## Voltage Divider for 3.3v?

10 posts / 0 new
Author
Message

I have a board that primarily runs on 5v. I would like to add an accelerometer that runs off of 3.3v. I figured using a voltage divider would be the simplest way of getting 3.3v.

Is this a sound way of getting 3.3v or should I break down and put a small voltage regulator on it?

Thanks.

## Attachment(s):

Did you realize that accelerometer has some current too?
For example MMA7260Q could have almost 800uA - and your regulator has unloaded current 1.66mA
It could cause insignificant voltage drop

Computers don't make errors - What they do they do on purpose.

The accelerometer consumes 400uA.

Would reducing the resistor values allow for more current to be made available to the accelerometer?

Lots of approaches available.
For a one-off project for yourself you have many options.

The divider will work if the accel chip draws a small current. Note that the current drawn by the accel chip will change the current being drawn through the upper resistor, and hence the nodal voltage.

As the accel chip current draw varies, so will the nodal voltage. This may or may not matter to the readings you are obtaining.

At the least put a 0.1 uF and 1 uF cap from the node to ground.

The smaller the resistors used, the less impact the (small) amount of current drawn by the accel chip will have on the nodal voltage, but the higher the current draw for the overall project.

Another option is to use a diode or two to give you a "more" regulated voltage drop, or a resistor / zener diode combination.

You also could stack several LEDs in series: V+ to R to LED1 to LED2 to LED3 to ground. Select the LEDs so that their total forward voltage is what you are looking for.

Best option is to shift the entire circuit to 3 V if possible. A second, regulated, 3V supply would be option #2. The other options are a (poor) third choice.

JC

With a fixed load the divider can be optimized so very little current goes through the lower resistor. But bandgap voltage drops are more stable. 1.7 volts is the drop across most red LEDs http://www.theledlight.com/LED101.html

Your decision depends on the power consumption, annual production quantity and other factors.

If it is a research project, give the divider a 50mA current and your accelerometer won't even notice a 1% voltage drop.

But if this is an energy constrained app, use DC/DC.

And Mr. Zener? Somewhere in between resistor divider and linear regulator.

No RSTDISBL, no fun!

Hey thanks guys for the info but I think I'm going to break down an put on a small voltage regulator. I found one in a sot 23-5 package, its a Microchip MCP1801.

LDO regulator is just a P-FET in resistance/linear mode. Didn't know you needed a small package.

Wow, That cleared my thought. I'am also having problem with voltage divider for 3.3v, thank god I found your forum.