DRV8873: DRV8873H OLP

Hi Ishiwata-san,

How is the customer testing OLP? Is the DRV8873H detecting an OLP or OLA?

Is there enough current in the load to create a 450mV voltage (Vola) across one of the high side body diodes? 

The OLP test is performed:

1)  at power-up when the nSLEEP pin is set to a logic high or

2) when the nSLEEP pin transitions from a logic low to a logic high.

OLA may not operate properly at low currents.

OLA does not operate properly during low side recirculation.

Ishiwata,

"Customers say that OLP is detected even when the motor is connected"

OLP is Open-Load Detection in Passive Mode. It is different than OLA. The note in your post is for OLA.

For DRV8873H OLA, if low-side current recirculation is done with independent PWM control, an open load condition is detected even though the load is connected. To avoid this false trip, the OLD must be disabled by taking the nOL pin high.

Since we are talking about Hardware version of DRV8873, OLA can also pull down the nFault pin which makes that customer think OLP doesn't work. So, can you keep all FETs and check OLP status?

Hi　Team,

Thank you for your reply.

Customers do not understand the difference between OLA and OLP.

As OLD, is there a way for customers to know either OLA or OLP?

In the case of OLA, detection is not performed if the nOL pinl is set to high.
If so, does OLP detect it?
Is it possible to set by register (IC4 Control Register (address = 0x05))?

Also, if you want to detect motor open, do you use OLA or OLP?

Best Regards,
Ishiwata

Hi Ishiwata-san,

OLD is Open Load Detection. Open load detection works in both standby mode (OLP) and active mode (OLA). OLP detects the presence of the motor prior to commutating the motor. OLA detects the motor disconnection from the driver during commutation. 

OLP is Open Load Passive detection. Please refer to Figure 19 and the last 3 paragraphs of section 7.3.2.4.1 for a description of how OLP operates when the device is exiting sleep or powering up with nSLEEP set to a logic high. It is expected that the outputs are not enabled at this time.

OLA is Open Load Active detection. The operation of OLA is described in section 7.3.2.4.2.

----------------------------------------------------------------------------------------------

As OLD, is there a way for customers to know either OLA or OLP?

If nFAULT asserts as the device exits sleep and the outputs are disabled, the fault is OLP. If nFAULT asserts when the outputs are active, the fault may be OLA. Other faults could cause nFAULT to assert during normal operation.

In the case of OLA, detection is not performed if the nOL pinl is set to high.
If so, does OLP detect it?

If nOL is set to high, both OLA and OLP are disabled. From section 7.3.2.4.2 of the datasheet:

In the hardware version of the device, OLA mode is active when the nOL pin if left as a no-connect pin or tied to ground. If low-side current recirculation is done with independent PWM control, an open load condition is detected even though the load is connected. To avoid this false trip, the OLD must be disabled by taking the nOL pin high; however, both OLA and OLP diagnostics will be disabled.

Is it possible to set by register (IC4 Control Register (address = 0x05))?

The register is not available in the DRV8873H.  The DRV8873S provides access to this register and others..

Also, if you want to detect motor open, do you use OLA or OLP?

For the DRV8873H, you must use both. nOL enables both OLA and OLP.

Hi Rick -san

Thank you for your reply.

I have not understood it yet. I'm sorry.

The customer wants to detect no motor connection, but detects OLD even if the motor is connected.
In that case, how should I cope? Do you have any content to confirm with the customer?

Does the customer need to change the nOL pin high and low on the CPU as needed?

Best Regards,
Ishiwata

Hi Ishiwata-san,

How is the motor connected to the DRV8873H?

Is a single motor connected across OUT1 and OUT2?

When the device exits sleep does nFAULT assert? If so, this is OLP.

Is the fault asserting as the motor is being driven? This is OLA.  If so, please provide scope captures of the inputs and outputs. This could be a case where the current is too low to create a voltage greater than Vola when PWM'ing.

If the problem is an OLA fault, a compromise is to perform OLP when the DRV8873H exits sleep. If no fault the motor is connected. Then connect nOL to a logic high while running the motor.

Hi Rick -san

Thank you very much for your reply.

I confirm with the customer whether an error occurs while driving the motor.
I also understand your response content.

"This could be a case where the current is too low to create a voltage greater than Vola when PWM'ing."

There is an additional confirmation.
There is with perform OLP in your plan, but

"a compromise is to perform OLP when the DRV8873H exits sleep. "

What does perform OLP mean specifically?

Best Regards,
Ishiwata

Hi Ishiwata-san,

OLP is Open Load Passive detection. Please refer to Figure 19 and the last 3 paragraphs of section 7.3.2.4.1 for a description of how OLP operates when the device is exiting sleep or powering up with nSLEEP set to a logic high. It is expected that the outputs are not enabled at this time.

OLP is used to detect no motor before trying to drive the motor.

If you have a specific question about the description in the datasheet, please ask.

Hi Ishiwata-san,

Was your question answered by the last reply?

Hi Rick -san

I'm sorry without contacting you.

I answered the customer.
We are currently waiting for customers to contact you.
I will contact you When a customer comes in contact.

Best Regards,
Ishiwata

Hi Ishiwata-san,

No worries. Please let us know if the customer has additional questions.

Hi Rick -san

I appreciate your support.

I made a confirmation to the customer.

OLD is occurring in OLA. Also, VOLA is less than 450mV.
I think OLA is occurring because VOLA is less than 450mV.

In this case, how to make VOLA more than 450mV?
Is it a measure to increase the current flowing to the high side?

Best Regards,
Ishiwata

Hi Rick -san

Thank you very much for your support.

I will contact the customer. Please wait for a reply.

Best Regards,
Ishiwata

Hi Ishiwata-san,

We will await the customer reply.

Hi Ishiwata-san,

The thread has been marked "TI Thinks resolved" for tracking purposes. If the customer has additional questions, please reply to this thread or if locked select "+ Ask a related question"

Thank you.