ABS MAX and 6V battery, is it safe?

[dsicon]

Hi all
i am working on a new design, power source to be a small 6V alkaline battery.
Open circuit voltage on fresh battery samples shows a consistent 6.50V
(there is also a lithium in the same package, same open ckt voltage).

ABS MAX VDD for the PIC is 6.5V.
The part will mostly be sleeping and quiescent current is planned to be 1-5uA.

There will be a series diode both for polarity protection and to trim the top off the open ckt voltage.

So the question is this:
am I living too dangerously close to the ABS MAX ?
a 4148 diode data sheet says that Vf will be only about 275mV at 1uA so the part will be sitting for most of it's life just a few hundred millivolts under the ABS MAX.

Does anyone have any experience or advice with this ? Long term experience with similar designs, good or bad ?

Using 2 diodes in series that does not appear to be an option for certain reasons.

[Dave]

dsicon, What are the "certain reasons"? Most pic processors will operate down to about 3.3 volts quite nicely...

Dave Purola,
N8NTA

[dsicon]

Hi Dave,
the system needs around 4.0V minimum for reasons other than the PIC itself, also the PIC will be doing a low battery detect with the on board PLVD module which is internally tied to VDD, so an extra diode drop will mess with the limited trip points available

i sort of suspect there must be a lot of PIC systems using a series diode and 4 alkalines which are in the same boat

BTW, a regulator is not an option either due to the requirement of several uA standby
oh and did i mention the board is about 0.8 inches sq with about 25 parts including a sw reg? :)
so frankly i am not sure i could even fit another diode

[Dave]

dsicon, You mention a switching regulator. What is it for? Does it have a shutdown input that you are using?

Dave Purola,
N8NTA

[dsicon]

the sw reg is fed from the battery before the series diode to the pic
and yes it has a shutdown but interestingly since it is a boost reg the full battery dc voltage appears at the output (thru the inductor) during shutdown so the shutdown pin is worthless to me
so i had to put a load sw (P+N mosfet) in front of the sw reg to fully shut it down
this also has the benefit of reverse protection to the sw reg due to the body diode in the P fet

it is really a challenge to get this thing down to a few uA's of standby!

but i still welcome any comments from people who run their pics for years near 6.5V abs max

[Dave]

dsicon, I have used a MAX1724 in the past with 2 diodes on the output (anodes), 1 cathode going to the load and 1 cathode going to the PIC VCC. Connected to the PIC VCC I had a second diode cathode connected. The anode of this second diode went to the battery source. The battery source was 2 AA batterys in series to give about 3 volts. This provided low power to the PIC and when the PIC awakened it placed + 3 volts on the shutdown pin of the MAX1724 which turned it on. This then produced +5 volts to the high current circuitry (100 Ma.) @ 4.5 volts (thru a diode drop). This switching regulator will work down to 2.6 volts on the input. There are ways to shin a cat at 2.6 volts.....

Dave Purola,
N8NTA

[Jerson]

I have a design thats up and running for years now. One version of it is based on 16F628A and the other around a 16f917. Both of them run off 4xA sized dry/alkaline cells. In series with the cells is a 1N4007 for polarity reversal protection. The device usually lies in sleep mode and gets woken by a keypress. So, I guess, your app would work Ok.

[dsicon]

thanks for your comment
can you confirm that with fresh batteries your sleep voltage post diode is about 6.2V ?
ds

[dsicon]

Jerson
did you use such a high voltage rating (1N4007) for any particular reason (higher Vf ?)
ds