PGA450-Q1: distance measurement offset

Hi Eyal,

The most accurate readings are made possible when the threshold is drawn as close to the noise floor so that the base of the echo is intercepted for the TOF readout. If the top/peak of the echo is intercepted instead, which is more likely to be the case when the object is detected within the decay profile, then there will be a positive-offset to the time-of-flight readout. This is because the peak representation of the echo indicates where the transducer's generated sound pressure level saturates, which may be after 16 pulses. For example, if the peak is at 16 pulses for a 58kHz transducer, then the TOF offset equivalent is [(16/58kHz)*343m/s]/2 = 4.7cm.

Positive-offset readings within the decay ramp-down should be precise, but less accurate. Are your short range readings fluctuating? You can use the PGA450-Q1 Threshold Mapper for EVM-S Example ( e2e.ti.com/.../PGA450_2D00_Q1-THRESHOLD-CALC-for-UART-OTP-EXAMPLE.xlsx ) to help optimize your threshold settings to improve the TOF readout accuracy. You could also create a function to extrapolate the full echo profile partially masked within the decay to estimate where the base of the echo begins to more closely capture the real TOF.

Unless the demo code was modified to reset the free running timer after the saturation decay time was logged, the TOF readout should not have to be offset. Can you send a copy of your FIFO readout and the TOF equivalent for your short range object?

7536.PGA450-Q1 THRESHOLD CALC for UART OTP EXAMPLE.xlsxHi Akeem,

Thanks for the quick response!

Please see attached excel file with a test I performed.

I’ve placed an object exactly 21cm from the transducer and recorded a short distance measurement.

Results are:

• Response for short distance  measurement - TOF == 1562.

- Response Printed in HEX value:

- Byte[0] = 0x06

- Byte[1] =  0x1A

- Byte[2] = 0xDF (checksum)

I’ve also printed the thresholds to make sure the configurations are OK (command (6) in demo firmware):

• Threshold– short distance

0 255

8 116

16 84

32 68

40 52

48 36

49 8

• threshold– long distance

0 255

32 132

192 116

224 92

25 30

Distance calculation and offset:

• Distance according to returned TOF = (TOF/1000000) * 343/2 = 100*( (1562/1000000) * 343/2) = 26.78cm

• Fifo location according to returned TOF  (TOF/1us)/(DOWN_SAMPLE/1us) = (1562/1000000) / (25/1000000) =  0.001562 / 0.000025 ~= 62

• Fifo cross point according to excel graph is 52. According to distance equation  dist(cm) = 100* ((fifo_index * DOWN_SAMPLE * 1us)*343/2) = 100*((52*25*1us)*343/2 ) ~= 22cm

Any idea why do I see this offset? Are my calculation correct or do I need to consider any other paraments?

Thanks,

Eyal

0513.7536.PGA450-Q1 THRESHOLD CALC for UART OTP EXAMPLE.xlsx

Hi Akeem,

I'm using UART connection.

I've measured the distance to an object at 2[m] and as you assumed, the offset does reduce to ~1cm.

Calculated fifo location was 237 and actual fifo location according to fifo graph was 235 (see attached file).

At ~1m the offset is about 2cm so it seems that the offset gradually reduces as the object is further away from the sensor and very small when the distance is above 2m.

if I understood correctly, the problem is somewhere in the free running timer logged time or because the demo is trying to calc the TOF in real-time.

If this is the case, I will try to calculate the distance after the fifo has been filled using the distance equation.

Thanks,

Eyal

Hi Akeem,

If possible, I would like to consult with you on another issue,

We are working on an application for tanks fill level monitoring and tested our solution using the attached Murata transducer.

since Murata are working strictly with the automotive industry, we are looking for a transducer that will best fit our application.

distance measurements should be for a range of at least 4m using the shortest possible beam angel to avoid reflections of the tank walls.

Any suggestions?

Thanks,

Eyal

Hi Eyal,

I have had much success using the Prowave 400EP250 for long-range narrow field-of-view detection. This will be most ideal for your tank-level application to avoid false positives from the walls.

If you notify Murata that you are using TI's PGA450 or PGA460 ultrasonic reference design as your working solution for a non-automotive application, they usually make an exception to their automotive-only transducer supply rule.

You can find a full listing of transducers which you can filter and sort by parameter at:
e2e.ti.com/.../620399

Hi Eyal,
An additional note about reducing/focusing the field-of-view of a transducer:
You can also use a mechanical horn to enclose any transducer to focus its ultrasonic echo propagation in the forward facing direction only.
The concept of the horn is explained / depicted on the PGA460 FAQ #4.4.9: www.ti.com/.../slaa733.pdf