200807902 UMCD-2006-0025 19504twf.doc/g 九、發明說明： 【發明所屬之技術領域】 田本發明是關於一種混頻器，且特別是關於一種使用摺 豐式電感電容疊接之次諧波混頻器。 【先前技術】 直接 ~ 頻接收為（direct_conversion receiver 或 homodyne receiver)在架構上採用一次的降頻動作，直接將 射頻訊唬降至基頻訊號，因此又稱為零中頻接收器（zero-IF receiver)。此種接收器所產生的多相位本地振盪訊號之振 盈頻率’與射頻訊號的振盪頻率非常接近，因此不僅避免 了鏡像雜訊(image noise)的干擾，且伴隨鏡像雜訊干擾的 消失’射頻訊號與多相位本地振盪訊號進行混波之前就不 需預置一鏡像抑制濾波器（image rejecti〇n mter)。如此一 來’相較於其它接收器架構-例如差外差接收器 (superheterodyne receiver)，直接降頻接收器具有架構簡 單單as片化專優點’而逐漸地在現今的收發器(transceiver) K J 中嶄露頭角。 直接降頻接收器雖具有單晶片化等優點，但其架構仍 存在一些缺點必須去克服。其中直流偏移(DC offset)為其 衍生的問題之一。直流偏移的產生主要是由於混頻器 (mixer)和低雜訊放大器（i〇w n〇ise ampiifier，簡稱LNA)的 輸入端(在此統稱為射頻訊號的輸入端），與混頻器接收多 相位本地振盪訊號的輸入端之隔絕度(isolation)並非無限 大’因此當多相位本地振盈訊號經由穿隧(feedthr〇ugh)效 200807902 UMCD-2006-0025 19504twf.d〇c/g 應’出現在射頻訊號的輸入端時，將和原本的多相位本地 振盪訊，進行自我混波(self_mixing)進而形成直流偏移。此 外，偶-人1¾失真(even_orcjer dist〇rti〇n)也是一個值得注意的 問題，σ因為直接降頻接收器進行混波之前不具有鏡像抑制 濾波，因此伴隨在射頻訊號附近的干擾訊號，經由非線 f生電路所產生的偶次階失真，也會經由穿隧效應而直接傳 送到混頻器的輸出，進而影響真正想接收的射頻訊號。 、 為了解決上述問題，直接降頻接收器採用如圖1所示 的傳統次諧波混頻器細1>}^1111〇11丨(：11^61'，簡稱811]\4)，來 提供良好的隔絕度給多相位本地振盪訊號與射頻訊號的輸 入女而。傳統次譜波混頻器是由如圖2所示之吉伯特混頻器 (Gilbert mixer)衍生而來的。繼續參照圖1與圖2，其中多 相位本地振盪訊號L01包括相移量分別為〇g、9〇g、180〇、 270G之本地振盪訊號L01—〇◦〜l〇1_2700，而多相位本地振 盪訊號L02則包括相移量分別為〇G與i8〇Q之本地振蘆訊 號L02一0G與L02一 180◦。吉伯特混頻器中的每一個n型電 晶體MN9〜MN12，若分別以兩個相互並聯的n型電晶體 取代，並將提供給吉伯特混頻器之多相位本地振盪訊號 L02之振盪頻率，降低至原本振盪頻率的〇.5倍，以形成 由相互並聯的NMOS電晶體所接收的多相位本地振盪訊 號L01 ’將可演變成如圖1所示的傳統次諧波混頻器。例 如MN9由MN1與MN2取代，且MN1與MN2接收‘差 為180G的本地振盪訊號L01J)G與l〇1_180g之振盪頻率， 是MN9所接收的本地振盪訊號l〇2_0g的2倍。如此一來， 200807902 UMCD-2006-0025 19504twf.doc/g 採用傳統次谐波此頻為之直接降頻接收器，不僅可以將多 相位本地振盪訊號之振盪頻率操作在〇·5倍的射頻訊號之 振盛頻率，且還可保有吉簡混_所具有的良好隔絕度。 然而，上述之傳統次諧波混頻器在實際的晶片化過程 中’往往會因製私技術所造成的元件不匹配，進而造成電 路在不對稱的情況下，影響傳統次譜波混頻器所能提供的 隔絕度。因此，如何利用原有的電路架構來提高傳統次諧 波混頻為輸入端的隔絕度，以降低多相位本地振盪訊號的 >W>i(leakage)訊號’與干擾訊號在射頻訊號輸入端所造成 的偶次階失真，已是直接降頻接收器在應用上所面臨的最 大隱憂。 【發明内容】 即有鑑於此，本發明的目的是在提供一種次諧波混頻 為，利用諧振電路將洩漏訊號導向至第一電壓或第二電壓 的f式，讓次諧波混頻器具有良好的隔絕度，進而解決使 2次谐波混頻器之直接降頻接收器在應用上，所面臨的直 流偏移與偶次階失真的問題。 、、曰并ί發明的另一目的是提供一種下轉換器，利用次諧波 所具有的良好隔絕度，減少茂漏訊號所造成自我混 波與偶次階失真，進而提昇使用下轉換器之工作特性。 器，為達成上述及其他目的，本發明提出一種次諧波混頻 x將多相位本地振盪訊號與射頻訊號進行混頻，以 緩衝2頻訊號。該次諧波混頻器包括差動放大單元、電流 早元、以及切換單元。差動放大單元將射頻訊號放大。 7 200807902 UMCD-2006-0025 19504twf.doc/g 電流緩衝單元耦接至差動放大單元，用以將差動放大單元 之輸出δίΐ號之增盈放大。切換單元粞接至電流緩衝單元， 用以依據多相位本地振盪訊號，而將電流緩衝單元之輸出 祝號切換至基頻訊號。 1 依照本發明的較佳實施例，上述之差動放大單元包括 第一諧振電路，電流緩衝單元則包括第二諧振電路。為了 讓次諧波混頻器達到將多相位本地振盪訊號盥射 行混頻，以產生基頻訊號的目的。首先，差動放大單元將 射頻訊號放大，且放大後的射頻訊號利用第一諧振電路導 向至電流緩衝單元，而此時本地振盪訊號所產生的洩漏訊 號，也將利用第一諧振電路導向至第一電壓。之後，耦接 至差動放大單元的電流緩衝單元，將差動放大單元之輸出 訊號之增益放大，並_第二諧振電路將電流緩衝單: 輸出訊號導向至切換單元，衫二譜振電路此時也將本地 振盪訊號所產生的洩漏訊號，導向至第二電壓。最後，士 換早兀依據乡相位本地減歸u，而將電流緩衝 二 出訊號切換至基頻訊號。 备 上述之第一諧振電路與第二諧振電路之1捃頻 射頻訊號之振盘頻率相同，且第—電壓可為二操作、1人 第二電壓可為一接地電壓。 ’、乍电£， 在一較佳實施例中，上述之次諧波混 降頻接收器。 、如通用於直接 從另一觀點來看，本發明另提出—種 將射頻訊號轉換至基頻訊號。下轉換器包括訊號 200807902 UMCD-2006-0025 19504twf.doc/g 次諧波混頻器。其中次諧波混頻器包括差動放大單元、電 流緩衝單元、以及切換單元。訊號產生器用以提供一多相 位本地振盪訊號。叙接至訊號產生器的次譜波混頻器則用 以將多相位本地振盪訊號與射頻訊號進行混頻，以產生基 頻θίΐ號。其中次諧波混頻產生基頻訊號的過程包括，利 用差動放大單元將射頻訊號放大。接著，耦接至差動放大 單元的電流缓衝單元，在將差動放大單元之輸出訊號之掩 益放大。最後，利用耦接至電流緩衝單元與訊號產生器^ 切換單元，依據多相位本地振盪訊號，而將電流緩衝單一 之輸出訊號轉換成基頻訊號。藉此，下轉換器就可達到將 一射頻訊號轉換至一基頻訊號的目的。 上述之下轉換器，在一較佳實施例中，訊號產生器包 括本地振盪器與相位偏移器。本地振盪器用以產生一^地 振盪§fl號，讓串接在本地振盪器與切換單元之間的相位偏 移器，可以將本地振盪訊號轉換成數個不同相移量的本地 振盪訊號，以輸出作為多相位本地振盪訊號。 在一較佳實施例中，上述之下轉換器適用於直 接收器。 貝 本發明因採用差動放大單元與電流緩衝單元組合之架 構，讓次諧波混頻器可利用第一諧振電路與第二諧振電 路，達到將洩漏訊號導向至第一電壓或第二電壓之功效， 如此一來，隨著次諧波混頻器隔絕度的提升，使用次諧 混頻器之直接降頻接收器，所面臨的直流偏移與偶次阡 真也將大幅度地降低。 9 200807902 UMCD-2006-0025 19504twf.doc/g 目的、特徵和優 並配合所附圖式 點能更明顯 ’作詳細說 為讓本發明之上述和其他 易懂’下文特舉較佳實施例， 明如下。 【實施方式】 _ j 3為根據本發明—較佳實施例之次錢混頻器結構 ，匕括差動放大單元30卜電流緩衝單元302、以及 刀換早兀303。電流緩衝單元3〇2 _接於差動放大單元训200807902 UMCD-2006-0025 19504twf.doc/g IX. Description of the Invention: [Technical Fields of the Invention] The present invention relates to a mixer, and more particularly to a subharmonic using a folded-type inductor-capacitor Mixer. [Prior Art] Direct-to-frequency reception (direct_conversion receiver or homodyne receiver) uses a frequency reduction action on the architecture to directly reduce the RF signal to the fundamental frequency signal, so it is also called zero-IF receiver (zero-IF). Receiver). The oscillation frequency of the multi-phase local oscillation signal generated by such a receiver is very close to the oscillation frequency of the RF signal, so that not only the image noise interference but also the disappearance of the image noise interference is eliminated. There is no need to preset an image rejection filter (image rejecti〇n mter) before the signal is mixed with the multi-phase local oscillation signal. As a result, compared to other receiver architectures, such as the superheterodyne receiver, the direct down-converter has a simple architecture and a single advantage. It is gradually in today's transceivers. The slap in the face. Although the direct down-conversion receiver has the advantages of single-chip, etc., its architecture still has some shortcomings that must be overcome. Among them, DC offset is one of the problems derived from it. The DC offset is mainly generated by the mixer and the input of the low noise amplifier (LNA), which is collectively referred to as the input of the RF signal, and is received by the mixer. The isolation of the input of the multi-phase local oscillator signal is not infinitely large. Therefore, when the multi-phase local oscillation signal is via the tunneling effect (200807902 UMCD-2006-0025 19504twf.d〇c/g should be ' When it appears at the input of the RF signal, it will self-mix with the original multi-phase local oscillation signal to form a DC offset. In addition, even-orbit distortion (even_orcjer dist〇rti〇n) is also a significant problem. σ because the direct down-conversion receiver does not have image suppression filtering before mixing, so the interference signal accompanying the RF signal passes through The even-order distortion generated by the non-wired-generation circuit is also directly transmitted to the output of the mixer via the tunneling effect, thereby affecting the RF signal that is actually intended to be received. In order to solve the above problem, the direct down-converter receiver is provided by a conventional subharmonic mixer as shown in FIG. 1}}^1111〇11丨(:11^61', referred to as 811]\4). Good isolation gives the input of multi-phase local oscillation signals and RF signals. The conventional sub-spectral mixer is derived from the Gilbert mixer as shown in Figure 2. 1 and FIG. 2, wherein the multi-phase local oscillation signal L01 includes local oscillation signals L01_〇◦~l〇1_2700 whose phase shift amounts are 〇g, 9〇g, 180〇, 270G, respectively, and multi-phase local oscillation. The signal L02 includes local vibration signals L02-0G and L02-180◦ with phase shift amounts of 〇G and i8〇Q, respectively. Each of the n-type transistors MN9 to MN12 in the Gilbert mixer is replaced by two n-type transistors connected in parallel with each other, and the multi-phase local oscillation signal L02 supplied to the Gilbert mixer is provided. The oscillation frequency is reduced to 〇.5 times of the original oscillation frequency to form a multi-phase local oscillation signal L01' received by the NMOS transistors connected in parallel with each other, which can be evolved into a conventional subharmonic mixer as shown in FIG. . For example, MN9 is replaced by MN1 and MN2, and MN1 and MN2 receive the oscillation frequency of "local oscillation signal L01J of 180G difference" G and l〇1_180g, which is twice the local oscillation signal l〇2_0g received by MN9. In this way, 200807902 UMCD-2006-0025 19504twf.doc/g uses the traditional subharmonic frequency as the direct down-conversion receiver, which can not only operate the oscillation frequency of the multi-phase local oscillation signal at 〇·5 times the RF signal. The vibration frequency, and can also maintain the good isolation of the Ji simple mixed _. However, the above-mentioned conventional subharmonic mixers often suffer from component mismatch caused by the manufacturing technology in the actual wafer formation process, thereby causing the circuit to affect the conventional sub-spectral mixer in the case of asymmetry. The degree of isolation that can be provided. Therefore, how to use the original circuit architecture to improve the isolation of the traditional subharmonic mixing to the input end, to reduce the >W>i (leakage) signal of the multi-phase local oscillation signal and the interference signal at the input of the RF signal The even order distortion caused by it is the biggest concern of the direct down-conversion receiver in application. SUMMARY OF THE INVENTION In view of the above, an object of the present invention is to provide a subharmonic mixer that uses a resonant circuit to direct a leakage signal to a first voltage or a second voltage, and a subharmonic mixer. It has good isolation and solves the problem of DC offset and even order distortion caused by the direct down-conversion receiver of the 2nd harmonic mixer. Another purpose of the invention is to provide a down converter that utilizes the good isolation of the subharmonics to reduce the self-mixing and even-order distortion caused by the leakage signal, thereby improving the use of the down converter. Working characteristics. To achieve the above and other objects, the present invention provides a subharmonic mixing x mixing a multiphase local oscillator signal with an RF signal to buffer the 2 frequency signal. The subharmonic mixer includes a differential amplifying unit, a current early element, and a switching unit. The differential amplifying unit amplifies the RF signal. 7 200807902 UMCD-2006-0025 19504twf.doc/g The current buffer unit is coupled to the differential amplifying unit for amplifying the gain of the output δίΐ of the differential amplifying unit. The switching unit is connected to the current buffer unit for switching the output buffer of the current buffer unit to the base frequency signal according to the multi-phase local oscillation signal. In accordance with a preferred embodiment of the present invention, the differential amplifying unit comprises a first resonant circuit and the current buffering unit comprises a second resonant circuit. In order to make the subharmonic mixer achieve the mixing of the multi-phase local oscillation signal, the fundamental frequency signal is generated. First, the differential amplifying unit amplifies the RF signal, and the amplified RF signal is guided to the current buffer unit by the first resonant circuit, and the leakage signal generated by the local oscillation signal is also guided to the first resonant circuit by using the first resonant circuit. A voltage. Thereafter, the current buffer unit coupled to the differential amplifying unit amplifies the gain of the output signal of the differential amplifying unit, and the second resonant circuit directs the current buffer: the output signal to the switching unit, and the second spectrum circuit The leakage signal generated by the local oscillation signal is also directed to the second voltage. Finally, the switcher switches the current buffered output signal to the baseband signal according to the local phase local subtraction u. The first resonant circuit and the second resonant circuit have the same resonant frequency of the first frequency RF signal, and the first voltage can be two operations, and one second voltage can be a ground voltage. In a preferred embodiment, the above-described subharmonic mixed down frequency receiver. As is apparent from another point of view, the present invention further proposes to convert an RF signal to a baseband signal. The down converter includes the signal 200807902 UMCD-2006-0025 19504twf.doc/g subharmonic mixer. The subharmonic mixer includes a differential amplifying unit, a current buffering unit, and a switching unit. The signal generator is used to provide a multi-phase local oscillation signal. The sub-spectral mixer connected to the signal generator is used to mix the multi-phase local oscillation signal with the RF signal to generate a fundamental frequency θίΐ. The process of generating the fundamental frequency signal by the subharmonic mixing includes amplifying the RF signal by using the differential amplifying unit. Then, the current buffer unit coupled to the differential amplifying unit amplifies the concealment of the output signal of the differential amplifying unit. Finally, the current buffering single output signal is converted into a baseband signal by means of a multi-phase local oscillation signal coupled to the current buffer unit and the signal generator switching unit. Thereby, the down converter can achieve the purpose of converting an RF signal to a fundamental frequency signal. In the above preferred converter, in a preferred embodiment, the signal generator comprises a local oscillator and a phase shifter. The local oscillator is used to generate a θfl number, and the phase shifter connected between the local oscillator and the switching unit can convert the local oscillation signal into a plurality of local oscillation signals with different phase shift amounts. The output acts as a multi-phase local oscillator signal. In a preferred embodiment, the lower converter described above is suitable for a straight receiver. By adopting a combination of a differential amplifying unit and a current buffering unit, the present invention allows the subharmonic mixer to utilize the first resonant circuit and the second resonant circuit to direct the leakage signal to the first voltage or the second voltage. Efficacy, as a result, with the increase in the isolation of the subharmonic mixer, the direct current down-converter using the subharmonic mixer will face a significant reduction in DC offset and even time. 9 200807902 UMCD-2006-0025 19504 twf.doc/g The objects, features, and advantages of the present invention can be more clearly understood. See below. [Embodiment] _j 3 is a secondary money mixer structure according to the preferred embodiment of the present invention, including a differential amplifying unit 30, a current buffer unit 302, and a knife switching 303. Current buffer unit 3〇2 _ connected to differential amplification unit training
與切換單元303之間。差動姑士置分ώ 是動放大早兀301將射頻訊號放大 ，:认‘波混頻器300利用電流緩衝單元3〇2將差動放大 早兀302之輸出訊號之增纽大，以讓切換單元3〇3依據 多相位本地振盪訊號，將電流緩衝單元3Q2之輸出訊號切 換至基頻訊號。如此一來，次諧波混頻器3〇〇就可以達到 將多相位本地振盪訊號與射頻訊號進行混頻，進而產生基 頻訊號的功效。 土 圖4為根據本發明較佳實施例之次諧波混頻器詳細電 路圖。對照圖3來看，其中輸入端31〇a與31〇b對應圖3 中，差動放大單元301之差動輸入端31〇。輸出端320a與 320b對應圖3中，動放大單元301之差動輸出端32〇。輸 出端330a與330b對應圖3中，電流緩衝單元302之差動 輸出端330。輸出端340a與340b對應圖3中，切換單元 3〇3之差動輸出端340。輸入端350aa與350ab、以及輸入 端350ba與350bb則對應圖3中，切換單元303之多相位 本地振盪訊號輸入端35〇a與350b。 10 200807902 UMCD-2006-0025 19504twf.doc/g 如圖4所示，差動放大單元3〇1包括譜振電路4〇i i 型電晶體MN41與MN42、電阻R41與R42、電容C4i與 C42、以及電感L41。電流緩衝單元3Q2包括p型電晶徵 M41與MP42、以及諧振電路4〇2。切換單元3〇3包括& 型電晶體MN43〜MN410、以及電阻R43與R44。電阻尺^ 與R42之第一端耦接至第一電壓（比如操作電壓VcC4)。n 型電晶體M41與MP42之汲極耦接至諧振電路4(π，N型 電晶體M41與MP42之源極耦接至第二電壓（比如接地電 壓），N型電晶體M41與MP42之閘極則分別輕接至電^ C41與C4之第一端。電感L41串接在電容C41之第二端 與電容C4之第二端之間。電容C41與C42之第一端分^ 耦接至電感L41之第一端與第二端，且電容C41與c幻 之弟一端分別拉線構成輸入端310a與310b。P型電晶# M41與MP42之源極分別叙接至n型電晶體M41與]VIP42 之汲極，P型電晶體Μ 41與Μ P 4 2之閘極耦接至第二電壓。 諧振電路402串接在ρ型電晶體Μ41與ΜΡ42之閘極與第 二電壓之間。電阻R43與R44之第一端耦接至第一電塵， 電阻R43與R44之第二端則分別拉線構成輸出端如與 340b。Ν型電晶體ΜΝ43與ΜΝ44、以及ΜΝ47與ΜΝ48 之没極耦接至電阻R43之第二端。N型電晶體MN45與 MN46、以及MN49與MN410之汲極耦接至電阻R44之第 二端。且N型電晶體MN43〜MN46之源極耦接至P型電晶 體M41之汲極。N型電晶體MN47〜MN410之源極耦接至 P型電晶體M42之汲極。 200807902 UMCD-2006-0025 19504twf.d〇c/g 上述之諸振電路4〇l包括電感L42與L43、以及電容 C45與C46。諧振電路402則包括電感L44與L45、以及 電容C47與C48。電感L42與L43、以及電容C45與C46 之第一端耦接至第一電壓。電感L42與電容C45之第二端 搞接至N型電晶體MN41之没極。電感L43與電容C46 之第二端耦接至N型電晶體MN42之汲極。電感L44與電 谷C47之第一端輕接至p型電晶體MP41之;:及極。電感L45 與電容C48之第一端耦接至P型電晶體MP42之汲極。電 感L44與L45、以及電容C47與C48之第二端麵接至第二 電壓。 繼續參照圖4來看本實施例之工作原理。差動放大單 元301中的電感L41與L42 ’在此提供低阻抗路徑，以形 成N型電晶體MN41與MN42的直流偏壓電流。此時，經 由電容C43與C44之第二端所接收的射頻訊號，藉由操作 上相當於差動轉導的N型電晶體]VIN41與]VIN42放大。為 了將放大後的射頻訊號經由輸出端320a與320b傳送至電 ， 流緩衝單元302，本實施例將諧振電路4〇1之共振頻率操 作在射頻訊號之振盪頻率。由於諧振電路4〇1操作在共振 頻率下相當於一咼阻抗，操作在共振頻率外相當於一低阻 抗。因此，放大後的射頻訊號將可被導向至電流緩衝單元 302。不僅如此，由於次諧波混頻器3〇〇之多相位本地振盪 訊號之振盪頻率為射頻訊號之振盪頻率的〇·5倍。因此， 由多相位本地振盪訊號經由穿隧效應出現在輸入端 12 200807902 UMCD-2006-0025 19504twf.doc/g 與310b的凜漏訊號，或是由干擾訊號經由非線性電路所產 生的偶次階失真，也將被諧振電路401導向至第一電壓。 之後的電流緩衝單元302利用P型電晶體M41與 MP42之源極接收放大後的射頻訊號。電感[44與L45提 供低阻抗路徑，形成P型電晶體MP41與MP42的直流偏 壓電流。此時，連接成共閘極組態的p型電晶體Μρ41與 MP42，除了有助於電流緩衝單元3〇2之隔絕度，並單增益 ' 放大由差動放大單元3〇1所輸出的訊號。為了將由P型電 晶體MP41與MP42所放大的射頻訊號，導向至切換單元 303 ’在此諧振電路402採取如同諧振電路401 —樣的做 法，將共振頻率操作在射頻訊號之振盪頻率，讓電流緩衝 單元302之輸出訊號導向至切換單元3〇3的同時，諧振電 路402也可將由多相位本地振盪訊號經由穿隧效應出現在 輸入端310a與31〇b的洩漏訊號，或是由干擾訊號經由非 線性電路所產生的偶次階失真，導向至第二電壓。 最後，多相位本地振盪訊號所包含的本地振盪訊號 "L04—0〇、L04—90°、L04—180°、以及 L04—270°，分別經由 切換單元303内的N型電晶體MN43與MN49之閘極、 MN45與MN48之閘極、MN44與MN410之閘極、以及 MN46與MN47之閘極所接收。此時，操作特性相當於開 關的N型電晶體MN43〜MN410，則依據本地振盪訊號 L04一00、L04—900、L04—1800、以及 LO4—2700，將電流缓 衝單元302之輸出訊號切換至基頻訊號。其中本地振盪訊 13 200807902 UMCD-2006-0025 〗9504tWf.d〇c/g 號 L〇4—00、L04—900、L04一 1800、以及 L〇4—2700 之相移 量分別為0度、90度、180度、以及270度。 圖5〜圖8為本實施例實現在現今CMOS製程技術下 的實際量測結果。本實施例在操作電壓VCC4為IV、射頻 訊號之振盪頻率為5.2GHz、多相位本地訊號之振盪頻率為 2.6GHz之條件下。如圖5所示，次諧波混頻器3〇〇在頻率 為10MHz下的雜訊指數(Noise Figure)為17.3dB。且如圖6 所示，次諧波混頻器300在接收功率為i5.5dBm之多相位 本地訊號下，於輸入端310a與310b測量到功率為 -65· 154dBm的洩漏訊號。此洩漏訊號將造成次諧波混頻器 300輸出功率約為-l〇〇.7dBm的直流偏移，但此直流偏移 卻未超過無線區域網路(WLAN)接收器所規範的背景雜訊 (noise floor)。換而言之，此時次譜波混頻器3⑻所形成的 直流偏移將掩蓋在背景雜訊中，而不會影響到電路本身的 工作性能。此外，如圖7所示的，於次諧波混頻器3〇〇之 輸入端310a與310b，只測量到非常地微小的2倍泡漏訊 v ’ 5虎(功率僅為-l〇9.934dBm) ’由此可正明本發明所提出的次 諧波混頻器300之輸入端具有良好的隔絕度。為了更進一 步了解本發明之電路性能’圖8列出了本實施例與電機電 子工程師協會(IEEE，Institute of Electrical and Electronic Engineers)於1998年固態電路會刊第33卷第期 (Solid-state Circuits，V0L.33，N0.12)中所發表的期刊（圖 8 中以期刊[1]表示）、2000年射頻暨無線會刊第219頁至第 222頁(11八\^0凡？？.219-222)中所發表的期刊（圖8中以 14 200807902 UMUD-2UU6-0025 19504twf.doc/g 期刊[2]表示）、2004年微波與無線元件會刊第14卷第7 期（Microwave and Wireless Components Letters,VOL.14, ΝΟ·7)中所發表的期刊（圖8中以期刊[习表示）、以及2〇〇4 年固態電路會刊第39卷第6期(Solid-stateCircuits，VOL.39, ΝΟ·6)中所發表的期刊（圖8中以期刊[4]表示）之比較結 果。由圖8可顯示出本發明之次諧波混頻器具有良好的隔 絕度，以至於與現今期刊所發表的論文相比較下，本發明 、不論是在輸入端三階交錯點（Input 3rd order intercept point，IIP3)、輸入端二階交錯點(Inpm 2rd 〇rder point ’ IIP2)、LOR(l〇cal oscillator rejection)、或是針對所 產生的洩漏訊號與直流偏移，都具有良好特性。其中圖8 中所附註的符號*表示該期刊所發表的次諧波混頻器具有 線性化電路，符號+表示該期刊所發表的次諧波混頻器假 設射頻訊號輸入端之隔絕度為5〇dB。 從另一觀點來看，圖9為依據本發明較佳實施例之下 轉換器結構示意圖，包括次諧波混頻器3〇〇與訊號產生器 501。次谐波混頻益300則包括差動放大單元3〇1、電流緩 衝單元302、以及切換單元303。其中次譜波混頻器3〇〇 耦，至訊號產生器501。電流緩衝單元3〇2耦接至差動放 大單元3(Π。切換單元3〇3編妾至電流緩衝單元3〇2與訊 號產生器50卜下轉換器在達到將射頻訊號轉換至基頻气 號的過程中’包括先利用差動放大單元3〇1將所接收的射 頻訊號放大。之後，再將經由電流緩衝單元3〇2單增益 大差動放大單元301之輸出訊號。藉此，讓切換單元地 15 200807902 uivil.u-zuuo-0025 19504twf.doc/g 依據訊號產生器501所提供的多相位本地振盈訊號，將電 流緩衝單元302之輸出訊號轉換成基頻訊號。 上述之訊號產生器501包括本地振盪器51〇與相位偏 f器520。相位偏移器52〇串接在本地振盪器51〇與切換 单το 3 0 3之間。其中本地振盪器5 i 〇用以產生本地振盪訊 號’以便讓相位偏移器' 520贿接收到的本地振盛訊號， 轉換成數個不_移量的本隸盪訊號，讀出作為多相 位本地振里訊號。至於圖9實施例中，次諧波混頻器300 之工作原理、電路_、以及相關電路特性，則包含在圖 3〜圖8實施例中，在此就不多加敘述。 _綜上所述，本發明因採用差動放大單元與電流緩衝單 2合之架構，讓次譜波混頻器在利用第-諧振電路與第 、咱振電路’將Ά漏訊號導向至第—電壓或第二電壓之情 =下/，有效地提高次諧波混頻器之隔絕度。如此一來，使 諧ί混頻器的相關電路’例如下轉換器、直 关f1牛須接收态，..荨，將隨著次諧波 而大幅地提升本身雷&度的k升， 而4讀對直接降頻接收器 5雖/= 偏移與偶:請失真將Α幅度地降低。 =發明已以較佳實施例揭露如上，然其並非用以 和二二=此;;者’在不脫離本發明之精神 ，圍當視後附之申請專:範二者=本斷 【圖式簡單說明】 ’ 圖1為傳統次諧波混頻器之結構示意圖。 16 200807902 uiviuu-zu〇6-0025 19504twf.doc/g 圖2為傳統吉伯特混頻器之結構示意圖。 一圖3為根據本發明一較佳實施例之次諧波混頻器結構 示意圖。 圖4為根據本發明較佳實施例之次諧波混頻器詳细 路圖。 、电 圖5圖7為用以說明圖4實施例電路特性之實際量測 結果。 、 圖8為圖4實施例與現今期刊之相關特性比較表。 圖9為根據本發明較佳實施例之下轉換器結構示音 圖。 μ 【主要元件符號說明】 301 ·差動放大單元 302 :電流緩衝單元 303 :切換單元 310、350a與350b:差動輸入端 320、330、340 ··差動輸出端 310a 與 310b、350aa 與 350ab、350ba 與 350bb :輪入 端 320a 與 320b、330a 與 330b、340a 與 340b、：輪出端 401、402:諧振電路 而 MN1 〜MN12、MN41 〜MN410 : N 型電晶體 MP41與MP42 : P型電晶體 R11 與 R12、R41 〜R44 ··電阻 C41〜C48 :電容 L41〜L45 :電感Between the switching unit 303. The differential auspicious splitter is amplifying the RF signal by the amplifier 301, and recognizes that the wave mixer 300 uses the current buffer unit 3〇2 to amplify the output signal of The switching unit 3〇3 switches the output signal of the current buffer unit 3Q2 to the fundamental frequency signal according to the multi-phase local oscillation signal. In this way, the subharmonic mixer 3〇〇 can achieve the function of mixing the multi-phase local oscillation signal and the RF signal to generate the fundamental frequency signal. Figure 4 is a detailed circuit diagram of a subharmonic mixer in accordance with a preferred embodiment of the present invention. Referring to FIG. 3, the input terminals 31A and 31B correspond to the differential input terminal 31 of the differential amplifying unit 301 in FIG. The output terminals 320a and 320b correspond to the differential output terminal 32 of the dynamic amplifying unit 301 in Fig. 3. Outputs 330a and 330b correspond to differential output 330 of current buffer unit 302 in FIG. The output terminals 340a and 340b correspond to the differential output terminal 340 of the switching unit 3〇3 in Fig. 3. The input terminals 350aa and 350ab, and the input terminals 350ba and 350bb correspond to the multi-phase local oscillation signal input terminals 35A and 350b of the switching unit 303 in Fig. 3. 10 200807902 UMCD-2006-0025 19504twf.doc/g As shown in FIG. 4, the differential amplifying unit 3〇1 includes a spectral circuit 4〇ii type transistors MN41 and MN42, resistors R41 and R42, capacitors C4i and C42, and Inductor L41. The current buffer unit 3Q2 includes p-type electric crystal signs M41 and MP42, and a resonance circuit 4〇2. The switching unit 3〇3 includes & type transistors MN43 to MN410, and resistors R43 and R44. The first end of the resistor ruler ^ and R42 are coupled to a first voltage (such as the operating voltage VcC4). The anodes of the n-type transistors M41 and MP42 are coupled to the resonant circuit 4 (π, the source of the N-type transistors M41 and MP42 is coupled to a second voltage (such as a ground voltage), and the gates of the N-type transistors M41 and MP42 The poles are respectively connected to the first ends of the capacitors C41 and C4. The inductor L41 is connected in series between the second end of the capacitor C41 and the second end of the capacitor C4. The first ends of the capacitors C41 and C42 are coupled to The first end and the second end of the inductor L41, and the ends of the capacitors C41 and c are connected to form input terminals 310a and 310b respectively. The sources of the P-type electro-crystals M41 and MP42 are respectively connected to the n-type transistor M41. The gates of the P-type transistors Μ 41 and Μ P 4 2 are coupled to the second voltage. The resonant circuit 402 is connected in series between the gates of the p-type transistors Μ41 and ΜΡ42 and the second voltage. The first ends of the resistors R43 and R44 are coupled to the first electric dust, and the second ends of the resistors R43 and R44 are respectively drawn to form an output end such as 340b. The 电-type transistor ΜΝ43 and ΜΝ44, and ΜΝ47 and ΜΝ48 The pole is coupled to the second end of the resistor R43. The N-type transistors MN45 and MN46, and the drains of the MN49 and MN410 are coupled to the second end of the resistor R44. And the N-type transistor M The source of N43~MN46 is coupled to the drain of P-type transistor M41. The source of N-type transistor MN47~MN410 is coupled to the drain of P-type transistor M42. 200807902 UMCD-2006-0025 19504twf.d〇 c/g The above-mentioned vibration circuits 4〇1 include inductors L42 and L43, and capacitors C45 and C46. Resonant circuit 402 includes inductors L44 and L45, and capacitors C47 and C48. Inductors L42 and L43, and capacitors C45 and C46 The first end is coupled to the first voltage, and the second end of the inductor L42 and the capacitor C45 is connected to the anode of the N-type transistor MN41. The second end of the inductor L43 and the capacitor C46 is coupled to the N-type transistor MN42. The first end of the inductor L44 and the electric valley C47 is lightly connected to the p-type transistor MP41; and the pole is connected. The first end of the inductor L45 and the capacitor C48 are coupled to the drain of the P-type transistor MP42. The inductor L44 and L45, and the second end faces of the capacitors C47 and C48 are connected to the second voltage. The operation of the embodiment will be continued with reference to Fig. 4. The inductors L41 and L42' in the differential amplifying unit 301 provide a low impedance path here. To form a DC bias current of the N-type transistors MN41 and MN42. At this time, the second end of the capacitors C43 and C44 are received. The RF signal is amplified by an N-type transistor VIN41 and VIN42 which are equivalent to differential transduction. In order to transmit the amplified RF signal to the power via the output terminals 320a and 320b, the stream buffer unit 302, this embodiment The resonant frequency of the resonant circuit 4〇1 is operated at the oscillation frequency of the RF signal. Since the operation of the resonant circuit 4〇1 corresponds to a 咼 impedance at the resonant frequency, the operation is equivalent to a low impedance outside the resonant frequency. Therefore, the amplified RF signal will be directed to the current buffer unit 302. In addition, the oscillation frequency of the multi-phase local oscillation signal of the subharmonic mixer is 〇·5 times of the oscillation frequency of the RF signal. Therefore, the multi-phase local oscillation signal appears through the tunneling effect at the input terminal 12200807902 UMCD-2006-0025 19504twf.doc/g and 310b, or the even order generated by the interference signal via the nonlinear circuit. Distortion will also be directed by the resonant circuit 401 to the first voltage. The subsequent current buffer unit 302 receives the amplified RF signal using the sources of the P-type transistors M41 and MP42. Inductors [44 and L45 provide a low impedance path to form the DC bias current of P-type transistors MP41 and MP42. At this time, the p-type transistors Μρ41 and MP42 connected to the common gate configuration, in addition to contributing to the isolation of the current buffer unit 3〇2, and the single gain 'amplify the signal output by the differential amplifying unit 3〇1 . In order to direct the RF signal amplified by the P-type transistors MP41 and MP42 to the switching unit 303', the resonant circuit 402 takes the same operation as the resonant circuit 401, and operates the resonant frequency at the oscillation frequency of the RF signal to allow current buffering. While the output signal of the unit 302 is directed to the switching unit 3〇3, the resonant circuit 402 can also transmit the leakage signal from the multi-phase local oscillation signal to the input terminals 310a and 31〇b via the tunneling effect, or the interference signal via the non-interference signal. The even order distortion produced by the linear circuit is directed to the second voltage. Finally, the local oscillation signals "L04_0〇, L04-90°, L04-180°, and L04-270° included in the multi-phase local oscillation signal are respectively passed through the N-type transistors MN43 and MN49 in the switching unit 303, respectively. The gates, the gates of MN45 and MN48, the gates of MN44 and MN410, and the gates of MN46 and MN47 are received. At this time, the operation characteristics are equivalent to the N-type transistors MN43 to MN410 of the switch, and the output signals of the current buffer unit 302 are switched to the local oscillation signals L04-00, L04-900, L04-1800, and LO4-2700. Baseband signal. Among them, the local oscillation signal 13 200807902 UMCD-2006-0025 〗 9504tWf.d〇c / g L〇4 00, L04-900, L04 1800, and L 〇 4-2700 phase shift amount is 0 degrees, 90 Degree, 180 degrees, and 270 degrees. FIG. 5 to FIG. 8 are actual measurement results of the present embodiment under the CMOS process technology. In this embodiment, the operating voltage VCC4 is IV, the oscillation frequency of the radio frequency signal is 5.2 GHz, and the oscillation frequency of the multi-phase local signal is 2.6 GHz. As shown in Fig. 5, the noise figure of the subharmonic mixer 3〇〇 at a frequency of 10 MHz is 17.3 dB. As shown in Fig. 6, the subharmonic mixer 300 measures a leakage signal having a power of -65·154 dBm at the input terminals 310a and 310b under a multi-phase local signal having a receiving power of i5.5 dBm. This leakage signal will cause the subharmonic mixer 300 to output a DC offset of approximately -1〇〇.7dBm, but this DC offset does not exceed the background noise specified by the wireless local area network (WLAN) receiver. (noise floor). In other words, the DC offset formed by the sub-spectral mixer 3 (8) at this time will be masked in the background noise without affecting the performance of the circuit itself. In addition, as shown in FIG. 7, at the input terminals 310a and 310b of the subharmonic mixer 3, only a very small 2 times bubble leakage v '5 tiger is measured (the power is only -l〇9.934) dBm) 'It is thus apparent that the input of the subharmonic mixer 300 proposed by the present invention has good isolation. In order to further understand the circuit performance of the present invention, FIG. 8 lists the present embodiment and the Institute of Electrical and Electronic Engineers (IEEE), 1998, Solid State Circuits, Vol. 33 (Solid-state Circuits) Journals published in V0L.33, N0.12) (indicated by the journal [1] in Figure 8), pp. 219 to 222 of the 2000 Radio and Wireless Journal (11 8 \^0凡??. Journal published in 219-222) (in Figure 8, 14 200807902 UMUD-2UU6-0025 19504twf.doc/g Journal [2]), 2004 Microwave and Wireless Components Journal, Vol. 14, No. 7 (Microwave and Journals published in Wireless Components Letters, VOL.14, ΝΟ·7) (in Journal 8), and in the 4th year of Solid State Circuits, Vol. 39, No. 6 (Solid-state Circuits, VOL) .39, ΝΟ·6) The results of the comparison of the journals published in the journal (indicated by the journal [4] in Figure 8). It can be seen from Fig. 8 that the subharmonic mixer of the present invention has good isolation so that the present invention, whether compared to the third-order interleaving point at the input end (Input 3rd order), is compared with the paper published in the current journal. Intercept point, IIP3), Inpm 2rd 〇rder point 'IIP2', LOR(l〇cal oscillator rejection), or both for the generated leakage signal and DC offset. The symbol * in the note of Figure 8 indicates that the subharmonic mixer published in the journal has a linearization circuit, and the symbol + indicates that the subharmonic mixer published in the journal assumes that the isolation of the RF signal input is 5 〇dB. From another point of view, Figure 9 is a schematic diagram of the structure of the converter in accordance with a preferred embodiment of the present invention, including a subharmonic mixer 3〇〇 and a signal generator 501. The subharmonic mixing benefit 300 includes a differential amplifying unit 3〇1, a current buffer unit 302, and a switching unit 303. The sub-spectral mixer 3 is coupled to the signal generator 501. The current buffer unit 3〇2 is coupled to the differential amplifying unit 3 (Π. The switching unit 3〇3 is programmed to the current buffer unit 3〇2 and the signal generator 50 to convert the RF signal to the base frequency gas. In the process of the number, the first step is to amplify the received RF signal by using the differential amplifying unit 3〇1. Thereafter, the output signal of the differential amplifying unit 301 is amplified by the current buffer unit 3〇2. The switching unit ground 15 200807902 uivil.u-zuuo-0025 19504twf.doc/g converts the output signal of the current buffer unit 302 into a fundamental frequency signal according to the multi-phase local vibration signal provided by the signal generator 501. The above signal is generated. The device 501 includes a local oscillator 51A and a phase offset device 520. The phase shifter 52 is connected in series between the local oscillator 51 and the switching unit το 3 0 3. The local oscillator 5 i is used to generate local The oscillation signal 'to allow the phase shifter' 520 to receive the local vibration signal, which is converted into a plurality of non-shifting local sway signals, and read out as a multi-phase local oscillating signal. As shown in the embodiment of FIG. Subharmonic mixer 300 The working principle, circuit_, and related circuit characteristics are included in the embodiment of FIG. 3 to FIG. 8 and will not be described here. _ In summary, the present invention uses a differential amplifying unit and a current buffering unit 2 The architecture allows the sub-spectral mixer to effectively increase the subharmonic mixture by using the first-resonant circuit and the first and second oscillating circuits to direct the leakage signal to the first voltage or the second voltage. The isolation of the frequency converter. In this way, the relevant circuit of the harmonizing mixer, such as the down converter, the direct off f1 cow must receive state, .., will greatly enhance its own thunder & The degree of k is increased, while the 4th reading is for the direct down-conversion receiver 5, although /= offset and even: please reduce the distortion to a large extent. = The invention has been disclosed above in the preferred embodiment, but it is not used and 2 = this;; the person's without leaving the spirit of the invention, the application of the application after the inspection: Fan = = broken [schematic description] Figure 1 is a schematic diagram of the structure of the traditional subharmonic mixer 16 200807902 uiviuu-zu〇6-0025 19504twf.doc/g Figure 2 shows the structure of a traditional Gilbert mixer. 3 is a schematic structural diagram of a subharmonic mixer according to a preferred embodiment of the present invention. Fig. 4 is a detailed road diagram of a subharmonic mixer according to a preferred embodiment of the present invention. To illustrate the actual measurement results of the circuit characteristics of the embodiment of Fig. 4. Fig. 8 is a comparison table of the characteristics of the embodiment of Fig. 4 and the current journal. Fig. 9 is a structural diagram of the structure of the converter according to a preferred embodiment of the present invention. μ [Main component symbol description] 301 · Differential amplifying unit 302: Current buffering unit 303: switching units 310, 350a and 350b: differential input terminals 320, 330, 340 · Differential output terminals 310a and 310b, 350aa and 350ab 350ba and 350bb: wheel terminals 320a and 320b, 330a and 330b, 340a and 340b, wheel terminals 401, 402: resonant circuit and MN1 to MN12, MN41 to MN410: N-type transistors MP41 and MP42: P-type Crystal R11 and R12, R41 to R44 · Resistor C41~C48: Capacitor L41~L45: Inductance