One of the biggest challenges facing semiconductor manufacturers is to balance function and performance in the semiconductor process of product manufacturing. Perhaps many people have heard of 90nm process, 45nm process, and so on. These are the specific processes used by semiconductor manufacturers to make new products. Each manufacturer has its own process, and each process has its own advantages.
In order to embed more digital processing functions in a small, low-cost chip, all process designers focus on making digital transistors smaller. But as the process nodes are smaller and the circuits are denser, the analog performance of the system seems to be affected.
It is becoming more and more difficult to implement ADCs and DACs with a dynamic range greater than 96dB on small silicon chips filled with many ASICs. Look at a high-performance set-top box (STB) processor. The advanced STB chip contains a real-time video decompression processor, a system microprocessor, an Ethernet subsystem, a hard drive interface, and many other functions. These modules are mostly digital, and they will all benefit from the reduction of semiconductor process nodes.
But the mixed-signal and analog system modules do not feel so good. Digital processing focuses on speed and volume, and the analog performance of the processor is second. Many times, because the digital part has a very low signal-to-noise ratio (SNR), this can only put a heavy burden on the analog part. In the real world, this means that it is extremely difficult to obtain higher audio performance from the SoC. Many SoCs arrange a little more than 90dB of dynamic range, but because many use differential outputs, this eats up another 3dB of performance.
In the mobile phone system, 90dB can be regarded as a relatively good quality. However, in today's home entertainment environment, 90dB is an ancient technology. Modern audiovisual systems usually provide a minimum performance of 105dB, and better systems require performance up to 120dB or higher.
Moving higher-performance audio converters to the outside is good for home and portable environments. In the low-level 96dB digital-to-analog converter, in order to maximize the fixed quantization noise and audio output difference, each of the 16-bit buckets needs to work at full scale.
Many systems on the market actually need to show users how much decibels the volume has been attenuated (usually seen on AV receivers).
What exactly does the 10dB change mean, corresponding to the increase and attenuation, the user will tell that the sound has increased or decreased by a factor of 2. And 20dB is equivalent to a change of 4 times the volume. This is a considerable change!
In DAC or DSP, any kind of digital volume control (attenuation) will use fewer bits to represent the signal, but there is also an equivalent level of noise. A good enough DAC suddenly became the level as early as the 1980s.
Figure 1 shows a full-scale CD signal being fed into a 100dB converter. The noise level lies below the dynamic range of the audio signal. In other words, the quantization noise on the audio signal is higher than the inherent noise of the converter. Now, let's attenuate the signal by 20dB by DSP or digitally in the DAC. This is equivalent to 10 times the signal attenuation. Now we see that the lower digits of 16-bit 96dB audio will be overwhelmed by the inherent noise in the converter.
The reason why I encourage designers to buy converters with higher performance than the original chip, the reason is to provide a margin for the signal, in other words, it reserves a certain space between the signal and the lowest noise level in the converter . If you need to attenuate 20dB, you can attenuate without affecting the audio. Figure 2 shows this principle. As you can see, even with 20dB of attenuation, you can still experience high-quality CDs.
Figure (a): Digital volume control (attenuation) in a 100dB digital-to-analog converter. (b): Digital volume control (attenuation) in the 118dB digital-to-analog converter.
How does this affect market sales?
At present, most home audio products that use audio DACs have poor volume control in the digital domain, or use analog volume control that cannot be remotely controlled.
The advantage of analog volume adjustment is to reduce both signal and noise, so the signal-to-noise ratio can be maintained. For higher performance audio systems, a series of programmable gain amplifiers can be used. A simple serial word is fed into the system, and the gain is changed at the next zero crossing.
However, there are significant cost pressures for most consumer electronics systems, and positioning the signal outside the noise floor of the system is the key to achieving high performance. A high-performance DAC (with a range of more than 110dB) can provide a high-quality listening experience for all users, and on the premise of ensuring that no difference can be heard, the CD signal has 14dB and higher attenuation space.
If this reason is not sufficient for the DAC, the concept of using a higher-performance ADC instead of the ADC in the SOC is mandatory. In today's home entertainment market, 2VRMS (5.6Vpp) is usually used in the system to transmit audio. However, many systems still use + 4dBu (1.23VRMS) and -10dBv (0.316VRMS). The difference from 0.316VRMS to 2VRMS is as high as about 22dB! Suppose you calibrate the ADC signal chain to receive 2VRMS input. Any old -10dBv signal sounds like it has been reduced by a factor of 4 (or about -20dB below full scale). Under normal circumstances, you can use digital automatic gain control algorithm in this case, and increase the digital gain of about 20dB. But the disadvantage of this algorithm is that along with the signal, the ADC noise floor is also increased by 20dB.
There are two ways to solve this problem, just like DAC. One is to use a better ADC (about 107dB or higher). The other is to use a programmable gain amplifier at the front end to adjust the gain as needed. Higher performance ADCs will provide a sufficiently low noise level, even when a gain of 10 or 15dB is added.
Many ADCs currently on the market include a multiplexer and a programmable gain amplifier. For example, Texas Instruments' PCM185xA series ADCs have a PGA with a programming range of 22dB and a step value of 0.5dB. When the system detects the presence of a -10dBv peak signal, the gain will gradually increase.
What is the goal?
The next step is to look at the price difference. A typical 105dB DAC (PCM1754) will add an additional $ 1.30 in small batches (about 100 units). However, this will allow the designer to provide a clean single-ended output that can be sent directly to a line-driven output and headphone amplifier without the need for additional differential to single-ended amplifier conversion. This saves the designer $ 0.41 (small batch of RC4580).
Provide a good system for everyone?
With a higher quality output, the price difference of only 0.80 US dollars (100 batch units) can truly provide your products with the opportunity to differentiate and stand out in the market. Launching really good products can make you walk in front of competitors. At the same point, the only difference is the price.
Therefore, try to slow down the competition at the bottom of the price ... Please put some thoughts and concerns into the signal chain you designed. Both the customer ’s ears and your sales team will benefit, and most importantly, your company ’s bottom line will benefit greatly from increased revenue due to better products than its competitors.
Author: Dafydd Roche
Audio Marketing Manager
TI
Solar Energy System Related Products
Wuxi Lerin New Energy Technology Co.,Ltd. , https://www.lerin-tech.com