本文選題：靈敏放大器　切入點(diǎn)：低功耗　出處：《安徽大學(xué)》2015年碩士論文

【摘要】：靈敏放大器對(duì)SRAM (Static Random Access Memory)電路的性能提升起到無(wú)可替代的作用。它可以檢測(cè)位線上的小擺幅信號(hào)并快速有效地進(jìn)行放大,從而極大地提高SRA M的整體速度。因此,靈敏放大器廣泛地被各類不同的SRAM電路使用。一般來(lái)說(shuō),靈敏放大器的結(jié)構(gòu)可分為兩大類：電壓型和電流型靈敏放大器。電壓型靈敏放大器依靠檢測(cè)位線上的電壓差值來(lái)進(jìn)行放大輸出,它結(jié)構(gòu)簡(jiǎn)單,穩(wěn)定性好,但受到位線負(fù)載電容的限制速度較慢;電流型靈敏放大器檢測(cè)電流差值而進(jìn)行放大,一般速度較快,但功耗高,穩(wěn)定性差。靈敏放大器設(shè)計(jì)中需考慮多種指標(biāo),如速度,功耗,面積以及良率等,且?guī)追N因素彼此影響。在當(dāng)今環(huán)境工藝不斷進(jìn)步、電源電壓不斷降低的情形下,速度和功耗會(huì)逐漸成為較為重要的兩點(diǎn)因素。本文從上述兩點(diǎn)要求出發(fā),首先分析了靈敏放大器設(shè)計(jì)中的重點(diǎn)以及難點(diǎn),提出研究的背景和意義。然后介紹了常用的幾種靈敏放大器的結(jié)構(gòu)和優(yōu)缺點(diǎn),最后通過(guò)對(duì)它們的總結(jié)和分析設(shè)計(jì)了一種低功耗電流型靈敏放大器。該放大器通過(guò)移除電路在非放電時(shí)間存在的的直流通路來(lái)較大幅度地降低功耗。電路的仿真在cadence中使用spectre仿真器進(jìn)行,工藝設(shè)定為65nm,并使用多種工藝角分別仿真。仿真結(jié)果表明,在TT、SS、FF三種工藝角下,同傳統(tǒng)的WTA型靈敏放大器相比,低功耗電流型靈敏放大器所費(fèi)的功耗分別減小了64%、88%和68%；同時(shí),放大器的放大速度相比于WTA型靈敏放大器也有一定的提升,通過(guò)對(duì)瞬態(tài)輸出的分析得到,其放大速度在三種工藝角下分別增快了40%、38%和31%。另外,本文針對(duì)提出的靈敏放大器還給出了兩種可用的改進(jìn)方案電路：輸入端補(bǔ)償電路和抗閩值電壓波動(dòng)電路。主要設(shè)計(jì)于針對(duì)當(dāng)前工藝條件下容易發(fā)生的不穩(wěn)定因素及它們對(duì)電路可能造成的不良影響。其中,輸入端補(bǔ)償電路在原有的電路基礎(chǔ)上增加了兩個(gè)反相器和兩個(gè)控制管,而抗閾值波動(dòng)電路沒有更多的器件添加。兩種方案互有利弊,使用這兩種方案可以有效的防止功能性失效的出現(xiàn),提高放大器的準(zhǔn)確度,但同時(shí)也增加了整體的功耗和面積。
[Abstract]:Sensitive amplifiers play an irreplaceable role in improving the performance of SRAM static Random Access memory circuits. They can detect small swing signals on bit lines and amplify them quickly and efficiently, thus greatly increasing the overall speed of SRA M. Sensitive amplifiers are widely used in various SRAM circuits. Generally speaking, The structure of sensitive amplifier can be divided into two categories: voltage mode and current type sensitive amplifier. Voltage mode sensitive amplifier amplifies the output by detecting voltage difference on the bit line. It has simple structure and good stability. However, the speed limited by the bit-line load capacitance is relatively slow, and the current mode sensitive amplifier amplifies by detecting the current difference, which is generally faster, but with high power consumption and poor stability. In the design of the sensitive amplifier, a variety of indexes should be considered, such as speed, power consumption, etc. Area, yield and so on, and several factors affect each other. With the continuous progress of environmental technology and the constant reduction of power supply voltage, speed and power consumption will gradually become more important two factors. This paper first analyzes the key points and difficulties in the design of sensitive amplifiers, puts forward the background and significance of the research, and then introduces the structure, advantages and disadvantages of several common sensitive amplifiers. Finally, a low power current-mode sensitive amplifier is designed by summing up and analyzing them. The amplifier can greatly reduce the power consumption by removing the DC path of the circuit in non-discharge time. Using the spectre emulator in cadence, The simulation results show that, compared with the conventional WTA sensitive amplifier, the power consumption of the low-power current-mode sensitive amplifier is reduced by 64% and 68%, respectively, in comparison with the conventional WTA type sensitive amplifier, and the simulation results show that the power consumption of the low-power current-mode sensitive amplifier is reduced by 64% and 68%, respectively. The amplification speed of the amplifier is also higher than that of the WTA sensitive amplifier. By analyzing the transient output, it is found that the amplification speed increases by 40% and 31% respectively at the three process angles. In this paper, two kinds of improved circuits for the sensitive amplifier are presented: the input compensation circuit and the anti-threshold voltage fluctuation circuit, which are mainly designed to deal with the unstable factors which are easy to occur under the current technological conditions. And their possible adverse effects on circuits. The input compensation circuit adds two inverters and two control tubes to the original circuit, but no more devices are added to the anti-threshold fluctuation circuit. The two schemes have advantages and disadvantages. Using these two schemes can effectively prevent the appearance of functional failure, improve the accuracy of amplifier, but also increase the overall power consumption and area.
【學(xué)位授予單位】：安徽大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TN722

【參考文獻(xiàn)】

相關(guān)碩士學(xué)位論文 前1條

1 吳維奇;亞閾值SRAM PVT波動(dòng)檢測(cè)與補(bǔ)償設(shè)計(jì)[D];安徽大學(xué);2012年

，

本文編號(hào)：1685648