Dale Botkin wrote:
> On Mon, 27 Jan 2003, Robert Rolf wrote:
>
>> What do you read when you short and then reverse the meter leads?
>
> Same thing, but with a minus sign.  -.006uA.
>
>> What do you get when you hook your meter to the battery with a
>> 10M series resistor? It should of course be something near 3V/10M
>> (0.3uA)
>
> Meter says the resistor is 9.77M, and current is .309uA.  So at least
> it's pretty close to being consistent...  the different lead
> arrangement between measuring the resistor and the current could
> account for some error as well.


Well, for the test purpose you could install the 9.77M resistor in series
with the PIC, then use a jumper wire with alligator clips to short circuit
the resistor and leave the PIC running free.

Make a software routine to store known values at the internal RAM.
Make it enters in SLEEP mode.
Remove the jumper from the resistor (it will insert the resistor in series
with the PIC).
Wait 10 minutes
Insert the jumper over the resistor again.
Wakeup the chip and make sure the previously store data in RAM still there.

THe 9.77Mohms resistor will offer only 58mV Vdrop at 6pA current.

You can also o some tests using capacitive charge to supply power to the
PIC while in sleep mode.
Considering 6pA in 3V, the PIC should be offering an impedance of around
500Mohms, so, once charged with 3Vdc, a 100nF capacitor should hold enough
charge to supply 6pA to the PIC in sleep mode for around 20 to 30 seconds.

Install a 100nF capacitor in parallel with the coin cell, use the same
technique above, store a know value to RAM, enter sleep mode, remove the
coin cell leaving only the 100nF cap charged, 30 seconds later reapply coin
cell, wakeup the chip, check the RAM contents.

6pA is really strangely low.

Wagner.

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics