Servo Torque for RC Planes: How Much Do You Need?

A 9 g micro servo flies a foam park flyer all day, and the same servo on a fast sport plane will not hold the aileron at full throttle. The difference is the load on the control surface, and that load grows with surface area and with the square of the airspeed. This guide works out how much torque your plane needs: the published formula, two worked examples, and which servo to pick from the answer.

The short answer: a slow park flyer aileron needs well under 1 kg-cm on paper, a sport plane aileron at 70 mph needs about 1.6 kg-cm before any margin, and giant scale planes need 20+ kg-cm.

How much torque does an RC plane need?

Start with the plane, not the servo. Published RC guides agree on the shape of the answer even though their exact numbers differ, so treat the bands below as starting points and run the calculation below for your own surfaces.

Plane Servo class Typical torque
Indoor and micro Sub-micro, 2 to 5 g 0.2 to 0.8 kg-cm
Park flyers up to about 1 m Micro 9 g 1.5 to 2.5 kg-cm
Sport planes 1 to 1.5 m Mini or standard 3 to 8 kg-cm
Planes 1.5 to 2 m Standard digital, metal gears 8 to 15 kg-cm
Giant scale and 3D High torque, high voltage 20 to 40+ kg-cm

Source: SwellRC, “How to Choose Servos for an RC Plane”, which labels this a rough guide.

A second guide sorts by weight and gives lower numbers: a micro plane under 1 lb needs about 8 oz-in (0.6 kg-cm), a 1 to 2 lb park flyer about 15 oz-in (1.1 kg-cm), and a 3 to 5 lb trainer about 40 oz-in (3 kg-cm). Source: ServoMotorStore. The two guides disagree on the park flyer band, and that disagreement is the point of this guide: a table can only guess at your surface size and top speed.

What does a servo torque rating actually mean?

Torque is the turning force a servo can apply, measured in kilogram-centimetres (kg-cm) or ounce-inches (oz-in). A rating of 2 kg-cm means the servo can hold 2 kg at a 1 cm arm, by the definition of the unit. Move the load out to a 2 cm arm and the same servo holds 1 kg. That is the arithmetic in our servo torque math guide: torque equals force times arm length. Convert between units with 1 kg-cm equal to about 13.9 oz-in. Source: SwellRC RC servo guide.

Three facts change how you read a rating. Torque is a stall figure. The listing number is the most the servo can hold before it stops moving, so a servo working at its stall rating in flight is already failing. Torque and speed are both rated at a voltage. The same servo gives more of both at 6.0V than at 4.8V: the MG90D is rated 2.1 kg-cm at 4.8V and 2.4 kg-cm at 6.6V. Compare two servos only at the same voltage. Holding torque is not the listing figure. A digital servo holds position harder than an analog servo of the same stall rating.

How do you calculate the torque a control surface needs?

The load on a control surface comes from the air. The published method for model planes is the Chuck Gadd servo torque formula, used by RC builders for over 15 years. The formula returns torque in oz-in:

Torque (oz-in) = 8.5E-6 x (C squared x V squared x L x sin(S1) x tan(S1) / tan(S2))

C is the control surface chord in cm, V is the speed in mph, L is the control surface length in cm, S1 is the maximum control surface deflection in degrees, and S2 is the maximum servo deflection in degrees. Source: Chuck Gadd servo torque calculator, mnbigbirds.com. The Scale Aviators paper “Determining Servo Torque Requirements” verifies the formula against its own derivation.

The formula’s notes state its assumptions: sea level, moderate temperature, zero hinge offset, frictionless linkages, and mass-balanced surfaces. The result is completely theoretical, and the maximum required torque may occur at less than maximum throw. Use it as a sizing estimate and add the safety margin below.

Worked example 1: park flyer aileron

A foam park flyer aileron, 30 cm long with a 4 cm chord, flying at 40 mph, deflected 30 degrees, servo travelling 60 degrees: 8.5E-6 x (16 x 1,600 x 30 x 0.5 x 0.577 / 1.732) = 1.09 oz-in, and 1.09 / 13.9 = 0.08 kg-cm.

The answer is 0.08 kg-cm of air load only. A 9 g micro servo rated at 1.6 kg-cm has roughly twenty times that in hand, which is why a park flyer flies fine on small servos. For a slow foam plane the real sizing question is the margin for gusts, friction, and hinge stiffness, not the air load.

Worked example 2: sport plane aileron

A sport plane aileron, 50 cm long with an 8 cm chord, flying at 70 mph, deflected 30 degrees, servo travelling 60 degrees: 8.5E-6 x (64 x 4,900 x 50 x 0.5 x 0.577 / 1.732) = 22.2 oz-in, and 22.2 / 13.9 = 1.6 kg-cm.

Doubling the chord and raising the speed from 40 to 70 mph multiplied the load by twenty.

Cross-check and safety margin

A second published method estimates the same load from the hinge moment. For the sport aileron above it gives about 0.50 N-m at the hinge, or 5.1 kg-cm, and the linkage ratio (30 over 60 degrees) halves that at the servo, giving about 2.5 kg-cm. Source: Micro Servo Motor hinge moment guide.

The two methods agree within a factor of two: 1.6 kg-cm from the Gadd formula, 2.5 kg-cm from the hinge moment method. That spread is why you never size to the exact calculated number. The hinge moment guide recommends a 50 percent safety factor: the sport aileron at 1.6 kg-cm becomes 2.4 kg-cm with margin, and the cross-check at 2.5 kg-cm becomes about 3.8 kg-cm, so a servo rated at 3.5 kg-cm covers both methods with margin.

What changes the answer?

Speed matters most. Load rises with the square of the airspeed, so doubling the speed quadruples the load. Size for the fastest speed the plane will reach, not the cruise speed.

Surface size matters next. Load rises with the square of the chord and linearly with the length.

Deflection matters more than most builders expect. The formula carries both sin and tan of the deflection angle, so the load grows faster than the angle itself. A 3D plane with 45 degree throws loads its servos far harder than a sport plane with 30 degree throws.

Linkage geometry changes the torque at the servo. A long servo arm gives more travel but less force at the surface. Torque equals force times arm length, so halving the arm length doubles the force at the horn. Set the arm and horn lengths first, then run the formula with the servo travel that linkage actually gives.

Speed vs torque: which surface needs which?

Ailerons want speed first, then torque. A slow aileron servo makes the plane feel mushy in turns, and published guidance puts aileron speed at 0.10 sec per 60 degrees or faster for aerobatic planes. Run the formula on the aileron first: it is usually the largest moving surface on a sport plane.

Elevators want a balance, with torque as the floor. An elevator failure is unrecoverable in a way an aileron failure often is not, so the hinge moment guide advises going one size up from the aileron servo for the elevator on any plane over 1.5 m.

Rudders want torque first. The rudder works against the airflow in knife edge and in crosswind landings, and it does not need to move fast. Published guidance recommends a minimum of 20 to 30 oz-in (about 1.4 to 2.2 kg-cm) for the rudder on a mid-sized plane, rising with size. Source: ServoMotorStore.

Flaps want torque and reliability, not speed. Flaps sit deflected for the whole landing approach, so holding torque and metal gears matter more than transit speed. Size them with the formula at full flap deflection, often 45 degrees or more, where the load is more than double the load at 30 degrees.

Digital or analog servos for a plane?

An analog servo receives its position signal about 50 times per second. A digital servo updates its motor far more often internally, typically 200 to 300 times per second or more, which gives faster response, more precise centring, and stronger holding power.

Analog suits trainers, slow scale models, gliders, and any plane where the surfaces move slowly and the loads are light. Analog servos cost less and draw less current.

Digital suits aerobatic planes, 3D foamies, fast sport planes, and any plane where the surfaces must hold against a load. Digital holding power is the difference between a surface that stays where you put it in a high-G pullout and one that blows back toward neutral.

Current draw is the hidden cost of digital. Digital servos draw more current under load, and four moving together can exceed a small BEC. The micro servo guide warns that digital servos can draw 500 mA to 1 A under load, and the TowerPro MG996R datasheet states a stall current of 1400 mA.

Which servo should you buy from your answer?

Run the formula, add the 50 percent margin, and match the result to the smallest servo that clears it. All four servos below were verified live on 2026-11-21.

Your answer after margin Servo pick Torque Weight Price
Under 0.5 kg-cm SA-37 sub-micro 0.5 kg-cm 3.7 g $10.49
0.5 to 1.6 kg-cm DSM-44 digital micro 1.6 kg-cm at 6.0V 5.8 g $22.49
1.6 to 2.4 kg-cm MG90D digital micro 2.1 kg-cm at 4.8V, 2.4 at 6.6V 13 g $14.49
2.4 to 3.5 kg-cm MG92B digital micro 3.5 kg-cm at 6.0V 13.8 g $17.49

Torque figures are the stall torque from each listing. Re-check the price at order time.

Under 0.5 kg-cm: SA-37. The SA-37 3.7g Metal Gear Sub-Micro Servo (ID 12046) at $10.49 is the lightest pick in the store at 3.7 g, with metal gears and 0.5 kg-cm of torque, for indoor foam planes where weight is the whole job.

0.5 to 1.6 kg-cm: DSM-44. The DSM-44 Digital Metal Gear Micro Servo (ID 12065) at $22.49 gives 1.6 kg-cm at 6.0V and moves 60 degrees in 0.07 sec at 6.0V, the fastest figure in this guide, at 5.8 g with metal gears and digital control. It is the pick when the plane needs speed: 3D foamies and quick ailerons. Run it at 6.0V to get the listed figures.

1.6 to 2.4 kg-cm: MG90D. The MG90D Digital Metal Gear Micro Servo (ID 12068) at $14.49 gives 2.1 kg-cm at 4.8V and 2.4 kg-cm at 6.6V, moves in 0.10 sec per 60 degrees at 4.8V and 0.08 sec at 6.0V, and weighs 13 g with metal gears, digital control, and a double ball bearing. It is the tough pick for larger park flyers and small sport planes.

2.4 to 3.5 kg-cm: MG92B. The Towerpro MG92B Digital Metal Gear Micro Servo (ID 16605) at $17.49 gives 3.5 kg-cm at 6.0V, moves in 0.11 sec per 60 degrees at 6.0V, and weighs 13.8 g with metal gears and digital control. It is the strongest micro servo in the store per its listing.

Above 3.5 kg-cm: step out of micro. No micro servo in the store clears 3.5 kg-cm. The next step is a mini servo such as the 1250MG, which gives 3.5 kg-cm at 6.0V in a 25 g case, or a standard servo such as the MG996R. The TowerPro MG996R datasheet lists 9.4 kg-cm at 4.8V and 11 kg-cm at 6V, and the QRC listing itself states no torque figure.

What are the common mistakes?

Sizing by span alone. Two 1.2 m planes can need very different servos if one is a slow trainer and the other is a fast warbird with wide-chord ailerons. Surface size and top speed are the answer.

Sizing to the exact calculated number. The formula is an estimate that does not allow for hinge friction, pushrod friction, or gusts. Add the 50 percent margin first.

Using a long servo arm to get more travel. A long arm gives more surface travel per degree of servo rotation, but it costs force at the same rate. Set the arm length for the force you need, then set the travel in the radio.

Mixing different servos on paired surfaces. Two aileron servos of different speed or torque move at different rates, and the plane rolls unevenly. Use matched servos on every paired surface.

Running high voltage servos on a low voltage supply. A servo rated for 7.4V gives less torque and speed at 4.8V than its headline figure suggests.

Forgetting the current draw. Four digital servos at stall can pull several amps together, and brownouts from many servos moving together are a common hidden cause of crashes.

Buying the cheapest servo for a valuable plane. Very cheap servos from unknown brands have inconsistent quality control. A servo failure is one of the most common causes of crashes.

How do you fix soft controls, blowback, and flutter?

The surface blows back toward neutral at speed. Blowback is the surface deflecting less than commanded because the air load beats the servo. Fix it by fitting a stronger servo, shortening the servo arm, or reducing the deflection.

The controls feel soft or mushy at full deflection. Soft controls at speed usually mean the servo is near stall, the linkage flexes, or the pushrod bends under load. Check the pushrod first.

The surface flutters at speed. Flutter is a rapid vibration of a control surface at speed that can destroy it in seconds. Model Aviation names improper servo choice, wrong linkage geometry, incorrect hinging, unsealed hinge gaps, and slop as the causes. The fix is mechanical before it is a servo swap: eliminate play in the linkages, seal the hinge gaps, and tighten the hinges.

The servo buzzes and gets hot on the bench. A servo that buzzes at full travel is pushing against a mechanical stop. Reduce the end points so the arm never binds. A stalled servo draws high current: the MG996R datasheet figure is 1400 mA at stall.

The gears strip on a landing. Stripped gears come from too small a torque rating, linkage geometry that lets the servo move over centre and lock, or a sharp knock through the surface.

The plane rolls unevenly. Mismatched servos on the two ailerons, different arm lengths, or different end points. Match the servos and the travel in the radio.

Parts used in this guide

Prices were verified live on qualityrcparts.com on 2026-11-21. All four items are published in the store. Stock status is not exposed by the product read used for this draft; the micro-servos guide verified all four in stock the same day.

Shipped fast and securely from our secure facility in Queensbury, NY. QRC ships within the USA only.

Related guides

FAQ

How much torque do I need for an RC plane servo? From about 2 kg-cm for park flyers to 20+ kg-cm for giant scale, per the SwellRC plane guide FAQ. The exact number depends on your control surface size and top speed, so run the formula in this guide. Source: SwellRC plane guide FAQ.

How do you calculate servo torque for a control surface? Use the Chuck Gadd formula: torque in oz-in equals 8.5E-6 times chord squared times speed squared times length times sin and tan of the surface deflection, divided by tan of the servo deflection. Work it for your largest, fastest surface, then add a 50 percent margin. Source: mnbigbirds.com.

What happens if my servo torque is too low? The surface blows back toward neutral at speed, the controls feel soft, and in the worst case the surface flutters and the gears strip. Source: Micro Servo Motor guide.

Should I use a digital or analog servo for my RC plane? Digital servos offer more precise and rapid responses for high-performance RC planes, while analog servos are more budget-friendly and adequate for general use. Source: ServoMotorStore FAQ.

Can I use the same servo for all the surfaces on my plane? Not necessarily. Ailerons want speed, rudders want torque, elevators want a balance with holding power, and flaps want torque and reliability. Source: ServoMotorStore FAQ.

Why does my servo buzz? It is holding position under load or pushing against an end stop. Reduce the end points if it buzzes at full travel. Source: SwellRC RC servo guide FAQ.

Do I need metal gears on an RC plane? On the main surfaces of larger or faster planes, yes. Metal gears handle shock loads and last longer than plastic gears under load. Source: SwellRC plane guide FAQ.

Sources

Last updated: 2026-11-21

RC Servo Brackets and Mounts: How to Pick and Install Them

A servo bracket is the mount that holds a servo motor square to the frame so the output shaft stays lined up with the load it drives. Get the bracket right and the servo steers clean, the arm repeats its position, and the screws stay put. Get it wrong and the servo shifts, the horn binds, or the mount cracks.

What is a servo bracket, and why does it matter?

A servo bracket is a shaped mount, usually a U-shaped holder, a flat plate, or a two-piece pan-tilt frame, that bolts to a chassis, frame, robot arm, or project board and cradles the servo body. The servo’s mounting ears screw to the bracket, and the bracket screws to the frame. It holds the servo body rigid so the case cannot flex under load, and it sets the output shaft in line with the linkage it drives.

Without a solid mount the servo can shift under load: a loose servo shows up as sloppy steering or a wandering arm.

Which bracket type fits your job?

The five brackets in the store cover the four shapes most RC and robotics builds need.

| Bracket type | What it does | Store bracket | |—|—|—|—| | Flat install holder | Bolts a standard servo straight to a flat surface | CNC install holder, ID 13458, $23.99 | | U-shaped holder | Cradles the servo body in a rigid U frame | CNC U-shaped holder, ID 19665, $8.99 sale | | U-shaped 2-set | Two U-shaped holders in one pack | CNC 2-set, ID 19666, $14.99 sale | | Pan-tilt bracket | Pairs two standard servos for two-axis movement | Pan tilt 2-set, ID 22925, $20.99 | | Micro plate bracket | Flat acrylic plate with a cutout sized for a 9 g servo | SG90 acrylic 10-pack, ID 13465, $12.99 |

Flat install holder (ID 13458). It bolts flat to the chassis. CNC aluminum, hardware included, for S3003 and MG995 class servos.

U-shaped holder (ID 19665). The U cradles the servo body. CNC aluminum, 55 x 26 mm, adjustable plate, screws included, for S3003, MG995, and MG996R class servos.

U-shaped 2-set (ID 19666). The same U-shaped holder, two in one pack with M3 screws, nuts, washers, and hex socket screws. For two-servo builds or as a spare: at $14.99 on sale it costs less per bracket than two singles.

Pan-tilt bracket (ID 22925). Two brackets that work as a pair: one servo pans, the second tilts, building a 2-degree-of-freedom head with a ball bearing on the pan axis. Each set includes the multi-function bracket, the short U-type bracket, a ball bearing, metal horns, and the screw set. It fits standard 40 x 20 x 36 mm servos including MG995, MG996R, and S3003.

Micro plate bracket (ID 13465). A flat acrylic plate with a cutout sized for the SG90 body plus a mounting slot and holes, so the servo sits flush and bolts to a chassis, frame, or project board. Ten brackets with screws, nuts, and washers.

Which material should you pick: aluminum, acrylic, or plastic?

The bracket material decides how much load the mount can take before it flexes, cracks, or loosens.

| Material | Strength | Best for | |—|—|—|—|—| | CNC aluminum alloy | High, rigid, resists vibration | Standard servos, steering, robot arms, high-torque joints | | Acrylic plate | Low, stiff but brittle | SG90 micro servos, light indoor builds, bulk packs | | Injection-molded nylon or ABS | Medium, some flex | Micro pan-tilt kits, light camera heads |

CNC aluminum alloy is the right pick for any standard servo that works hard: steering, robot arm joints, a pan-tilt head carrying a camera. It resists vibration and keeps the servo aligned run after run.

Acrylic plate is the right pick for 9 g micro servos in light builds: a 9 g servo cannot load it hard. Acrylic cracks rather than bends, so tighten only until snug and never use threadlocker on it: Kpower’s guide warns it can crack the plastic.

Nylon and ABS sit between the two. Adafruit’s Mini Pan-Tilt Kit uses injection-molded nylon. For a standard servo, plastic is the wrong pick: Kpower’s guide describes a plastic pan-tilt bracket that cracked after 20 minutes on a six-axis arm, and the fix was an aluminum bracket.

Which bracket fits your servo size?

Standard and micro mounts are not interchangeable: a standard servo will not fit a micro bracket, and a micro servo rattles loose in a standard bracket. Measure your servo, then match it to the bracket listing.

| Servo class | Body size | Stall torque | Bracket to buy | |—|—|—|—|—| | Micro, SG90 9 g | 23 x 12.2 x 29 mm | 1.8 kg·cm at 4.8 V | SG90 acrylic 10-pack, ID 13465 | | Standard, S3003 / MG995 class | About 40 x 20 x 36 mm | Check the servo datasheet | CNC U-shaped holder, ID 19665, or install holder, ID 13458 | | Standard, MG996R | 40.7 x 19.7 x 42.9 mm | 9.4 kg·cm at 4.8 V, 11 kg·cm at 6 V | CNC U-shaped holder, ID 19665, or pan-tilt 2-set, ID 22925 |

Sources: SG90 figures are TowerPro’s; cheap SG90 clones vary, so measure your own servo. MG996R figures are TowerPro’s datasheet figures, mirrored by Cornell University. The 40 x 20 x 36 mm fit is the pan-tilt listing (ID 22925).

Measure before you order. The numbers that decide the fit are the servo body length, width, and height, and the mounting-hole pattern on the servo’s ears. Measure the body with calipers, then the hole spacing on each ear. The listing states the servo class it fits, so the class match is the check that matters. Kpower’s bracket guide gives the pattern rule for micro servos: most follow a standard pattern of two holes spaced 26 to 28 mm center to center, with M2 or M2.5 screws, most commonly 27 mm. Verify it on your own servo.

Check the flange thickness too. A metal-gear micro servo can share a body size with its plastic-gear twin but carry a thicker mounting flange. Protosupplies warns that MG90 metal-gear servos do not fit its SG90 pan-tilt kit because the flange is thicker. Test-fit a metal-gear micro before tightening anything.

How do you install a servo bracket?

The steps below work for every bracket in the store. Read the horn clearance section before you tighten anything.

Step 1: Dry-fit the servo. Set the servo into the bracket without screws. The body should sit square and flush, the ears flat on the mounting face, and the shaft pointing where the linkage needs it. If the servo rocks, the bracket is the wrong size.

Step 2: Center the servo, then fit the horn. Power the servo and drive it to center, 90 degrees on a standard 180-degree servo. Push the horn onto the splined shaft at the centered position, then fix it with the short screw that came with the servo. Use the short screw: Adafruit’s guide warns that the longer mounting screws packed with some servos will damage the servo if screwed in too far.

Step 3: Screw the servo to the bracket. Insert the screws through the servo’s mounting ears and tighten evenly, finger-tight plus a quarter turn and no more. Kpower’s guide caps micro-servo screw torque at about 0.2 N·m: more strips plastic threads or deforms the bracket. A drop of medium threadlocker on metal screw threads stops vibration loosening. Threadlocker only on metal-to-metal screws, never on the acrylic SG90 plate.

Step 4: Bolt the bracket to the frame. Make sure the bracket sits level and the servo shaft lines up with the linkage axis. Do not mount the bracket reversed: test the full range of motion before final fastening.

Step 5: Assemble a pan-tilt bracket as a pair. Mount the pan servo in the base bracket first, with its horn fitted at center, then attach the base to the horn so the pan axis turns the whole head. Then mount the tilt servo in the tilt bracket, fit its horn, and join the tilt bracket to the pan base on the pivot pin. If the tilt bracket cannot move through the full 180 degrees, take it off the horn, rotate it, and re-install it.

Step 6: Test the full travel, then check the screws after the first run. Power the servo and sweep it slowly from one end to the other. The horn must clear the bracket at both ends, the linkage must not bind, and the wires must not snag. If anything touches, stop and fix it before running the build: a servo stalled against its own bracket draws stall current and heats up fast. After the first hour of running, check every bracket screw again.

Servo horn clearance: the step everyone skips

Most bracket failures are horn failures: the horn hits the bracket at one end of the travel, the servo stalls against it, and the gears strip or the bracket cracks. Check clearance before you tighten the bracket to the frame.

Center the servo first, then fit the horn. A horn fitted off-center gives uneven travel: full range one way, binding the other.

Trim the horn if it has to fit inside the bracket. Adafruit’s Mini Pan-Tilt Kit guide shows exactly this: the outer horns are trimmed with a hobby knife or wire cutters so they fit inside the pan-tilt base, while the center horns stay unmodified.

Keep the tilt travel inside the bracket. Limit the tilt range so the top servo cannot crash into the bracket: the Thinking Robot SG90 guide limits tilt to 30 to 150 degrees.

Route the wires clear of the horn. The servo lead must not cross the horn’s sweep. Route it along the bracket base, away from the moving parts, with enough slack for the full travel.

Which bracket should you buy?

Standard servo in an RC plane, boat, or car: the CNC install holder (ID 13458), $23.99. Machined aluminum, hardware included, for S3003 and MG995 class servos.

Standard servo in a robot arm or upright build: the CNC U-shaped holder (ID 19665), $8.99 on sale, regular $10.99. The cheapest metal mount in the store. Its adjustable plate fits S3003, MG995, and MG996R class servos. Size 55 x 26 mm.

Two standard servos: the CNC 2-set (ID 19666), $14.99 on sale, regular $17.99. Two U-shaped holders with M3 screws, nuts, washers, and hex socket screws per set. It costs less per bracket than two singles, and the second bracket is a spare.

A camera head, sensor mount, or robot joint that needs two axes: the pan-tilt 2-set (ID 22925), $20.99. Two full pan-tilt sets, each with a ball bearing, metal horns, and hardware, for standard 40 x 20 x 36 mm servos including MG996R, MG995, and S3003.

SG90 micro servos: the SG90 acrylic 10-pack (ID 13465), $12.99. Ten plates with screws, nuts, and washers, sized for the SG90 body. One pack covers a whole smart car or small robot build.

If you are not sure which servo you need yet: read the servo chooser guides first, then come back. Buying the bracket first and the servo second is the order that causes most fit problems.

What are the common mistakes?

  • Wrong size class. A standard servo in a micro bracket, or a micro servo in a standard bracket. Match the servo class to the listing before ordering.
  • Over-tightening the screws. Finger-tight plus a quarter turn is enough. More strips plastic threads or warps the bracket. Kpower’s guide calls over-tightening the number one cause of bracket failure.
  • Threadlocker on plastic. Medium threadlocker belongs on metal-to-metal screws only. On acrylic or nylon it can crack the bracket.
  • Mounting the bracket reversed. The servo shaft ends up on the wrong side of the linkage axis. Test the range of motion before final fastening.
  • Fitting the horn off-center. The servo binds at one end of travel and stalls. Center the servo at 90 degrees first.

How do you fix a servo bracket that does not work?

The servo jitters or stalls in the bracket. Check for an obstruction first: a misaligned screw, a wire caught under the servo, or a horn touching the bracket. Then check power: a servo starved of current jitters even in a perfect bracket. A single MG996R can draw 1,400 mA at stall, per TowerPro’s datasheet, so a small board regulator or a USB port cannot feed it.

The screws keep loosening. Clean the screws, apply a drop of medium threadlocker, and re-tighten, but only on metal-to-metal joints. Kpower’s guide describes exactly this case: screws loosened after one hour of RC driving, fixed by cleaning, threadlocker, and a re-torque, with no loosening after 10 hours. In an acrylic plate, do not use threadlocker: check for cracks instead.

The bracket cracked. Replace it, and go up a material: acrylic to aluminum for a standard servo, or nylon to aluminum for a micro servo that works hard. A cracked bracket has already told you the load is too much for the material. Check the servo size match at the same time.

The servo will not fit the bracket. Do not force it. Measure the servo body and the bracket opening with calipers and compare the numbers. If the bracket is the wrong size class, order the right one. Shimming a too-small servo into a too-big bracket with strips of plastic is a stopgap, not a fix.

The horn hits the bracket at one end. Take the horn off, re-center the servo, and refit the horn one spline over. If it still hits, trim the horn or limit the travel in code so the servo never reaches the contact point.

Parts used in this guide

Prices and stock verified live on qualityrcparts.com on 2026-11-21 via the public Store API. All five are in stock. Servos are not included.

Shipped fast and securely from our secure facility in Queensbury, NY. QRC ships within the USA only.

Related guides

FAQ

What is a servo bracket? A servo bracket is a shaped mount, usually a U-shaped holder, a flat plate, or a two-piece pan-tilt frame, that bolts to a chassis or frame and cradles the servo body so the output shaft stays lined up with the load. Source: QRC servo-mounts category description and Kpower micro servo bracket guide.

How do you mount a servo to a bracket? Dry-fit the servo first, center it at 90 degrees, fit the horn with the short screw, screw the servo to the bracket finger-tight plus a quarter turn, then sweep the full travel to check clearance. Source: Kpower micro servo bracket guide and Kpower servo mounting guide.

What size bracket do I need for an SG90 servo? An SG90 needs a micro bracket sized for its 23 x 12.2 x 29 mm body; the acrylic SG90 10-pack (ID 13465) is cut for that body. A standard servo bracket will not hold it. Source: TowerPro SG90 datasheet figures and the QRC servo-mounts category description.

What size bracket do I need for an MG996R servo? An MG996R needs a standard bracket. Its body is 40.7 x 19.7 x 42.9 mm and it weighs 55 g, so it fits the standard brackets in the store: the CNC U-shaped holder (ID 19665), the CNC 2-set (ID 19666), and the pan-tilt 2-set (ID 22925). Source: TowerPro MG996R datasheet and the QRC pan-tilt bracket listing.

Can I use a standard servo bracket for a micro servo? No. A micro servo rattles loose in a standard bracket: an SG90 is 23 mm long against a standard servo at about 40 mm. Source: QRC servo-mounts category description.

Do servo brackets come with screws? The brackets in the store do: screws, nuts, and washers with the CNC brackets and the SG90 10-pack, plus ball bearings and metal horns with the pan-tilt set. Servos are not included. Source: QRC product listings for IDs 22925, 19666, 19665, 13465, and 13458.

Why does my servo buzz or jitter in its bracket? Check for an obstruction first: a horn, wire, or screw blocking travel. Then check power: a servo starved of current jitters even in a perfect bracket, and a single MG996R can draw 1,400 mA at stall. Source: Kpower micro servo bracket troubleshooting and TowerPro MG996R datasheet.

Sources

Last updated: 2026-11-21

RC Servo Sizes Chart: Dimensions for Every Class

An RC servo size class tells you roughly how big the case is, not exactly how strong the servo is. A standard servo case measures about 40 by 20 by 38 mm. A micro servo case measures about 23 by 12 by 29 mm. A sub-micro servo can weigh less than 4 g. The class name alone will not tell you whether the servo fits your bay, because the lines between classes are drawn in different places by different sources, and two servos in the same class can differ by several millimetres. Measure your bay, read the listing figures at your voltage, and match the numbers, not the label.

What are the RC servo size classes?

There is no single official standard for servo size classes. The hobby community uses five rough classes, plus a low profile variant of the standard size. Every source below draws the lines in a slightly different place, which is why this guide quotes the listing figures next to the class bands.

  • Sub-micro, also called nano. About 2 to 5 g. Indoor planes, 1/24 to 1/18 cars, micro helicopters. Source: SwellRC.
  • Micro. About 5 to 14 g in the store listings. Park flyers, foam planes, 1/18 cars. Source: RCMods and QRC listings.
  • Mini. About 20 to 30 g. 450 to 500 class helicopters, sport planes, small cars. Source: SwellRC and RCMods.
  • Standard. About 40 to 65 g. Most 1/10 and 1/8 cars, larger planes, boats. Source: EuroRC and RCMods.
  • Large, also called 1/5 scale or XL. About 120 to 200 g. 1/5 cars, giant scale planes, big boats. Source: EuroRC.

A low profile servo is a standard footprint servo with a shorter height, about 40 by 20 by 30 mm per EuroRC, made for tight touring car chassis. There is no single official definition: manufacturers use “micro” differently depending on their market, per MicroServoMotor, so check the actual weight and dimensions, not the category name.

What is the full RC servo sizes chart?

This is the reference chart. Typical figures are the bands the cited guides agree on. Torque figures are stall torque at the voltage the listing states, and torque always rises with voltage, so never compare a 6 V figure against a 4.8 V figure as if they were the same test.

Size class Typical case dimensions (mm) Typical weight Typical torque (kg-cm) Fits
Sub-micro / nano About 18 to 22 long, 8 to 11 wide, 15 to 22 tall 2 to 5 g Under 1 Indoor foam planes, 1/24 to 1/18 cars, micro helicopters
Micro About 22 to 23 long, 11 to 12 wide, 22 to 31 tall 5 to 14 g 0.5 to 3.5 Park flyers, small foam planes, 1/18 cars, pan-tilt mounts
Mini About 29 to 31 long, 13 to 16 wide, 26 to 30 tall 20 to 30 g 3 to 8 450 to 500 class helicopters, sport planes, 1/16 to 1/12 cars
Standard About 40 to 41 long, 20 wide, 36 to 41 tall 40 to 65 g 3 to 21.5 1/10 and 1/8 cars, larger planes, boats, robot arms
Low profile (standard footprint) About 40 long, 20 wide, 30 tall 35 to 55 g 8 to 15 1/10 touring cars with tight chassis
Large / 1/5 scale About 60 to 66 long, 30 wide, 55 to 60 tall 120 to 200 g 20 to 40 1/5 cars, giant scale planes, big boats

Bands attributed: SwellRC, EuroRC, RCMods, and MicroServoMotor. The standard torque range of 3 to 21.5 kg-cm is the span of the live listings below, not a guide band: the KS-201 lists 3 kg/cm at 4.8 V and the DS3218MG lists up to 21.5 kg-cm at 6.8 V.

What do real servos measure? Live listing figures by class

Class bands are rough. Listing figures are exact. The table below anchors every class except large scale with servos actually sold in the store, checked live on 2026-11-21. Dimensions are printed exactly as each listing states them. Note that listings do not use one consistent axis order: the SA-37 lists 20 x 25 x 8.2 mm while the SG92R lists 23.0 x 12.2 x 29 mm, so read the three numbers as the case size and measure the case yourself before you order.

Servo (ID) Class Dimensions as listed (mm) Weight Torque as listed Price
SA-37 (ID 12046) Sub-micro 20 x 25 x 8.2 3.7 g 0.5 kg/cm $10.49
SG92R (ID 12070) Micro 23.0 x 12.2 x 29 9 g 2.5 kg/cm at 4.8V $8.49
MG90D (ID 12068) Micro 22.8 x 12.2 x 28.5 13 g 2.1 kg/cm at 4.8V, 2.4 kg/cm at 6.6V $14.49
MA-90 (ID 12066) Micro 24 x 12.2 x 30 Not stated 1.3 kg/cm (18.1 oz/in) $8.49
1250MG (ID 12064) Mini 31.0 x 16.3 x 29.5 25.0 g 3.0 kg-cm at 4.8V, 3.5 kg-cm at 6.0V $18.49
KS-201 (ID 12057) Standard 40.5 x 20 x 36.5 42 g 3 kg/cm at 4.8V, 4.5 kg/cm at 6.0V $8.99
DS3218MG (ID 18825) Standard 40 x 20 x 40.5 60 g Up to 21.5 kg.cm at 6.8V $23.99
SPT4412 (ID 19668) Standard 41 x 20.3 x 28 About 68 g 10 kg.cm at 4.8V, 12 kg.cm at 6.0V $34.99

Full names, gear types, and links for all eight are in the Parts block below. All eight were in stock and purchasable on 2026-11-21. Prices are the live store prices checked that day. Re-verify at publish.

Two things stand out. First, the standard class spans 3 kg/cm to 21.5 kg-cm, so a standard case tells you almost nothing about torque: the KS-201 and the DS3218MG share the same roughly 40 by 20 mm footprint. Second, the SPT4412 weighs about 68 g, above the 40 to 65 g band: high torque standard servos run heavy.

What is a sub-micro servo?

A sub-micro servo, also called a nano servo, is the smallest class in the chart: about 2 to 5 g, with torque under 1 kg-cm per the SwellRC class table. The live example is the SA-37 (ID 12046) at $10.49: 3.7 g, metal gears, 0.5 kg/cm, 20 x 25 x 8.2 mm as listed. It suits indoor foam planes with small, light control surfaces only. Do not put a sub-micro servo on car steering or on a larger outdoor surface: at 0.5 kg/cm it will stall, buzz, and strip gears.

What is a micro servo?

A micro servo is the 9 g class most builders start with: about 22 to 23 mm long, 11 to 12 mm wide, 22 to 31 mm tall, and 5 to 14 g. The SG92R (ID 12070) at $8.49 is the 9 g reference: 23.0 x 12.2 x 29 mm, 2.5 kg/cm at 4.8 V, plastic composite gears. The MG90D (ID 12068) at $14.49 is the tough micro: 22.8 x 12.2 x 28.5 mm, 13 g, metal gears, digital control, double ball bearing, 2.1 kg/cm at 4.8 V and 2.4 kg/cm at 6.6 V. The MA-90 (ID 12066) at $8.49 is the value metal gear micro: 24 x 12.2 x 30 mm, 1.3 kg/cm, metal gears, with weight not stated on its listing.

Micro servos suit park flyers, small foam planes, 1/18 cars, and pan-tilt mounts. A micro servo suits 1/16 and 1/18 cars only: a 1/10 car needs a standard servo. For the ranked micro picks, the torque by job table, metal versus plastic gears, analog versus digital, and the foam plane install steps, read the micro servo guide linked below. This chart does not repeat that ground.

What is a mini servo?

A mini servo sits between micro and standard: about 29 to 31 mm long, 13 to 16 mm wide, 26 to 30 mm tall, and 20 to 30 g, with typical torque of 3 to 8 kg-cm per SwellRC. The live example is the 1250MG (ID 12064) at $18.49: 31.0 x 16.3 x 29.5 mm, 25.0 g, metal gears, 3.0 kg-cm at 4.8 V and 3.5 kg-cm at 6.0 V. It is the step up when a micro servo is too weak but a standard servo will not fit the bay or is too heavy for the model. Mini servos suit 450 to 500 class helicopters, sport planes, and 1/16 to 1/12 cars.

What is a standard servo?

A standard servo is the default size in the hobby: about 40 to 41 mm long, 20 mm wide, 36 to 41 mm tall, and 40 to 65 g, with standard servos sharing roughly the same 40 by 20 mm mounting footprint, per SwellRC. That shared footprint is what makes standard servos easy to swap: a replacement usually drops into the same tray. The KS-201 (ID 12057) at $8.99 is the light standard: 40.5 x 20 x 36.5 mm, 42 g, nylon gears, 3 kg/cm at 4.8 V and 4.5 kg/cm at 6.0 V, listed for 1/10 RC cars. The DS3218MG (ID 18825) at $23.99 is the heavy standard: 40 x 20 x 40.5 mm, 60 g, full metal gears, waterproof seals, 270 degrees of controlled rotation, up to 21.5 kg.cm at 6.8 V. The SPT4412 (ID 19668) at $34.99 sits between them on torque and above them on weight: 41 x 20.3 x 28 mm, about 68 g, metal gears, dual ball bearings, 10 kg.cm at 4.8 V and 12 kg.cm at 6.0 V, with a JR connector and a 260 mm lead.

One common standard servo is missing from the table on purpose. The MG996R (ID 3479) at $15.99 is listed as a metal gear standard servo, but its listing states no dimensions and no torque numbers, so it cannot anchor a dimensions chart. Our general servo guide states the MG996R datasheet lists 9.4 kg/cm at 4.8 V and 11 kg/cm at 6 V; that is a datasheet figure as cited in that guide, not a listing figure. For 1/10 car steering, the general guide states at least 9.4 kg/cm of torque at 4.8 V, which the SPT4412 and DS3218MG clear on listing figures and the KS-201 does not.

What is a large servo?

A large servo, also called a 1/5 scale or XL servo, is the biggest class in the chart: about 60 to 66 mm long, 30 mm wide, 55 to 60 mm tall, and 120 to 200 g, with torque of 20 to 40 kg-cm per the EuroRC reference table. SwellRC puts the large scale class at 150 g and up with 30 to 60+ kg-cm. Large servos suit 1/5 cars, giant scale planes, and big boats. The store holds no large scale example, so no listing figure is given here and none is invented.

How do you measure a servo?

Measure the servo you are replacing, not the class name. The full method, with all seven measurements and what goes wrong when each one is skipped, is in the servo replacement measurements guide linked below. The short version is four numbers plus the spline.

  1. Measure the case length, width, and height with a caliper, across the case body only, without the mounting ears.
  2. Measure the mounting hole spacing, centre to centre across the mounting ears.
  3. Count the spline teeth on the output shaft and measure the shaft diameter across the outside of the teeth. A 25 tooth spline is common on standard servos, and the DS3218MG listing states a 25T aluminum horn is included, but the shaft diameter must match too. Micro servo listings state no spline tooth counts, so count the teeth on the old output before you order a micro horn.
  4. Measure the bay, not just the servo. Measure the opening in the wing, fuselage, or chassis tray, and check clearance for the lead exit and the arm swing. The micro guide gives a fit rule of cutting the bay opening 1 mm larger per side than the servo body.

Write the numbers down in that order and compare them against the listing figures, not the class name.

How do you read a servo listing?

Three listing habits cause most wrong orders. First, dimension order varies. Listings state three numbers but do not always say which is length, width, or height, and the order differs between listings in the same store. Treat the three numbers as the case size and confirm the fit with a caliper. Second, torque without a voltage is not a comparison. The same servo reads higher at the higher voltage: the 1250MG lists 3.0 kg-cm at 4.8 V and 3.5 kg-cm at 6.0 V. Compare candidates only at the voltage your receiver or BEC actually supplies. Third, listings sometimes leave a figure out. The MA-90 listing states torque, speed, and dimensions but no weight, and the MG996R listing states no dimensions or torque at all. A missing figure is a gap, not a zero: do not guess it.

Torque is measured in kg-cm or oz-in: one kg-cm is about 13.9 oz-in, per SwellRC. Speed is measured in seconds per 60 degrees: a lower number is faster, and 0.10 sec/60deg is fast while 0.20 sec/60deg is slow, per SwellRC.

Which size fits my model?

Match the size to the model first, then the torque to the job. Model scale is a guide, not an exact measurement: two 1/10 cars from different brands can differ in weight and tire size, and the weight of the vehicle, its tire size, and how it is driven matter as much as the scale on the box, per RC Hop Up Parts.

Model Job Size to start with Torque note
Indoor foam plane, under 250 g Small light surfaces Sub-micro, 3.7 to 4.4 g 0.5 to 0.8 kg/cm
Small foam plane, park flyer Ailerons, rudder, elevator Micro, 9 g 0.8 to 1.3 kg/cm; 2.5 kg/cm for the largest micro jobs
1/24 to 1/18 car or crawler Steering Micro Light vehicles need little torque; match the mounting style
1/16 to 1/12 car Steering Mini, or standard where the chassis takes it Mini covers most; some 1/12 cars use low profile or standard
1/10 car, touring or buggy Steering Standard, or low profile for tight chassis At least 9.4 kg/cm at 4.8 V, per the general servo guide
1/10 crawler or trail truck Steering Standard, high torque Torque matters more than speed; big tires, heavy loads
1/8 buggy, truggy, or monster truck Steering Standard, high torque, high voltage High voltage servos run on 2S LiPo BEC voltage
1/5 car Steering, throttle Large Standard servos will not survive the loads
RC boat Rudder Standard, waterproof where the listing states it DS3218MG lists waterproof sealing
Robot arm or hexapod Loaded joints, then light joints Standard for base and shoulder, micro for wrist and gripper Size loaded joints first

Plane torque rows are from the live micro servo guide. Car scale rows are from RC Hop Up Parts and SwellRC. Our hexapod build uses 18 servos across six legs, and our joystick robot arm guide runs four axes through a PCA9685 driver board.

Which size should you buy?

Four profiles cover most orders. Buy the size that fits the bay with torque to spare, not the biggest servo in the chart.

Small foam plane, first build. Buy micro. The SG92R (ID 12070) at $8.49 covers the largest micro jobs at 2.5 kg/cm, and the MA-90 (ID 12066) at $8.49 covers the same jobs with metal gears at 1.3 kg/cm. If the plane is indoor only, buy sub-micro: the SA-37 (ID 12046) at $10.49 at 3.7 g.

1/10 car steering. Buy standard. The SPT4412 (ID 19668) at $34.99 clears the 9.4 kg/cm steering figure at 10 kg.cm at 4.8 V, and the DS3218MG (ID 18825) at $23.99 clears it further at up to 21.5 kg.cm at 6.8 V. Do not buy the KS-201 (ID 12057) at $8.99 for steering: at 3 kg/cm at 4.8 V it is a light standard servo, not a steering servo.

A replacement for a failed servo. Buy the same case size first, then torque. Measure the old servo with the four steps above, match the case size and spline, and step up in torque at the same case size only where the bay allows it.

A robot arm or hexapod. Buy standard for the loaded joints and micro for the light ones. The DS3218MG (ID 18825) at $23.99 suits a loaded joint: 60 g, up to 21.5 kg.cm at 6.8 V, and 270 degrees of controlled rotation, which suits robot joint work per its listing. The SG92R (ID 12070) at $8.49 suits a wrist, gripper, or pan-tilt axis.

What are the common mistakes?

  • Trusting the class name instead of the listing. Two servos in the same class can differ by several millimetres and grams. Check the weight and dimensions on the listing, per MicroServoMotor.
  • Comparing torque at different voltages. A servo rated at 6 V always looks stronger than the same servo rated at 4.8 V. Compare torque only at the same voltage.
  • Buying a standard servo for a micro bay. A standard servo is about 40 by 20 mm and will not fit a micro bay. Measure the bay first.
  • Buying a micro servo for a 1/10 car. Micro servos suit 1/16 and 1/18 cars only. A 1/10 car needs a standard servo.
  • Forgetting the horn and the cable. The installed footprint is larger than the bare servo: the horn sweeps an arc.
  • Assuming a 25T horn fits a micro servo. Standard servo listings state 25T fitment, but micro listings state no spline counts. Count the teeth on the old output first.
  • Ordering around a missing figure. The MA-90 listing states no weight and the MG996R listing states no dimensions or torque. A missing figure is a gap.

How do you fix a servo that does not fit or does not work?

Check the fit before you suspect a bad servo. Most size problems are measurement problems, and most performance problems are power or linkage problems.

The servo will not fit the bay. Do not force it. A case 1 mm larger than the bay can bind in a molded opening or shift the arm plane enough to bind the linkage. Pick a listing whose stated dimensions match the bay, or fit a mount bracket that holds the larger case.

The servo fits but the horn will not press on. The spline does not match. Count the teeth on the servo output and measure the shaft diameter, then match the horn listing fitment line. A 25T horn fits a 25T output of the matching diameter.

The servo buzzes at full travel. It is stalled against a mechanical stop. Reduce the end points on the radio so the arm never pushes against the stop. A stalled servo draws high current and overheats. The MG996R datasheet lists 10 mA at idle and 1,400 mA at stall, as cited in our general servo guide: size the power supply for stall current, not idle current.

The servo jitters or the receiver resets when all the surfaces move together. Check the power supply first. Three or four servos moving at once on a small pack can pull the voltage down and brown out the receiver. Keep the leads short, and use a separate supply or a larger BEC.

The servo does not reach full range. The pulse width is usually the cause. The standard servo signal is 50 Hz PWM with a 1 to 2 ms pulse, per the general servo guide. Check the radio end points before you replace the servo.

Parts used in this guide

Prices and stock were verified live on qualityrcparts.com on 2026-11-21 through the public Store API and the store REST API. All four items below are in stock and purchasable. The four class examples in the live listing table above were also verified live in stock and purchasable on 2026-11-21.

Shipped fast and securely from our secure facility in Queensbury, NY. QRC ships within the USA only.

Related guides

FAQ

What size is a standard RC servo? About 40 by 20 by 38 mm and 40 to 65 g. The live examples here run 40.5 x 20 x 36.5 mm at 42 g for the KS-201 and 40 x 20 x 40.5 mm at 60 g for the DS3218MG. Standard servos share roughly the same 40 by 20 mm mounting footprint, which makes swapping easy. Source: EuroRC, SwellRC, and QRC listings.

What size is a micro servo? About 22 to 23 mm long, 11 to 12 mm wide, and 22 to 31 mm tall, and 5 to 14 g. The familiar 9 g servo, such as the SG92R at 23.0 x 12.2 x 29 mm and 9 g, is the reference size. Some sellers call lighter servos micro too, so check the weight and dimensions on the listing, not the class name. Source: MicroServoMotor, RCMods, and QRC listings.

What is the difference between a micro servo and a mini servo? A micro servo weighs 5 to 14 g and measures about 23 by 12 mm across the case. A mini servo weighs 20 to 30 g and measures about 29 to 31 mm long and 13 to 16 mm wide. A mini gives more torque, typically 3 to 8 kg-cm per SwellRC, against 0.5 to 3.5 kg-cm for micro servos in the store listings. Source: SwellRC and QRC listings.

What size servo do I need for a 1/10 RC car? A standard servo. The general servo guide states at least 9.4 kg/cm of torque at 4.8 V for steering. A micro servo suits 1/16 and 1/18 cars only. For tight touring car chassis, a low profile standard servo at about 40 by 20 by 30 mm fits where a full height standard servo will not. Source: QRC general servo guide, RC Hop Up Parts, and EuroRC.

How do I measure a servo to replace it? Measure the case length, width, and height with a caliper without the mounting ears, then the mounting hole spacing centre to centre, then count the spline teeth and measure the shaft diameter, then measure the bay the servo sits in. Compare those numbers against the listing figures, not the class name. Source: QRC servo replacement measurements guide.

Can I use a micro servo instead of a standard servo? Only if the bay, the mount, and the torque all match. A micro servo is about half the length and a quarter of the weight of a standard servo, and at 0.5 to 3.5 kg-cm it cannot do a steering job that needs 9.4 kg/cm or more. Source: QRC listings and the general servo guide.

Sources

Last updated: 2026-11-21

Hexapod Spider Robot Kit Build: 18 Servos, Six Legs

Hexapod spider robot kit with six legs and 18 servos

Hexapod Spider Robot Kit Build: 18 Servos, Six Legs

Build the hexapod spider robot kit with 18 servos across six legs, driven by PCA9685 boards over I2C. Three legs lift and swing while the other three hold the body up, which is the tripod gait. The servos need a 5 to 6 V supply rated at 10 A, because 18 servos can draw more than any USB port supplies.

Quick facts

Item Value
Legs 6 legs, 3 servos per leg, 18 servos total
Gait Tripod gait: 3 legs lift, 3 legs support
Drivers PCA9685 boards, 16 channels each; two boards cover 18 servos
Controller Freenove Crawling Robot Controller in the kit, Arduino IDE compatible; wireless remote included
Servo supply Separate 5 to 6 V, sized for stall current
Battery Not included; 2x 3.7 V 18650 flat-top cells plus a charger, bought separately

Hexapod spider robot kit: how does it walk?

A hexapod walks with a tripod gait. At any moment, three legs touch the ground and hold the body up, while the other three lift, swing forward, and set down. Then the two tripods swap. The body never has fewer than three legs on the ground, so it stays stable without balancing.

Each leg has three servos: a hip servo that swings the leg forward and back, a knee servo that lifts the leg, and an ankle servo that sets the foot angle. Three servos per leg times six legs is 18 servos. No Arduino has 18 PWM pins to spare, which is why the build uses PCA9685 boards.

The two tripods are the front left, middle right, and rear left legs in one group, and the front right, middle left, and rear right legs in the other. Lift one tripod, swing it forward, set it down, then lift the other. That is the whole walk. Turning shifts the swing direction, and strafing shifts the foot targets sideways.

Why do 18 servos need two PCA9685 boards?

One PCA9685 drives 16 servos. Eighteen servos need two boards, with 16 channels on the first and 2 on the second. The boards share one I2C bus, so the wiring stays at two wires plus power. Set the first board to address 0x40 and the second to 0x41 with its A0 solder pad, then pass each address to its own library object.

Both boards share the same 50 Hz servo frequency, and each channel is independent. The sketch below creates two driver objects, one per address, and a helper that maps a servo number from 0 to 17 onto the right board and channel. The gait code then thinks in legs and joints, not in boards and channels.

The kit as shipped needs no PCA9685 at all: its own Freenove Crawling Robot Controller drives all 18 servos directly from ports 22 to 39, with the yellow servo wire to S, red to +, and black to -. The kit ships with a remote, and the kit controller plus its provided library handle the first walks. The PCA9685 wiring in this post is the path for builders who drive the servos from their own board instead of the kit controller. The kit tutorial also requires the batteries to be fully charged before assembly, with about 8.0 to 8.4 V across the two 18650 cells, and warns that wrong batteries can damage the servos.

How do you size power for 18 servos?

Size for stall current, not idle current. Eighteen servos starting or stalling together draw several amps, far more than any USB port or board regulator supplies. No single authoritative figure exists for the stall current of the servos in this kit, and the common SG90 stall figures disagree across resellers, so this post states no per-servo current number. Size generously and meter the rail on your own build.

Use a separate 5 to 6 V supply rated at 10 A or more for the servo rail, and feed it to the PCA9685 V+ terminals. A 5 V 10 A bench supply works on the bench. That 10 A rating is a planning rating with headroom, not a computed stall total. The kit itself runs from 2x 3.7 V 18650 cells, which is a separate power path from the PCA9685 servo rail in this post; do not feed the 18650 pack straight into the servo rail.

Add a large capacitor across each V+ terminal, 1000 microfarads or more, close to the board. Eighteen servos starting at once draw a sharp spike, and the capacitor absorbs it. Use thick wire for the V+ feed, because thin jumper wires drop voltage at 10 A and heat up. Never route servo current through the Arduino 5 V pin.

How do you assemble and test the legs?

Assemble one leg at a time and test it before you fit the next. Center each servo at 90 degrees with the sketch before you fit the horn, so the leg starts at its middle position. A horn fitted 30 degrees off center steals a third of that joint travel and the leg will never reach its full step.

Test the tripod gait with the body held up in your hand first. Lift the robot, run the gait, and watch the feet. Three feet should lift and swing while the other three stay still, then swap. If a leg swings the wrong way, its hip servo is reversed in the mapping, so flip its direction sign in the code, not the wiring.

Set the robot down only after the lifted test looks right. On the floor, start with short steps and low speed. A hexapod that steps too far tips forward, because the support tripod moves behind the center of mass. Short steps keep the body over its feet.

Wiring

Wire the build with the power off. Check every connection in the table before you connect any battery or USB cable.

From To Notes
PCA9685 board 1 VCC Controller 5 V or 3V3 Logic only; use 3.3 V on an ESP32 so the I2C pull-ups stay at 3.3 V
PCA9685 board 1 GND Controller GND and servo supply negative Common ground for all three
PCA9685 board 1 SDA Controller SDA I2C data; address 0x40, A0 pad open
PCA9685 board 1 SCL Controller SCL I2C clock
PCA9685 board 2 VCC Controller 5 V or 3V3 Logic only
PCA9685 board 2 GND Controller GND and servo supply negative Common ground
PCA9685 board 2 SDA Controller SDA Shared I2C bus; address 0x41, A0 pad bridged
PCA9685 board 2 SCL Controller SCL Shared I2C bus
Both boards V+ terminals Separate 5 to 6 V supply, 10 A or more Servo power only; thick wire, one feed per board
Board 1 channels 0 to 15 Servos 0 to 15 Legs 1 to 5 full sets plus one joint of leg 6
Board 2 channels 0 and 1 Servos 16 and 17 Remaining two joints of leg 6
Capacitor 1000 uF or more Across each V+ terminal and ground Absorbs the 18-servo start-up spike; mind the polarity

Power budget

Eighteen micro servos draw several amps while walking and far more if they all stall at once, so the supply must cover the walking current with margin, and the fuse or pack must survive the stall spike without cutting out mid-step. A 5 V 10 A bench supply is the bench choice. A battery pack needs the same current rating. The kit battery is 2x 3.7 V 18650 flat-top cells plus a charger, none of which are included.

The logic side draws only a few milliamps. The controller, the two PCA9685 chips, and the remote receiver run from the controller supply. The servo rail is separate, and only the grounds join. Never power the servos from the controller 5 V pin, and never feed the servo supply into the controller.

Heat matters at 10 A. Use thick wire for the V+ feeds, keep the runs short, and check the terminal screws after the first hour of walking. A loose screw at 10 A heats up and drops voltage, and the legs will jitter and stall.

Build steps

  1. Unbox the kit and count the parts: frame, six legs, 18 servos, horns, screws, the Freenove Crawling Robot Controller, and remote. Confirm the battery is not included: buy 2x 3.7 V 18650 flat-top cells plus a charger, and fully charge them before assembly.
  2. Run the kit as shipped first if you can: load the provided code on the kit controller, calibrate the servos with the calibration graph in the kit tutorial, and walk it on the remote. The PCA9685 steps below are the own-board alternative, not the kit default.
  3. Center every servo at 90 degrees with the sketch before you fit any horn. Fit each horn at the middle position.
  4. Assemble one leg at a time: hip, knee, then ankle. Test each leg by hand for free movement before you fit the next.
  5. Fit the six legs to the body in the tripod groups: front left, middle right, rear left in one group, and the other three in the second group.
  6. Wire the two PCA9685 boards from the wiring table. Set board 1 to 0x40 and board 2 to 0x41 with the A0 pad.
  7. Connect the servos to the boards in order, 0 to 17, and connect the separate servo supply to both V+ terminals. Join the grounds.
  8. Upload the sketch below. Hold the body up and run the tripod gait. Confirm three legs lift and swing while the other three hold still, then swap.
  9. Set the robot on the floor and walk it forward in short steps. Shorten the step if it tips forward.

Code

Library: Adafruit PWM Servo Driver library. Board: Arduino IDE compatible controller. The sketch maps servo numbers 0 to 17 across two PCA9685 boards and runs a basic tripod gait. Tune the step size and the per-leg direction signs for your kit.

// Hexapod spider robot: 18 servos, 2 x PCA9685, tripod gait
// Board: Arduino IDE compatible controller.
// Library: Adafruit PWM Servo Driver (Adafruit_PWMServoDriver).
#include 
#include 

Adafruit_PWMServoDriver pwmA = Adafruit_PWMServoDriver(0x40);  // Servos 0-15
Adafruit_PWMServoDriver pwmB = Adafruit_PWMServoDriver(0x41);  // Servos 16-17

#define SERVOMIN 150   // About 730 microseconds at 50 Hz; calibrate per servo
#define SERVOMAX 600   // About 2930 microseconds at 50 Hz; calibrate per servo
#define SERVO_FREQ 50

// Legs 0-5, joints 0=hip, 1=knee, 2=ankle. Servo number = leg*3 + joint.
// Tripod 1: legs 0, 2, 4. Tripod 2: legs 1, 3, 5.
const uint8_t TRIPOD1[3] = {0, 2, 4};
const uint8_t TRIPOD2[3] = {1, 3, 5};

uint16_t angleToPulse(int angle) {
  return map(constrain(angle, 0, 180), 0, 180, SERVOMIN, SERVOMAX);
}

void setServo(uint8_t servo, int angle) {
  uint16_t pulse = angleToPulse(angle);
  if (servo < 16) pwmA.setPWM(servo, 0, pulse);
  else pwmB.setPWM(servo - 16, 0, pulse);
}

void setLeg(uint8_t leg, int hip, int knee, int ankle) {
  setServo(leg * 3 + 0, hip);
  setServo(leg * 3 + 1, knee);
  setServo(leg * 3 + 2, ankle);
}

void setTripod(const uint8_t legs[3], int hip, int knee, int ankle) {
  for (uint8_t i = 0; i < 3; i++) setLeg(legs[i], hip, knee, ankle);
}

void setup() {
  pwmA.begin();
  pwmA.setOscillatorFrequency(27000000);
  pwmA.setPWMFreq(SERVO_FREQ);
  pwmB.begin();
  pwmB.setOscillatorFrequency(27000000);
  pwmB.setPWMFreq(SERVO_FREQ);
  delay(10);
  // Park every servo at 90 degrees: legs centered, feet down.
  for (uint8_t s = 0; s < 18; s++) setServo(s, 90);
  delay(500);
}

void loop() {
  // Tripod gait, one step per cycle. Tune the angles for your kit.
  // Lift tripod 1, swing forward, set down.
  setTripod(TRIPOD1, 110, 60, 90);  // Lift and swing forward
  delay(300);
  setTripod(TRIPOD1, 110, 90, 90);  // Set down
  setTripod(TRIPOD2, 70, 90, 90);   // Push body forward
  delay(300);
  // Lift tripod 2, swing forward, set down.
  setTripod(TRIPOD2, 110, 60, 90);
  delay(300);
  setTripod(TRIPOD2, 110, 90, 90);
  setTripod(TRIPOD1, 70, 90, 90);
  delay(300);
}

Test and tune it

Hold the body up and run the gait. Three legs should lift and swing while the other three stay still, then swap. If a leg in the wrong tripod moves, its leg number is wrong in the TRIPOD table.

If a leg swings backward instead of forward, flip its hip direction sign in the gait code. Do not rewire the servo, because the mapping is the fault.

Set the robot down and walk it 3 feet. If it tips forward, shorten the swing from 110 degrees toward 100 degrees. Short steps keep the body over the support tripod.

FAQ

How many servos does a hexapod need?

Eighteen servos: three per leg times six legs. Each leg has a hip servo, a knee servo, and an ankle servo. One PCA9685 drives 16 servos, so this build uses two boards.

What is the tripod gait?

Three legs hold the body up while the other three lift, swing forward, and set down. Then the two tripods swap. The body always has three legs on the ground, so it stays stable without balancing.

Why does my hexapod jitter when it walks?

The servo supply sags under the start-up spike of 18 servos. Add a 1000 microfarad capacitor across each V+ terminal, check the supply current rating, and tighten the terminal screws.

Can I power 18 servos from USB?

No. Eighteen micro servos together draw several amps while walking and far more at stall. Use a separate 5 to 6 V supply rated at 10 A or more for the servo rail.

Does the kit include a battery?

No. The kit needs 2x 3.7 V 18650 flat-top cells plus a charger, none included, and the remote takes 1x 9 V battery or USB power. For the PCA9685 servo rail in this post, use a bench supply for indoor testing.

Parts used

Build it with the Hexapod Spider Robot Kit with Remote, Six Legs with 18 Motors, Arduino IDE Compatible, Walking and Crawling STEM Project (Battery Not Included) and the PCA9685 16-Channel PWM Servo Driver Board, 12-Bit IIC Interface, Arduino + Raspberry Pi, 2-Pack (two boards cover 18 servos; one board drives only 16). For a servo arm build without the driver board, see How to Build a Robot Arm with ESP32 and Servo Motors, and the arm build uses the same driver in Joystick Robot Arm: 4-Axis Control with PCA9685. For the servo signal behind the positions, see Digital vs Analog Servo Guide: Which Servo Holds Center.

Sources

Last updated: 2026-11-21. Search volume: not verified. Product links use the store short link form with the product ID; confirm the product slug at publish time.

Joystick Robot Arm: 4-Axis Control with PCA9685

4-axis joystick controlling a 4-DOF acrylic robot arm with claw

Joystick Robot Arm: 4-Axis Control with PCA9685

Control a joystick robot arm with a 4-axis joystick and a PCA9685 servo driver. The joystick gives three analog axes, X tilt, Y tilt, and Z twist, plus a momentary button on top, which the listing counts as the fourth axis. X maps to base rotation, Y maps to the shoulder, and Z twist maps to the elbow.

Quick facts

Item Value
Arm 4-DOF acrylic arm kit with claw, 4 SG90 servos in the With SG90 option
Joystick 4-axis potentiometer, 10 k, X Y tilt plus Z twist, top button; 3 analog outputs plus the button
Driver PCA9685, 16 channels, I2C address 0x40
Controller Arduino Uno R3, 5 V logic
Joints Base, shoulder, elbow, plus claw; 4 servos, no wrist joint
Servo supply Separate 5 to 6 V, sized for stall current

Joystick robot arm: what does each joystick axis control?

The joystick gives three analog signals, one per potentiometer, plus the top button. The X tilt controls the base rotation, so pushing the stick left and right spins the arm around its vertical axis. The Y tilt controls the shoulder, so pushing forward and back raises and lowers the upper arm. The Z twist controls the elbow, so twisting the stick bends the forearm. There is no fourth analog output on this joystick: the fourth axis in the product name is the momentary top button, and this arm kit has no wrist joint, so no wrist mapping exists. The button toggles the claw instead.

That mapping is a choice, not a law. If the twist feels wrong for the elbow, swap the elbow and shoulder assignments in the code. The mapping lives in three lines, one per joint, so changing it takes a minute. Keep the mapping you choose written on a card next to the joystick until it becomes habit.

The top button toggles the claw. Press once and the claw closes, press again and it opens. A toggle suits a claw, because holding a button down while the arm moves is awkward. The sketch debounces the button with a 200 ms debounce interval, so one press never registers as two.

How do you map a joystick reading to a servo angle?

Each joystick axis is a 10 k potentiometer, so the Uno analog pin reads 0 to 1023. Center is near 512. The sketch reads the pin, subtracts the center, and applies a dead zone of about 30 counts so a resting stick sends no motion. Outside the dead zone, the reading maps to a small angle step per loop, not to an absolute angle.

Relative control suits an arm. The stick sets the speed and direction of each joint, and the joint holds its position when the stick centers. Absolute control, where the stick position equals the joint angle, is simpler to code but twitchy in practice, because a small stick move swings the joint a long way.

Each joint has a safe angle range in the sketch. The base runs 0 to 180 degrees, the shoulder 30 to 150 degrees, and the elbow 30 to 150 degrees. The claw runs 20 to 160 degrees. Those limits keep the arm from driving its own joints into the table or past the servo end stops. Adjust them for your arm after the first test, and write the final numbers down.

How do you wire the joystick and the PCA9685?

The joystick has three analog outputs, one per potentiometer, plus a button output. The X and Y potentiometers end in solder tabs on the sides of the base, so solder wires to them and meter each wire before you connect it. The Z potentiometer comes prewired: red and white are the reference ends and black is the wiper. Wire each potentiometer as a voltage divider across the 5 V supply and read the wiper on an ADC pin: X wiper to A0, Y wiper to A1, and Z wiper (black wire) to A2. The top button has two blue wires: connect one to digital pin 2 and the other to ground, and use INPUT_PULLUP, so a press reads LOW.

The PCA9685 connects over I2C. Connect its VCC to the Uno 5 V pin, its GND to ground, its SDA to A4, and its SCL to A5. Connect the separate 5 to 6 V servo supply to the V+ screw terminal, and join the supply ground to the Uno ground. The servo plugs go on channels 0 to 3: base, shoulder, elbow, and claw.

Keep the joystick wires short and away from the servo power wires. Servo current spikes can couple into long analog wires and make the joints jitter. If a joint trembles with the stick centered, widen the dead zone from 30 to 50 counts before you change anything else.

How do you tune the arm so it moves smoothly?

Start with slow steps. The sketch moves each joint by at most 1 degree per 20 ms loop, which is about 50 degrees per second at full stick deflection. That is fast enough to feel responsive and slow enough to stay controllable. Raise the step to 2 degrees only after the arm tracks well at 1 degree.

Load changes everything. An empty claw moves freely, but a loaded claw sags the shoulder and elbow servos. Test with the load you plan to lift, not empty. If the shoulder sags under load, raise the servo supply from 5 V to 6 V, which adds torque, or reduce the load. Do not raise the supply above the servo 6 V rating for SG90 servos.

If the arm oscillates, the steps are too large or the dead zone is too small. Lower the step size first. If one joint alone oscillates, check its pulse calibration, because a pulse past the servo end stop makes the servo fight its own gears and buzz.

Wiring

Wire the build with the power off. Check every connection in the table before you connect any battery or USB cable.

From To Notes
Joystick VCC Uno 5 V 10 k potentiometers, 5 V logic
Joystick GND Uno GND Common ground
Joystick X wiper Uno A0 Base rotation, analog 0 to 1023; solder tabs on the base, meter before soldering
Joystick Y wiper Uno A1 Shoulder, analog 0 to 1023; solder tabs on the base, meter before soldering
Joystick Z wiper, black wire Uno A2 Elbow, analog 0 to 1023; red and white are the reference ends
Joystick top button, blue wires Uno digital pin 2 and GND Claw toggle; INPUT_PULLUP, press reads LOW
PCA9685 VCC Uno 5 V Logic only, a few milliamps
PCA9685 GND Uno GND and servo supply negative Common ground for all three
PCA9685 SDA Uno A4 I2C data
PCA9685 SCL Uno A5 I2C clock
PCA9685 V+ screw terminal Separate 5 to 6 V supply positive Servo power only; size for stall current
PCA9685 channel 0 Base servo Signal, V+, GND
PCA9685 channel 1 Shoulder servo Signal, V+, GND
PCA9685 channel 2 Elbow servo Signal, V+, GND
PCA9685 channel 3 Claw servo Signal, V+, GND; toggled by the top button

Power budget

The logic side draws only a few milliamps from the Uno 5 V pin: the joystick potentiometers and the PCA9685 chip. The servo side draws whatever the servos demand. Four micro servos moving together draw on the order of 1 A under light load, and far more when they start or stall together.

Use a separate 5 to 6 V supply rated at 3 A or more for the V+ terminal. Never power the servos from the Uno 5 V pin. The Uno regulator supplies about 500 mA from USB, and a single stalled micro servo can exceed that. The board will reset and the arm will drop. No authoritative TowerPro datasheet publishes an SG90 stall current, so do not size the supply from a single stall number: the 3 A rating above is a planning rating with headroom, and meter your own servos if you need an exact figure.

Add a 470 to 1000 microfarad capacitor across V+ and ground at the terminal. The servos draw sharp spikes when they start moving, and the capacitor absorbs them. Without it, the rail sags and every joint jitters at once.

Build steps

  1. Assemble the 4-DOF acrylic arm kit with the claw. Fit the 4 SG90 servos and confirm each joint moves freely by hand before you power anything. Buy the With SG90 option: the Arm Only option ships as acrylic parts and screws with no servos.
  2. Mount the PCA9685 near the arm base. Connect the four servos to channels 0 to 3 in base, shoulder, elbow, and claw order.
  3. Wire the joystick to the Uno analog pins and the button to digital pin 2 from the wiring table.
  4. Wire the PCA9685 I2C pins and the separate servo supply. Join the supply ground to the Uno ground. Leave the servo supply off for now.
  5. Install the Adafruit PWM Servo Driver library. Upload the sketch below with the servo supply still off.
  6. Switch the servo supply on. The arm should park at its home position: base 90, shoulder 90, elbow 90, claw open.
  7. Push the stick in each direction and confirm the right joint moves. Press the top button and confirm the claw toggles. Calibrate the pulse limits per joint until no servo buzzes at its ends.

Code

Library: Adafruit PWM Servo Driver library. Board: Arduino Uno R3. Each joystick axis drives one joint in relative steps, and the top button toggles the claw. The joystick has three analog axes, so the claw is the only joint on the button; the kit has no wrist joint.

// Joystick robot arm: 4-axis joystick + PCA9685, 4 servos
// Board: Arduino Uno R3. Library: Adafruit PWM Servo Driver.
#include 
#include 

Adafruit_PWMServoDriver pwm = Adafruit_PWMServoDriver(0x40);

#define SERVOMIN 150   // About 730 microseconds at 50 Hz; calibrate per servo
#define SERVOMAX 600   // About 2930 microseconds at 50 Hz; calibrate per servo
#define SERVO_FREQ 50

const int JOY_X = A0;  // Base
const int JOY_Y = A1;  // Shoulder
const int JOY_Z = A2;  // Elbow
const int BTN_PIN = 2; // Claw toggle, INPUT_PULLUP

const uint8_t CH_BASE = 0, CH_SHOULDER = 1, CH_ELBOW = 2;
const uint8_t CH_CLAW = 3;

// Safe joint ranges in degrees. Adjust for your arm after testing.
const int LIM[4][2] = {{0, 180}, {30, 150}, {30, 150}, {20, 160}};
int angle[4] = {90, 90, 90, 20};  // Start parked, claw open
bool clawClosed = false;
unsigned long lastBtnMs = 0;

uint16_t angleToPulse(int a) {
  return map(constrain(a, 0, 180), 0, 180, SERVOMIN, SERVOMAX);
}

void setServo(uint8_t ch, int a) {
  pwm.setPWM(ch, 0, angleToPulse(a));
}

// Read one joystick axis and return a step of -1, 0, or +1.
int readStep(int pin) {
  int v = analogRead(pin) - 512;  // Center is near 512
  if (v > 30) return 1;           // Dead zone of 30 counts
  if (v < -30) return -1;
  return 0;
}

void moveJoint(uint8_t ch, int step) {
  if (step == 0) return;
  angle[ch] = constrain(angle[ch] + step, LIM[ch][0], LIM[ch][1]);
  setServo(ch, angle[ch]);
}

void setup() {
  Serial.begin(115200);
  pinMode(BTN_PIN, INPUT_PULLUP);
  pwm.begin();
  pwm.setOscillatorFrequency(27000000);
  pwm.setPWMFreq(SERVO_FREQ);
  delay(10);
  for (uint8_t ch = 0; ch < 4; ch++) setServo(ch, angle[ch]);
}

void loop() {
  moveJoint(CH_BASE, readStep(JOY_X));
  moveJoint(CH_SHOULDER, readStep(JOY_Y));
  moveJoint(CH_ELBOW, readStep(JOY_Z));

  // Top button toggles the claw, debounced by 200 ms.
  if (digitalRead(BTN_PIN) == LOW && millis() - lastBtnMs > 200) {
    lastBtnMs = millis();
    clawClosed = !clawClosed;
    angle[CH_CLAW] = clawClosed ? LIM[CH_CLAW][1] : LIM[CH_CLAW][0];
    setServo(CH_CLAW, angle[CH_CLAW]);
  }
  delay(20);  // About 1 degree per 20 ms at full deflection
}

Test and tune it

Center the stick and confirm all joints hold still. A joint that creeps with a centered stick has a joystick reading outside the dead zone. Widen the dead zone from 30 to 50 counts.

Move each joint to both ends of its range. A servo that buzzes at an end is past its mechanical limit. Lower the limit in the LIM table for that joint until the buzz stops.

Pick up a light object with the claw and hold it for 30 seconds. If the shoulder sags, raise the servo supply from 5 V to 6 V or reduce the load. Do not exceed the SG90 6 V rating.

FAQ

Which joystick axis controls which joint?

X controls the base, Y controls the shoulder, and Z twist controls the elbow. The top button toggles the claw. There is no fourth analog axis and no wrist joint on this kit: the fourth axis in the product name is the top button. Swap the assignments in the code if another mapping feels better.

Why does the arm jitter when the stick is centered?

The joystick reading sits outside the dead zone, or servo current spikes are coupling into the analog wires. Widen the dead zone to 50 counts and keep the joystick wires away from the servo power wires.

Why does the arm drop when I power it on?

The servos have no holding torque until the sketch parks them, and the Uno 5 V pin cannot hold them under load. Power the servos from the separate V+ supply and let the sketch park the arm at startup.

Can I use MG996R servos on this arm?

Only if the arm frame and the supply suit them. An MG996R draws up to 2.5 A at stall on 6 V, so four of them need a far larger supply than four SG90 servos. Check the arm kit servo mounts first.

How do I stop a joint from hitting the table?

Set its safe range in the LIM table in the sketch. The shoulder and elbow limits of 30 to 150 degrees keep the arm from driving its joints into the table or past the servo end stops.

Parts used

Build it with the 4-Axis Joystick Potentiometer with Momentary Top Button, 10K, X Y Tilt plus Z Twist, for Arduino, ESP32, and Robot Projects, the PCA9685 16-Channel PWM Servo Driver Board, 12-Bit IIC Interface, Arduino + Raspberry Pi, 2-Pack, and the 4-DOF Acrylic Robot Arm Kit with Claw for Arduino, Unassembled, SG90 / UNO / Rocker Options. For a servo arm build without the driver board, see How to Build a Robot Arm with ESP32 and Servo Motors, and the servo signal behind the positions is Digital vs Analog Servo Guide: Which Servo Holds Center.

Sources

Last updated: 2026-11-21. Search volume: not verified. Product links use the store short link form with the product ID; confirm the product slug at publish time.

L298N vs TB6612 vs DRV8833: Which Motor Driver for Your Robot

L298N, TB6612FNG and DRV8833 motor driver boards side by side

L298N vs TB6612 vs DRV8833: Which Motor Driver for Your Robot

Pick the TB6612FNG for most battery robots. It drives two DC motors from 4.5 V to 13.5 V at 1.2 A average per channel, and it loses only about 0.5 V inside the chip at 1 A. Pick the DRV8833 when your battery stays under 10.8 V, and pick the L298N only when your motor supply runs above 13.5 V.

L298N vs TB6612 vs DRV8833 at a glance

The table below uses the ST L298 datasheet DS0218 Rev 5 (October 2023), the Toshiba TB6612FNG datasheet dated 2014-10-01, which prints no revision number, and the Texas Instruments DRV8833 datasheet SLVSAR1E Rev E (July 2015). Module ratings vary with board layout and cooling, so treat the continuous current figures as planning values, not guarantees.

Spec (unit) L298N TB6612FNG DRV8833
Switch type Bipolar Darlington MOSFET MOSFET
Motor supply (V) Up to 46 chip, 5 to 35 on common modules 4.5 to 13.5, 15 absolute max 2.7 to 10.8
Continuous current per channel (A) 2.0 1.2 average 1.5 RMS, PWP and RTY packages only (0.5 RMS in PW package), with good cooling
Peak current per channel (A) 3.0 non repetitive, 100 ms 3.2 single 10 ms pulse 2.0
Voltage lost at 1 A (V) 1.8 to 3.2 About 0.5, from 0.5 ohm on resistance About 0.36, from 0.36 ohm on resistance
Logic supply (V) 5, module regulator or Arduino 2.7 to 5.5 None needed, internal
Standby or sleep current 13 mA typ on VS, 24 mA typ on VSS (36 mA max on VSS) 1 uA max in standby 1.6 uA typ in sleep (2.5 uA max)
Channels (count) 2 DC motors or 1 stepper 2 DC motors or 1 stepper 2 DC motors or 1 stepper
Built in 5 V regulator Yes on common modules No No
Standby pin No, enable pins only Yes, STBY pin Yes, sleep via nSLEEP

Why does voltage drop decide this comparison?

A motor driver sits between the battery and the motor, and every volt it keeps is a volt the motor never sees. The L298N uses bipolar transistors, and the ST datasheet lists a total saturation drop of 1.8 V to 3.2 V at 1 A. On a 7.4 V battery with a motor drawing 1 A, the motor sees about 4.2 V to 5.6 V, and the missing 1.8 V to 3.2 V becomes heat in the driver. Heat is power lost: at 1 A and a 2 V drop, the driver burns 2 W per channel.

The MOSFET drivers lose far less. The TB6612FNG has about 0.5 ohm total on resistance, so at 1 A it drops about 0.5 V, and the heat is current squared times resistance, or 0.5 W per channel. The DRV8833 has about 0.36 ohm total on resistance, so at 1 A it drops about 0.36 V and burns about 0.36 W. On the same 7.4 V battery at 0.25 A per motor, the TB6612FNG delivers about 7.28 V and the DRV8833 about 7.31 V, while the L298N loses more than a volt at that current, so the motor sees less than the MOSFET drivers deliver. That is why a robot car runs faster and longer on a MOSFET driver with the same battery.

The practical rule is to size for stall current, not running current. A TT motor that draws 0.25 A while rolling can draw over 1 A when stalled against a wall. A driver rated for 1.2 A average per channel handles that stall briefly, but a driver run at its headline current for minutes overheats. Plan for roughly half the headline number for sustained running, and always check the stall current of your exact motor.

When should you pick the L298N?

Pick the L298N when the motor supply runs above 13.5 V. The TB6612FNG tops out at 15 V absolute maximum and the DRV8833 at 10.8 V, so a 24 V gear motor or a 12 V motor with a fully charged 4 cell LiPo that peaks above 16 V needs the L298N range. The L298N chip tolerates up to 46 V, and common modules are rated for 5 V to 35 V motor supplies.

The L298N module also suits a first build. It has screw terminals that accept thick motor wires without soldering, a heatsink, a built in 5 V regulator that can power an Arduino when the motor supply is under 12 V, and an enable jumper that every tutorial covers. Its 5 V jumper rule matters: with the jumper on and a motor supply under 12 V, the on board regulator outputs 5 V. Above 12 V, remove the jumper and power the logic separately, or the regulator overheats.

The cost of the L298N is heat and wasted battery. On a 7.4 V pack, about a quarter of the battery voltage never reaches the motor. On a 12 V supply the loss is a smaller share, so the L298N is at its best on 12 V builds and at its worst on small 2 cell battery robots.

When should you pick the TB6612FNG?

Pick the TB6612FNG for most small battery robots, including TT motor cars, N20 motor robots and 2WD chassis kits on 6 V to 12 V supplies. Its 4.5 V to 13.5 V range covers 4 AA cells, a 2 cell LiPo at 7.4 V and a 3 cell LiPo at 11.1 V. Its 1.2 A average and 3.2 A peak per channel cover TT and N20 stall currents with margin. Its STBY pin stops both motors with one logic line, which is the clean way to implement an emergency stop or a sleep mode.

The TB6612FNG logic supply runs from 2.7 V to 5.5 V, so it works directly with a 3.3 V ESP32 or a 5 V Arduino without a level shifter on the logic pins. Its low standby current, 1 uA max, suits a battery robot that sleeps between runs. Its only weakness is the 13.5 V ceiling: a 4 cell LiPo at full charge exceeds it, so a 4 cell build needs the L298N or a different driver.

When should you pick the DRV8833?

Pick the DRV8833 for compact, low voltage robots. Its 2.7 V to 10.8 V range starts at a single Li-ion cell, so a 1 cell robot at 3.7 V runs on it where the TB6612FNG needs at least 4.5 V. Its 1.5 A RMS and 2 A peak per channel in the PWP and RTY packages are the highest continuous current in this group, and its 0.36 ohm on resistance gives the lowest voltage drop and the least heat.

The DRV8833 needs no separate logic supply, because its logic runs from the motor supply internally, and it accepts 3.3 V and 5 V logic inputs. Its sleep current is 1.6 uA typ, close to the TB6612FNG standby current. Its weakness is the 10.8 V ceiling: a 3 cell LiPo at 11.1 V nominal and 12.6 V full charge exceeds it, so a 3 cell build belongs on the TB6612FNG or the L298N. It also has no built in 5 V regulator and no screw terminals on most breakout boards, so it suits a soldered build more than a first screw terminal build.

How do you wire each driver?

All three drivers use the same pattern: motor supply to VM or VS, logic supply where needed, ground shared with the microcontroller, two direction pins and one PWM enable pin per motor. On the L298N, the enable pins are ENA and ENB, and the direction pins are IN1 to IN4. On the TB6612FNG, the pins are PWMA and PWMB for speed, AIN1 and AIN2 plus BIN1 and BIN2 for direction, and STBY to enable the chip. On the DRV8833, the pins are AIN1 and AIN2 plus BIN1 and BIN2, and PWM goes on one input of each pair.

Three wiring rules apply to all three drivers. First, share the ground between the motor supply, the driver and the microcontroller, or the logic signals have no reference. Second, never power a motor from the microcontroller 5 V pin: use a separate battery or supply, and size it for the stall current. Third, add a 100 uF capacitor across the motor supply close to the driver on the TB6612FNG and DRV8833 boards, because motor current spikes pull the supply down and cause brownout resets.

Which driver for TT motors, N20 motors and 12 V builds?

Motor or build Pick this driver Why
TT motors, 3 to 6 V, 2WD car TB6612FNG Low drop on a small battery, 1.2 A covers stall, STBY pin
N20 motors, 3 to 6 V, small robot DRV8833 or TB6612FNG DRV8833 for 1 cell builds, TB6612FNG for 2 cell builds
12 V gear motors L298N Voltage headroom above 13.5 V, screw terminals, heatsink
3 cell LiPo, 11.1 V car TB6612FNG 13.5 V ceiling covers 12.6 V full charge, low drop
1 cell LiPo, 3.7 V robot DRV8833 Only driver here that starts at 2.7 V
First robot car, screw terminals L298N Screw terminals and 5 V regulator, simplest first wiring

For the L298N in practice, see the L298N robot car build.

FAQ

Can the L298N run on 3.3 V logic from an ESP32?

Yes, with care. The L298N logic high threshold is about 2.3 V, so a 3.3 V ESP32 output registers as high. Power the L298N logic from 5 V where the module expects it, share the ground, and keep the motor supply separate.

Why does my L298N get hot?

The L298N drops 1.8 V to 3.2 V at 1 A, and that power becomes heat. At 1 A per channel, the chip burns 1.8 W to 3.2 W per channel. Keep the heatsink on, improve airflow, or move to a MOSFET driver such as the TB6612FNG.

What is the STBY pin on the TB6612FNG?

STBY is the standby pin. Pull it low to stop both motors and drop the chip to 1 uA max. Pull it high to enable the driver. Tie it high through your code at startup, never leave it floating.

Can the DRV8833 drive a stepper motor?

Yes. The DRV8833 is a dual H-bridge, so it drives one bipolar stepper motor as well as two DC motors. It does not microstep; a stepper driver such as the A4988 is the pick when microstepping matters.

Do I need a flyback diode with these drivers?

No external flyback diodes are needed on these three boards, because the H-bridge transistors include body diodes that handle the motor inductive kick. A bulk capacitor across the motor supply is still good practice.

Parts used

L298N Dual H-Bridge Motor Driver Module: the rugged bipolar driver for 12 V and higher voltage builds. TB6612FNG Dual Motor Driver, 3-Pack: the MOSFET driver for most small battery robots. DRV8833 DC Motor Driver, 3-Pack: the low voltage MOSFET driver for 1 cell and 2 cell robots.

Sources

Last updated: 2026-11-21. Search volume: not verified. Product links use the store short link form with the product ID; confirm the product slug at publish time.