PDA

View Full Version : PIC power drain



Ron Marcus
- 26th April 2006, 19:24
Does anyone know what the power drain of a PIC is with an external clock vs. HS setting. I assume it is less in general. What about if the clock is stopped? I am running a 16F819 and a 18F2420 with a clock pulse derived from a Chipcon CC1100 and was trying to figure what I can expect from a tank of gas. The clock speed is 8 MHz.

Thanks,
Ron

Melanie
- 26th April 2006, 20:37
Datasheet - Electrical Characteristics Section. For the 16F819 it's Section 15.

Don't look at my next paragraph... go read the Datasheet first...

OK, now you've read it, you've wasted your time (but admit it was fun!), because the current drain will depend on your I/O's (ie what's sourcing and what's sinking), what internal hardware you've enabled (ADC's weak pull-up's internal voltage references etc etc). You get the drift. You just can't calculate it that easy.

You build your circuit, stick your DMM on Amps and measure... If you want to really feel good, you measure on the 10A range and it'll show zero... Note that when measuring on very low (single figure) uA ranges, your meter will impact quite a high series resistance which can throw the operation of your circuit as well as give you a false reading.

Ron Marcus
- 26th April 2006, 21:19
Datasheet - Electrical Characteristics Section. For the 16F819 it's Section 15.

Don't look at my next paragraph... go read the Datasheet first...

OK, now you've read it, you've wasted your time (but admit it was fun!), because the current drain will depend on your I/O's (ie what's sourcing and what's sinking), what internal hardware you've enabled (ADC's weak pull-up's internal voltage references etc etc). You get the drift. You just can't calculate it that easy.

.
I "RTFMed" three times prior to posting. There is no info on the EC vs any other mode. Now, there is HS vs LP vs XT and even RC! But no mention of EC?!? I will build the proto with the little QLP package, go blind from squinting, and stick my handy dandy wide range constant voltage microammeter of my design (thankyou very much!) in the appropriate places to check current drain, but, my question was sparked as much from the lack of info in "TFM", as the chioce of including an 8 Meg resonator and shutting down the crystal oscillator on the CC1100 or running everything from the 8.666 MHz output from the RF chip. The 1100 runs at 90 uA with the crystal running in standby, so I am leaning toward it's oscillator.

Ron Marcus
- 27th April 2006, 19:17
Some of the 18F devices have comparisons of EC vs. other modes. It appears that there is about a 30% power savings over using the HS mode . Does this translate to the '819? I believe it will have 'some' savings in power. It has answered the question that it is not less power efficient, not to mention space saving, by using an external oscillator and adjusting the PBP routines for the 9% frequency difference. I'll have the boards back next week, and report my findings.

Thanks for the feedback,
Ron

Bruce
- 28th April 2006, 01:58
Some 18F parts have "power managed modes" that work with EC oscillators, so the data sheet indicates current consumption in EC mode as well as other oscillator options.

In the 18F2420 data sheet, look in the power managed modes section for details.

Note that this does not indicate whatever current your external TTL clock osc is going to eat up. Only what's needed/used by the PIC.

The 819 doesn't have anywhere near the bells & whistles the 18F series does, and all the 819 data sheet can really tell you about power consumption is what the PIC itself requires.

TTL clock type oscillators have their own current specs. You would need to factor that in with what's shown in the 819 data sheet as Melanie indicated.

Ron Marcus
- 1st May 2006, 00:23
Some 18F parts have "power managed modes" that work with EC oscillators, so the data sheet indicates current consumption in EC mode as well as other oscillator options.

In the 18F2420 data sheet, look in the power managed modes section for details.

Note that this does not indicate whatever current your external TTL clock osc is going to eat up. Only what's needed/used by the PIC.

The 819 doesn't have anywhere near the bells & whistles the 18F series does, and all the 819 data sheet can really tell you about power consumption is what the PIC itself requires.

TTL clock type oscillators have their own current specs. You would need to factor that in with what's shown in the 819 data sheet as Melanie indicated.
The oscillator output is one of the of the CC1100 programmable functions. It draws 90 uA at 3 volts with the clock running while in "wake on radio mode". Of course, if it is feeding the PIC also, this value will increase. I have it breadboarded right now with a GPS module and on screen display, so current measurement is a bit tricky. I am running separate oscillators while I check out the GPS functionality. The next test will be to make sure the oscillator output from the 1100 will reliably run the PIC.

Thanks again for the input...
Ron