“Initially, video filters were passive LC circuits surrounded by amplifiers. By combining amplifiers with RC filters, smaller and more efficient designs can be achieved. In addition, the sensitivity analysis and predistortion methods developed in the 1960s overcome the poor performance, which caused the early video filters to lose their good reputation.
“
Initially, video filters were passive LC circuits surrounded by amplifiers. By combining amplifiers with RC filters, smaller and more efficient designs can be achieved. In addition, the sensitivity analysis and predistortion methods developed in the 1960s overcome the poor performance, which caused the early video filters to lose their good reputation.
This document introduces the basic principles and characteristics of filters. More importantly, it shows the filter design guidelines for video applications and how to improve them.
High-performance operational amplifiers and special software for PCs can realize the design of broadband active filters, but these advantages cannot meet the requirements of any specific application. For video filters, specific applications and signal formats add nuances to each circuit design.
Phase linearity and group delay
The phase linearity of the filter is specified as the envelope delay or group delay (GD) versus frequency. A flat group delay means that all frequencies are delayed by the same amount, thereby preserving the shape of the waveform in the time domain. Therefore, the absolute group delay is not as important as the change in group delay. Separate specifications called inter-channel differences should not be designated as “time coincidence” and should not be confused with group delay.
Although the video does not want to do this, how many sets of delay changes can be accepted and why? The answer depends on the application and video format. For example, ITU-470 very loosely regulates the group delay of composite video. However, ITU-601 strictly regulates it to ensure the stability of generations, including MPEG-2 compression and controlling phase jitter before serialization.
Design of anti-aliasing filter For anti-aliasing filters, the selectivity is determined by the ITU-601 template, as shown in Figure 1. The specified bandwidth is 5.75MHz±0.1dB, the insertion loss at 6.75MHz is 12dB, the insertion loss at 8MHz is 40dB, and there is a group delay variation of ±3ns on the 0.1dB bandwidth. For analog filters only, such performance is too difficult, but 4 times oversampling will require modification to 12dB at 27MHz and 40dB at 32MHz.
Using software or normalization curve 8, it can be found that a 5-pole Butterworth filter with a -3dB bandwidth of 8.45MHz can meet the requirements of selectivity, but cannot meet the requirements of group delay. For the latter, a delay stage is required. For this, the important operational amplifier parameter is 0.1dB, 2VP-P bandwidth9. This number is applied to Equations 1 and 2 to obtain an accurate design. The schematic of this application is based on 4x oversampling (Figure 2), and the curve shows its gain and group delay characteristics.
Next consider PC video. VESA does not specify a template for anti-aliasing or reconstruction filters. XGA resolution (1024 x 768 at 85 Hz) has a sampling rate of 94.5MHz and a Nyquist frequency of 47.25MHz. For an attenuation greater than 35dB at the Nyquist frequency, a 20MHz 4-pole Butterworth filter can be used to achieve an attenuation (Rauch). Similarly, the MAX4450/MAX4451 were chosen because of their excellent transient response and large signal bandwidth (175MHz for 2VP-P).
The Links: TFD58W03-MM2 SKIIP13AC12T4V1
