TI PGA460-Q1：集成超声波解决方案，满足AEC-Q100标准

需积分: 10 171 浏览量更新于2024-06-28 4 收藏 1.87MB PDF 举报

身份认证购VIP最低享 7 折!

领优惠券(最高得80元）

资源详情

资源推荐

AFE

GAIN

Initial Gain

(0, GAIN_INIT)

TVG Start Time

(TVG_T0, GAIN_INIT)

TVG Point 1

(TVG_T1, TVG_G1)

TVG Point 2

(TVG_T2, TVG_G2)

TVG Point 3

(TVG_T3, TVG_G3)

TVG Point 4

(TVG_T4, TVG_G4)

TVG Point 5

(TVG_T5, TVG_G5)

End of Record Time

(P1_REC or P2_REC)

0.5 dB

PGA460-Q1

ZHCSFZ7B –FEBRUARY 2017 –REVISED JANUARY 2019

www.ti.com.cn

Feature Description (接接下下页页)

图图 6. Time-Varying Gain Assignment Example

The time value, TVG start time, is expressed in terms of absolute time, and all following TVG point times

(TVG_Tx parameters) are expressed as a delta time between the current and previous point. All gain values are

expressed in an absolute gain value in dB and are unrelated from each other. The final gain setting of TVG Point

5 (TVG_G5) will be kept constant until the end of the echo record time. The time-varying gain assignment is the

same for both presets. A linear interpolation scheme is used to calculate gain between two TVG points. The AFE

gain resolution is 0.5 dB typical.

注注

The time-varying gain changes during the recording are applied only on the record cycle

that follows. If the TVGAIN[0:6] registers programmed to 0x00, the time-varying gain

function of the PGA460-Q1 device is disabled and a fixed gain defined by the INIT_GAIN

during the recording.

The offset on the time-varying gain is controlled through the two AFE_GAIN_RNG bits in DECPL_TEMP register.

For each of the four settings as defined in the Register Maps section, the gain can be varied from 0 to 32 dB

added on top of the offset.

7.3.4 Digital Signal Processing

The DSP block of the PGA460-Q1 device processes the digital data from the ADC to extract the peak profile of

the echo after which the output of the DSP is compared against the programmed threshold to measure the time

of flight for object distance calculation.

图 7 shows the data path of the DSP. Also, the output of the comparator can be deglitched by the

THR_CMP_DEGLTCH[7:4] bits in the DEADTIME register.

Scale _Exponent

if (t Time_Offset)

Output Input;

else

Output (Input Noise_Level) ;





PGA460-Q1

ZHCSFZ7B –FEBRUARY 2017 –REVISED JANUARY 2019

www.ti.com.cn

7.3.4.1 Ultrasonic Echo—Band-Pass Filter

The ultrasonic echo signal is an amplitude-modulated signal with an underlying carrier frequency equal to the

drive frequency of the ultrasonic transducer. The DSP band-pass filter block allows the frequencies outside of the

observed frequency band to be filtered out and therefore reducing the amount of noise influencing the ultrasonic

echo signal.

The center frequency of the filter is automatically adjusted based on the driving frequency set by the FREQ bit

while the bandwidth of the filter can be adjusted from 2 kHz to 8 kHz in steps of 2 kHz by setting the BPF_BW bit

in the INIT_GAIN EEPROM register.

The band-pass filter is a second-order Butterworth IIR type filter. On power up, the PGA460-Q1 device calculates

the coefficients and places them in the BPF_A2_xSB, BPF_A3_xSB, and BPF_B1_xSB registers. These

registers can be overwritten by the user to reconfigure the filter. However, if the FREQ or BPF_BW bit is

changed, the coefficient calculation sequence is rerun and the device rewrites these registers. In case the

FREQ_SHIFT bit is set to 1 (80- to 480-kHz driving frequency range), the band-pass filter coefficients are not

calculated automatically by the PGA460-Q1 device. In this case the MCU is required to write these values

through the UART or USART interface.

7.3.4.2 Ultrasonic Echo–Rectifier, Peak Hold, Low-Pass Filter, and Data Selection

The rectifier, peak extractor, and low-pass filter DSP blocks demodulate the echo signal while outputting a base-

band representation to be compared against the programmed thresholds. These blocks are defined as:

Rectifier This block outputs the absolute value of the input signal since the input signal can be positive and

negative in amplitude.

Peak hold This block holds the peak value of the rectified signal for a specific amount of time required for the

low-pass filter to detect the peak amplitude of the signal.

Low-pass filter (LPF) This block removes any noise artifacts from the echo signal. The LPF is realized as a

first-order IIR type filter. The user can set the cutoff frequency by setting the LPF_CO bit in

CURR_LIM_P2 register from 1 kHz to 4 kHz with a step of 1 kHz.

On power up the PGA460-Q1 device calculates the values of the filter coefficients and places them in the

LPF_A2_xSB and LPF_B1_xSB registers, respectively. The user can overwrite the values in these registers and

reconfigure the filter. In this case, the PGA460-Q1 device does not take any action. However, if the LPF_CO bit

is changed, the coefficient calculation sequence must be rerun and the device repopulates these registers.

7.3.4.3 Ultrasonic Echo—Nonlinear Scaling

The nonlinear scaling block in the DSP data path provides exponential scaling (digital nonlinear amplification) for

the echo signal to achieve a higher SNR. This feature is useful for detecting long distance object where the

amplitude of the echo signal is very attenuated and close to the noise floor.

The nonlinear scaling block performs the following algorithm:

where

• t is the current record time.

• Time_Offset is set by the SCALE_N parameter and is used to select one of the time points corresponding to

threshold points, TH9, TH10, TH11, or TH12 defined in the Ultrasonic Echo—Threshold Data Assignment

section.

• Scale_Exponent is the nonlinear exponent (1.5 or 2) and defined by the SCALE_K bit.

• Noise_Level is the user-set noise level between 0 and 31 in 1 LSB step and defined by the NOISE_LVL bit. (3)

The SCALE_N, SCALE_K, and NOISE_LVL bits are EEPROM parameters in the DSP_SCALE register.

Threshold

Level

(TH_Px_T1,

TH_Px_L1)

End of Record Time

(Px_REC)

247

255

239

±8

. . . . . . . . . . . . . . . . . . . . . . . . . .

Offset Adjustment Range (TH_Px_OFF)

(TH_Px_T2,

TH_Px_L2)

(TH_Px_T3,

TH_Px_L3)

(TH_Px_T4,

TH_Px_L4)

(TH_Px_T5,

TH_Px_L5)

(TH_Px_T6,

TH_Px_L6)

(TH_Px_T7,

TH_Px_L7)

(TH_Px_T8,

TH_Px_L8)

(TH_Px_T9,

TH_Px_L9)

(TH_Px_T10,

TH_Px_L10)

(TH_Px_T11,

TH_Px_L11)

(TH_Px_T12,

TH_Px_L12)

Initial

Segment

I II III

±8

PGA460-Q1

www.ti.com.cn

ZHCSFZ7B –FEBRUARY 2017 –REVISED JANUARY 2019

注注

The nonlinear scaling block can be applied to Preset1 and Preset2.

7.3.4.4 Ultrasonic Echo—Threshold Data Assignment

The PGA460-Q1 threshold assignments are organized in two presets: Preset1 and Preset2. Both of these

presets have an independent memory map for threshold segment allocation. The PGA460-Q1 device supports

up to 12 threshold segments for each preset defined by the threshold segment points (TSP) in the

P1_THR_[0:15] registers for Preset1 and P2_THR_[0:15]registers for Preset2.

图 8 shows an example of a threshold assignment.

图图 8. Threshold Assignment Example

As shown in 图 8, each TSP is described in the (time, level) format while Px is the preset number (P1 for

Preset1, P2 for Preset2). Additionally, only the initial segment time parameter (TH_Px_T1) value is expressed in

terms of absolute time, while all following TSP times (TH_Px_Tx parameters) are expressed as a delta time

between the absolute time value of the previous TSP and the absolute time value of the current TSP. The level

values of each TSP (TH_Px_Lx parameters) are all expressed in an absolute LSB-level value and are unrelated

from each other. The TSP level threshold value at any given time moment is determined by the PGA460-Q1

device as a linear interpolation function between the two neighboring threshold segment points

As shown in 图 8, the initial segment has a constant threshold value determined by the TH_Px_L1 parameter

until reaching the start of the first segment and also the 12th segment will have a constant threshold value

determined by the TH_Px_L12 parameter until reaching the end of record time defined by the Px_REC

parameter.

The TH_Px_L1 through TH_Px_L8 threshold parameters are 5-bits wide and the TH_Px_L9 through TH_Px_L12

parameters are 8-bits wide. These sizes help save memory space and at the same time allow higher resolution

for long-range detection of weak echo signals in presence of noise while keeping the range constant across all

TSPs. Because the TH_Px_L1 through TH_Px_L8 resolution is an 8 LSB, a threshold offset is defined to allow

finer adjustment of the threshold map for short-range detection.

PGA460-Q1

ZHCSFZ7B –FEBRUARY 2017 –REVISED JANUARY 2019

www.ti.com.cn

注注

• All calculated values of TSP after adding offset, if negative, are clamped to 0 before

linear interpolation which causes the slope of threshold curve to deviate from expected

value.

• Both Preset1 and Preset2 threshold map parameters are protected by a CRC

calculation algorithm (公式 6).

• At power up or wakeup from low power mode, all threshold registers (Px_THR_XX)

and threshold CRC register (THR_CRC) are not initialized to the default value which

causes a CRC error and sets THR_CRC_ERR bit to 1. This occurrence indicates to

the MCU that the configuration is not loaded properly. Writing to threshold registers

reruns CRC calculation and updates the error bit.

7.3.4.5 Digital Gain

A digital gain feature after the low-pass filtering is implemented to improve the SNR of the received echo without

lowering the threshold values. Because this gain is applied after the band-pass and low-pass filtering, the digital

gain does not amplify the out of band noise. This gain feature can help in suppressing false detection such as

ground reflection and detecting farther objects with more accuracy.

Two sets of digital gain ranges are available: short range (SR) and long range (LR). The SR and LR gain levels

are set using the Px_DIG_GAIN_SR and Px_DIG_GAIN_LR parameters, respectively, in the Px_GAIN_CTRL

by Px_DIG_GAIN_LR_ST parameter to the end of the record period. The SR gain is applied from time zero to

the start of the selected LR-threshold level point.

To prevent false detection of an echo at the point in time where the digital gain is applied, the defined thresholds

are also changed as shown in the example plot in 图 9. Here, the LR gain is applied starting from the threshold

level point 9. If the LR gain is different than the SR gain at threshold level point 8, the threshold level 8 is

multiplied by the ratio between the LR gain and SR gain (DIG_GAIN_LR/DIG_GAIN_SR) 1 µs after the end of

the SR threshold level 9 point. Although this creates a discontinuity in the threshold level, the object detection is

not affected (a false threshold crossing is prevented) because the echo signal is also scaled by the same gain

ratio. After this point, the threshold level is changed to the next set threshold level (point 9 in example below)

using a linear interpolation scheme. The threshold levels should be adjusted by taking the digital gain and the

ratio between the LR and SR gains into account.

剩余120页未读，继续阅读

不觉明了

粉丝: 3420
资源: 5635

TI PGA460-Q1：集成超声波解决方案，满足AEC-Q100标准

TI-PGA460.pdf

fs-md-pga.zip

pga411-q1 替代芯片

单亲遗传算法的优缺点

电脑各硬件之间都有哪些接口或类型

docker 安装pgagent

oracle startup报错ORA-27102: out of memory OSD-00026: ???????????? O/S-Error: (OS 1455) ????????????????????????????

ora-04036，实例使用的pga

ORA-04036: PGA memory used by the instance exceeds PGA_AGGREGATE_LIMIT

oracle查看sga和pga设置成最佳

Shure PGA48与Shure PGA58的区别

WARNING: Heavy swapping observed on system in last 5 mins. pct of memory swapped in [104.35%] pct of memory swapped out [0.06%]. Please make sure there is no memory pressure and the SGA and PGA are configured correctly. Look at DBRM trace file for more details.

SOC_VALUE_ENUM_SINGLE(ES8388_ADCCONTROL2, 6, 3, ARRAY_SIZE(es8388_pga_sel), es8388_pga_sel, es8388_line_values),/* Left PGA Mux */

oracle查看sga、pga使用率

pga放大器配置需要什么

ORacle SGA PGA

conda安装PGA

具体介绍下PGA模块的控制寄存器是PGACR怎么使用

设计基于PGA204/205设计程控增益放大器电路

PGA411 状态机

最新资源