DRV10866EVM: Start-up sequence intermittent with high inertia

Q1) Is the external 6 ohm phase to phase resistance needed or can we use the the current control

A1) You can use either method, but I would recommend the Rcs so you are not wasting extra power on the resistances. 

CS resistor = 3.8k ohms means that max current (A)  = 3120/3.8kohms = 821mA. Keep in mind that the driver when starting from standstill, applied 100% duty cycle in open loop regardless of the pwm input duty cycle, so if the motor isnt spinning, it is either because the current is too little, or the motor is not detecting bemf properly.

Q2) Would this additional resistance be detrimental to the start-up bemf detection?

A2) I am not sure if it would be detrimental to startup bemf detection, so the best thing would jsut be to try out startup without the resistors and using the current limit resistor to limit current.

Q3) How small can I make the current sense resistor? 

A3) It depends on the torque load required to spin the motor. Thus it needs to be determined experimentally. 

When the chip goes into current limit mode does it still supply the positive current drive? I thought I read on the forum that in limit mode the high-side is turned off and the low-side is limited to the short circuit current.

Presently I have no capacitance on the winding's. Would it help to add 3.9nf on each of A, B and C to ground?

1) When the chip goes to current limitting mode, it doesnt shut off the driver entirely to my knowledge. It basically does cycle-by-cycle current limiting to my knowledge, but I can confirm. According to the datasheet "During overcurrent protection, the DRV10866 only limits the current to the motor; it does not shut down the device.". 

2) Capacitances may help to lower noise on the phases to read bemf.

I added 3900 pf from each motor winding to ground. It may have helped a little but not much. Here is a photo of the Fg pin and a motor winding at start-up. The Fg has the pull-up on the DRV8066 pin. The Fg pin goes to a input pin on an FPGA which is why it probably gets pulled to +3.8V. If I give the motor a hand-spin it will start up and Fg toggles normally (at 360Hz for this motor).

Note: Fg goes high and does not toggle. [Q] What bemf level does the DRV8066 look for? what amplitude & hysteresis? Didn't see this in data sheet or app note.

Hey brian,

1)      FG will stay high during open loop and during lock. So if fg doesn’t change, it could mean your stuck in open loop or stuck in fault. Since FG isn’t changing, the driver doesn’t seem to go to closed loop.

2)      It could be that there isn’t enough current to spin the higher torque motor in open loop . It would be useful to see how much current is applied to spin the motor. I know the customer added 6 ohms phase to phase so full current should be 5V/7 ohms phase to phase = 714mA, so I want to see if this is the current applied by looking at an oscilloscope capture of the motor phase current.

Let me discuss with the team and follow up with you further.

Here is an update on my issue. First, here is a YouTube link showing the problem: 

The video is for a motor with 3.3 ohms added to each winding (total ~7.6 ohms phase-phase). The pseudo center tap is 3.0k ohms.

I now have built 3 motor board setups and I got time to test an array of configurations. I was looking at the premise that if the driver goes into current limit mode the start sequence won't be successful. I varied the current limit resistor and the series winding resistance to get current that were at the limit value or slightly below. I did not get a clear indication this was the root problem. Even with 821mA limit current and 658ma motor current the start-up was slightly better but needed a hand "push-start".

I then started looking at the center tap values. I had previously started at 10k and worked my way down to 3k. Now I tried 2k and 1k, I was able to get the motor to self-start with the center-tap resistor 1k.

To summarize, my present best configuration is:

1. Center tap 1k.

2. CS 3.8k (Ilimit 821mA).

3. R series 3.9 ohms in each winding (phase-phase 8.8 ohms). Note Imax/Ics = 0.69

I did this testing on one board and need to bring the other 2 boards up to the same level. I will leave this issue open until I test more boards and motors.

Brian

Thanks for the more details brian. Let me come back to this tomorrow.

Hey Brian,

     It is great to hear that decreasing the center tap resistors until the motor is able to spin, as instructed in the datasheet, helps. Let me know if this solutions works.

If it doesnt, let us actually look at the motor phase current waveform to verify how much current is required to spin the motor at startup

Sanmesh,

Do you have any details of the circuit at the pseudo center-tap? I created a spice simulation of a DC motor for looking at the start sequence. In my model changing the center-tap resistors does not change the center tap voltage.

I will be doing more testing to day looking for why the startup failed when I added the light inertia and mounted the motor in its enclosure. These were unexpected results.

I do not have any details immediately, I think theoretically, the center tap resistance shouldnt change the center tap voltage, but I will check with my coworkers why decreasing it might allow the motor to spin.