PGA460PSM-EVM: Poor detection results

Hi Josh,

I suspect you are close to getting working results, but there are a few items to be optimized:

1) For the hardware connection, you only need the MSP-EXP430F5529LP to connect to the UART of the PGA460PSM-EVM. You can remove the BOOSTXL-PGA460. This will ensure you are only evaluating the PSM, and that there will be no data collision between to two PGA460 devices. Other than the UART pins, just ensure there is a common ground between the PSM board and the LaunchPad.

2) From your GUI General page snapshot, I see you are mostly using the factory settings from the PGA460. In the factory settings, the problem is that Preset 1 tends to be too weak, while Preset 2 tends to be too strong. Attached is the configuration used as part of the Design Guide results for the PGA460PSM-EVM with the 40kHz PUI Audio transducer. This should serve as a better starting point for the hardware solution of the PSM.

;GRID_USER_MEMSPACE
00 (USER_DATA1),41
01 (USER_DATA2),11
02 (USER_DATA3),11
03 (USER_DATA4),10
04 (USER_DATA5),FF
05 (USER_DATA6),FF
06 (USER_DATA7),00
07 (USER_DATA8),00
08 (USER_DATA9),08
09 (USER_DATA10),20
0A (USER_DATA11),C6
0B (USER_DATA12),30
0C (USER_DATA13),38
0D (USER_DATA14),50
0E (USER_DATA15),80
0F (USER_DATA16),00
10 (USER_DATA17),00
11 (USER_DATA18),00
12 (USER_DATA19),00
13 (USER_DATA20),00
14 (TVGAIN0),9D
15 (TVGAIN1),EE
16 (TVGAIN2),EF
17 (TVGAIN3),2D
18 (TVGAIN4),B9
19 (TVGAIN5),EF
1A (TVGAIN6),DC
1B (INIT_GAIN),03
1C (FREQUENCY),32
1D (DEADTIME),80
1E (PULSE_P1),12
1F (PULSE_P2),12
20 (CURR_LIM_P1),72
21 (CURR_LIM_P2),32
22 (REC_LENGTH),09
23 (FREQ_DIAG),00
24 (SAT_FDIAG_TH),EE
25 (FVOLT_DEC),7C
26 (DECPL_TEMP),8F
27 (DSP_SCALE),00
28 (TEMP_TRIM),00
29 (P1_GAIN_CTRL),29
2A (P2_GAIN_CTRL),29
2B (EE_CRC),76
40 (EE_CNTRL),04
41 (BPF_A2_MSB),89
42 (BPF_A2_LSB),52
43 (BPF_A3_MSB),FC
44 (BPF_A3_LSB),CE
45 (BPF_B1_MSB),01
46 (BPF_B1_LSB),99
47 (LPF_A2_MSB),7F
48 (LPF_A2_LSB),33
49 (LPF_B1_MSB),00
4A (LPF_B1_LSB),67
4B (TEST_MUX),00
4C (DEV_STAT0),80
4D (DEV_STAT1),00
5F (P1_THR_0),41
60 (P1_THR_1),11
61 (P1_THR_2),11
62 (P1_THR_3),10
63 (P1_THR_4),FF
64 (P1_THR_5),FF
65 (P1_THR_6),00
66 (P1_THR_7),01
67 (P1_THR_8),F8
68 (P1_THR_9),20
69 (P1_THR_10),C6
6A (P1_THR_11),30
6B (P1_THR_12),38
6C (P1_THR_13),50
6D (P1_THR_14),80
6E (P1_THR_15),00
6F (P2_THR_0),41
70 (P2_THR_1),11
71 (P2_THR_2),11
72 (P2_THR_3),10
73 (P2_THR_4),FF
74 (P2_THR_5),FF
75 (P2_THR_6),00
76 (P2_THR_7),01
77 (P2_THR_8),F8
78 (P2_THR_9),20
79 (P2_THR_10),C6
7A (P2_THR_11),30
7B (P2_THR_12),38
7C (P2_THR_13),50
7D (P2_THR_14),80
7E (P2_THR_15),00
7F (THR_CRC),F9
EOF

3) The PUI Audio transducer on the PSM has a very wide field of view, so you will likely detect many unwanted objects from the sides are other parts of the room. If you want to focus the beam to the chair only, you have two options. A: Use a different transducer with a narrower field of view, or B: Use a transducer wave-guide / horn around your existing PUI Audio transducer to force the field of view to be more narrow. Option B will be easiest for you to implement and test in the short term since you can create a horn using common items such as the cardboard cylinder of a toilet paper roll or 3D printed plastic model. You can find more details about the concept of the transducer horn on the PGA460 E2E FAQ # 4.4.9 at: PGA460 Frequently Asked Questions (FAQ) and EVM Troubleshooting Guide. For a list of alternative transducer part numbers and list of specifications, see the: PGA460 Transducer & Transformer Listing.

Hi Josh,

The result we collected in the PGA460PSM-EVM Design Guide report were of a 1m tall 75mm diameter plastic pole up to 6 meters away, so I would suspect detecting a person at 2m should be feasible. There must be a reason why your echo results are weaker than usual.

Other items to check:

• Is the excitation driver actually generating up to ~120Vpp at the transducer? To check this, use an oscilloscope to probe the positive terminal of the transducer. You should see a sinusoidal voltage of +/-60V. If not, it is likely that either the transducer, transformer, or part of the PGA460 has become damaged.
• Looking at your setup, I see you have a long power cable routed to the PGA460PSM-EVM. I wonder if noise is coupling onto the power cable, causing distortion of the internal regulators post-VPWR. For example, if the regulated output of AVDD is noisy, then the AFE performance diminishes. You could spot-check a shorter power cable /  more robust ground setup to see if makes a difference. Alternatively, try using a 9V battery to supply the PGA460 to see if this helps.

I agree that the MaxBotix sensor will perform better due to the open-top transducer, which leads me to my next point. The PUI Audio transducer on the PGA460PSM-EVM is intended for harsh environmental / outdoor conditions. Assuming your application is exclusively indoors, you could replace the transducer with an open top for improved sensitivity. Note: the PGA460PSM-EVM is a transformer driven board, and the transformer driver isn't necessarily required for an open-top transducer. We typically recommend a direct-driver for open-tops, but it is still possible to use a transformer with an open-top. Just be sure to limit your preset driver current limit between 50-100mA and try to minimize the voltage at the center-tap of the transformer. 9V should still be OK as long as you limit the driver current limit. Here is a list of transducer part numbers with their specifications. Filter parts by construction type = open:  PGA460 Transducer & Transformer Listing. If you want to experiment with the direct driver, you could use the BOOSTXL-PGA460 EVM since this option is available on the included daughtercard.

Hi Josh,

I've observed similar excitation waveforms for this particular transducer, such that a notch can be present in each cycle of the sinusoidal burst.

The PGA460PSM-EVM transducer matching component values were selected based on a combination of minimizing decay time and maximizing detectable range. The sinusoidal driver integrity may appear to be compromised due to the notches, but the transducer is still able to generate sufficient sound pressure level. If you'd like to clean-up the sinosoidal waveform, you can change the C_Tune matching component (capacitor in parallel to the transducer). This will require some trial-and-error experimentation, but may also impact your min/max range performance.You can see the impact of the inclusion or omission of specific matching pair combinations in Figure 16 of the Ultrasonic Proximity-Sensing Module (PSM) Reference Design for this particular transducer.

The driver current limit definitely has a more significant impact on the peak-to-peak excitation voltage than the center-tap voltage. Spot checking the transducer performance as a sweeping function of current limit versus center-tap voltage, the peak-to-peak barely changed regardless if the center-tap was 7V or 15V. Meanwhile, the peak-to-peak voltage nearly doubles from a current limit of 100mA to 300mA. Note, the center-tap should be limited to a max of 15V because the center-tap value doubles at the OUTA/B pins, which have max ratings of 30V. You can continue using your 9V supply.

Since you need a closed-top transducer for your application, you may want to consider a more directional transducer to better detect human presence, such as the Prowave 400EP250. This transducer will eliminate the need for you to have to use a mechanical horn / wave-guide, and will  give strong forward facing results due to its narrow beam pattern. I think the problem right now is more so the 40kHz PUI Audio transducer and your intended target combination, not the PGA460 device.