Quick answer: an OLED makes a voltage readout easy to read, but it does not make it accurate or safe. Treat the analog input as a 10-bit, reference-limited ADC, size the divider so the tap always stays inside the reference window, and protect the pin separately from the display.
What the display does and does not change
On a classic UNO R3, analogRead() uses a 10-bit ADC, and the board table identifies A0 through A5 as the usable analog inputs. The reference also notes roughly 100 microseconds per analog read on ATmega-based boards, which is useful for reasoning about sampling but is not a bandwidth guarantee for a finished application. The OLED only renders the number your code computed, so the screen is a readability feature and nothing more. Accuracy is set by the divider and the ADC reference, which makes this a compact visual readout for low-voltage DC rather than a calibrated instrument. If you would rather compare several inputs on a character LCD, a four-channel LCD voltmeter reference covers that layout; if display-driver or module identity is your concern, the companion OLED voltmeter display-verification guide addresses it.
Divider orientation and reference uncertainty
A resistive divider scales the input down so the tap sits inside the ADC input window. The tap fraction is the lower resistor divided by the sum of both resistors. Orientation matters: swap the two resistors and the tap can rise toward the source instead of staying low, potentially exceeding the reference and driving the pin outside its intended range. Check the orientation against your own board before applying power.
The single-ended conversion result is referenced to the selected VREF, so the reference you choose sets both full scale and the value of every count. Assume a reference value instead of knowing it, and the whole scale shifts. Build an accuracy budget that lists divider tolerance, reference tolerance, ADC quantisation, and source impedance as separate terms rather than blaming one of them.
Why a screen is not input protection
Nothing about an OLED limits what reaches the pin. Absolute maximum ratings in the device datasheet are stress limits, not guaranteed functional operating conditions, so a pin can sit inside its absolute maximum rating and still be outside guaranteed ADC operation. Treat over-voltage, reverse polarity, and transients as upstream design problems. The datasheet ADC input discussion recommends source impedance of roughly 10 kilohms or less, because the sample-and-hold needs time to charge; a divider that looks correct can still be too high-impedance to sample well.
Comparison: what each choice actually affects
| Factor | What it affects | Practical check |
|---|---|---|
| 10-bit ADC | Quantisation step | Is one count small enough at your full scale? |
| VREF choice | Full scale and every LSB | Known and stable, or assumed? |
| Divider ratio | Scaling factor | Orientation correct? Tap always in range? |
| Divider tolerance | Gain error | Is worst-case spread acceptable? |
| Source impedance | Settling and sample accuracy | Near the 10 kilohm guidance? |
| Display | Readability only | Does the screen change accuracy? No. |
Ordered evaluation process
- Define the highest voltage you will ever apply, and confirm it is isolated low-voltage DC.
- Pick a reference you can defend, because full scale follows it.
- Size the divider so the maximum tap stays inside the reference window with margin.
- Verify the layout: lower resistor to ground, upper resistor to the source, tap to the analog pin.
- Check that source impedance is near the datasheet ADC input guidance.
- Plan protection ahead of the pin, such as series limiting and clamping to a defined rail. Do not rely on the MCU alone for protective shutdown.
- Convert counts using the chosen reference and state the measurement limits beside the readout.
- Cross-check one reading against an independent known reference before trusting the scale.
Reader questions
Does a higher-resolution display make the reading more accurate?
No. Pixel count is not measurement resolution. The number of counts is fixed by the 10-bit ADC, and the scale is set by the reference, regardless of how finely the screen can draw.
Can I use the ATmega328P datasheet for my specific board?
Use it as a reference for ADC behaviour, but note that it is the 7810D automotive edition. Check the exact device on your board rather than assuming every edition behaves identically.
Why does a reading drift over time?
Reference uncertainty and divider tolerance are the first suspects, followed by source impedance if the sample-and-hold cannot charge in time. Work through them in that order on your own hardware.
Is it fine to measure 50 V this way?
No. This guide makes no high-voltage hookup recommendation. Restrict use to isolated low-voltage DC within verified input limits, with the pin protected.
Documentation
- Arduino analogRead() reference (official reference-en source)
- Microchip ATmega328P datasheet, 7810D automotive edition
Documentation-based reference only: no bench testing, calibration, or code execution was performed, so divider values, display pinout, SPI operation, and any accuracy figure still need to be confirmed for a given build.
