Fig. 07 · Robotics Lab · Set the current limit, then step

A4988 stepper driver

Arduino Nano

Loading the figure…

Bench

No reading yet

VREF (meter)

No reading yet

Current draw

0 mA

Build and break

1BuildNext up
Strip to the bare board
Start from the bare green carrier: 15.24 x 20.32 mm, square corners, no mounting holes, sixteen plated holes along the two long edges. Everything goes back on in factory order.
2
Place and reflow the surface-mount parts
3
Put in the two headers
4
Solder the 16 header pins
5
Fit the RST-SLP cap and the 100 µF
6
Wire the Nano
7
Plug in the motor
8
Wire the LiPo
9
Put the meter on the pot
10
Power the bench
11
Set the current limit
12
Full step
13
Half step
14
1/16 step
15
Put the LiPo on backwardsBreak it
16
Put the 100 µF in backwardsBreak it
17
Leave the RST-SLP cap offBreak it
18
Run on a tired packBreak it
19
Mix up the coil pairsBreak it
20
Turn the limit up to 1.68 ABreak it
21
Unplug the motor while it runsBreak it
22
Exploded view

Break it on purpose

Each fault is one that catches real builders. Trigger it, read the symptom on the bench, then fix it.

LiPo connected backwards
reverse polarity
Why it happens
Symptom: A pop from the chip, the pack's leads get hot, and the driver is dead afterwards.
Pololu lists no reverse-voltage protection. Every bridge switch has a body diode that points from GND to VMOT; reversed, those diodes conduct and short the pack through the chip. A LiPo can push tens of amps, far more than the diodes survive.
Fix: Check red to VMOT before you connect, use a keyed connector, or add a series diode or a reverse-protection MOSFET on the supply.
100 µF electrolytic in backwards
part in backwards
Why it happens
Symptom: The capacitor gets hot, bulges and vents with a hiss.
An electrolytic only holds voltage one way. Backwards, the oxide layer breaks down, it conducts, heats up and the pressure vent in its top opens.
Fix: Stripe (minus) to GND, the long leg to VMOT. Replace a capacitor that has been in backwards even if it looks fine.
RST floating
missing part
Why it happens
Symptom: The power is on and the Nano sends pulses, but the coils are off: the shaft turns freely by hand.
On this carrier RST has no pull-up of its own. Floating, it sits low or wanders, and while RST is low the translator is held at Home and the outputs are off.
Fix: Put the jumper cap back across RST and SLP, or drive RST high from a pin.
VMOT under 8 V
brown-out
Why it happens
Symptom: The motor stutters and falls behind the pulses: it misses steps and ends up somewhere other than where the sketch thinks it is.
The A4988 is specified from 8 V on VMOT. A 2S pack is 8.4 V full but falls under 8 V long before it is empty. Below that the chip is out of its rating: the outputs drop out and steps are lost. A stepper has no sensor, so nothing notices.
Fix: Use a 3S pack or a 12 V supply, and stop running before the pack sags.
Coil pairs mixed
wrong setting
Why it happens
Symptom: The motor buzzes and shakes in place but does not turn.
Each H-bridge has to drive both ends of one coil. With green and red swapped, each bridge drives one end of each coil, the two fields fight, and the rotor rocks between two positions.
Fix: Find the pairs with a meter: a few ohms between the two leads of one coil, open between coils. Then one pair on 1A and 1B, the other on 2A and 2B.
Current limit too high
overheating
Why it happens
Symptom: The board gets too hot to touch, then the motor stops, restarts a few seconds later and stops again.
At 1.68 A the two coil paths burn about 1.8 W in the chip, I² times about 0.64 Ω, more as it warms. The bare carrier cannot shed that: the chip reaches 165 °C, thermal shutdown switches the outputs off, it cools 15 °C and starts again. Pololu puts the limit without a heatsink at about 1 A per coil.
Fix: Turn VREF back to 0.544 V (1.0 A). For more current, add a heatsink and airflow.
Motor unplugged while powered
wrong voltage
Why it happens
Symptom: A tiny spark at the plug, the motor goes slack, and the driver never moves it again.
A coil's current cannot stop at once. Opening the circuit forces it to, and the coil answers with a voltage spike, L x di/dt, far past the 35 V the A4988's outputs can take. Pololu warns that this can destroy the driver, often with no visible mark.
Fix: Only plug or unplug the motor with the power off. Check motor plugs and header joints: a loose one does the same thing.

Meet the bench

build it, meter the pot, step a NEMA 17

A stepper motor turns in fixed steps, 200 to a turn for a NEMA 17, and holds its place with both coils on. The A4988 does the hard part: it switches the two coils through two H-bridges in the right order, one step for every pulse on STEP, in the direction DIR sets, and it limits the current in each coil with a chopper that watches the voltage on a 0.068 Ω sense resistor. You set that limit before anything turns: the voltage on the trimpot's metal rotor, VREF, gives I_MAX = VREF / (8 x 0.068 Ω), so 0.544 V on a meter means 1.0 A, about as much as the bare carrier can take without a heatsink. Build the Pololu carrier, wire a Nano, the motor and a 2S LiPo, meter the pot while a hold sketch keeps the coils on, then step the motor in full, half and 1/16 steps. The breaks are the classic ones: a pack backwards, the reset pin left floating, a tired pack under 8 V, the coil pairs mixed, the limit turned up to the motor's 1.68 A, and the motor unplugged while it runs.

Sources

On the bench

Selected

Sketch uses 1,580 bytes of program storage

Code: Hold, set Vref

// A4988: one pulse on STEP is one step, DIR picks the way
const int STEP = 3, DIR = 4;           // STEP and DIR
const int MS1 = 5, MS2 = 6, MS3 = 7;   // MS1, MS2 and MS3 pick the step size

// n pulses, each one HIGH then LOW for halfUs microseconds
void pulses(long n, int halfUs) {
  for (long i = 0; i < n; i++) {
    digitalWrite(STEP, HIGH); delayMicroseconds(halfUs);
    digitalWrite(STEP, LOW); delayMicroseconds(halfUs);
  }
}

void setup() {
  for (int p = 3; p <= 7; p++) pinMode(p, OUTPUT);
  Serial.begin(9600);
  digitalWrite(MS1, LOW); digitalWrite(MS2, LOW); digitalWrite(MS3, LOW);   // all low: full step, 200 steps a turn
  digitalWrite(DIR, HIGH);
  delay(1);                     // the chip needs 1 ms after waking before the first step
  Serial.println("hold: coils on, no steps");   // no pulses: both coils stay on and the shaft holds
}

void loop() {
}

Practice tasks

0/2 tasks

Write the code, then run the checks. The run uses a virtual clock and a scripted world, so the same code always gives the same result.

One turn and back

Arduino C++

Full step, MS pins open. Turn the shaft exactly one turn forward (DIR on D4 HIGH, 200 pulses on D3, 2500 µs high and 2500 µs low each), print "one turn", wait 500 ms, then set DIR LOW and step 200 pulses back to where it started and print "home".

Checks

Not run yetRuns for 3 s
Not run yetDIR (D4) is HIGH on the way out
Not run yetSTEP (D3) pulses every 5 ms, about 400 times
Not run yetPrints "one turn" after the first second
Not run yetDIR is LOW on the way back
Not run yetPrints "home"

Hints

200 full steps of 1.8° make 360°. Each pulse: digitalWrite(STEP, HIGH); delayMicroseconds(2500); digitalWrite(STEP, LOW); delayMicroseconds(2500);
Put the pulse loop in a function, steps(int n), and call it twice.
Change DIR between the two moves, before the first pulse of the way back.

Code for One turn and back

A quarter turn a second, in 1/16 steps

Arduino C++

Set MS1, MS2 and MS3 (D5, D6, D7) HIGH for 1/16 step. Then pulse STEP at 800 pulses a second, 625 µs high and 625 µs low, so the shaft turns a quarter of a turn every second, and print "1 turn" after each 3200 pulses.

Checks

Not run yetRuns for 4.3 s
Not run yetMS1 (D5) is HIGH
Not run yetMS2 (D6) is HIGH
Not run yetMS3 (D7) is HIGH
Not run yetSTEP (D3) pulses every 1.25 ms, at least 3000 times
Not run yetPrints "1 turn" after 4 s

Hints

1/16 step needs all three MS pins HIGH, set once in setup().
A quarter turn is 800 microsteps; 800 a second is one every 1250 µs, so 625 µs high and 625 µs low.
Count 3200 pulses in a for loop, then print. 3200 x 1.25 ms = 4 s a turn.

Code for A quarter turn a second, in 1/16 steps

Build and break

1BuildNext up

Strip to the bare board

Start from the bare green carrier: 15.24 x 20.32 mm, square corners, no mounting holes, sixteen plated holes along the two long edges. Everything goes back on in factory order.

Place and reflow the surface-mount parts

Put in the two headers

Solder the 16 header pins

Fit the RST-SLP cap and the 100 µF

Wire the Nano

Plug in the motor

Wire the LiPo

Put the meter on the pot

Power the bench

Set the current limit

Full step

Half step

1/16 step

Put the LiPo on backwardsBreak it

Put the 100 µF in backwardsBreak it

Leave the RST-SLP cap offBreak it

Run on a tired packBreak it

Mix up the coil pairsBreak it

Turn the limit up to 1.68 ABreak it

Unplug the motor while it runsBreak it

Exploded view

Meet the bench

build it, meter the pot, step a NEMA 17

On the bench

Selected

Sketch uses 1,580 bytes of program storage

Code: Hold, set Vref

// A4988: one pulse on STEP is one step, DIR picks the way
const int STEP = 3, DIR = 4;           // STEP and DIR
const int MS1 = 5, MS2 = 6, MS3 = 7;   // MS1, MS2 and MS3 pick the step size

// n pulses, each one HIGH then LOW for halfUs microseconds
void pulses(long n, int halfUs) {
  for (long i = 0; i < n; i++) {
    digitalWrite(STEP, HIGH); delayMicroseconds(halfUs);
    digitalWrite(STEP, LOW); delayMicroseconds(halfUs);
  }
}

void setup() {
  for (int p = 3; p <= 7; p++) pinMode(p, OUTPUT);
  Serial.begin(9600);
  digitalWrite(MS1, LOW); digitalWrite(MS2, LOW); digitalWrite(MS3, LOW);   // all low: full step, 200 steps a turn
  digitalWrite(DIR, HIGH);
  delay(1);                     // the chip needs 1 ms after waking before the first step
  Serial.println("hold: coils on, no steps");   // no pulses: both coils stay on and the shaft holds
}

void loop() {
}

A4988 stepper driver

The bench

Build and break

Strip to the bare board

Place and reflow the surface-mount parts

Put in the two headers

Solder the 16 header pins

Fit the RST-SLP cap and the 100 µF

Wire the Nano

Plug in the motor

Wire the LiPo

Put the meter on the pot

Power the bench

Set the current limit

Full step

Half step

1/16 step

Put the LiPo on backwardsBreak it

Put the 100 µF in backwardsBreak it

Leave the RST-SLP cap offBreak it

Run on a tired packBreak it

Mix up the coil pairsBreak it

Turn the limit up to 1.68 ABreak it

Unplug the motor while it runsBreak it

Exploded view

Break it on purpose

LiPo connected backwards

100 µF electrolytic in backwards

RST floating

VMOT under 8 V

Coil pairs mixed

Current limit too high

Motor unplugged while powered

Live status

Current draw

Check yourself

After: Fit the RST-SLP cap and the 100 µF

After: Set the current limit

After: 1/16 step

After: Unplug the motor while it runs

End of bench quiz

Meet the bench

Sources

On the bench

Practice tasks

One turn and back

Checks

A quarter turn a second, in 1/16 steps

Checks

A4988 stepper driver

The bench

Build and break

Strip to the bare board

Place and reflow the surface-mount parts

Put in the two headers

Solder the 16 header pins

Fit the RST-SLP cap and the 100 µF

Wire the Nano

Plug in the motor

Wire the LiPo

Put the meter on the pot

Power the bench

Set the current limit

Full step

Half step

1/16 step

Put the LiPo on backwardsBreak it

Put the 100 µF in backwardsBreak it

Leave the RST-SLP cap offBreak it

Run on a tired packBreak it

Mix up the coil pairsBreak it

Turn the limit up to 1.68 ABreak it

Unplug the motor while it runsBreak it

Exploded view

Break it on purpose

LiPo connected backwards

100 µF electrolytic in backwards

RST floating

VMOT under 8 V

Coil pairs mixed