Using this magnetic sensor as a wakeup sensor

[kellogs]

for the rest of the circuit. Here is the sensor's datasheet

However, looking at the diagram of interest makes me think this is going to drain the batteries quite fast due to that LMV324N comparator and its voltage divider on the positive terminal.

Is there any way to tweak this design for better battery conservation ?

Thanks

 

[Alec_t]

There's no way you can reduce the IC's10mA per bridge current requirement significantly. Can't you use a different IC? It doesn't look suitable as a wake-up sensor, unless you are really forced to sense very low magnetic field strengths.

 

[kellogs]

I have overlooked that aspect...

Yes, I need to sense a couple of uT; there are alternatives but only in minute packages with no leads. Extremely hard to hard to hand solder.

One thing I do not understand, the 4 resisors of that Wheatstone bridge are 1.1k each. Should the drain not be more around 5 mA at 5 V Vdd ? If you look further down in that table there is another parameter given for I_bridge = 5 mA. The parameter is

Sensitivity Tempco TA= -40 to 125°C, Vb=5V
TA= -40 to 125°C, Ibridge=5mA

Or is it that the resistance varies with drawn current, and at 10 mA it is 100 Ohm ? Which still I can not grasp :|
Thank you

 

[Alec_t]

the 4 resisors of that Wheatstone bridge are 1.1k each.

Typically, but only for the 1021/1022. Which IC do you have?

Should the drain not be more around 5 mA at 5 V Vdd ?

Probably. Then the 10mA figure would relate to a supply voltage closer to the maximum.

 

[(*steve*)]

Depending on the rate of change of the magnetic field, you might turn the sensor on intermittently.

If you can operate it for 100ms every second, the average current drain will be 10% of what it is now.

 

[kellogs]

@Alec_t it is the 1022.

Good idea @(*steve*) however, thinking more about it I don't think I should be using it anyway since it needs to operate around 10 seconds per minute. But out of curiosity, how would I make it work 100ms / sec ? With some combo of a 555 and a transistor maybe ?

 

[Alec_t]

I think R19/20 could each be at least 10k without any adverse effect.
If you use the LMV324S, it has a shut-down feature which would conserve a bit of energy.

 

[(*steve*)]

But out of curiosity, how would I make it work 100ms / sec ? With some combo of a 555 and a transistor maybe ?

I was thinking a microcontroller. You may have to deal with unusual outputs as the device is powered up or down and while it is powered off.