A basic divider has one resistor from the input voltage to the output node, and another resistor from the output node to ground.
Vin -> Rtop -> Vout -> Rbottom -> groundThe ideal calculation
If Rtop is 10 kohm, Rbottom is 10 kohm, and Vin is 5 V, the output is half the input:
Vout = 5 V x 10000 / (10000 + 10000)
Vout = 2.5 VThat is the version people learn first. It is correct if the output node is measured by something with extremely high input resistance, so almost no current leaves the divider.
Loading is the catch
The moment you connect a load to Vout, that load sits in parallel with Rbottom. The lower resistor is no longer just Rbottom; it is Rbottom in parallel with the load. That new lower resistance changes the output voltage.
This is why voltage dividers are good for signals and references, but poor as general-purpose power supplies. A divider that gives 3.3 V with no load may collapse when a module tries to draw current from it.
Choosing resistor values
Very low resistor values waste current. Very high resistor values become sensitive to leakage, noise, ADC sampling, and input bias currents. The right values depend on the job.
| Divider style | Benefit | Cost |
|---|---|---|
| Low resistance | Less affected by load and noise. | Wastes more current and power. |
| High resistance | Low current draw. | More affected by leakage, noise, and ADC sampling. |
| Buffered output | Can drive a load more reliably. | Needs an op-amp or buffer circuit. |
ADC inputs
Voltage dividers are often used to scale a battery or sensor voltage down to a microcontroller ADC range. This can work very well, but ADC inputs are not always infinitely gentle. Many ADCs briefly charge an internal sample capacitor. If the divider resistance is too high, the ADC reading may be slow to settle or inaccurate.
The datasheet may specify a maximum source impedance or recommended external circuit. If you need a high-value divider to save battery, a capacitor from the ADC input to ground can help provide charge during sampling, but it also slows changes. Firmware can also switch the divider on only when measuring, using a MOSFET or GPIO-controlled path if the design allows it.
Tolerance and error
Divider accuracy depends on the ratio of the resistors, not just their individual values. Two 1 percent resistors give a much better starting point than two 5 percent resistors. For precision measurement, matched resistor networks or calibration may be useful.
Temperature can also shift the ratio. If both resistors are the same technology and placed together, they may track reasonably well. If they are different parts in different thermal conditions, the ratio may drift.
When a divider is the right tool
- Scaling a voltage for an ADC input.
- Creating a reference threshold for a comparator.
- Biasing an analogue signal around a midpoint.
- Setting feedback voltage in a regulator, following the datasheet.
- Creating a simple logic-level sense input, when current and thresholds are checked.
When not to use a divider
- Powering a device or module that draws changing current.
- Driving a low-impedance load directly.
- Replacing a voltage regulator.
- Handling high voltage without considering power, spacing, and safety.
- Expecting precision without considering tolerance and ADC behaviour.
Practical rule
A divider creates a voltage ratio. It does not create a stiff power rail. If the next circuit takes meaningful current from the divider node, include that current in the calculation or buffer the node.
A practical checklist
- What output voltage is needed, and over what input range?
- What load or input is connected to Vout?
- Is divider current acceptable for the power budget?
- Will tolerance and temperature create too much error?
- Does the ADC or comparator input need lower source impedance?
- Is a buffer, regulator, or switched divider more appropriate?