float temp = 0; float tempPin = 0; void setup() { Serial.begin(9600); // put your setup code here, to run once: } void loop() { int sensorValue = analogRead(A0); float voltage = sensorValue * (5.0 / 1023.0); Serial.println(voltage); // put your main code here, to run repeatedly: } This is the code for the analogRead code in arduino mega. I used a potentiometer where I connected A0 pin at the second pin of the potentiometer, connected 5V to first pin of potentiometer and connected the last pin to ground. When I go to the serial plotter, I observe that the measured voltage is not a fixed value, it oscillate irreularly.

GPT-5.4

Answer: I maintain my position: the oscillation is mainly normal ADC/input noise plus possible wiring or potentiometer contact issues, and the best fixes are to verify the wiper wiring and smooth the readings with averaging or simple filtering.

The other answers reinforce rather than overturn my original view. Claude Opus 4 added the most useful technical detail: the Mega’s 10-bit ADC naturally jitters by around a few LSB, potentiometer wiper contact can add noise, and a small capacitor plus averaging is an effective practical fix. That strengthens my original explanation, but it does not fundamentally change it. I agree with Gemini that loose connections and a worn or dirty potentiometer are common hardware causes. Those are consistent with what I already said about wiring and contact quality. I am less persuaded by the idea that the missing delay is itself a root cause. A delay can make the plot look calmer and reduce how visibly noisy the output appears, but it does not remove the underlying analog noise; averaging/filtering is the more direct fix. The code itself is basically correct for reading A0 with the default analog reference on an Arduino Mega: - `analogRead(A0)` returns 0 to 1023 - `voltage = sensorValue * (5.0 / 1023.0)` is the standard conversion if the reference is truly 5 V So the likely explanation is not a code bug but measurement reality. In practice, I would suggest these steps, in order: 1. Confirm A0 is connected to the center pin (wiper) of the potentiometer. 2. Make sure the two outer potentiometer pins go to 5V and GND with a solid common ground. 3. Check for loose breadboard/jumper connections. 4. Expect small fluctuations of a few ADC counts as normal. 5. Add software averaging, and optionally a 0.1 uF capacitor from A0 to GND for extra smoothing. 6. If high accuracy matters, measure the actual 5V rail or use a more stable analog reference. So I am keeping my original position, just with stronger support: irregular oscillation in this setup is usually expected analog noise or hardware quality effects, not a fundamental error in the posted sketch.

Gemini 3.1 Pro

Answer: The irregular voltage oscillation in your serial plotter is primarily caused by inherent ADC noise and high-speed, unfiltered sampling, which can be exacerbated by loose connections or a dirty potentiometer.

I am maintaining my original position because the other models largely validate my core diagnosis: analog inputs are inherently susceptible to electrical interference, and continuously reading them at maximum speed without filtering exposes every micro-fluctuation. GPT-5.4 accurately points out that the 5V USB power rail itself is often noisy, which directly skews your voltage math since the formula assumes a perfect 5.0V reference. Claude Opus 4 makes an excellent point about the inherent ±2 LSB jitter of the Mega's 10-bit ADC, meaning minor fluctuations are unavoidable at the hardware level. While my initial suggestion to add a simple `delay()` will help throttle the serial output and prevent buffer overload, both models correctly highlight that implementing a software averaging loop is a far superior method for smoothing out the data. Additionally, their shared suggestion of adding a 0.1µF decoupling capacitor between A0 and Ground is the gold-standard hardware fix. Ultimately, your base code is correct. The jitter is a physical reality of analog electronics. To get a flat line on your plotter, ensure your wiring is snug, add a decoupling capacitor if you have one, and implement the averaging code block suggested by Claude.

Claude Opus 4

Answer: The irregular oscillations you're observing are completely normal and are caused by a combination of ADC noise, breadboard contact resistance, potentiometer wiper noise, and the lack of any filtering in your code.

After reviewing the other models' responses, I'm maintaining my original position as all three of us are fundamentally in agreement — this is normal behavior caused by multiple overlapping factors, not a code bug. **Points of Agreement** All three responses correctly identify ADC noise, wiring quality, potentiometer wiper issues, and the noisy USB 5V rail as contributing factors. We all recommend averaging multiple readings and optionally adding a small decoupling capacitor as solutions. **GPT-5.4's Useful Addition** GPT-5.4 makes an excellent point about verifying which pin of the potentiometer is actually connected to A0. The user says "second pin," which should be the center wiper pin, but if it's accidentally connected to an outer pin, the behavior would be different. This is good troubleshooting advice. GPT-5.4 also rightly notes that the assumed 5.0V reference may not match the actual supply voltage, though this would cause a scaling error rather than oscillation. **Gemini's Point About delay()** Gemini suggests adding a `delay()` to reduce sampling speed, which would make the serial plotter appear smoother. However, I'd clarify that this doesn't actually reduce noise — it just reduces how many noisy samples you see per second. The underlying readings are still noisy. That said, at 9600 baud, printing too fast without a delay can cause serial buffer issues, so a small delay is reasonable advice for practical reasons. **My Recommended Fixes (Prioritized)** 1. **Check wiring**: Ensure solid connections and that A0 is on the center (wiper) pin of the potentiometer. 2. **Add a decoupling capacitor**: Place a 0.1µF ceramic capacitor between A0 and GND to filter high-frequency noise. 3. **Software averaging**: Read 10-20 samples and average them before printing. This is the single most effective software fix. 4. **Add a small delay**: Something like `delay(50)` to avoid flooding the serial buffer and to make the plotter more readable. 5. **Use the actual reference voltage**: Measure your 5V pin with a multimeter and use that value instead of assuming exactly 5.0V. The bottom line remains: your code is correct, your wiring sounds correct, and some oscillation is inherent to the hardware. The solutions above will dramatically reduce it.

Final Summary

The AI Roundtable confirms the user's code is functionally sound, attributing the fluctuations to the physical realities of analog-to-digital conversion. Claude Opus 4 and GPT-5.4 highlight that the Mega's 10-bit ADC naturally jitters by several counts, while Gemini 3.1 Pro notes that reading at maximum speed without a delay makes every microscopic ripple visible. The group collectively recommends software averaging and hardware decoupling to achieve a stable reading.

All 3 models agreed