Author Topic: 52233 for low power use  (Read 7803 times)

Offline neil

  • Sr. Member
  • ****
  • Posts: 438
    • View Profile
52233 for low power use
« on: November 26, 2010, 07:42:04 PM »
Hi Mark,
  I have a project that will be runnng from solar panel and battery, so I need a low'ish power processor. I haev a tray of 52233 processors which I hoped to use for this. 

 1. I dont need the PHY, so I assume I can switch this off?
 2. Also I dont need to run at the full speed, can I reduce this, and what speed can I reduce this to?
 3. If I can reduce the processor speed, can I still use the defines that you have (example, serial 9600 baud will still be valid)?

Many thanks
Neil

Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3236
    • View Profile
    • uTasker
Re: 52233 for low power use
« Reply #1 on: November 26, 2010, 09:02:34 PM »
Hi Neil

1. If you don't enable Ethernet the PHY will be automatically powered down.
2. You can adjust the seep settings with the defined in app_hw_M5223x.h

        #define CRYSTAL_FREQ                   25000000                  // 25 MHz oscillator / crystal input
        #define PRE_DIVIDER                    5                         // warning: in first silicon this cannot be changed!
        #define OSCCLK                         (CRYSTAL_FREQ / PRE_DIVIDER)
        #define PLL_MUL                        12                        // 60MHz operation

Depending on which crystal speed you have (you don't need 25MHz without Ethernet) and the settings chosen (invalid settings will result in compiler error messages so it shouldn't be possible to do anything wrong) you can specify the speed of operation.
By setting DISABLE_PLL you can also run directly at the crystal speed.

3. The UART defines adjust them selves depending on the PLL output speed so these don't need to be changed. Note however that 25MH is not a very good frequency for generatin fast baud-rates, especially when the PLL is a low value. It may be best to choose something like 8.388MHz or 16.777MHz (and similar) so that the standard Baud rates can be accurately achieved.

Regards

Mark

P.S: Also try the low power mode (#define SUPPORT_LOW_POWER) since this sets the processor to sleep mode when it has nothing to do, which reduces the average current drain.

Offline svl

  • Jr. Member
  • **
  • Posts: 52
    • View Profile
Re: 52233 for low power use
« Reply #2 on: December 21, 2010, 08:45:04 AM »
Hi Mark & Neil
I am facing almost the same problem.

We have a case where our unit gets more warm that we like.(Not critical at all).

In our case we do not use fully potential of the 80MHz in 95% of the time.
My first idée was to use the SUPPORT_LOW_POWER operation, but can see that my system never will be able to get in to low power state do to 100 ms timers.

The next idée was to change the clock as runtime. This was are very fine from the theoretical point of view, but my problem are that it looks like many fundamental things in the system must be change, do you agree on this.
Any other suggestions are welcome.

Best regards
Steen Larsen



Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3236
    • View Profile
    • uTasker
Re: 52233 for low power use
« Reply #3 on: December 21, 2010, 03:25:14 PM »
Hi Steen

The use of the low power mode should help slightly (although most heat comes from the internal PHY and so it may not be very noticeable) - it will stop the CPU clock whenever the CPU is not needed (this is often >95% or the time). It won't stop peripherals (like timers) but unused peripherals will usually be powered down anyway.

If clock speeds are changed during operation (probably won't save any more than the low power mode since the CPU uses most power - after the PHY) it is necessary to recalculate various settings (timers, UARTs, etc.) and there are also risks of losing data during changes. This is why the uTasker project uses fixed values - the low power mode is usually the simplest method and has no side-effects. [At 80MHz clock it may save about 40mA consumption but this is only about 10..15% of the total power due to the PHY at 100MHz].

Regards

Mark