs.
    String Manipulation Library - Don't allow strings to be the bane of the Propeller, bend them to your will!
    Fast Inter-Propeller Comm - Fast communication between two propellers (1.37MB/s @100MHz)!
  • LawsonLawson Posts: 870
    edited 2010-07-09 18:54
    Another couple tips I've heard is to disable the brown out detector. It consumes 3-4 micro amps by itself, a lot more than the <1uA minimum draw of the prop. (RCSLOW, only one cog active and waiting for CNT or a pin transition) I think the EEPROM adds ~1uA to the system and a low power linear regulator would add ~1uA. If I remember right, the regulators on the protoboard spec at ~250uA idle current.

    I've also remember a thread with someone running a Propeller at 2.0v and a couple of megahertz? I seem to remember that RCSLOW was fine down to 1.8v? I'd recommend finding this thread or testing this all yourself. I DO know that many of the CMOS logic chip families will "run" at voltages well below spec. A 74HC86 I used for an oscillator would run fine at 0.8v but ran at about ~4Hz. It would normally run at ~1Hz when powered with the 2.0v minimum operating voltage for 74HC logic. Likewise a 74AC164 chip I'm using will retain it's state down to <0.5v but the chips' output drivers quit at ~0.73 volts. (I'm not expecting full speed performance from these chips though)

    Lawson

    *edit* The low quiescent current regulator I've played with is the MCP-1700 series. Ultra low drop out, 250mA max, ~1uA quiescent current. OnSemi's MC78LC is similar. [noparse]:)[/noparse]

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Lunch cures all problems! have you had lunch?

    Post Edited (Lawson) : 7/10/2010 4:54:38 AM GMT
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2010-07-09 19:58
    More on the regulator: I use this regulator: LP2989IMM-3.3. It's ultra-low dropout (in the single-digit mV -- if you get the current draw low enough). According to a graph, it looks like about 180µA quiescent current when the input is above 2.6V. It recommends an additional C0G capacitor to help stabilize the power, but it can output 500mA continuous and comes in really small packages.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!

    Some of my objects:
    MCP3X0X ADC Driver - Programmable Schmitt inputs, frequency reading, and more!
    Simple Propeller-based Database - Making life easier and more readable for all your EEPROM storage needs.
    String Manipulation Library - Don't allow strings to be the bane of the Propeller, bend them to your will!
    Fast Inter-Propeller Comm - Fast communication between two propellers (1.37MB/s @100MHz)!
  • Christof Eb.Christof Eb. Posts: 1,285
    edited 2010-07-10 12:16
    Thanks alltogether,

    I will do some experiments.

    Christof
Sign In or Register to comment.