PGA460-Q1: Bit widths of digital data processing vs. threshold settings/UART response data

user3263001

Part Number: PGA460-Q1

I am confused as to the bit widths of the digital data processing, the threhold settings, and the UART response data of PGA460-Q1.

The ADC of the device has the 12-bit bit width and all the digital data processing are done in the same 12 bits, I've been assuming.

However, in section 7.3.4.4 (Ultrasonic Echo—Threshold Data Assignment) of the datasheet for PGA460-Q1 (SLASEC8B), the thresholds are seemingly set within the-8 bit bit width (or actually 5 bits for TH_Px_L1 through TH_Px_L8). The UART response details in section 7.3.6.2.1.5 of the same datasheet also shows that the command 'Ultrasonic measurement result' (address 5) receives 8 bit data as the 'Object peak amplitude.'

Do the 12-bit data simply converted to 8-bit data by dropping their 4 LSBs, before being compared to the threshold values and sent out via communication?

Thank you.

over 4 years ago

0 user3263001 over 4 years ago

Prodigy 60 points

I forgot to add that, since there is a rectification process, the 12-bit data should become essentially 11-bits. Maybe three MSBs are dropped to compare against threshold etc.?

0 Mekre Mesganaw over 4 years ago in reply to user3263001

TI__Expert 7495 points

Hi user3263001,

The following video shows example outputs from the different stages of the DSP block:

https://training.ti.com/ti-precision-labs-ultrasonic-sensing-ultrasonic-signal-processing

The digital gain also can play a part in selecting which bits are output, as is mentioned in Section 7.3.7.2 of the datasheet.

Regards,

Mekre

0 user3263001 over 4 years ago in reply to Mekre Mesganaw

Prodigy 60 points

I'm sorry I'm still not sure. Could you elaborate? There does not seem to be any mentioning of bit widths in the video.

0 Mekre Mesganaw over 4 years ago in reply to user3263001

TI__Expert 7495 points

Hi user3263001,

Which bits are kept in the result is dependent on the digital gain settings. The possible digital gain values are 1, 2, 4, 8, 16, and 32. Multiplying by these values is the same as performing a left shift operation by 0, 1, 2, 3, 4, and 5, respectively. The more you shift left, the more MSBits are discarded from the final result and the more LSBits are included in the result. As a result, the selection of the digital gain value determines which bits are kept in the results. Lower gain values would have more MSBits in the results and higher gain values would have more LSBits in the results.

Regards,

Mekre

0 user3263001 over 4 years ago in reply to Mekre Mesganaw

Prodigy 60 points

Thank you for the reply.

I think I understand now, but let me make sure with an example. If the digital gain is 1, then 8 MSBits are selected out of 12-bit A/D data for the purposes of echo dump/comparison to threshold, etc. Am I right?

+1 Mekre Mesganaw over 4 years ago in reply to user3263001

TI__Expert 7495 points

user3263001,

Yes, that is correct.

Regards,

Mekre

Sensors

Sensors forum

PGA460-Q1: Bit widths of digital data processing vs. threshold settings/UART response data