DRV8825

Four things. First the current formula: the A4988 uses ITripMAX = VREF / (8 × RS), the DRV8825 uses ITripMAX = VREF / (5 × RS), and carrier sense resistors usually change from 0.05 ohm to 0.25 ohm — so the same VREF produces a very different current and you must recalibrate. Second the microstep truth table above 1/8 step: A4988 all-high is 1/16, DRV8825 all-high is 1/32, and DRV8825 needs MODE2 high with MODE1 and MODE0 low for 1/16. Third the voltage and current headroom improves from 35 V and 2 A to 45 V and 2.5 A. Fourth the DRV8825 adds an explicit DECAY pin, a 3.3 V reference output and a 250 kHz STEP limit, and drops the ROSC pin.

Use ITripMAX = VREF / (5 × RS). With the 0.25 ohm sense network on a standard carrier this becomes VREF = 1.25 × ITripMAX, so 1.5 A needs about 1.88 V and 2 A needs 2.5 V. Measure the voltage at the VREF pin (or the wiper of the trimmer) with a multimeter while turning it, and set it to 70 to 80 percent of your motor's rated current. You can source VREF from the built-in V3P3OUT pin through a divider; keep VREF between 1 V and 3.5 V for full accuracy.

Because nSLEEP and nRESET have internal pulldowns. nSLEEP has a 1 MΩ pulldown and nRESET a 100 kΩ pulldown, so if either floats low the bridges stay off and STEP is ignored. Commercial modules tie them together and pull them up; on a bare design, drive both from GPIOs or strap them to logic high. Also remember wake-up: after nSLEEP goes high the device needs about 1.7 ms before it accepts a STEP edge.

No. It is a bipolar driver with two full H-bridges and an internal current chopper designed for a four-wire motor. A five- or six-wire unipolar motor has centre taps and is driven by sinking one half-winding at a time. Either rewire the unipolar motor as bipolar by ignoring the centre taps (halves the current rating, doubles the inductance), or use a Darlington sink array such as the ULN2803 or ULN2003, or an L298N with the centre taps tied to the motor supply, and generate the phase sequence in the MCU or with an L297.

Mixed decay is the default recommendation and what the DECAY pin gives when left open — it balances torque smoothness against audible noise. Fast decay (DECAY high) reverses the bridge after the chopping threshold and gives faster current decay, useful at high step rates where winding inductance prevents the current from falling quickly enough, but it produces more ripple and noise. Slow decay (DECAY low) recirculates through both low-side FETs, is quietest and most efficient, but can cause sluggish current decay and audible resonance at some speeds. Experiment with your specific motor and load.

The 2.5 A figure is TI's maximum drive current at 24 V and 25 °C with proper heatsinking, not a free-air number. In an HTSSOP-28 the exposed PowerPAD must be soldered to a large copper pour with a via array into inner layers; with a modest pour and no airflow you should design for about 1.5 A to 2 A continuous. Note also that 2.5 A is the chop current limit, and the RMS winding current at microstepping is lower than full scale. Derate further above 25 °C ambient and check the die temperature against the 150 °C to 180 °C thermal shutdown point.

Program a known VREF, measure the resulting winding current, and compare with ITripMAX = VREF / (5 × RS): genuine TI parts track the 660 mV ± 25 mV internal trip reference within the specified accuracy, while remarked or cloned dies miss badly. Then verify all six MODE combinations and check that the device survives a 36 V VM supply — clones remarked from lower-voltage drivers often fail or clamp well below 45 V. Visually, look for laser-etched TI marking, the correct HTSSOP-28 PowerPAD, and a date code you can trace. We X-ray and electrically sample-test every reel.

MOQ is one reel of DRV8825PWPR (2000 pcs, HTSSOP-28) for volume pricing, or 50 pcs loose and carrier modules in packs of 10 for prototyping. Shenzhen stock ships in one to two business days; TI authorized-channel lead time is ten to eighteen weeks and can extend during allocation on stepper drivers.