0
\$\begingroup\$

I'm getting back into electronics as a hobby and doing some refreshers having been away from it for a long time. I understand Ohm's law and voltage dividing, however, I'm not sure I understand what the proper ranges of resistors to use are when dividing voltages.

For example, if I have a 5 V source and I want two 2.5 V divisions, I can use two 10 kΩ resistors to get that.

But I can ALSO use two 100 kΩ resistors, or two 1 MΩ resistors. What I'm struggling with is: what is the determining factor or calculation that I should be using to determine which pair is the best way to do this? Is it based on current requirements?

On a side/similar note, as a project I'm going to build a digital thermometer. I have a TM346 sensor and will be converting its output to digital and use a microprocessor to convert to a temperature, but to use the A/D converter (ADC0804LCN) and get the full dynamic range of conversion, I need to supply this chip some reference voltages.

What's the best way to provide accurate voltages to this chip? I figure voltage dividing is not the best way to handle it. Aa a low reference I need to supply 322 mV and as a high reference I need to supply 628 mV. I'm assuming voltage dividing is NOT the best way to accurately get these values.

\$\endgroup\$
0

2 Answers 2

3
\$\begingroup\$

When you build a simple resistor voltage divider, you're also creating a path to ground for whatever is sourcing current. For example, if you're just dividing 5 volts in half with two 10k resistors, you've also essentially added a 20k load to the supply, adding 250 µA. That might not seem like much, but if you're building a device that's meant to be battery-operated and last for years, that matters a lot.

On the other hand, if you have too little current, for example, to charge the gate of a capacitive input, it may be inadequate because it takes too long to measure, etc.

As you suspected, the range of values should be based on current: enough to drive whatever is downstream effectively, but not so much that it burdens something upstream.

\$\endgroup\$
2
\$\begingroup\$

As already said, one variable is how much current you want to use/"waste" on the voltage divider. That suggests going with high values. But high values will introduce more noise and be more sensitive to electromagnetic interference, since the voltage can be varied by less induced current. There is no correct value as such, it depends on your application and sometimes doesn't matter much at all. FWIW, I often use 100k when I don't much care about the finer things and just want a bog standard reference voltage buffered by an op amp.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.