
Centre-Positive vs Centre-Negative DC Jacks: Protection Circuit Design

The connector cannot prevent it, so the circuit has to
Centre-positive is the dominant convention for 12 V and 24 V equipment. It is not universal. Centre-negative appears on guitar pedals, on some audio equipment, and on a scattering of industrial and telecom hardware. A barrel jack cannot tell the difference, and it will accept either.
That means reverse polarity protection is not a luxury feature. For any product where the supply is supplied by someone other than the factory, it is the difference between a support call and a destroyed board.
This article compares the practical protection topologies, with the cost and loss numbers that determine which one belongs in your design.
What reverse polarity actually damages
| Component | What happens with reversed input | Recoverable? |
|---|---|---|
| Electrolytic capacitor | Vents or fails short; may catch fire | No |
| Linear regulator | Junction reverse breakdown, often short input to output | No |
| Buck converter IC | Body diode conducts, input rail pulled down, IC destroyed | No |
| TVS diode | Conducts as a forward diode, may survive but does not help | Usually |
| Microcontroller | Latch-up through substrate diodes | Sometimes |
| Reverse protection MOSFET | Blocks; circuit simply does not run | Yes — that is the point |
The failure is usually not graceful. A reversed electrolytic can vent, and a reversed regulator often fails short, which then applies the reversed rail to everything downstream.
The four practical topologies
| Topology | How it works | Voltage loss | Part cost |
|---|---|---|---|
| Series Schottky diode | Blocks reverse current | 0.3–0.5 V, varies with current | Lowest |
| Series P-channel MOSFET | Body diode blocks; channel conducts forward | RDS(on) × I, typically tens of millivolts | Low |
| Ideal diode controller plus MOSFET | Actively drives the FET to minimise drop | Very low, similar to the MOSFET alone | Moderate |
| Crowbar (SCR plus fuse) | Shorts the input and blows the fuse on reversal | None in normal operation | Low, but one-shot |
Each has a different failure profile, which matters more than the cost difference.
Loss comparison at realistic currents
| Topology | Drop at 1 A | Drop at 3 A | Dissipation at 3 A | Efficiency impact at 12 V |
|---|---|---|---|---|
| Schottky, 0.4 V | 0.4 V | 0.45 V | 1.35 W | 3.8% |
| MOSFET, 20 mΩ | 0.02 V | 0.06 V | 0.18 W | 0.5% |
| MOSFET, 50 mΩ | 0.05 V | 0.15 V | 0.45 W | 1.3% |
| Ideal diode | ~0.02 V | ~0.05 V | ~0.15 W | 0.4% |
| Crowbar | 0 | 0 | 0 (normal operation) | 0% |
The Schottky’s 1.35 W at 3 A is not just an efficiency number. It is heat inside the enclosure, which then has to be removed, and it may be more than the total loss budget for a small product.
At 0.5 A the same Schottky dissipates around 0.2 W, which is acceptable. The choice therefore depends on current, not on principle.
Schottky: when it is still the right answer
| Condition | Schottky acceptable? | Reason |
|---|---|---|
| Load below 0.5 A | Yes | Loss is small and the part is cheap |
| Input rail well above the regulator’s minimum | Yes | The drop does not matter |
| Cost is the dominant constraint | Yes, with the trade-off accepted | Simplest possible protection |
| Load above 2 A | No | Heat and efficiency |
| Battery-powered with tight efficiency targets | No | Every milliwatt counts |
| Input close to the regulator dropout | No | The drop may push the regulator out of regulation |
The P-channel MOSFET version, and its traps
The standard reverse-polarity FET circuit is simple: a P-channel device in the high side, source toward the load, gate pulled to ground through a resistor. With correct polarity the body diode conducts briefly until the channel turns on, then the channel carries the current with low drop. With reversed polarity the body diode blocks and the gate is not driven.
Three details determine whether it works.
| Detail | Correct | If wrong |
|---|---|---|
| Source and drain orientation | Body diode must oppose reverse current | Reversed device conducts the wrong way and provides no protection |
| Gate pull-down reference | To the source, so VGS is defined | Floating gate; the FET may not turn on |
| Gate-source clamp | Zener or resistor divider if the rail exceeds VGS rating | Gate oxide damaged at high input voltage |
The second row is more subtle than it looks. The gate must be pulled below the source by enough to turn the FET on, and the pull-down has to survive the full input voltage. A resistor from gate to ground works, but at 24 V it dissipates continuously and requires a divider to stay inside the gate rating.
Choosing the MOSFET
| Parameter | Requirement | Why |
|---|---|---|
| VDS rating | At least 2× the maximum input voltage | Input transients and load dump |
| VGS rating | Above the highest gate drive voltage | Gate oxide failure |
| RDS(on) at the actual gate voltage | Low enough that dissipation is acceptable | The datasheet value at 10 V may be much better than at 4.5 V |
| Body diode rating | Must survive the inrush current | The diode conducts until the channel turns on |
| Thermal resistance | Considered for continuous load | Small packages have limited dissipation |
The third row is where designs go wrong. A FET specified at VGS = 10 V may have three times the resistance at VGS = 4.5 V, and a 3.3 V gate drive may barely turn it on at all. Check the curve at your actual drive voltage.
Inrush and the body diode
At power-up, the FET’s body diode conducts before the channel is enhanced. The load capacitance charges through the diode, which has a forward drop and a current rating.
| Load capacitance | Inrush through the body diode | Concern |
|---|---|---|
| Small (tens of µF) | Brief, low energy | None |
| Moderate (hundreds of µF) | Higher, longer | Diode heating; check the rating |
| Large (thousands of µF) | High energy | May exceed the diode’s surge rating; add soft-start |
A soft-start circuit — a resistor bypassed by the FET after a delay, or a dedicated controller — limits the inrush and protects the diode. It is worth adding when the downstream capacitance is large.
Where the crowbar makes sense
A crowbar does nothing in normal operation, so it has zero loss. On reversed input it short-circuits the supply and blows a fuse, which then has to be replaced.
| Property | Implication |
|---|---|
| No steady-state loss | Attractive for high current |
| One-shot protection | Requires a serviceable fuse |
| Depends on the supply surviving the short | A stiff supply may be damaged |
| Common in automotive and industrial | Accepted practice where service is available |
For a sealed consumer product with no serviceable fuse, a crowbar is a poor choice. For equipment where a technician will replace a fuse, it is efficient and effective.
Decision table
| Product profile | Recommended |
|---|---|
| Low current, cost sensitive, sealed | Series Schottky |
| Moderate to high current, sealed | P-channel MOSFET |
| High current, efficiency critical | Ideal diode controller |
| Industrial, serviceable fuse | Crowbar plus fuse |
| Supply shipped with the product only | Schottky for insurance; the risk is lower but not zero |
FAQ
Is reverse polarity protection really necessary? If the supply is supplied by anyone other than your factory, yes. A barrel jack accepts centre-negative plugs without complaint.
Can a TVS diode provide reverse polarity protection? Not reliably. It conducts in the forward direction and does not block sustained reverse current.
Why not just use a Schottky diode? Because at 2 A it drops about 0.45 V and dissipates roughly 1 W inside the enclosure. That is often more thermal cost than the part saves.
How do I choose the MOSFET’s RDS(on)? From the curve at your actual gate drive voltage, not from the headline figure measured at 10 V.
What happens to the FET at power-up? The body diode conducts until the channel turns on. Check the diode’s surge rating against your inrush current.
Is a crowbar suitable for a sealed product? No, because it requires a serviceable fuse. It suits equipment where a technician can replace one.






