DRV8428: Motor driver doesn't work

Dong,

Would you share the output current, STEP pin and VREF pin waveform?

Regards,

Wang Li

Hello, it happened like this. When I set the EN pin and nsleep pin of the motor to the high level, the motor starts to vibrate, and then input a stepping signal to the motor, which still keeps the vibration unchanged. Since there is no current monitoring equipment, I put a 3 ohm observation voltage in series between the stepping motor and the aout1 pin of the drive board. The observed voltage is shown in Figure 1, In the same way, observe that there is no output of bout1 and there is no change after the stepping signal is given. The stepping signal is shown in Figure 2. As for VREF, I input a voltage of about 2.9v through the control board DAC (because the driving current of my motor is 1a), but even if I leave the VREF pin empty, the motor will still vibrate after enabling.

I attached the design screenshot of the circuit. If there is any error, please point it out.

Hello, Dong,

Why is the low voltage of the stepping signal not at ground? Is the controller floating from the motor driver ground?

BTW, on the layout, please make solid connection on VM capacitor and DVDD capacitor, instead of hatched connection.

Regards,

Wang Li

Hello, Wang,

I find that I both the low voltge and high voltage was wrong, so I chang the stepping signal as figure1 shows.And I also make a new experiment with it.The result are as shown in the following picture.

figure1 stepping signal

figure2 vref

At the beginning, i only set the EN pin and nsleep pin of the motor to the high level without stepping signal input, the output of aout1 is shown in figure3, and there is no output from bout.

figure3 aout1

Then i input the stepping signal, and the motor worked strangely as shown in the following video.And the outout of aout and bout is shown as the figure4 and figure5.

figure4 aout1

figure5 bout1

Finally, I want to ask some questions:

Q1: According to the vref voltage i gived,  the high voltage of xOUT should be like bOUT,why the aout is 4V which means the current should be 4 / 3 A.

Q2:In your reply you said I'm ought to make solid connection on VM capacitor and DVDD capacitor, instead of hatched connection. I'm not sure if you mean the  connection between the groud pad of the two capacitor and the groud copper laying.

Q3:The 8848 chip was very hot when i enable it, is there any other problem in my circuit design?

I‘m looking forward to your answer and solution!

attached information:

Schematic diagram

pcb design

Hello, Dong,

"why the aout is 4V"

What is the input voltage? Would you make the input voltage to 5V? 

Would you share VM; VOUT and STEP pin voltage waveform in a same scope picture? 

Regards,

Wang Li 

Hello, Wang，

You misunderstood what the 4V means. As I've said, since there is no current monitoring equipment, I put a 3 ohm observation voltage in series between the stepping motor and the aout1 pin of the drive board.The 4V is the votage on the 3 ohm resistance，which means the current is 4 /3 A. And  the vref value I set is 2.4V, so the current of the aout pin shoould be 2.4 / 3 = 0.8 A, which the bout current is. And the VM supply is 24V, STEP pin voltage waveform is shown in the figure I've given.

Hello, Dong,

Please show VM voltage and Aout1 voltage waveform on the same scope picture. The previous test waveform used 5V/div. But, Aout1 peak voltage is much lower than 24V. Please make sure all grounds be connected together.

Regards,
Wang Li 

Hello,Wang,

I think there must be some misunderstanding about the way I test, and I show it in the figure1.As the figure shows, I want to use the voltage on this resistor to calculate the current in the circuit， it is not the output of the Aout1, so the peak voltage is much lower than 24V.

figure1  the way I used to test

And I also make a new test as you required, the test point is shown in the figure2.

figure2 the way I test this time

Since there are only 4 channels, so when I don't input the steppping signal, I connect ch4 with vref input.

At the beginning, I don't input the stepping signal.

And the low voltage of the ch1  fluctuates between 0 and -2V.

Then I connect the ch4 with step pin and input the stepping signal.

Hello, Wang,

 CH1+ is the signal input of the chanel1, and CH1- is the groud connection of chanel1, and so is other channel.

Hello, Dong,

CH1- or CH2- cannot be put on AOUTx (even has 3ohm) because the the scope ground could be connected to the your power source's ground. Would you put CH1- to the schematic ground?

Regards,
Wang Li

Hello,Wang,

This time I put the CH1- and CH2- to the schematic ground as you said, and there is no 3ohm both this time and last time.

Hello, Wang,

I did the same test on the other board, and this is the result:

Hi Dong,

Could you try to drill small holes in places where I have marked with white circles and connect bottom GND layer

with top GND layer with some pieces of wire (to make something like GND vias)? 

It will limit current return paths and might help to solve the problem.

IC is getting hot because of bad pcb thermal design.

Regards,

Grzegorz

Hello, Dong,

Without current probe, it is hard to imagine what happens on BOUTx. Would change the M0 and M1 setting to have a full step setting? And then, run a lower Step pulse frequency to see if the switching waveform is normal or not.

Regards,

Wang Li

Hello, Wang,

I bought the current probe this week,.And the result of test is shown as the figures follow(CH3).Also, the probe was set to the multimeter range to 1mV/10mA.

Figure1 Aout1

Figure2 Bout1

Hello, Grzegorz,

Thanks for your advice.According to it, am i supposed to make some via at the palce you marked to connect the bottom GND and top GND when i design a new board?

Hi Dong,

These 4 vias would be just a quick fix to the existing pcb to check if long return current paths cause problems. If I were to make a new board I would place more stiching vias and made a few more changes like:

- moving C3 and C4 about 1,5mm left so they are connected directly to IC pins PGND and VM the shortest possible way,

- adding GND pin to connector P1,

- eliminating thermal reliefs under IC - very important!

- placing as many thermal vias under IC as possible,

- changing to 4-layer pcb for a start if I wanted to use IC close to its current capacity, inner layers used just as ground planes to dissipate heat.

Regards,

Grzegorz

Hello,Wang,

I'm so sorry that I found I bought a wrong chip on taobao.The mark is 8428e. And also, could I buy the 8428 chip in small quantities from TI?

I apologize to you again for my carelessness and thanks for your patience during this time.