powerStep01 Constantly tripping Overcurrent.

[Kurtnelle]

My powerStep01 is constantly triggering the Overcurrent alarm whenever I give a movement command, even if there is no stepper connected. What could possibly be causing this? Previously I was getting a similar effect on the L6470 chip. When I turned off the overcurrent shutdown feature the chip fried and apparently took out GPIO25 of the Raspberry Pi that was command it. In both cases, the motors never moved.

[Dario CUCCHI]

Hi @Kurtnelle​ , finally I did the tests !

Here a step by step sequence that you can try.

I think its only a matter of providing the supply voltages and sending the right SPI commands.

So, you can use any control hardware you want: can be the Raspberry Pi or the NUCLEO F401 board.

First, check the connection between the control board and the IHM03A1; you need:

1. Vs motor voltage and main supply
2. 3.3V for the digital part and pull-up resistors
3. A common GND reference
4. The SPI lines (CS, CK, SDI, SDO)
5. RESET line (this is active low, so it must be set high for normal operation)

Here is the operation sequence:

1. Supply the VS = 18V (as you mentioned)
2. Supply the 3.3V and initialize the communication lines
3. Clear the status flags sending the command “GetStatus�?. Once is executed, you should see the red LED (D2) turning off.
4. Send the commands to write all the registers in order to configure the device (see the list below)
5. Send again a “GetStatus�? command
6. Send a movement command, for example a “Run�? at 100 step/s

I suggest to try this routine without the motor first and check that no Overcurrent or other error occurs.

If it works, stop the power stage setting it in high impedance, connect your motor and try to run it in the same way.

Consider that I tried the configuration on my motor. Probably yours will move as well, but then you will have to fine tune the configuration according to the target motor and application.

Here is the list registers to be written: the configuration is in current mode, about 0.85A of phase current on IHM03A1.

Addr. 0x05       ACC               = 0x15

Addr. 0x06       DEC               = 0x15

Addr. 0x07       MAX_SPEED   = 0x106

Addr. 0x08       MIN_SPEED    = 0x000

Addr. 0x09       TVAL_HOLD    = 0x0A

Addr. 0x0A      TVAL_RUN      = 0x0A

Addr. 0x0B      TVAL_ACC      = 0x0A

Addr. 0x0C      TVAL_DEC      = 0x0A

Addr. 0x0E      T_FAST          = 0x35

Addr. 0x0F       TON_MIN        = 0x08

Addr. 0x10       TOFF_MIN      = 0x29

Addr. 0x13       OCD_TH         = 0x08

Addr. 0x15       FS_SPD          = 0x3FF

Addr. 0x16       STEP_MODE  = 0x0C

Addr. 0x17       ALARM_EN     = 0xFF

Addr. 0x18       GATECFG1     = 0xA6

Addr. 0x19       GATECFG2     = 0x20 

Addr. 0x1A      CONFIG          = 0x3F88

[Kurtnelle]

The red light never goes out.

Clear the status flags sending the command “GetStatus�?. Once is executed, you should see the red LED (D2) turning off. <- It doesn't turn off.

I haven't tried setting any registers beyond that.

[Kurtnelle]

Also, the status register:

0xE003 : 1110000000000011 <= bit 2 is set! The device is busy?!

(EDIT: Busy is active low. So no the device is not busy.)

[Dario CUCCHI · Accepted solution]

Yes, looking at the status register I only see that the device has detected an UVLO and and UVLO on the ADC (both bits = 0). Also OCD is active low, so OCD =1 means no overcurrent detected. UVLO and UVLO_ADC are the ones that keep the red LED on.

An UVLO_ADC is detected because the X-NUCLEO-IHM03A1 by default does not have any component mounted on the ADC pin, so the voltage on it is around zero. In my board I've already some components on ADC, I forgot to specify it :grinning_face_with_sweat: .

I see two possibilities:

1. If you don't need the ADC in your application (it is not mandatory to run a motor), you can just disable the related alarm using bit4 in Reg. 0x17 (ALARM_EN).
2. If you need ADC, just mount a proper couple of resistor on R27 and R28: UVLO threshold is max. 1.35V, the voltage on ADC pin must not exceed 3.3V.

The other bit UVLO is referred to Vs: at the start up there is always a detection because the voltage supply is ramping-up

All these bits are latched, meaning that even if the UVLO condition is gone, the bits stay in the error condition until a "GetStatus�? command is sent.

I expect that if 18V are applied on Vs and you send a "GetStatus�? , the UVLO bit becomes 1 (error released).

[Kurtnelle]

So then:

• Write to the Alarm Register to disable ADC_ULVO flag
• Get Status => Flag LED should go out?

[Dario CUCCHI]

This is the expected behavior. Let me know if it works.

[Kurtnelle]

The motor is humming. The configuration works! Turning off the ADC_ULVO is required for the operation of this eval board.