Quick Answer
A 12V camera on copper two-wire loses voltage over distance. Current makes a round trip, so a 150 ft run is 300 ft of wire in the circuit. Keep the camera at 10.8 V or higher. Design for under 5% drop. Once you push past 10% (or under 10.8 V), you get reboots, no power-up, or weak night IR.
Locked example: a 12V bullet drawing 0.5 A over 150 ft of 18-gauge copper drops about 8%, landing near 11.0 V at the camera. That’s caution territory, not a pass. Step the same run up to 14-gauge and you’re comfortably under 5%.
Four real fixes when the numbers don’t work: thicker (lower-gauge) copper, a shorter run, the power supply closer to the camera, or PoE. Guessing gauge by “it should be fine” is how jobs come back.
Why 12V drops over distance
Copper isn’t a perfect conductor. Every foot has resistance. Ohm’s law is simple: voltage drop = current x resistance. Longer wire, thinner wire, or more amps means more drop.
The part people miss is the round trip. Power leaves the supply on one conductor and comes back on the other, so a 150 ft cable is 300 ft of copper in the circuit. Count one-way only and your math is half wrong, and the camera is the one that pays for it.
12V systems have little headroom. The calculator judges the result against the minimum voltage you enter for your camera — there is no default. 10.8 V, the 10% drop point, is a common floor, but check your camera spec; many need more. Treat that floor as the limit, not the target. Below it the regulator inside can’t hold steady power, and that gets worse at night when the IR LEDs switch on and current jumps. (PoE is a different animal: higher voltage, different cable, different rules. Out of scope here.)
What low voltage looks like
You rarely get a clean “low voltage” error. You get behavior that looks like a bad camera or bad cable:
- Night reboots. The camera is fine all day, then cycles after dark. Daytime current is low enough to stay above the floor. At dusk the IR fires, current rises, drop increases, and the camera browns out. It reboots, the IR drops during boot, voltage recovers, IR comes back on, and it browns out again. A camera that loops at night and behaves at noon points at the power path. Start there before you blame the DVR.
- Dim or short-range night vision. The camera stays up but the illuminators can’t run at full output. Night-vision range may shrink or become uneven. People blame the camera and swap it. Same wire, same problem.
- Won’t power up at all. On a long, thin run the camera never comes up. No LED, no signal, looks like a dead camera. Bring it inside on a short cable and it works fine. That’s your answer.
- Intermittent video or dropouts can come from marginal voltage too, though those have other causes. Rule out power first. It’s cheap to check.
If a camera is solid on a short jumper at the DVR and flaky at the far end, the supply voltage at the head end doesn’t tell you anything. You have to measure what the camera actually sees.
Field test, in order:
- Meter the power leads at the camera, not at the supply.
- Leave the camera connected and drawing current.
- Test with the IR active if you can — that’s the real load.
- Compare against a short, known-good connection.
- Check power before you replace the camera or the DVR.
An unloaded reading is misleading. With no current flowing there’s no drop, so you’ll read 12 V at the far end of a run that fails the second it’s under load.
Use the calculator
Don’t freehand this if you don’t have to. Plug in run length, wire gauge, and the camera’s real current draw into the 12V Camera Voltage Drop Calculator on GuardSourceHQ. Use the camera’s maximum draw from the label or spec sheet, not idle, and remember IR pushes it higher.
The calculator does the round-trip math and shows both the percent drop and the volts at the camera. Your job is still to pick the target: stay under 5% when you can, and treat anything heading toward 10% or 10.8 V as a redesign, not a hope.
How to size the wire
Two numbers, and they do different jobs.
Under 5% is the design target. That leaves margin for a supply that sags a little under load, connections that corrode over a few winters, cold weather, and the day someone adds a second camera to that circuit. Nothing in a real install stays exactly as drawn.
10% is the ceiling, not a goal. Around 10.8 V you’re sitting on the failure floor with nothing left over. It might pass on the bench. It won’t hold up once anything changes.
The locked example shows how narrow the window is. 0.5 A, 150 ft, 18-gauge: about 8% drop, roughly 11.0 V at the camera. That’s above the 10.8 V floor, so it isn’t an automatic no, but there’s no headroom for a cold night, a weak supply, or an aging connection. The same run in 14-gauge comes in under 5%. The wire costs a little more and is stiffer to pull. That’s the whole downside, and it beats going back into a hard-access run later.
The judgment call: if the calculator lands you between 5% and 10%, you can proceed only if the run is easy to service. If it’s a hard-access location, size it conservatively now.
Max distance by gauge
One-way run lengths for a single camera drawing 0.5 A at 12 V on solid copper, sized to stay under 5% drop. Planning numbers, floored.
| Gauge (AWG) | Max one-way run at 5% (0.5 A) |
|---|---|
| 22 | 37 ft |
| 20 | 59 ft |
| 18 | 93 ft |
| 16 | 149 ft |
| 14 | 237 ft |
| 12 | 377 ft |
| 10 | 600 ft |
Read it right:
- These are one-way distances. The current travels that distance twice through the complete circuit — the table already accounts for it.
- The 10% ceiling is roughly double each figure. 18-gauge can technically limp to about 187 ft at 0.5 A. Don’t design there. That’s the same territory the 150 ft, 18-gauge example sits in.
- More current, less distance. A 1 A camera gets about half these numbers. Run your own figures for the load you actually have.
- These assume solid copper. Copper-clad aluminum and unknown wire change everything (see mistakes below).
When to switch to PoE
Stay on 12V copper when the runs are short, the cameras are already 12V analog, and thicker wire or a closer supply solves it without drama.
Switch to PoE when the copper math stops making sense: 12- or 14-gauge for one camera, runs long enough that the wire costs more than the camera, several cameras fed from one spot, or the numbers keep pushing you toward ugly workarounds. PoE sends power at a higher voltage over network cable and steps it down at the camera. Higher voltage means less current for the same power, and less current means far less drop. It’s not the same cable with a different brick.
Use PoE when the calculator shows poor margin, impractically heavy wire, or difficult shared-power runs. If you’re extending an existing 12V system, fixing the copper or moving the supply usually wins on cost and labor.
Multi-camera and shared runs
Shared power is where good installs go bad.
Current adds. Three cameras at 0.5 A on one shared trunk is 1.5 A through that section. Triple the current, triple the drop on that segment. A trunk that’s fine for one camera fails for three. Two 0.5 A cameras on a single 18-gauge pair is already 1 A, which cuts the usable 18-gauge distance to about 46 ft at the 5% target.
Size the shared portion for the total current, then add each branch’s own drop on top. The camera at the far end of a chain sees the worst of it, because it eats the drop from the whole trunk plus its own spur. If any camera on the circuit is going to fail, it’s that one.
Two layouts beat daisy-chaining, and they’re not the same thing. A true home run gives each camera its own dedicated pair all the way back to the supply — nothing shared, every leg calculated on its own. Local distribution runs one properly sized shared feeder to a protected distribution point near the cameras, then short individual spurs from there; it saves wire but the feeder carries the combined current, so size it for the total. Either one beats daisy-chained power, which makes every camera depend on the load and length ahead of it. The calculator’s Advanced mode models only the simple case — one shared pair with identical loads. Mixed loads and branched circuits still need to be checked leg by leg.
Common mistakes
- Counting one-way length. Round trip, every time.
- Measuring at the supply. The supply can read fine at the head end. Meter at the camera, under load, IR on.
- Measuring with the camera disconnected. No current, no drop. You’ll read 12 V on a wire that can’t carry it.
- Using idle current. Night is the load case. Size for the maximum draw with IR on.
- “18-gauge is what I have” on a 150 ft run at half an amp. That’s the 8% / 11.0 V caution case, not a comfortable design.
- Undersized shared legs for multiple cameras.
- Aluminum or unknown CCA wire. Copper-clad aluminum has higher resistance for the same gauge, so every number moves the wrong way. If the packaging doesn’t clearly say copper, treat it as a smaller gauge, or don’t use it. Thin loses.
- Designing right against 10.8 V with nothing left for connections, temperature, or a future camera that draws more.
- The thin power pair in a premade combo cable. Video-and-power cable often pairs decent coax with light power conductors. Check the actual gauge of the power pair against your length.
- Swapping the camera first. If the problem is voltage, the new camera fails the same way. Meter first, buy second.
FAQ
How much drop is too much? Under 5% is the practical target. Around 10%, or below 10.8 V at the camera, is failure territory: reboots, no power-up, dim IR.
Why did it work in daylight and fail at night? IR LEDs raise current. More current, more drop. The night load is the real test.
How do I find my camera’s current draw? The label or spec sheet, in amps or milliamps (500 mA is 0.5 A). If it only lists watts, divide by 12. Use the peak figure when a range is given.
Can I just turn the supply up to 13.5 or 14 V? Do not raise the supply voltage unless the camera, every device sharing that supply, and the regulated power supply explicitly permit it. A higher supply can damage a short-run device even while helping the far camera. Fix the wire size, distance, or supply location first.
Is 16-gauge good enough? Depends on length and current. It’s better than 18. It’s not magic. Run the numbers for your longest, hungriest camera.
What about leftover alarm or thermostat wire? Only if the gauge and copper content are real and the length-and-current math works. A lot of leftover wire is thin, and thin loses.
Do I need the calculator if I have this article? The article teaches the decision. The calculator does the arithmetic for your exact length, gauge, and current so you’re not redoing Ohm’s law on every job. Use both.
Does the video coax affect voltage drop? No. The power pair and the video conductor are separate circuits. Drop is a function of the power conductors, their length, and the current. Power the camera correctly first.
Thirty years in the trades teaches the same lesson: a lot of cameras replaced as “bad” were never tested under load at the installed location. The camera may be fine — test the installed voltage under load before you replace it. Often the wire was undersized, the supply was too far away, or nobody counted the return leg.
Count the round trip, size for under 5%, respect 10.8 V as a floor, and treat that 150 ft, 18-gauge, 0.5 A case as a warning, not a target. Thicker wire, shorter run, closer supply, or PoE when copper stops making sense. Measure at the camera when something’s flaky. Do that and you stop replacing cameras that were never the problem.