DRV8873: Interfacing with MCU

Hey Dmitry,

Dmitry Marukh said:

1. In documentation for DRV8245 it is suggested to add capacitor to IPROPI pin  (8.3.4.1 Over Current Protection). There is no such suggestion in DRV8873 documentation. Does this mean the capacitor will not be beneficial for DRV8873?

In DRV8245 the CIPROPI is listed as optional, and that its purpose is to slow down the ITRIP regulation loop.  Since both devices use current mirror for IPROPI and were released near the same time I would guess that you could achieve the same effect using a capacitor on DRV8873 as well.  However, it is not necessary on either device, and only testing would really show whether it had that effect on DRV8873 or not.  So maybe beneficial depending on your use case but not needed.  

Dmitry Marukh said:

2. For low speed ADC it is necessary to have low-pass filter on IPROPI output. What is recommended solution, simply put a capacitor on Rsense, or add RC filter after Rsense?

Recommend a RC filter, per this E2E post. Putting a capacitor on Rsense seems like it would be the same as the CIPROPI discussed above. 

Dmitry Marukh said:

3. If we use PH/EN mode and connect two IPROPI pins of DRV8873 to the same Rsense resistor, will the output voltage behave the same way as on DRV8245? Specifically, since one of the high side FETs is always active, does this mean we will always see motor current on the output regardless of specific state of PH and EN signals?

Let me ask a coworker this one, give me a day or two to get back to you. 

Dmitry Marukh said:

4. The DISABLE (DRV8873), DRVOFF (DRV8245) and nSCS pins are internally pulled to 5V VDD. However DRV8245 documentation says "Internal pull-up resistance to VDD (reverse current blocked) ", while DRV8873S simply says "Internal pullup to DVDD". Does this mean DRV8245 can be directly connected to MCU pins and DRV8873S requires some kind of overvoltage protection on the MCU?

No, DRV8873 doesn't need additional overvoltage protection for MCU side, though we do often recommend putting a ~1k resistor in series with the signal to/from MCU for current limiting.  Though if there is a concern for reverse polarity you might want to read this article:  Protecting automotive motor-drive systems from reverse polarity conditions

Regards,

Jacob

Hello,

For #3, you can just connect the two IpropIx pins together to a single resistor and then you will get the sum of the currents in HS1 and HS2.  As long as one high-side is active, you will get this current on the single resistor connected to the parallel connected IpropIx pins.

Regards,

Ryan

Dmitry Marukh said:

1. In documentation for DRV8245 it is suggested to add capacitor to IPROPI pin  (8.3.4.1 Over Current Protection). There is no such suggestion in DRV8873 documentation. Does this mean the capacitor will not be beneficial for DRV8873?

Hi Dmitry,

Do you plan to use ITRIP current limit feature to protect your load? If not then OCP is guaranteed to trigger in a short circuit event. If you do use ITRIP, then the additional cap will be useful for both 8245 and 8873 drivers to ensure OCP wins over ITRIP service.

Brian

Hi Jacob,

thank you for answering my questions.

Regarding #4, do I understand correctly that the "DRV8873 doesn't need additional overvoltage protection for MCU side" statement relies on high values (100-400k per documentation) of internal pull-up resistors to not damage 3.3V MCU outputs by 5V VDD voltage?

Also, the documentation for DRV8245S says (in 7.3) that maximum nSCS voltage is VnSleep + 0.5V i.e. 3.8V with nSleep at 3.3V. However in 7.5.3 and 8.2.2 it also says that it is pulled up to 5V VDD. This does not look right to me.

Hi Ryan,

As it happens, it seems I did not express main point of the question clear enough. I know from the documentation that there will be a sum of currents on the connected IPROPIx. What I asked about is the relationship between the current and PH/EN control mode. In this mode a) at least one of the high side FETs is always conducting and b) they never conduct in the same direction. To me this means that connected IPROPIx should ALWAYS represent motor current, not a "sum" of anything. The documentation for DRV8245 says precisely that at the end of 8.3.3.2.

What I wanted to know is if two DRV8873 IPROPI pins connected together will behave exactly as a single pin of DRV8245. Basically, if I need any synchronization between ADC sampling and PH/EN switching, or I can use simple RC filter and ~10 sps ADC to always read correct motor current.

Hi Brian,

I am planning on

a) dynamically configure ITRIP level for specific load

b) use ADC to monitor IPROPI for unexpected changes, e.g. load disconnected from the motor due to mechanical failure

c) use OCP to protect the driver chip itself

I hope all three can work together as required.

Dmitry Marukh said:

I am planning on

a) dynamically configure ITRIP level for specific load

b) use ADC to monitor IPROPI for unexpected changes, e.g. load disconnected from the motor due to mechanical failure

c) use OCP to protect the driver chip itself

Hi Dmitry,

Then a cap should be added to IPROPI pin on these drivers.

Brian

Hi Brian,

Yes, I already added 47nF cap parallel to 2k resistor on the schematic.

However I also added 5.1k and 470nF RC filter between IPROPI and ADC input. I wonder if this additional capacitance will affect the functionality of ITRIP or OCP.