Low current 08M

[moey]

Hi Technical,
I'm keen to reduce the power consumption of 08Ms in Sleep mode, (for battery powered RF jobs).

The lowest I can reduce current consumption in Sleep is 0.65mA using a low current LM2936 regulator and have followed the advice of previous posts - such as pulling unused 08M inputs to ground etc.

Is there an update to the 08M firmware as mentioned below?

--------------------------------
Technical Posted - 14 November 2005 15:5 Further research has shown that the 12F683 (used for the 08M) has a unique brownout option (not found on any other chip) that we were not previously aware of - a fuse setting that enables the BOD during normal use but automatically disables it during the sleep command. This would overcome this sleep current issue on 08M chips (not available to 08 chips).

We will do some tests and if this operates as expected we will implement this in a new release firmware update early in 2006. However this will a 'new' feature not available on existing chips.
________

 

[Technical]

This is a planned future release that we are testing at present. It should be available by around Easter.

 

[frankeltham]

Moey, also have you looked for other micropower regulators?
e.g. Nat Sem LPC2950 or Zetex ZMR500 for LOW power circuits e.g. 100mA or less.

Or how about switched-modes e.g. I have found the MAX639 is happy supplying up to 200mA. Care required to reduce ripple/sparkles. I used one of these on a PICAXE controlled 434 radio reciever. Its quiescent consumption claims to be about 10 microamps. Its been running for 8 months perfectly.
And switched modes will give better battery usage efficiency too.

Remember, even if the new PICAXE uses 10 microamps its somewhat lost if you have an old-fashioned regulator using gallons.

Obviously there are loads of regulators and everyone has My Favourite Regulator (they really should get out more!).

And finally, in the meantime, also have a look at reducing clock speed too - if your project can stand it.

I'm looking foward to the new version.

 

[moey]

frankeltham,
The LM2936 is supposed to draw just a few microamps. Although I had not thought of switchmode, will see what I can buy locally.

 

[frankeltham]

I've just looked. That LM2936 is a nice low power spec. I think I'll use it myself. Well spotted Moey.

The ONLY couple of advantages of the Switched mode chip would be:
1) Has a low battery warning output.
2) It does power conversion rather than the voltage conversion of a linear device.
i.e. (@ 90% efficiency) if you have a load of 20mA @ 5V, you would have a 9V battery draw of less than 13mA. This would increase as the battery 'flattens'.Whereas with linear it would be 20mA+quiescent etc. So you'd gain a bit on battery life if that is important.

Downsides:
1. Rippley output* - amount depends a lot on physical circuit design. PCB best.
2. Extra components. (increases Cost & size)
3. Limited maximum input voltage (12V for 639).

But with the extra components and the circuit layout required to minimise ripple then I think I'd stick to the LM2936.

* 'Smoothed' spikes from the switched mode process unavoidable but can be minimised by good track layout. On some SM ICs e.g. LM1303 a change in track length of 2mm can make a big differene. The 639 doesn't seem so fussy.

 

[moey]

Just had a slap on forehead moment.
I am using one of the 08M inputs to enable the "low current" mode, ie. Sleep period between temp readings.

The input is held low via 10k then the switch takes it high.

Current draw when switch closed = 0.65mA
Current draw when switch open = 0.15mA
So I just reversed the action in software.

So the 08M does draw very little in Sleep.

Thanks for listening.