The right way to drop 6v to 5.5v

Hi,

I am thinking of using 2 x CR2032 (for space considerations and increased mAh) which would give me 6v to run a 328p chip. I was thinking of using a voltage divider (1MΩ and 100KΩ resistors) that would drop it to 5.45v.

The project would be for a simple countdown timer that would then trigger two 5V active piezo buzzers.

I was wondering if this would be the right approach?

Thanks for any guidance in advance.

It is going to be hard to get enough current through 1 megohm and 100 kohm resistors unless you are planning to put very many in parallel. Of course I don't know what you are planning because you did not explain. Remember that your Arduino acts similar to an additional resistor except that it is variable.

The voltage divider approach is not practical using batteries as an input.

A series diode (e.g. 1N4001) may be a more practical approach.

ardandrew:
I am thinking of using 2 x CR2032 (for space considerations and increased mAh) which would give me 6v to run a 328p chip.

Only connecting batteries in parallel will increase mAh.
mAh will stay the same when connecting them in series.
Leo..

I was wondering if this would be the right approach?

NO.

Voltage dividers are ok for signals or analog input but not for power input voltage reduction.

You would be better to use 3 batteries and use the external dc barreljack as the input.

Also, I would not choose a 2032 coin cell to run a microprocessor , even if it is only an ATMega328.
Find some other battery, like litheum-ion.
Two 18650 batteries in series is 2*3.7=7.4V which is ideal input voltage for a 5V regulator.
So get a 5V regulator, two 18650 batteries , a dual series battery holder, one or two TP4056 lithium-ion battery chargers and then your set.

An ATmega328 can run for years on a single CR2032 cell, if you do it right. Excellent low power tutorial here.

To get usable volume from the piezo buzzers, you probably need higher voltage, but they could be powered separately.

raschemmel:
NO.

Voltage dividers are ok for signals or analog input but not for power input voltage reduction.

You would be better to use 3 batteries and use the external dc barreljack as the input.

Also, I would not choose a 2032 coin cell to run a microprocessor , even if it is only an ATMega328.
Find some other battery, like litheum-ion.
Two 18650 batteries in series is 2*3.7=7.4V which is ideal input voltage for a 5V regulator.
So get a 5V regulator, two 18650 batteries , a dual series battery holder, one or two TP4056 lithium-ion battery chargers and then your set.

Thanks very much for all the replies and I think will go with raschemmel's suggestion!

vaj4088:
It is going to be hard to get enough current through 1 megohm and 100 kohm resistors unless you are planning to put very many in parallel. Of course I don't know what you are planning because you did not explain. Remember that your Arduino acts similar to an additional resistor except that it is variable.

The voltage divider approach is not practical using batteries as an input.

A series diode (e.g. 1N4001) may be a more practical approach.

I may also give this a go just as an experiment. Thanks for the idea :slight_smile:

+1 for a single CR2032 if you really want to go low power. Use the internal 8 MHz clock; sleep most of the time; and indeed your project will be able to run for a really long time. Use an RTC with alarm for the countdown part; super low power use, accurate countdown and the MCU can sleep until it's time to sound the buzzer. Well, it doesn't even have to wake up for that, the alarm output of the RTC can activate the buzzer.

2x LiPo makes you lose LOTS of power. About 32% over the regulator - not counting its quiescent current (which can be much higher than what the ATmega328p uses while asleep).

wvmarle:
+1 for a single CR2032 if you really want to go low power. Use the internal 8 MHz clock; sleep most of the time; and indeed your project will be able to run for a really long time. Use an RTC with alarm for the countdown part; super low power use, accurate countdown and the MCU can sleep until it's time to sound the buzzer. Well, it doesn't even have to wake up for that, the alarm output of the RTC can activate the buzzer.

2x LiPo makes you lose LOTS of power. About 32% over the regulator - not counting its quiescent current (which can be much higher than what the ATmega328p uses while asleep).

Thats a brilliant idea!!! Thanks so much for that :0) +1 Karma

For rough timing use WDT. For precise timing may be used Timer2 clocked by a watch crystal.

I have no idea how loud the buzzers need to be but driving a passive piezzo directly from ATMega pins may do decent sound even at 3V.

This is also interesting low power "tutorial".

Hi,
What is the application?

Thanks.. Tom.. :slight_smile:

ardandrew:
Hi,

I am thinking of using 2 x CR2032 (for space considerations and increased mAh) which would give me 6v to run a 328p chip. I was thinking of using a voltage divider (1MΩ and 100KΩ resistors) that would drop it to 5.45v.

The project would be for a simple countdown timer that would then trigger two 5V active piezo buzzers.

Likely to be beyond the ability of CR2032s to power. They are spec'd to produce a few milliamps only.

I was wondering if this would be the right approach?

You need to use a battery capable of producing enough current, perhaps 3x LR44 alkaline coin cells.
CR2032 lithium coin cells are designed to power calculators and other devices that require a few milliamps or less. They have a 10 year shelf life

Alkaline coin cells can supply many 10's of milliamps happily and 3 of them produce a nice safe 4.5V, no need to reduce voltage.

Thanks for any guidance in advance.

Smajdalf:
For rough timing use WDT. For precise timing may be used Timer2 clocked by a watch crystal.

I have no idea how loud the buzzers need to be but driving a passive piezzo directly from ATMega pins may do decent sound even at 3V.

This is also interesting low power "tutorial".

Thanks for all the replies. I have done a project using a passive 3v buzzer using a transistor to get the full 3v from the battery whereby the base is connected to the ATtiny85 pin and it is plenty loud (not sure the exact SPL) for what I need. This runs from one Cr2032 and has been running for 6 months or so but with light usage.

Well, heck, if you're really going to do that, why mess around with 9400uF when you can use ONE FARAD !

Because it will take ages to charge up from the puny coin cell, and use about 1mAh of capacity to do so, and has
a leakage current rating of upto 1.5mA ?

I actually can't find the leakage current on the datasheet

but I wouldn't charge it with a coin cell. (for that reason). The object was to charge it with a regular supply and then install it. You just need to remove the connection to the load until it's intalled and then install the
jumper, or you can just use an on off switch for that purpose.

What's wrong with the classic solution of running at 8MHz, direct from a rechargeable LiPo battery, eg, 18650? This gives you plenty of capacity (unlike a coin cell), very long standby life (unlike a supercap), a flat discharge curve (unlike a supercap), and you can get little recharging boards with a microusb connector for ease of recharging. LiPo battery packs are available in a wide range of sizes, so you can make the tradeoff between size and capacity.

And since the device will be sleeping most of the time anyway, the load is negligible so you can leave it connected while charging

ardandrew:
Hi,

I am thinking of using 2 x CR2032 (for space considerations and increased mAh) which would give me 6v to run a 328p chip. I was thinking of using a voltage divider (1MΩ and 100KΩ resistors) that would drop it to 5.45v.

The project would be for a simple countdown timer that would then trigger two 5V active piezo buzzers.

I was wondering if this would be the right approach?

Thanks for any guidance in advance.

OP, please read this link on Batteries. zBattery.com | Connecting-Batteries-in-Series-or-Parallel

Romonaga:
OP, please read this link on Batteries. zBattery.com | Connecting-Batteries-in-Series-or-Parallel

Thank you and will do! :slight_smile:

raschemmel:
Well, heck, if you're really going to do that, why mess around with 9400uF when you can use ONE FARAD !

I take the capacitor tutorial as a "basic research". Using such capacitor as a power source alone is an interesting exercise. Or a test setup - if you want to run years from a coin cell you must be able to run hours from a capacitor. There is little gain by increasing the capacity to 1F.

DrAzzy:
What's wrong with the classic solution of running at 8MHz, direct from a rechargeable LiPo battery, eg, 18650? ...

Cost, size, weight, complexity. 2032 primary cell has about 200mAh. Nice flat discharge curve. Rechargeable equivalent has about 1/3 or 1/4 of its capacity, has too high voltage when full, needs overdischarge protection etc. etc. If you aim for 10 years service life with the 200mAh cell the device must have less than 20uA average current. Which is plenty for many projects if done properly.