How to Determine Receiver Gain and Take Advantage of High SNR Receivers

The signal-to-noise ratio (SNR) performance of ultrasound receivers has been greatly improved in recent years. The main progress is reflected in low-power ADC technology, helping users to upgrade the system from 10-bit to 12-bit, or even higher ADC. At the same time, VGAs with low output-referred noise were introduced to take full advantage of these ADCs. When these new ADCs and VGAs are integrated into the ultrasound receiver, the SNR can be effectively improved. With the wide application of a new generation of high SNR receivers, the system can support B-mode harmonic imaging and pulse Doppler imaging to obtain beneficial system performance.

By John Scampini, Executive Director, Industrial Communications and Ultrasound Business Unit, Maxim Integrated

The signal-to-noise ratio (SNR) performance of ultrasound receivers has been greatly improved in recent years. The main progress is reflected in low-power ADC technology, helping users to upgrade the system from 10-bit to 12-bit, or even higher ADC. At the same time, VGAs with low output-referred noise were introduced to take full advantage of these ADCs. When these new ADCs and VGAs are integrated into the ultrasound receiver, the SNR can be effectively improved. With the wide application of a new generation of high SNR receivers, the system can support B-mode harmonic imaging and pulse Doppler imaging to obtain beneficial system performance.

To improve SNR, the gain range of newer ultrasound receivers must be changed, which poses problems for users of older, low SNR receiver designs.

This article describes how to determine the receiver gain and the negative impact on the receive SNR when the receiver gain is set too high. The article also discusses how to properly optimize the dynamic range of digital beamformers, filters, detectors, and compressed signal mapping. After the above optimizations are implemented, the system will maximize the advantages of the high SNR receiver and greatly improve the diagnostic performance.

Calculate Ultrasound Receiver Gain

Figure 1 shows a typical high-performance ultrasound receiver configuration consisting of an LNA, VGA, anti-aliasing filter (AAF), and ADC. The LNA buffers the input signal and provides enough gain to overcome noise from subsequent circuits. In a properly designed receiver, the noise of the LNA largely determines the noise of the overall receiver configuration. The VGA stage provides the necessary variable time gain control to adjust the large dynamic range of the input signal to the limited dynamic range of the ADC. The AAF provides the necessary filtering to ensure that out-of-band noise and signals do not alias into the signal bandwidth, thereby corrupting reception performance.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 1. Block diagram of a typical ultrasound receive path. Example taken from the MAX2082 Octal Ultrasound Receiver.

In Figure 1, the maximum and minimum gains of the receiver are 44.7dB and 5.9dB, respectively. The question now is how to choose the gain?

The minimum gain of the receiver is chosen to ensure that the LNA does not saturate the ADC near-field at maximum input. For the MAX2082 receiver, when the LNA gain is 18.5dB, the maximum input signal is 330mVP-P; the maximum input range of the 12-bit ADC is 1.5VP-P. Therefore, the minimum receiver gain requirement is no more than 20×log(1.5/0.33), or about 13.2dB. For the MAX2082, the minimum gain is actually 5.9dB, providing 7.3dB of additional headroom.

The maximum gain of the receiver is chosen so that the output noise contribution of the combined VGA, AAF, and ADC circuits does not significantly affect its noise figure. To ensure that this does not happen, the output noise of the receiver at maximum gain must be at least 10dB greater than the combined noise contribution of these noise sources. 10dB is a generally acceptable “empirical parameter”. When the above conditions are met, the noise contributions of the VGA, AAF and ADC generally reduce the receiver noise figure within 0.25dB, which is generally considered acceptable. Figure 2 shows the output noise versus gain of the MAX2082 receiver.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 2. MAX2082 total output noise versus gain.

Figure 2 shows that in the MAX2082 transceiver, the receiver noise floor at low gain is about 50nV/rtHz. This noise originates from the output noise of the 12-bit ADC, VGA, and AAF. With a properly designed receiver, the ADC is the main contributor to this noise. In the MAX2082, the noise floor of the ADC is approximately 42nV/rtHz; if the total output noise is 50nV/rtHz, the noise contribution of the circuit before the ADC is small. Assuming that the source impedance of the transmitter is 200Ω and the matching resistance of the receiver is 200Ω, the input referred noise is approximately 1.0nV/rtHz. Therefore, the receiver maximum gain needs to be at least 20×log(50/1)+10dB, or about 44dB. The MAX2082 was chosen for a maximum gain of 44.7dB, and the transceiver meets this criterion with 0.7dB of headroom. As can be seen from Figure 2, the measured noise at the maximum gain is 190nV/rtHz, which is 11dB higher than the 50nV/rtHz noise voltage at the minimum gain.

To further illustrate this concept, Figure 3 shows a graph of the receive gain for the MAX2082.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 3. MAX2082 gain.

In this example, it is worth noting that we assumed an LNA gain of 18.5dB, as this is usually the most commonly used LNA gain setting. This gain setting provides adequate LNA input range and very good noise figure. In most cases, higher LNA gain settings tend to reduce the LNA input range and limit near-field imaging with increased noise figure margin. For example, increasing the LNA gain by 6dB typically reduces the input range by a factor of two. However, lower LNA gain allows a larger input range but reduces noise performance to unacceptable levels.

Undesirable effects of too much receiver gain

For a typical receiver using a 12-bit ADC, such as the one integrated in the MAX2082 transceiver, it is not necessary to increase the maximum gain beyond 44.7dB. At this gain level, a good noise figure is obtained. Further increases in gain will not result in a corresponding increase in receive sensitivity or noise figure.

It is now easy to see why a low SNR receiver requires more gain. The ADCs in these receivers have a higher noise floor, assuming roughly the same maximum input range of the ADCs; therefore, to maintain a good noise figure, the receivers must have greater gain. In short, if the receiver’s SNR is 10dB lower, about 10dB additional maximum gain will be required to provide the same noise figure performance.

For users migrating from low SNR receivers to high SNR receivers with lower maximum gain, problems can arise if the system is not optimized for these changes. We will discuss the reasons for this below. But now consider why you need to fundamentally limit the maximum gain of a 12-bit high SNR receiver. There is no doubt that we have shown that in a 12-bit high-SNR receiver, the maximum gain as high as a 10-bit, low-SNR receiver is not required. The question remains: Why not increase the maximum gain and gain range of the 12-bit receiver to match the 10-bit receiver, thereby minimizing system problems when porting from low SNR to high SNR? This question is very good. The answer involves practical design limitations of VGA.

Increasing the maximum gain of the VGA essentially results in a corresponding increase in the output-referred noise of the VGA. In a properly designed receiver, the VGA output noise at medium and low gains should be appropriately lower than the ADC noise. If this is the case, the receiver SNR at medium and low gain should be about the same as the ADC SNR – which is what we want. Unfortunately, if we try to increase the VGA maximum gain, the VGA output noise at medium and low gains starts to increase accordingly. When the VGA output noise reaches the level of the ADC noise, the receiver’s SNR begins to degrade.

This phenomenon is easily seen in comparable ultrasound receivers with adjustable post VGA gain amplifiers (PGA) to allow the user to increase the maximum VGA gain output. A closer look at the SNR versus gain curves for these devices shows that the SNR deteriorates when the VGA is operated at a high PGA post-amp gain setting. Therefore, these post-gain amplifiers do little to improve noise figure, have limited benefit, and have a significant detrimental effect on receiver SNR.

system design

The entire ultrasound system is optimized to support the improved SNR of the new device, including the digital beamformer (digital delay and sum), digital filters, detectors, and compression mapping, as shown in the ultrasound receiver block in Figure 4 Sketch.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 4. Simplified block diagram of n-channel ultrasound receiver beamforming.

If the digital beamforming, filter, and compression circuits do not have sufficient dynamic range (ie, enough bits), and/or the compression adjustment of the detected signal used to Display grayscale is not properly adjusted, the high performance of these new receivers cannot be effectively utilized. SNR performance. Furthermore, if these critical circuits were optimized for older low SNR receivers, the results would be as if these newer high SNR receivers did not have sufficient gain or adjustment range.

For clarity, take the typical 64-channel system shown in Figure 5 as an example.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 5. Noise analysis of a simplified 64-channel ultrasound receiver system at minimum VGA gain.

In this example, we assume the MAX2082 transceiver is used. The SNR versus gain curve of a single receive channel is shown on the left side of Figure 5. As can be seen from the curve, the SNR is about 68dBFS at medium and low gain. As expected, the SNR deteriorates as the gain increases; the amplified receiver and transmitter element input noise is greater than the ADC noise. This can also be seen in the output noise versus gain plot of the MAX2082 shown in Figure 2.

The digital beamformer in this example delays and sums the digital outputs of the receiver to produce the digital beamformed output. When summing the ADC outputs in beamforming, SNR will increase by 3dB for every doubling of the number of channels. Therefore, for a 64-channel receiver, the SNR of the beamforming output at low gain will be 68dB + (3dB×log2(64))=86dBFS. The beamformer must maintain at least this dynamic range, so the output should be at least 16 bits to sum all 64 12-bit outputs. The output of the beamformer is typically filtered with a filter matched to the transmitter bandwidth and then detected. These circuits must also maintain the necessary dynamic range. The output of the detector then needs to be mapped to the dynamic range of displayed grayscales that can be used. Figure 6 shows a typical detector to grayscale mapping curve.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 6. Detector output to grayscale mapping curve showing the detector noise level with the VGA at minimum gain.

For proper system design, there is a critical setting: setting the minimum grayscale Display level or black level just above the detector output noise floor at minimum receiver gain. Setting the black level to this point ensures that the maximum dynamic range of the entire receiver, and the output noise of the receiver at medium and low gain, is not visible on the image.

Now, we consider the situation when the VGA is at maximum gain, as shown in Figure 7. At this time, the single-channel SNR is about 59dBFS, as shown in the single-channel SNR versus gain curve in the figure. Therefore, the output SNR of the 64-channel beamformer is 77dBFS. Therefore, the beamformer output noise at maximum VGA gain is approximately 11 dB higher than at minimum VGA gain.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 7. Noise analysis of a simplified 64-channel ultrasound receiver system at maximum VGA gain.

At the maximum VGA gain shown in Figure 7, the detector noise floor should be as shown in Figure 8 relative to a properly set compression curve. At this point, low-level signals close to the noise floor at high gain should be adjusted to levels clearly visible in B-mode. It is worth noting that for low-level detected signals, the adjustment curve should be quite steep in order to make them clearly visible and to enhance the differential grayscale of these low-level signals.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 8. Detector output to grayscale mapping curve showing detector noise levels with VGA at maximum gain.

Through analysis, it is easy to understand that if the beamformer, filter, detector, and grayscale mapping are all optimized for a lower SNR 10-bit receiver, one might think that a high SNR receiver must have a higher maximum gain. When using a low SNR receiver, the detector output noise is higher at lower VGA gains. Therefore, the black level of the grayscale mapping curve must be set high to ensure that this noise is not visible on the screen. However, if the receiver is changed to a high SNR 12-bit receiver, the small signal at the maximum VGA gain will be below the black level of the compression curve; this will manifest as insufficient receive gain.

Another problem associated with the use of high SNR receivers and the system’s temporal gain control (TGC) is that in a typical B-mode ultrasound image, the temporal gain control is adjusted so that the same type of tissue has a consistent grayscale from near to far in the image. Spend. To ensure consistent grayscale, the necessary TGC gain adjustment range is about 50dB. From our previous analysis, the required gain range for a high SNR transceiver like in the MAX2082 transceiver is only about 39dB. Then this analog gain adjustment range is clearly insufficient to provide the necessary TGC range.

Therefore, systems using high SNR receivers must employ digital gain adjustment techniques to provide additional TGC gain adjustment range. A software-controlled digital attenuator is usually installed after the beamformer to provide the necessary additional adjustment range. Figure 9 below shows a system block diagram with digital and analog gain circuits. The figure shows how the analog receiver VGA and digital gain adjustment can be used in combination to provide sufficient adjustment range. For lower TGC gains, adjust digitally using a digital attenuator. In this example, the lower 12dB adjustment range is achieved using this technique; for TGC gains that are 12dB lower than the TGC range, gain adjustment is achieved using the analog VGA in the receiver.

How to Determine Receiver Gain and Take Advantage of High SNR Receivers
Figure 9. Combined analog receiver VGA and digital TGC gain adjustment.

in conclusion

With the proliferation of new high SNR ultrasound receivers, users need to ensure that the system is properly designed to take full advantage of the improved technology.

If high PGA and LNA receiver gains are chosen to compensate for incorrect system design, the SNR and LNA input range of these advanced receivers will be lost. The user must ensure that the receive dynamic range is maintained throughout the digital beamforming, filtering, detection, and conditioning paths, and that the signal is correctly adjusted to the grayscale display range. As next-generation receivers continue to improve, designers must also utilize a combination of digital and analog gain adjustment techniques to ensure the necessary TGC range. It is hoped that this article will provide readers with a clearer understanding of these issues and make it easier for users to take full advantage of the performance benefits of new, high SNR receivers.

The Links:   TT250N16KOF LTM09C012

Related Posts