DRV8343H-Q1EVM: Can I use this EVM to test DRV8340H?

Hi Aaron, 

DRV8340H has no current shunt amplifiers, the equivalent on the DRV8343H would be to tie the inputs (SPx, SNx) together to GND and float the outputs (SOx). Since you are using the Hardware version, you cannot disable the CSAs via SPI. The only way to replace DRV8340H on the EVM would be to:

1) Cut the input traces SPx and SNx, and blue-wire those traces to GND
2) Cut the SOx traces so they are floating

An easier solution would be to evaluate the DRV8343S-Q1EVM, which I still see stock on ti.com. Then in SPI you can disable the CSAs through these registers:



Then set the settings the customer would use on the DRV8340H, and this should give the equivalent functionality of evaluating a DRV8340H using the DRV8343S-Q1EVM. 

Thanks,
Aaron

Hi Aaron,

1) Cutting traces on a PCB is using a sharp tool such as a razor, exacto knife, or tweezers to cut the copper trace connection. You would need to cut traces SPx/SNx and SOx near the DRV8343-Q1 IC to split those connections from the rest of their traces. This may be hard to do since the traces are thin and close to other traces, and it is hard to repair those connections once they are broken. 

2) DRV8340 and DRV8343 have same gate driver architecture, there is no functional difference between the gate drive outputs and configurable settings between devices. The EVM is meant to support DRV8343, not DRV8340, or else you'll have to do the steps from above to make it work without running into GUI issues or getting unnecessary overcurrent faults. If you follow the steps to disable the CSAs and set the same settings the customer would use in Hardware on the DRV8340H, then functionally the DRV8343S will act as the DRV8340H. 

Thanks,
Aaron

Hi Aaron, 

Happy New Year to you as well. 

If SPI interface is not used, fault debug is a bit more difficult to determine since the only indication of a fault will be the nFAULT pin going low. In this case, you'll need to use oscilloscope waveforms from ENABLE going high and look at timings and gate driver outputs to determine if the fault could be:
- VDS overcurrent
- VGS faults
- Diagnostic faults (i.e. Open Load passive, open load active, Short to Battery, or Short to GND).

Thanks,
Aaron

Hi Aaron,

Since DVDD only provides 30mA, external ICs should be minimal on this rail, or else you could overload the LDO and cause LDO undervoltage, which can unpower ICs on the DVDD rail (i.e. an MCU). 

You could use an external buck or LDO to generate 3.3V from VM. I don't know much about current amplifiers, that may be possible in this configuration. For recommendations, can you start another E2E thread with the appropriate team?

Thanks,
Aaron

Hi Aaron, 

Hardware files for DRV8343S-Q1EVM are here: https://www.ti.com/lit/zip/slvc710
You can find schematic in there called MD005A(001)_Sch.pdf. 

The DVDD is only used for VSDO (pullup for SDO push-pull driver), which requires less than 5mA current. MSP430 is powered via the USB, which is 5V. The USB powers the FT232RL, which is a USB-to-UART FTDI chip for GUI communication. It has a built in regulator that supplies 3.3V out (3.3V_COM), which is used to power the MCU (3.3V_MCU). This is standard across many of our EVMs that support USB. 

Thanks,
Aaron

Hi Aaron,

Best way to accomplish this is to depopulate 0-ohm resistors (red arrow) between MSP430F5529 MCU on the EVM and populate the 100-mil header row on the EVM (blue arrow). Then jumper over signals from customer's MCU to the header row. 



Thanks,
Aaron

Hi Aaron,

Yes, that's correct. Using 0-ohm resistors is common because it's easy to remove the connection without having to break the trace. The headers make for nice test points as well. 

Thanks,
Aaron

Hi Aaron,

On DRV8343S-Q1EVM, the DRV8343 device is SPI variant and the H/W variant headers are automatically depopulated (see red circles):




To use the hardware device on this EVM, you'll need to:
- Replace the DRV8343S-Q1 with a DRV8343H-Q1
- Depopulate the 0-ohm resistors for the SPI pins (SDI, SDO, SCLK, nSCS) between the MCU and header row: R29, R33, R34, R35  (see blue circle)
- Add headers for the hardware variant jumpers and use a shunt jumper to connect the correct settings as mentioned in the User's Guide

In my opinion, it would be easier to just use the DRV8343H-Q1EVM instead since it has everything already configured for the hardware variant of the device. 

Thanks,
Aaron

Hi Aaron,

Recall back to the beginning of the thread:

Aaron Barrera said:

An easier solution would be to evaluate the DRV8343S-Q1EVM, which I still see stock on ti.com. Then in SPI you can disable the CSAs through these registers:



Then set the settings the customer would use on the DRV8340H, and this should give the equivalent functionality of evaluating a DRV8340H using the DRV8343S-Q1EVM. 

Here's what you can do but use the customer's 6 PWM signals from their MCU, instead of the PWM signals from the MSP430F5529:

1) Populate the header row on the EVM
2) Depopulate these resistors: R23, R24, R25, R26, R30, R31



3) Jumper over the customers 6 PWM signals from their MCU to the associated test points on the header row. 
4) Power on the DRV8343S-Q1EVM and plug into the computer. 
5) Launch the DRV8343-Q1 EVM GUI. 
6) Click "Enable" to set the ENABLE signal high and enable the SPI logic. 
7) Set the SPI settings in the "Registers" tab to the same that the customer would like to set in DRV8340H-Q1. 

IDRIVE pin --> IDRIVEP_HA, IDRIVE_HB, IDRIVE_HC, IDRIVE_LA, IDRIVE_LB, IDRIVE_LC
MODE pin --> PWM_MODE
VDS pin --> VDS_LVL_HA, VDS_LVL_HB, VDS_LVL_HC, VDS_LVL_LA, VDS_LVL_LB, VDS_LVL_LC

You can disable the VSEN_OCP faults from the CSAs by setting DIS_SEN_A, DIS_SEN_B, DIS_SEN_C to 1

After these steps, you should be able to spin the motor using 6x PWM signals and hall sensors externally. This should be the same operation as DRV8340H-Q1, you should be able to ignore the CSA outputs. Recall that this method is the most ideal, since the other method would be to use DRV8343H-Q1EVM and cut the CSA input/output traces and tie the inputs to GND. 

Thanks,
Aaron

Hi Aaron,

Yes, this should work from a hardware and firmware aspect. Thank you. 

If the GUI runs into any software overcurrent limit faults due to the CSAs outputs, we can figure out which settings in the GUI we can toggle to get rid of those so there are no dependencies on the CSAs in the DRV8343-Q1, which would make performance similar to DRV8340H-Q1. 

Thanks,
Aaron

Hi Aaron,

The GUI should be in the Firmware folder on the EVM page. I linked below.

Hi Aaron,

In the registers tab on the left. 

Thanks,
Aaron