Roughly one in three enquiries that reach our engineering desk arrives with the power rating fixed and the voltage still open: "we need a 750W geared BLDC motor — should it be 24V or 48V?" It is a good question, and most of the answers online treat it as a vague trade-off between cost and performance. It is not vague. The decision comes down to one equation, and everything downstream — cable gauge, connector class, fuse rating, battery pack topology, controller temperature, and whether you need extra shock protection to get the machine certified — falls out of it.
Quick Answer
- Same power, half the current. A 48V system draws exactly half the amps of a 24V system for the same shaft output. Because resistive loss goes with current squared, wiring and connector losses drop to one quarter.
- Under ~300W → 24V. Currents are already low, and 24V is usually already present in the machine's control cabinet.
- Over ~500W → 48V. Above 500W the 24V peak current (75A+ during acceleration) starts dictating expensive cable, connectors and fuses.
- The motor is not more efficient at 48V. Same frame, same efficiency (88–90% at 750W). All the gains are outside the motor.
- 48V is the practical ceiling. A full 48V pack sits at ~54.7V, still under the 60V DC SELV limit (IEC 61140). 72V crosses it and triggers a much more expensive compliance path.

The One Equation That Decides Everything
Electrical input power is voltage times current. Rearranged for the current a motor pulls from the bus:
I = Pshaft ÷ (Vbus × η)
where η is the combined motor-plus-controller efficiency — realistically 0.82–0.88 for an industrial BLDC drive at rated load. We use 0.85 throughout this article. Double the bus voltage and the current halves. That is the whole story, and every practical consequence below is a second-order effect of it.
The reason it matters so much is that almost every loss outside the motor is resistive, and resistive loss is I²R. Halve the current and you quarter the loss in the battery interconnects, the fuse, the contactor, the connector pins, the DC cable and the controller's low-side MOSFETs. If you need help converting a mechanical requirement into a shaft power figure first, our motor torque and power calculation guide walks through the formulas.
Current Draw: 24V vs 48V by Power Rating
This is the table we send customers before we send a quotation, because it usually settles the argument on its own.
| Shaft power | 24V continuous | 24V peak (2.5×) | 48V continuous | 48V peak (2.5×) |
|---|---|---|---|---|
| 100 W | 4.9 A | 12 A | 2.5 A | 6 A |
| 200 W | 9.8 A | 25 A | 4.9 A | 12 A |
| 400 W | 19.6 A | 49 A | 9.8 A | 25 A |
| 750 W | 36.8 A | 92 A | 18.4 A | 46 A |
| 1,000 W | 49.0 A | 123 A | 24.5 A | 61 A |
| 1,500 W | 73.5 A | 184 A | 36.8 A | 92 A |
| 2,000 W | 98.0 A | 245 A | 49.0 A | 123 A |
| 3,000 W | 147 A | 368 A | 73.5 A | 184 A |
Calculated at η = 0.85. Peak column assumes a 2.5× acceleration/grade-climb overload, which is typical for AGVs, tuggers and machinery movers. Figures in red are where standard 50–80A connectors and fuses stop being practical.
Look at the 2,000W row. At 24V you are designing a 98A continuous, 245A peak DC system — that is welding-cable territory, a 250A contactor, and busbar-style terminals. At 48V the same machine is a 49A / 123A system that a standard 10 AWG harness and an SB120 connector handles comfortably. This is exactly why every motor in our 750W–3kW high-power range is offered in 48V and 72V only. Below 750W we offer 12V, 24V, 36V and 48V.
Looking for a reliable BLDC motor supplier? Get a sample motor shipped in 7 days.
Request a Quote →Cable Sizing and Copper Cost
Cable is the first place the voltage decision shows up in your bill of materials. The table below gives continuous DC ampacity for 105°C silicone-insulated single conductors in free air — the normal choice for battery and motor leads on mobile equipment — along with resistance at 20°C.
| AWG | Cross-section | Continuous DC current | Resistance | Copper mass per metre |
|---|---|---|---|---|
| 16 AWG | 1.31 mm² | 18 A | 13.2 mΩ/m | 12 g |
| 14 AWG | 2.08 mm² | 25 A | 8.29 mΩ/m | 19 g |
| 12 AWG | 3.31 mm² | 35 A | 5.21 mΩ/m | 30 g |
| 10 AWG | 5.26 mm² | 50 A | 3.28 mΩ/m | 47 g |
| 8 AWG | 8.37 mm² | 70 A | 2.06 mΩ/m | 75 g |
| 6 AWG | 13.3 mm² | 95 A | 1.30 mΩ/m | 119 g |
| 4 AWG | 21.2 mm² | 125 A | 0.82 mΩ/m | 190 g |
Ampacity figures are for single conductors in free air with 105°C insulation. Derate 20–30% for bundled harnesses or conduit runs — see IEC 60364-5-52 Annex B for the correction factors that apply to your installation.
Worked example: 750W drive, 3-metre battery run
A 3 m battery-to-controller run is 6 m of conductor once you count the return path. Here is what that costs you at each voltage:
| Design | Current | Cable | Loop resistance | Voltage drop | Power lost in cable |
|---|---|---|---|---|---|
| 24V, minimum gauge | 36.8 A | 10 AWG | 19.7 mΩ | 0.72 V (3.0%) | 26.7 W |
| 48V, minimum gauge | 18.4 A | 14 AWG | 49.7 mΩ | 0.91 V (1.9%) | 16.8 W |
| 48V, same 10 AWG cable | 18.4 A | 10 AWG | 19.7 mΩ | 0.36 V (0.8%) | 6.7 W |
Loop resistance = 6 m × resistance per metre. Voltage drop = I × R. Cable loss = I² × R.
The third row is the one to internalise. Run the same cable at 48V instead of 24V and you lose 6.7 W instead of 26.7 W — a quarter, exactly as the I²R relationship predicts. On a battery-powered AGV, 20 W of recovered cable loss on a 750 W drive is roughly 2.7% more runtime per charge, for free.
The copper itself: 6 m of 10 AWG is about 285 g; 6 m of 14 AWG is about 112 g. A typical dual-drive AGV has four to six power runs, so the 48V design usually removes 0.5–1 kg of copper from the chassis and drops the connector down one class — from an SB120 to an SB50, for instance. On a 3,000 W machinery mover the gap is much larger.
Not sure which voltage fits your machine? Send us your payload, duty cycle and battery constraints and our application engineers will run the current and cable numbers for both options. Start with the drive selection tool →
Battery Pack Design: 8S vs 15S
Bus voltage dictates cell count, and cell count dictates BMS cost, charger choice and pack geometry. Same energy, very different pack.
| Parameter | 24V system | 48V system |
|---|---|---|
| LFP configuration | 8S (25.6V nominal) | 15S (48.0V nominal) |
| LFP full-charge voltage | 29.2 V | 54.8 V |
| NMC configuration | 7S (25.2V nominal) | 13S (46.8V nominal) |
| NMC full-charge voltage | 29.4 V | 54.6 V |
| Lead-acid equivalent | 2 × 12V blocks | 4 × 12V blocks |
| Capacity for a 3 kWh pack | 125 Ah | 62.5 Ah |
| BMS continuous rating needed (750W drive) | 60–100 A | 30–60 A |
| Cell interconnect current | High — thick nickel/copper busbars | Half — standard nickel strip often adequate |
| Charger for 4-hour recharge | 24V / 32 A | 48V / 16 A |
Note that both fully charged packs — 29.2 V and 54.8 V — stay under 60 V DC. That matters, and we come back to it below. The practical takeaway is that the 48V pack uses the same cells and the same energy, but the current-carrying hardware inside it (BMS FETs, busbars, main fuse, contactor) is one size smaller and correspondingly cheaper. In our experience a 3 kWh 48V pack lands 8–15% below the equivalent 24V pack once the BMS and interconnects are priced in.
Controller and MOSFET Selection
Controller design rules of thumb, both of which our engineering team applies when specifying BLDC controllers and drivers:
- 24V bus → 55–75V MOSFETs. You need roughly 2× bus voltage of headroom to absorb regenerative spikes and switching ringing.
- 48V bus → 100–150V MOSFETs. Same rule, applied to a bus that also reaches 54.8V at full charge.
Higher-voltage MOSFETs have higher on-resistance for a given die area — a typical 60V TO-220 device sits around 3–5 mΩ, a 100V device around 8–12 mΩ. That sounds like it should cancel the advantage. It does not, because conduction loss is I²×RDS(on) and the current term is squared:
| 750W drive | Phase current | MOSFET RDS(on) | Conduction loss per device |
|---|---|---|---|
| 24V controller | 36.8 A | 4 mΩ (60V class) | 5.4 W |
| 48V controller | 18.4 A | 10 mΩ (100V class) | 3.4 W |
Simplified: conduction loss only, one device at full current. Switching loss scales with V×I×fsw and is roughly a wash between the two — the doubled voltage is cancelled by the halved current.
In bench testing on our own controller boards, the 48V version of a 750 W drive runs 8–15°C cooler on the heatsink than the 24V version at identical shaft output. That thermal margin is worth more than it sounds: MOSFET RDS(on) rises roughly 0.4%/°C, and electrolytic capacitor life halves for every 10°C, so a cooler controller is a longer-lived controller. If you are still choosing a controller topology, our BLDC motor controller selection guide covers sensored vs sensorless, protection features and CAN/RS485 options.
What Actually Changes Inside the Motor
This is the part most articles get wrong, so it is worth stating plainly: a 24V and a 48V motor of the same frame size and power rating are mechanically identical. Same stator lamination stack, same rotor, same magnets, same bearings, same housing. The only difference is the winding.
The 48V winding uses roughly twice the number of turns per slot at roughly half the wire cross-sectional area. The consequences:
| Parameter | 24V winding | 48V winding | Practical effect |
|---|---|---|---|
| Turns per slot | N | ≈ 2N | Winding change only |
| Wire cross-section | A | ≈ A/2 | Same slot fill |
| Torque constant Kt | K | ≈ 2K | Twice the torque per amp |
| Speed constant Kv | k rpm/V | ≈ k/2 rpm/V | Same no-load speed on its own bus |
| Phase resistance | R | ≈ 4R | Same I²R loss at half the current |
| Rated torque | Identical | Frame and magnets unchanged | |
| Rated speed | Identical | Kv × V is the same product | |
| Motor efficiency | Identical (88–90% at 750W) | Copper loss unchanged | |
So if a supplier tells you their 48V motor is "more efficient" than their 24V motor of the same rating, ask to see both efficiency curves. In a properly designed pair, they overlap. The system-level gain is real; the motor-level gain is marketing. This is also why rewinding a motor for a different voltage is a routine change in our factory rather than a redesign — it is one of the most common requests in our custom OEM motor development projects, and it does not affect tooling or lead time.
The 60V Line: Why 48V Is the Practical Ceiling
There is a hard regulatory boundary at 60 V DC that shapes almost every industrial low-voltage drive design:
- IEC 61140 / IEC 60364-4-41 set the SELV (safety extra-low voltage) limit at 60 V DC ripple-free. Below it, protection against electric shock is inherently satisfied by the voltage level itself.
- ISO 6469-3 (electrically propelled road vehicles) defines Voltage Class A as ≤30 V AC / ≤60 V DC. Class B systems require additional protection, insulation monitoring and marked cabling.
- UL 1004-1 and UL 2271 (rotating machines, and batteries for light electric vehicles) apply progressively stricter construction requirements above the extra-low-voltage band.
A 48V nominal lithium pack peaks at 54.6–54.8 V fully charged — comfortably inside the limit with about 5 V of margin. A 72V nominal pack (20S NMC) charges to 84 V and lands squarely in Class B: orange high-voltage cabling, higher creepage and clearance inside the controller housing, service-disconnect requirements, and a slower and more expensive certification path.
That single threshold is why 48V has become the default for AGVs, electric tuggers, cleaning robots and mobile industrial equipment worldwide. We do build 72V motors for customers who need 3 kW+ per axle and have accepted the compliance work — but if 48V will do the job, 48V is the cheaper machine to certify.
Where 12V, 36V and 72V Fit
| Bus voltage | Typical power band | Where it makes sense | Main drawback |
|---|---|---|---|
| 12V | 15–150 W | Automotive-derived equipment, small pumps and fans sharing a vehicle battery, portable devices | Current gets impractical above ~150W (12A+) |
| 24V | 30–750 W | Machine-mounted drives sharing the PLC/control supply, conveyors, automatic doors, lab and medical equipment | Peak current above 500W forces heavy cable |
| 36V | 200–1,000 W | Battery tools and light mobility platforms where an existing 10S pack is standard | Non-standard in industrial cabinets; limited off-the-shelf accessories |
| 48V | 250–3,000 W | AGVs, electric tuggers, PV cleaning robots, machinery movers, warehouse equipment | Needs its own pack; not a typical control-cabinet voltage |
| 72V | 1,500 W and up | Heavy tugs, 30–60 t machinery movers, high-speed traction | Crosses the 60V DC line — Class B compliance burden |
System Cost: A Complete 750W Drive Compared
Component-level FOB Ningbo pricing at 100-unit order quantity, for a complete drive: geared BLDC motor, controller, DC harness, connectors, fuse and BMS share. Motor pricing is consistent with our BLDC motor price guide.
| Item | 24V build | 48V build | Difference |
|---|---|---|---|
| 750W planetary geared BLDC motor | $120–280 | $120–280 | None — winding change only |
| Controller | $45–90 | $50–100 | +$5–10 (higher-voltage FETs) |
| DC harness (6 m) | $14–22 (10 AWG) | $6–10 (14 AWG) | −$8–12 |
| Main connector | $9–16 (120A class) | $4–7 (50A class) | −$5–9 |
| Fuse + holder | $5–9 (80A) | $3–5 (40A) | −$2–4 |
| BMS (3 kWh pack, per drive share) | $38–65 (100A) | $25–45 (60A) | −$13–20 |
| Total drive system | $231–482 | $208–447 | −$23–35 |
Indicative FOB Ningbo pricing at 100 units, September 2026. Actual quotations depend on gearbox ratio, brake and encoder options, and certification requirements.
The saving is modest per unit — roughly 7–10% of drive system cost — but it comes with 20 W less waste heat, a cooler controller, less copper in the chassis and easier cable routing. On a 2–3 kW machine the gap widens substantially, because 24V at those powers pushes you into 4 AWG cable and 250A contactors.
Decision Matrix by Application
| Application | Typical power | Recommended | Why |
|---|---|---|---|
| Automatic door / gate operator | 60–300 W | 24V | Shares the control-cabinet supply; low current already |
| Light conveyor / roller drive | 100–400 W | 24V | Mains-fed PSU, short cable runs, 24V is the plant standard |
| Medical pump / centrifuge | 50–250 W | 24V | SELV by design, standard medical PSU voltage |
| Lab / analytical instrument | 30–200 W | 24V | Low current, integrates with existing supply |
| AGV / electric tugger (1–5 t) | 400–1,500 W | 48V | Battery powered, high peak current, long harness runs |
| PV cleaning robot | 200–800 W | 48V | Battery weight matters; 48V halves interconnect mass |
| Machinery mover (5–30 t) | 1,500–3,000 W | 48V | 24V is impractical at 100A+ continuous |
| Heavy tug / mover (30–60 t) | 3,000 W+ | 48V or 72V | 72V only if you accept ISO 6469-3 Class B compliance |
| Industrial pump / fan (mains-fed) | 200–1,500 W | 48V | Rectified supply; 48V minimises rectifier and cable current |
Five Mistakes We See in Customer Specifications
These come up often enough in incoming enquiries that they are worth listing explicitly.
- Sizing cable for continuous current and forgetting the acceleration peak. A 750W 24V AGV drive is a 37A machine on paper and a 92A machine for the two seconds it takes to get moving. Undersized cable does not fail immediately — it sags the bus, trips the controller's undervoltage protection mid-launch, and the customer reports "the motor stutters when starting."
- Specifying a 24V motor because the machine already has a 24V control supply. A 24V PLC supply is usually rated 5–10A. It will not run a 750W drive. If the drive needs its own supply anyway, the shared-voltage argument disappears and 48V becomes the better choice.
- Assuming a 24V motor can be "upgraded" to 48V by changing the power supply. Speed is proportional to voltage. A 24V motor on a 48V bus runs at double rated speed, well past its bearing and rotor-retention design point. Order the correct winding — it costs nothing extra and has no lead-time impact.
- Choosing 72V to "future-proof" a machine that draws 1 kW. At 1 kW, 48V draws 24.5A — completely comfortable. Going to 72V buys nothing and moves you across the 60V DC line into ISO 6469-3 Class B, adding real certification cost.
- Ignoring voltage drop over long harness runs. On a 6 m tugger chassis the round-trip conductor length can reach 12 m. At 24V/37A with 10 AWG that is a 1.4 V drop — 6% of the bus, and it gets worse as the battery discharges. Keep total drop under 3% of nominal bus voltage as a design rule.

How to Decide in Four Steps
- Establish shaft power. From payload, gradient, wheel diameter and target speed — see our BLDC motor sizing guide for AGVs or the torque and power calculation guide.
- Calculate current at both voltages. I = P ÷ (V × 0.85), then multiply by your worst-case overload factor (2–3× for traction, 1.5× for pumps and fans).
- Check the peak against connector and fuse availability. If the 24V peak exceeds about 80A, the connector and fuse hardware becomes the deciding factor and you should move to 48V.
- Confirm the full-charge pack voltage stays under 60V DC. 15S LFP (54.8V) and 13S NMC (54.6V) both pass. Anything above that changes your compliance path.
Frequently Asked Questions
Is a 48V BLDC motor better than a 24V BLDC motor?
For the same shaft power a 48V motor draws exactly half the current, so resistive loss in cables, connectors, fuses and controller MOSFETs drops to one quarter. Above roughly 500W, 48V is almost always the better system choice. Below 200W the current is already low and 24V costs less. The motor itself is not more efficient at 48V — a 750W motor is 88–90% efficient either way. The savings are all outside the motor.
How much current does a 750W BLDC motor draw at 24V vs 48V?
At 85% combined motor-plus-controller efficiency, about 36.8A at 24V and 18.4A at 48V. Peak during acceleration is typically 2–3× continuous, so the 24V version can pull 75–110A momentarily. That peak is what forces 6 AWG cable, a 150A fuse and high-current connectors on 24V designs.
Can I run a 24V BLDC motor on 48V?
No. No-load speed is proportional to supply voltage, so a 24V motor on 48V spins at roughly twice rated speed. Bearing life drops sharply and the rotor sees higher eddy heating. If it also sees rated torque it runs at double rated power and overheats within minutes. A 48V motor on 24V is safe but halves both speed and power. Always match the winding to the bus.
Why do most AGVs use 48V rather than 72V?
60V DC is the SELV ceiling in IEC 61140 and IEC 60364-4-41, and the Voltage Class A limit in ISO 6469-3. A 48V lithium pack peaks at about 54.6–54.8V, safely under it. A 72V pack charges to 84V and crosses into Class B, requiring additional shock protection, marked cabling, greater creepage and clearance, and a more expensive certification path.
What cable size do I need for a 24V vs 48V BLDC drive?
For 750W: the 24V version at 36.8A needs at least 10 AWG (5.26 mm²) continuous, or 6–8 AWG if sized for a 2.5× peak. The 48V version at 18.4A needs only 14 AWG (2.08 mm²) continuous, or 10 AWG for peaks. On a 3 m run that is 285g of copper per conductor versus 112g. Across a full AGV harness the 48V design typically saves 0.5–1 kg of copper.
Does a 48V motor produce more torque than a 24V version?
No. Same frame size and power rating means the same torque. The 48V winding has roughly twice the turns at half the wire cross-section, so Kt (N·m/A) roughly doubles and Kv (rpm/V) roughly halves. Slot area and copper fill are unchanged, so copper loss at rated torque is identical. You get the same torque and speed at half the current.
Which voltage should I pick for a 200W motor?
24V. At 200W the current is only 9.8A, which 16–14 AWG cable and a standard 20A connector handle easily. Moving to 48V saves about 5A but forces a 13S/15S pack instead of a 7S/8S pack, or a boost converter. Below roughly 300W the wiring savings do not pay for the extra pack complexity, and 24V is already present in most machine control cabinets.
Do you stock both 24V and 48V windings, or is it a custom order?
Both are standard. Our geared and cylindrical BLDC motors up to 750W are built in 12V, 24V, 36V and 48V; the 750W–3kW high-power range is 48V and 72V. Changing the winding voltage on an existing frame is a routine production change with no tooling cost and no lead-time impact. MOQ for a standard winding is 10 pieces; non-standard voltages (such as 60V or 96V) need 100 pieces.
Still Deciding Between 24V and 48V?
Send us your payload, duty cycle, cable run length and battery constraints. Our engineers will run the current, cable and cost numbers for both options and quote the matching motor, gearbox and controller.
