A4988 vs DRV8825 vs TMC2209: Which Stepper Driver Wins

Buy the TMC2209 for any new 3D printer or CNC build that runs in a room with people in it. Buy the DRV8825 when the axis needs more than 29 V and noise does not matter. Buy the A4988 only when the budget is the whole decision, or when you are replacing a dead A4988 in a machine that already works. All three are bipolar stepper chopper drivers on the StepStick footprint. Source: QRC posts 16272 and 23584.

A4988 vs DRV8825 vs TMC2209 at a glance

Feature A4988 (Allegro) DRV8825 (TI) TMC2209 (Trinamic/ADI)
Max continuous current 1.0 A/phase (2.0 A max, cooled) 1.5 A/coil (2.5 A chip max, heatsinked) 2.0 A RMS, 2.8 A peak/coil
Max microstepping 1/16 1/32 1/256 interpolated
Noise Loud PWM whine Very loud harsh chopping Silent StealthChop2
Configuration Jumpers, VREF pot Jumpers, VREF pot UART or jumpers
Sensorless stall detection No No Yes, StallGuard4 via UART
Voltage range 8 to 35 V 8.2 to 45 V 4.75 to 29 V chip, 4.75 to 28 V listing
Best use Basic prototyping, low-cost plotters Standard 3D printers, light CNC, high voltage axes Silent operation, sensorless homing

Sources: QRC post 16272; Allegro A4988, TI DRV8825, and Trinamic TMC2209 datasheets; QRC listings 17896, 23573, and 19618.

When does the A4988 win?

The A4988 wins on price and on nothing else. It is the cheapest way into stepper control and the workhorse of a decade of RepRap printers. The QRC single module costs $6.99 and includes a heatsink. Its limits are real: 1 A per phase without a heatsink, 2 A max with cooling, per the Allegro datasheet, and the QRC listing states up to 1.2 A continuous per phase. It thermally throttles at roughly 1.2 A without a heatsink, it stops at 1/16 microstepping, its fixed-decay chopping buzzes at low speed, and it runs on 5 V logic, with 3.3 V unreliable, so an ESP32 build needs level shifting. Source: QRC post 16272; QRC listing 17896; ElectricalFlux ESP32 guide. Buy it when the motor is small, the budget is tight, and the noise does not matter. Do not buy it for a printer on a desk or any motor above 1.2 A per phase.

When does the DRV8825 win?

The DRV8825 wins when the axis needs voltage or current the A4988 cannot give, and the TMC2209 ceiling is in the way. Its motor supply runs 8.2 V to 45 V, the widest range of the three, so it is the only one that handles a 36 V axis or a high-speed CNC axis. It is rated 1.5 A per coil, with a chip maximum of 2.5 A per coil with proper heatsinking at 24 V and 25 C, per the TI datasheet. The QRC listing adds one more chip figure: 1.75 A RMS per coil with proper heatsinking. Source: QRC post 16272; TI DRV8825 datasheet SLVSA73F; QRC listing 23573. Its microstepping runs full step to 1/32 step, set with the M0, M1, and M2 pins. One mapping difference matters when swapping from an A4988: the setting that gives 1/16 step on the A4988 gives 1/32 step on the DRV8825. Buy it for a standard 3D printer axis, a light CNC axis, or any axis above 29 V. Do not buy it for a silent build: it is the loudest of the three.

When does the TMC2209 win?

The TMC2209 wins for any new build in 2026 where the machine runs near people. StealthChop2 removes the high-pitch whine of the older drivers, StallGuard4 lets the axis home with no endstop switch, and CoolStep cuts the current automatically when the load is light. Its current rating is 2 A RMS and 2.8 A peak per coil, per the Trinamic TMC2209 datasheet Rev 1.05, and the QRC 5-pack listing states the same figures. That is the highest continuous current of the three, and it is an RMS figure, not a peak figure. Source: QRC post 16272; Trinamic datasheet Rev 1.05; QRC listing 19618. Its limits are the voltage ceiling and the setup effort: the chip tops out at 29 V, and the store listing states 4.75 to 28 V, so a 36 V axis belongs on the DRV8825, and UART needs a wire and a firmware change. Once configured, current, microstepping, and mode change in firmware. Source: QRC post 16272; Sounds Electric TMC2209 guide. Buy it for a 3D printer, a CNC machine in a shared room, or any build that needs sensorless homing. Do not buy it for an axis above 28 V, or for a NEMA 23 motor over 2 A RMS per coil: that job belongs to a larger external driver such as the DM542 in post 16272.

Current compared: the RMS versus peak trap

The three current ratings look comparable in a table. They are not, because they are measured differently. The A4988 and DRV8825 formulas set a peak current: the top of the sine wave each coil sees during microstepping. The TMC2209 formula sets an RMS current: the heating value of that sine wave, about 70.7 percent of the peak. So 1 A on a TMC2209 is not the same setting as 1 A on an A4988: the TMC2209 at 1 A RMS pushes each coil to a 1.41 A peak. Source: The Science 360 VREF guide; Trinamic datasheet Rev 1.05 via BIGTREETECH. Size the driver from the motor rated phase current with headroom, not from the headline number. The rule from QRC post 16272 is: driver continuous current must be at least 1.5 times the motor continuous running current, and peak current must cover the stall current. A 1.5 A NEMA 17 needs a driver that holds 1.5 A continuously: the A4988 at 1 A cannot do it, the DRV8825 at 1.5 A is borderline and runs hot without a fan, and the TMC2209 at 2 A RMS handles it with margin.

Microstepping compared: resolution is not accuracy

Microstepping splits each full step into smaller steps, which makes motion smoother and quieter. It does not add holding torque, and it does not add positioning accuracy in the way the numbers suggest. A 1.8 degree stepper is accurate to about plus or minus 5 percent of a full step no matter how finely you slice it. The A4988 runs full step to 1/16 step (200 to 3,200 steps per rev on a 1.8 degree motor), the DRV8825 runs full step to 1/32 step (200 to 6,400), and the TMC2209 runs to 1/256 interpolated steps (up to 51,200) through MicroPlyer interpolation, which smooths incoming step signals up to 256 subdivisions. Full-step mode gives the most torque per step on every driver, so if an axis loses steps at high microstepping, dropping back to a coarser step setting is a legitimate fix. Source: QRC post 16272; Sounds Electric TMC2209 guide. Do not pick a driver for its microstepping number alone: a 1/256 driver on a machine with a loose belt and a wobbling frame still prints wobble.

Noise compared: why the old drivers whine

The A4988 and the DRV8825 chop the motor current with a fixed-decay chopper whose switching frequency falls in the human hearing range, typically 2 kHz to 20 kHz. The motor turns that switching into a constant whine during every move, with the pitch shifting with speed. Higher VREF settings and low microstepping make it worse. Source: ElectricalFlux stepper driver module guide; Sounds Electric TMC2209 guide. The TMC2209 fixes this two ways. StealthChop2 is a voltage-mode PWM chopper that keeps the motor near silent at low and medium speeds. SpreadCycle is a current-mode chopper that gives higher torque for rapid moves, at the cost of more noise. The driver switches between them, so print moves stay quiet and travel moves keep their torque. The A4988 and DRV8825 have no equivalent: their only noise fix is more microstepping, which helps a little and never reaches silence. Source: Sounds Electric guide; QRC post 16272.

Heat compared: where the current goes

Every amp a chopper driver regulates turns partly into heat in the chip. The A4988 thermally throttles at roughly 1.2 A without a heatsink. The DRV8825 needs a heatsink plus airflow above about 1.5 A per coil, or the chip throttles on temperature. The TMC2209 runs cooler at the same load because CoolStep cuts the current when the load is light. Source: ElectricalFlux stepper driver module guide; QRC listings 17896 and 23573; Sounds Electric TMC2209 guide. For all three: fit the heatsink, keep air moving across the driver bank, and never treat the headline current as a continuous rating. A driver that stops after three minutes, then works again after cooling, has hit thermal shutdown. Source: ElectricalFlux module guide.

How to set the current limit: VREF, worked examples

Set the current limit before connecting the motor. The A4988 and the DRV8825 set it with a trim potentiometer: measure VREF between the VREF pin and ground while turning the pot. The TMC2209 sets it in software over UART. Source: QRC post 16272; The Science 360 VREF guide.

Read the sense resistor first. The formula depends on the sense resistor printed on your board, not just on the chip. The sense resistors are two small black parts near the chip, printed with a code such as R068, R100, or R110: R068 is 0.068 ohm, R100 is 0.100 ohm, R110 is 0.110 ohm. Read your own resistors before copying a number from any guide. Source: The Science 360 VREF guide; Pololu carrier pages.

Driver Formula Worked example
A4988 VREF = 8 x Rs x I 1 A limit: 0.544 V on R068, 0.400 V on R050, 0.8 V on R100. Start at about 70 percent of the motor rated current
DRV8825 VREF = 5 x Rs x I; on a Pololu carrier with R100 resistors, current limit = VREF x 2 1.5 A NEMA 17: VREF = 0.75 V on a Pololu carrier. The x2 shortcut is Pololu-carrier specific
TMC2209 Over UART: set in software with the irun register. Standalone: VREF is about 1.41 x the RMS current on R110 boards 0.8 A RMS: VREF = 1.131 V; pot tops out near 1.77 A RMS

Sources: Allegro and TI datasheets via The Science 360 VREF guide; Pololu carrier pages; BIGTREETECH wiki; QRC post 16272.

How to wire all three

All three share the StepStick footprint. The full A4988 wiring table lives in QRC post 16272. The DRV8825 drops into the same socket, with the same STEP, DIR, VMOT, VDD, and coil pins, and these differences: M0, M1, and M2 instead of MS1, MS2, and MS3, an nFAULT pin, a DECAY pin, and SLEEP and RESET pins in place of SLP and RST. The TMC2209 drops into the same socket too: PDN_UART carries the single-wire UART link, and DIAG signals a stall for sensorless homing. In standalone mode, MS1 and MS2 set the microstepping and the pot sets the current. With UART, firmware sets the current, microstepping, chopper mode, and StallGuard4 threshold. Source: QRC listings 23573 and 19618; QRC post 16272.

DRV8825 microstepping table

M0 M1 M2 Resolution
LOW LOW LOW Full step
HIGH LOW LOW Half step
LOW HIGH LOW 1/4 step
HIGH HIGH LOW 1/8 step
LOW LOW HIGH 1/16 step
HIGH LOW HIGH 1/32 step
LOW HIGH HIGH 1/32 step
HIGH HIGH HIGH 1/32 step

Source: QRC product listing 23573. Note the trap: the M0-M2 setting that gives 1/16 step on the A4988 gives 1/32 step on the DRV8825.

Coil pairing and the bulk capacitor

Find the coil pairs with a multimeter: two wires that show a low resistance between them, typically 1 to 5 ohms, are one coil. Connect coil 1 to 1A and 1B, and coil 2 to 2A and 2B. Fit a bulk capacitor across the motor supply near the driver when the supply leads are long: for A4988 boards, fit at least a 47 uF electrolytic capacitor across VMOT and GND, close to the driver. Without it, voltage spikes can exceed 35 V at VMOT even from a 12 V supply and destroy the driver on power-up. Join the grounds, never unplug a motor cable while powered, and do not leave STEP or DIR floating. Source: QRC post 16272.

Which driver should I buy? The decision table

Your build Pick Why
New 3D printer, desk or bedroom TMC2209 Silent, sensorless homing
3D printer axis above 28 V DRV8825 45 V ceiling
Light CNC axis, 12 to 24 V DRV8825 1.5 A per coil, widest range
Classroom set of 10 drivers A4988 Cheapest per driver
ESP32 build at 3.3 V logic TMC2209 3.3 to 5 V native
NEMA 23 motor over 2 A per coil None of the three Step up to a DM542

What are the common mistakes?

  • Setting VREF from a guide without reading the sense resistor. A 0.544 V setting for a 1 A limit is correct only on a Pololu-style A4988 carrier with 0.068 ohm resistors. Read the R-code first.
  • Using the DRV8825 x2 shortcut on the wrong board. Current limit = VREF x 2 holds only on a Pololu carrier with 0.100 ohm resistors.
  • Comparing a TMC2209 RMS current with an A4988 peak current. 1 A RMS on a TMC2209 pushes each coil to a 1.41 A peak.
  • Swapping an A4988 for a DRV8825 without re-checking the microstep table. The setting that gives 1/16 step on the A4988 gives 1/32 step on the DRV8825.
  • Running an A4988 at the motor full rated current with no heatsink or airflow. It thermally throttles near 1.2 A without a heatsink.

How do I fix a stepper driver problem?

The motor hums and vibrates but does not spin. The coil pairs are mismatched, or VREF is too low. Rewire the coil pairs with a multimeter, then raise VREF in 0.1 V steps.

The driver is too hot to touch, or the motor stops after a few minutes and works again after cooling. The driver hit thermal shutdown. Lower VREF, fit the heatsink, and put a fan across the driver bank. Source: ElectricalFlux module guide.

The motor stalls or skips steps mid-move. The ramp is too aggressive, or the axis hit mid-band resonance at about 2 to 4 revolutions per second. Slow the ramp, change microstepping, or switch to a TMC2209.

The motor moves half as far per step as expected after a driver swap. You swapped an A4988 for a DRV8825 without changing the mode pins. Check the DRV8825 microstepping table above.

Parts used in this guide

Prices and stock were verified live on qualityrcparts.com on 2026-11-21. All three items are in stock and purchasable.

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

Related guides

FAQ

Which stepper driver is best for a 3D printer, A4988, DRV8825, or TMC2209? The TMC2209: near silent with StealthChop2, 2 A RMS per coil, and StallGuard4 sensorless homing. The DRV8825 is the pick only above 28 V. Source: QRC post 16272.

Can I replace an A4988 with a TMC2209? Yes, all three share the StepStick footprint. Re-check the microstepping pins and the current setting: TMC2209 figures are RMS, A4988 figures are peak.

What VREF should I set on an A4988? VREF = 8 x Rs x I. On a Pololu carrier with 0.068 ohm resistors, 0.544 V gives a 1 A limit; on a 0.100 ohm board, 0.8 V gives 1 A.

What VREF should I set on a DRV8825? On a Pololu carrier with 0.100 ohm resistors, current limit = VREF x 2, so a 1.5 A NEMA 17 needs 0.75 V. That shortcut is Pololu-carrier specific.

Does the TMC2209 need VREF? Not over UART: the current is set in software with the irun register. In standalone mode, on 0.110 ohm boards, VREF is about 1.41 times the RMS current you want.

Why does my stepper motor lose steps? The load torque passed what the motor can deliver at that speed, the ramp is too fast, or the axis hit mid-band resonance at about 2 to 4 revolutions per second. Slow the ramp, change microstepping, or switch to a TMC2209.

Sources

Last updated: 2026-11-21