200913789 九、發明說明： 【發明所屬之技術領域】 本發明涉及一種用於操作一放電燈的電路配置，此放電 燈包括：一整流器，其具有一第一和一第二輸入端以與一 交流電壓源相耦接；一第一和一第二輸出端，以提供一直 流操作電壓；一自由震盪的半橋式反相器，其包括二個電 壓控制的電子式開關所形成的串聯電路，其耦接在該整流 器之第一和第二輸出端之間，其中該第一和第二電子式開 關之間形成一種半橋-中點；一起動電路，用來起動一逆整 流器之自由震盪，其中該起動電路包括一控制電路，其具 有一第一和一第二輸出端，此控制電路之第一輸出端是與 第一和第二電子式開關的控制電極相耦接；以及一起動電 容器，其具有第一和第二終端。本發明亦涉及一種在上述 電路配置上的放電燈之操作方法。 【先前技術】 本發明一般而言涉及一種可獨立地起動一可自由震盪之 半橋式反相器之電路配置，其具有多個電壓控制的電晶 體，例如，MOSFET和IGBT，以便在施加一電源電壓時用 作開關。 EP 0 9 1 7 4 1 2 B 1用來形成本發明之申請專利範圍之獨立 項的前言，由此一文件中已知可藉由對一電容器的充電來 提供半橋式電晶體之第一次接通所需的電壓，該電容器串 聯至一種控制該半橋式反相器之電晶體用的控制電路。爲 了防止該串聯加入之電容器之一種不期望的穩定之充電狀 態且因此亦可防止該半橋式反相器之”停留，，狀態，則在上 -6 - 200913789 述專利文件中建議直接由主電源而經由一電阻來對該電容 器進行充電。經由該主電源終端，則能以電源頻率之一半 頻率來重複一起動測試一段時間，直至該半橋式反相器之 自我振盪開始爲止。在上述的文件中，只顯示一些實施例， 其中二個半橋式電晶體分別設有控制電路。 然而，爲了降低成本或使電路配置簡化，半橋式反相器 亦可由互補式電晶體（η-和p-通道）所構成，此時該二個電晶 體由相同的控制電路來控制。此原理由ΕΡ 0 78 1 077 Β1中 已爲人所知，該文件中爲了起動該半橋式反相器亦需一種 雙向二極體（Diac)。依據ΕΡ 0 917 412 Β1而得之槪念可轉 用至 EP 0 7 8 1 077 B1中已爲人所知的電路配置，此時該電 路起動所需的電容器串聯至一控制電路。於是，可完全設 定該”電源起動”之功能。 然而，上述文件所顯示的缺點在於，一起動電阻之一在 該起動電容器充電時所需的終端在該放電燈操作時實際上 位於該半橋-中點電位。藉由該起動電阻之另一終端之與二 條電源導線之一的互相連接，則該中點電位之高頻率的矩 形電壓可經由該起動電阻而處於電源電位處。這樣會使整 個電路配置出現高的無線干擾値，此乃因流經該起動電阻 的電流、且因此是一種在無線干擾濾波器上的高頻、近似 矩形的干擾信號已消逝而直接導引至電源上。爲了防止此 現象，若該起動電阻是與提供該正電源電壓（所謂中間電路 電壓）之整流器之終端相連接而不是與該二條電源導線之一 相連接，則這樣所顯示的缺點是：此種電路配置只能進行 -7- 200913789 一種起動測試，即，在一種失敗的起動測試中該電路保持” 停留”狀態。 【發明內容】 本發明目的是繼續提供一種上述形式的電路配置或方 法，以便一方面可防止高的無線干擾値且另一方面在施加 該主電源電壓之後可確保該電路配置之可靠的起動。 本發明以下述認知爲基礎，即：由EP 0 78 1 077 B 1和EP 〇 917 412 B1之原理之組合以及隨後將該起動電阻（不是電 源導線）固定至整流器之正輸出端時不能達成上述的目的， 此乃因在該起動電容器上可設定穩定的電壓比，其會妨礙 一種重複式的起動測試。因此，依據本發明，該起動電容 器是與該半橋-中點相耦接且該半橋-中點經由一下拉電阻 而與該整流器之第二輸出端相耦接。此種組態可使該半橋-中點之電位經由該下拉電位而在一失敗的起動測試之後又 被導引至由該整流器之第二輸出端所提供的負參考電位 處。因此，該起動電容器重新充電且可進行一種新的起動 測試。由於此種電路配置中該電阻不再與主電源之一導線 相連接而是與整流器之正輸出端相連接，則可防止無線干 擾傳送至該主電源中。該電容器經由該電阻而與一正電位 相連接。 在一較佳的實施形式中，該電路配置另外具有一種燈抗 流圈，其具有至少一主繞組，該主繞組是與放電燈用的一 終端相串聯，此處，一控制電路具有電感，此電感具有一 第一終端和一第二終端，此電感是此燈抗流圈之一個二次 200913789 繞組。該電路配置另具有一串聯電路，其由一第二歐姆電 阻和一並聯共振電路所構成，該並聯電路耦接在該電感之 第一和第二終端之間。特別有利的是將一種包括一電容器 之半橋式電路並聯至第二歐姆電阻。此種特別有利的實施 形式是以下述認知爲基準：爲了起動該電路配置，在施加 主電源電壓之後首先接通的電晶體之控制電壓必須保持足 夠大，當該起動電容器藉由該電晶體之接通而已放電且因 此該有效的控制電壓（即，由該起動電容器上的電壓與該控 制電路之輸出電壓相加之和）傾向於下降時亦如此。在該起 動電容器開始放電時該電晶體所需之足夠的控制電壓的維 持較佳是藉由下述方式來確保：該控制電路可快速地支配 一種夠大的輸出電壓。爲了此一目的，由於該電容器對各 導通脈衝之高頻光譜成份而言顯示出一種很低的電阻，則 該第二歐姆電阻成爲旁路（by-passed)，該第二歐姆電阻在操 作期間對該半橋式反相器之電子式開關之控制而言是需要 的。流經該電容器之電荷載體因此可對一設置在並聯共振 電路中的電容器進行很快速的充電。這樣可快速地對首先 待接通的半橋式電晶體提供一種足夠大的控制電壓。 該半橋式反相器之首先待接通的開關上的一種不足夠大 的控制電壓之危險性因此在該起動電容器放電時可有效地 受到限制。 該半橋式電路較佳是包括一第三歐姆電阻，其串聯至該 電容器。此一第三歐姆電阻可帶來二種優點：其中之一優 點是該燈抗流圏上的二次繞組所提供的電壓之矩形成份可 -9- 200913789 被衰減。這是與下述事實有關，即：該燈抗流圏一方面與 該半橋-中點（其上存在著一種高頻之矩形信號）相連接，且 另一方面是與該燈相連接，該燈施加一種很像正弦形式的 信號至該燈抗流圏之另一終端上。第三歐姆電阻可在操作 期間支配一種基本上是正弦形式的信號以控制該半橋式反 相器之二個電子式開關。另一優點在於，該第三歐姆電阻 可在該半橋式反相器之振盪期間使可能存在的靜止時間 (dead time)延長。於是，可使該半橋式反相器之開關達成一 種無損耗的切換。通常，適當的方式是使該第三歐姆電阻 之尺寸較第二歐姆電阻之尺寸小很多。 此外，當本發明的電路配置另外具有一個二極體時是有 利的，此二極體是與第一歐姆電阻並聯且須定向成”當第一 電子式開關接通時，該二極體可使該起動電容器經由第一 電子式開關而放電”。因此，該起動電容器在成功地起動之 後該二極體使該起動電容器放電，於是，不會有其它的起 動測試來干擾該電路配置的操作。 在一較佳的實施形式中，該起動電容器之第一終端經由 另一歐姆電阻而與該整流器之第二輸出端相耦接。在此一 實施形式中，不需設置一先前所述之二極體。該起動電容 器之放電是經由第一和另一歐姆電阻來達成。其原因在 於，在操作時該起動電容器之二個終端平均位於一種電位 處，此電位恰巧等於半橋-電壓之一半，這樣可使該起動電 容器在操作時放電。 於是，當設置另一電容器時是有利的，此另一電容器耦 -10- 200913789 接在該控制電路之第一輸出端和第二電子式開關之控制電 極之間。由於該起動電容器之純歐姆放電持續一段時間， 直至電壓經由該起動電容器而消失爲止，則該控制電路之 輸出電壓具有一種正的偏移（offset)成份。此種偏移成份可 阻止第二電晶體之接通，此乃因第二電晶體所可支配的控 制電壓可藉由該偏移成份而減少。所建議的另一電容器用 作耦合電容器且接收上述的偏移-電壓。該控制電路之輸出 電壓之純交流電壓成份經由可選擇的（optional)另一耦接在 該第二電子式開關之控制電極和工作電極之間的歐姆電阻 而下降’這樣可立即達成足夠負的電壓以控制該第二電晶 體。 該半橋式反相器之二個電子式開關可以是極性互補的電 晶體，但亦可以是極性相同的電晶體。使用極性互補的電 晶體所可提供的優點是，只須設置一個控制電路。 其它有利的實施形式可由申請專利範圍各附屬項得知。 參考本發明的電路配置所設置的較佳的實施形式及其優 點只要可供使用時亦適用於本發明的操作方法中。 【實施方式】 以下的描述中相同或作用相同的組件使用相同的參考符 號。 第1圖顯示本發明之電路配置之第一實施例的圖解，此 電路配置用來操作一種放電燈，此放電燈以負載EL來表 示。因此’一整流器GL經由一保險絲S I而被供應以主電 源電壓’其供電至一電解式電容器C1或保持著電壓。在該 -11- 200913789 電解式電容器c1上可經由一濾波器而測得二種電源分支， 該濾波器是由其中一分支中的繞組L1和一連接此二個分支 的電容器C2所構成。第1圖下方的電源分支具有負電位且 定義了該電路配置之整流側上的參考電位。第1圖上方的 電源分支是與下方的分支相反而爲具有正電位的正電源分 支。在此二個電源分支之間存在一種由N-通道-電晶體T1 和P-通道-電晶體T2所構成的半橋式電路。在該半橋式電 路之中點HB和正電源分支之間存在一種負載電路，其由一 與負載串聯的燈抗流圈L2、放電燈EL和一與負載串聯的耦 合電容器C7所構成。此外，一種與負載並聯的電路設有二 個共振電容器C8和C9以及一正溫度係數熱敏電阻KL以用 來點燃該燈。 爲了切換各MOSFET-電晶體T1和T2之放電現象，一電 容器C 6須與上方之半橋式電晶體T 1並聯。 在電晶體T 1和T2之作爲電晶體定向的參考電位用的源 極終端和個別的閘極終端之間存在一種控制電路AS。此控 制電路AS包括一由繞組L3、電容器C3和一串聯電路所構 成的並聯電路，該串聯電路由一控制變壓器之二次繞組 HW 1和一電阻R3所構成，其中該控制變壓器之主繞組是上 述之燈抗流圈L2。由電容器C4和電阻R4所構成的串聯電 路是與電阻R3並聯。 在電晶體Τ 1之閘極終端和源極終端之間，一種起動電容 器C5串聯至該控制電路AS。該控制電路AS和該起動電容 器C 5之間的連接點經由一歐姆電阻R1而與該正電源分支 -12- 200913789 相連接，該半橋-中點則經由一下拉電阻R2而與該負電源 分支相連接。 一個二極體D1並聯至該歐姆電阻ri。 爲了起動該電路’則該起動電容器C5須經由電阻R 1和 該下拉電阻R2而充電。在起動完成之後，即，在電晶體T1 第一次完全導通之後，該起動電容器C5基本上又經由該二 極體D1和該電晶體T1而放電。 在一種失敗的起動測試中，半橋-中點HB上的電位經由 該下拉電阻R2而又被導引至負參考電位。這樣使該起動電 容器C5又重新充電且可開始一種新的起動測試。 該起動測試的周期藉由電阻R1和R2以及電容器C5和電 容器C1上的中間電路電壓Uzw之大小來決定。由繞組L3 和電容器C3所構成的並聯共振電路經由電阻R3而與燈抗 流圈L2之二次繞組相連接。電阻R3以電容方式而被旁路。 此種電容方式的旁路可藉由一種差異特性而快速地對該並 聯共振電路之電容器C3進行充電，且因此可對該待接通的 半橋式電晶體T1提供一種足夠高的控制電壓。 然而，電阻R3之純電容方式的旁路會有以下的缺點：在 正常操作時，半橋式電路之整流相位中之控制電壓會太快 速地使個別待接通的電晶體被接通，這樣會產生損耗。爲 了使此效應減弱，則仍可將一電阻R4串聯至電容器C4。藉 由各構件R3, R4和C4之適當的大小’則可界定整個電路之 與起動和操作有關的最佳特性。 第2圖顯示本發明之電路配置之第二實施例的圖解’其 -13- 200913789 中該起動電容器C5之放電不是藉由二極體來達成而 歐姆電阻R1和R5來進行。 在操作時，該起動電容器C5之二個終端平均位於 處，該電位等於該半橋-電壓的一半。於是，電容器 電。當該半橋式電路不振盪時，該起動電容器C5充 乃因半橋側的一終端經由該下拉電阻R2而位於負電 處。 第3圖顯示第2圖之電路配置的一較佳的實施形 於該起動電容器C5之純歐姆放電持續一段時間，直 動電容器C5上的電壓消失爲止，則該控制電路AS 電壓在時間上具有一種正的偏移成份。此偏移成份 該電晶體T2之接通，此乃因該電晶體T2可支配的 由於該偏移電壓而下降。 因此，建議設置另一電容器C10，其作爲耦合電 接收上述之偏移成份-電壓。該控制電路AS之輸出 純交流電壓成份下降在該電阻R 1 0上，這樣可立即 種足夠的負電壓以控制該電晶體T2。 第4至7圖顯示本發明之不同的實施例中不同的 位時該起動電容器C5上的電壓、二個半橋式電晶體 電壓Um、半橋-電壓UHB以及流經該燈抗流圈L2之 流U的時間曲線圖。 第4圖顯示振盪開始時的電壓和電流特性。由第 可明顯看出，電晶體Tl，T2之控制電壓UCS逐漸增 然電容器C5上的電壓UC5由時間點tl時開始下降， 是經由 一電位 C5放 電，此 源電位 式。由 至該起 之輸出 可阻止 電壓會 容器而 電壓之 形成一 操作相 之控制 負載電 4圖中 加，雖 其中該 -14- 200913789 半橋-電壓Uhb可到達正電源電位。於是，該起動電容器C5 經由二極體D 1和電晶體T 1而放電。該控制電壓Ucs之上升 的梯度大於該負載電流U之上升的梯度。以電容器C4和電 阻R4所構成的串聯電路來使電阻R3旁路，則二次繞組HW 1 上的電壓之跳躍形式的上升可以差動（differential)方式傳 送至並聯共振電路C3，L3上。 第5圖以不同的時間解析度來顯示一種成功的振盪過 程，其是在失敗的測試之後進行。由第5圖可知，在一種 失敗的振盪測試時，該起動電容器C5上的電壓下降，此乃 因該起動電容器C5經由二極體D1而至少放電一次。由於 該半橋-電壓UHB藉由該下拉電阻R2而又被導引至負電源電 位的方向中，則可到達一種在施加主電源電壓之前已存在 的原來的電壓狀態且可自動進行一種新的起動測試。此種 特性可確保該電路配置之一種高的操作安全性。 第6圖以較高的時間解析度來顯示第5圖之失敗的振盪 測試。由第6圖可明顯看出，該控制電壓Ucs足夠使電晶體 T1接通，這樣可使該半橋-電壓Uhb到達該正電源電位之高 度處且該起動電容器C5可經由二極體D1而進行第一次放 電。當然，該控制電壓Ucs之負振盪不足以使電晶體T2接 通。整個振盪須進行一種漸進式的振盪停止過程。經由該 燈抗流圈L2之負載電流It之隨後所造成的振盪可周期性地 驅動該半橋·電壓Uhb至正電源電位，這樣可使該起動電容 器C 5達成一種多次的放電作用。 第7圖以類似於第4圖的方式來顯示電路配置的振盪， -15- 200913789 其使用第2圖所示的電路配置。由第7圖中明顯可看出， 當該半橋-電壓Uhb已到達正電源電位的高度時，該起動電 容器C5不再以突然的方式來放電。 第1至3圖之實施形式同時顯示互補的電晶體τι, T2 時，則本發明亦可用來實現半橋式電路，其具有二個獨立 的控制電路和相同極性的電晶體。當然，這相對於上述各 種電路配置而3需要更多的成本。 【圖式簡單說明】 第1圖本發明之電路配置之第一實施例的圖解。 第2圖本發明之電路配置之第二實施例的圖解。 第3圖本發明之電路配置之第三實施例的圖解。 第4圖顯示第1圖之實施例中一起動電容器上的電壓、二 個半橋式電晶體之控制電壓、半橋-電壓以及流經該燈抗流 圏之負載電流的時間曲線圖。 第5圖顯示第1圖之實施例中一起動電容器上的電壓、二 個半橋式電晶體之控制電壓、半橋-電壓以及流經該燈抗流 圈之負載電流在與第4圖不同的時間解析度時的時間曲線 圖’特別是在失敗的測試之後發生一種少見的振盪過程時 的時間曲線圖。 第6圖顯示第5圖中在較高的時間解析度時在失敗的振盪 過程之範圍中各種値之時間曲線圖。 第7圖以類似於第4圖的方式顯示第2圖之電路配置之相 對應的數値之時間曲線圖。 -16- 200913789 主要 元件之符號 說 明 Cl〜 C9 電 容 器 D 1 二 極 體 LI、 L3 繞 組 L2 燈 抗 流 圈 R1〜 R5 電 阻 GL 整 流 器 SI 保 險 絲 AS 控 制 電 路 HB 半 橋 1點 Hw 1 電 感 T1〜 T2 電 子 式 開關 II 負 載 電 流 EL 放 電 燈 KL 正 溫 度 係數熱敏電阻 U Η B 半 橋 -電壓 Ugs 控 制 電 壓 UC5 電 壓 -17-200913789 IX. Description of the Invention: [Technical Field] The present invention relates to a circuit arrangement for operating a discharge lamp, the discharge lamp comprising: a rectifier having a first and a second input for communicating with an The voltage source is coupled; a first and a second output to provide a DC operating voltage; a free-running half-bridge inverter comprising a series circuit formed by two voltage controlled electronic switches, It is coupled between the first and second output ends of the rectifier, wherein a first half-bridge-mid point is formed between the first and second electronic switches; and a moving circuit is used to start a free oscillation of an inverse rectifier The starting circuit includes a control circuit having a first output and a second output, the first output of the control circuit being coupled to the control electrodes of the first and second electronic switches; A capacitor having first and second terminals. The invention also relates to a method of operating a discharge lamp in the above described circuit arrangement. [Prior Art] The present invention generally relates to a circuit configuration that can independently activate a freely oscillating half-bridge inverter having a plurality of voltage controlled transistors, such as MOSFETs and IGBTs, for applying a Used as a switch when the power supply voltage. EP 0 9 1 7 4 1 2 B 1 is used to form the preamble of the independent item of the scope of the patent application of the present invention, whereby it is known in the document that the first half-bridge transistor can be provided by charging a capacitor. The required voltage is turned on in turn, and the capacitor is connected in series to a control circuit for controlling the transistor of the half-bridge inverter. In order to prevent an undesired stable state of charge of the capacitors connected in series and thus also to prevent the "stay" state of the half-bridge inverter, it is suggested in the patent document of the above - 6 - 200913789 to be directly The power source charges the capacitor through a resistor. Through the main power terminal, the test can be repeated for one time at one half frequency of the power frequency until the self-oscillation of the half bridge inverter starts. In the document, only some embodiments are shown, in which two half-bridge transistors are respectively provided with control circuits. However, in order to reduce the cost or simplify the circuit configuration, the half-bridge inverter may also be a complementary transistor (η- And p-channel), in which case the two transistors are controlled by the same control circuit. This principle is known from ΕΡ 0 78 1 077 Β1, in order to start the half-bridge inversion in this file. A bidirectional diode (Diac) is also required. The complication based on ΕΡ 0 917 412 Β 1 can be transferred to the well-known circuit configuration in EP 0 7 8 1 077 B1, in which case the circuit is activated. The required capacitor is connected in series to a control circuit. Thus, the "power-on" function can be fully set. However, the above document shows a disadvantage in that one of the common resistances required to charge the start capacitor is The discharge lamp is actually located at the half bridge-midpoint potential. By the other terminal of the starting resistor being connected to one of the two power supply wires, a high frequency rectangular voltage of the midpoint potential can be started via the start The resistor is at the power supply potential. This causes high radio interference in the entire circuit configuration due to the current flowing through the starting resistor and is therefore a high frequency, approximately rectangular interference signal on the wireless interference filter. Escaped and directed directly to the power supply. To prevent this, if the starting resistor is connected to the terminal of the rectifier that supplies the positive supply voltage (so-called intermediate circuit voltage) rather than to one of the two power leads, The disadvantage shown in this way is that this circuit configuration can only be performed on a -7-200913789 start-up test, ie, in a failure The circuit maintains a "stay" state during the start-up test. SUMMARY OF THE INVENTION It is an object of the present invention to continue to provide a circuit arrangement or method of the above-described form in order to prevent high radio interference on the one hand and to apply the main power supply on the other hand. The voltage can then ensure a reliable start-up of the circuit configuration. The invention is based on the recognition that the combination of the principles of EP 0 78 1 077 B 1 and EP 〇 917 412 B1 and the subsequent starting resistor (not the power supply conductor) The above purpose cannot be achieved when fixed to the positive output of the rectifier, since a stable voltage ratio can be set on the starting capacitor, which would hinder a repetitive start-up test. Therefore, according to the present invention, the starting capacitor is The half bridge-midpoint is coupled and the half bridge-midpoint is coupled to the second output of the rectifier via a pull-up resistor. This configuration allows the half-bridge-midpoint potential to be directed via the pull-down potential to a negative reference potential provided by the second output of the rectifier after a failed start-up test. Therefore, the starting capacitor is recharged and a new start-up test can be performed. Since the resistor is no longer connected to one of the main power sources but to the positive output of the rectifier in such a circuit configuration, wireless interference can be prevented from being transmitted to the main power source. The capacitor is connected to a positive potential via the resistor. In a preferred embodiment, the circuit arrangement additionally has a lamp choke having at least one main winding, the main winding being connected in series with a terminal for the discharge lamp, where a control circuit has an inductance, The inductor has a first terminal and a second terminal, and the inductance is a secondary 200913789 winding of the lamp choke. The circuit arrangement further has a series circuit comprising a second ohmic resistor and a parallel resonant circuit coupled between the first and second terminals of the inductor. It is particularly advantageous to connect a half bridge circuit comprising a capacitor in parallel to a second ohmic resistor. Such a particularly advantageous embodiment is based on the recognition that in order to activate the circuit arrangement, the control voltage of the first transistor to be switched on after the application of the mains voltage must be kept sufficiently large, when the starting capacitor is used by the transistor This is also the case when the switch-on has been discharged and thus the effective control voltage (i.e., the sum of the voltage across the start capacitor and the output voltage of the control circuit) tends to decrease. The maintenance of sufficient control voltage required for the transistor when the starting capacitor begins to discharge is preferably ensured by the fact that the control circuit can quickly dominate a sufficiently large output voltage. For this purpose, since the capacitor exhibits a very low resistance to the high frequency spectral components of the turn-on pulses, the second ohmic resistor becomes by-passed, the second ohmic resistor during operation This is required for the control of the electronic switch of the half bridge inverter. The charge carriers flowing through the capacitor thus enable very fast charging of a capacitor disposed in the parallel resonant circuit. This provides a sufficiently large control voltage for the half-bridge transistor to be switched on first. The risk of an insufficiently large control voltage on the first switch to be switched of the half-bridge inverter is therefore effectively limited when the starting capacitor is discharged. The half bridge circuit preferably includes a third ohmic resistor connected in series to the capacitor. This third ohmic resistor brings two advantages: one of the advantages is that the rectangular component of the voltage provided by the secondary winding on the lamp anti-flow can be attenuated. This is related to the fact that the lamp anti-flow is connected on the one hand to the half-bridge midpoint (on which a high-frequency rectangular signal is present) and on the other hand to the lamp. The lamp applies a signal that resembles a sinusoidal form to the other end of the lamp that is resistant to flow. The third ohmic resistor can dominate a substantially sinusoidal form of the signal during operation to control the two electronic switches of the half-bridge inverter. Another advantage is that the third ohmic resistor can extend the dead time that may be present during the oscillation of the half-bridge inverter. Thus, the switch of the half bridge inverter can be made to achieve a lossless switching. Generally, it is appropriate to make the size of the third ohmic resistor much smaller than the size of the second ohmic resistor. Furthermore, it is advantageous when the circuit arrangement of the invention additionally has a diode which is connected in parallel with the first ohmic resistor and which must be oriented "when the first electronic switch is switched on, the diode can The starting capacitor is discharged via the first electronic switch." Therefore, the diode discharges the start capacitor after the start capacitor is successfully activated, and thus, there is no other start-up test to interfere with the operation of the circuit configuration. In a preferred embodiment, the first terminal of the starting capacitor is coupled to the second output of the rectifier via another ohmic resistor. In this embodiment, it is not necessary to provide a previously described diode. The discharge of the starting capacitor is achieved via the first and another ohmic resistors. The reason for this is that, in operation, the two terminals of the starting capacitor are located at a potential which is exactly equal to one-half of the half-bridge voltage, so that the starting capacitor can be discharged during operation. Thus, it is advantageous when another capacitor is provided, the other capacitor coupling -10-200913789 being connected between the first output of the control circuit and the control electrode of the second electronic switch. Since the pure ohmic discharge of the starting capacitor continues for a period of time until the voltage disappears via the starting capacitor, the output voltage of the control circuit has a positive offset component. Such an offset component prevents the second transistor from being turned on because the control voltage attainable by the second transistor can be reduced by the offset component. Another capacitor suggested is used as a coupling capacitor and receives the offset-voltage described above. The pure AC voltage component of the output voltage of the control circuit is decreased by an optional ohmic resistance coupled between the control electrode and the working electrode of the second electronic switch, which can be sufficiently negative immediately Voltage to control the second transistor. The two electronic switches of the half-bridge inverter may be transistors of complementary polarity, but may also be transistors of the same polarity. The advantage of using a complementary transistor is that only one control circuit has to be provided. Further advantageous embodiments are known from the respective patents of the scope of the patent application. The preferred embodiment set forth with reference to the circuit arrangement of the present invention and its advantages are also applicable to the method of operation of the present invention as long as it is available. [Embodiment] The same reference numerals are used for the same or identical components in the following description. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a diagram showing a first embodiment of the circuit configuration of the present invention, which is configured to operate a discharge lamp which is represented by a load EL. Therefore, a rectifier GL is supplied with a mains voltage ‘ via a fuse S I to supply it to an electrolytic capacitor C1 or to maintain a voltage. On the -11-200913789 electrolytic capacitor c1, two kinds of power supply branches can be measured via a filter, which is composed of a winding L1 in one of the branches and a capacitor C2 connecting the two branches. The power branch below Figure 1 has a negative potential and defines the reference potential on the rectified side of the circuit configuration. The power supply branch above the first figure is a positive power supply branch with a positive potential opposite to the lower branch. Between the two power supply branches, there is a half bridge circuit composed of an N-channel-transistor T1 and a P-channel-transistor T2. Between the half-bridge circuit point HB and the positive power supply branch, there is a load circuit which is constituted by a lamp choke coil L2 connected in series with the load, a discharge lamp EL and a coupling capacitor C7 connected in series with the load. In addition, a circuit in parallel with the load is provided with two resonant capacitors C8 and C9 and a positive temperature coefficient thermistor KL for igniting the lamp. In order to switch the discharge phenomena of the MOSFET-transistors T1 and T2, a capacitor C 6 must be connected in parallel with the upper half-bridge transistor T 1 . There is a control circuit AS between the source terminals of the transistors T1 and T2 as the reference potential for the transistor orientation and the individual gate terminals. The control circuit AS comprises a parallel circuit composed of a winding L3, a capacitor C3 and a series circuit. The series circuit is composed of a secondary winding HW 1 of a control transformer and a resistor R3, wherein the main winding of the control transformer is The above-mentioned lamp choke coil L2. The series circuit composed of the capacitor C4 and the resistor R4 is connected in parallel with the resistor R3. Between the gate terminal and the source terminal of the transistor , 1, a starting capacitor C5 is connected in series to the control circuit AS. The connection point between the control circuit AS and the starting capacitor C 5 is connected to the positive power supply branch -12-200913789 via an ohmic resistor R1, and the half bridge-mid point is connected to the negative power supply via a pull-up resistor R2 The branches are connected. A diode D1 is connected in parallel to the ohmic resistor ri. In order to activate the circuit 'the starter capacitor C5 must be charged via the resistor R 1 and the pull-down resistor R2. After the start-up is completed, i.e., after the transistor T1 is fully turned on for the first time, the start capacitor C5 is substantially discharged again via the diode D1 and the transistor T1. In a failed start-up test, the potential on the half-bridge-midpoint HB is again directed to the negative reference potential via the pull-down resistor R2. This causes the starting capacitor C5 to be recharged again and a new start-up test can be initiated. The period of the start-up test is determined by the magnitudes of the resistors R1 and R2 and the capacitor C5 and the intermediate circuit voltage Uzw on the capacitor C1. A parallel resonant circuit composed of a winding L3 and a capacitor C3 is connected to the secondary winding of the lamp choke coil L2 via a resistor R3. Resistor R3 is bypassed capacitively. This capacitive bypass allows the capacitor C3 of the parallel resonant circuit to be quickly charged by a differential characteristic, and thus a sufficiently high control voltage can be provided to the half-bridge transistor T1 to be turned on. However, the pure capacitive bypass of resistor R3 has the following disadvantages: during normal operation, the control voltage in the rectified phase of the half-bridge circuit will cause the individual transistors to be turned on to be turned on too quickly, thus Will produce losses. To reduce this effect, a resistor R4 can still be connected in series to capacitor C4. By the appropriate size ' of each of the components R3, R4 and C4', the optimum characteristics of the entire circuit in relation to starting and operation can be defined. Fig. 2 is a view showing the second embodiment of the circuit configuration of the present invention. The discharge of the start capacitor C5 in the -13-200913789 is not performed by the diodes and the ohmic resistors R1 and R5. In operation, the two terminals of the starting capacitor C5 are located on average, which is equal to half the voltage of the half bridge. Thus, the capacitor is electrically charged. When the half bridge circuit does not oscillate, the start capacitor C5 is charged because a terminal on the half bridge side is at a negative power via the pull-down resistor R2. Figure 3 shows a preferred embodiment of the circuit configuration of Figure 2 in which the pure ohmic discharge of the start capacitor C5 is continued for a period of time, and the voltage on the linear capacitor C5 disappears, then the control circuit AS voltage has time in time. A positive offset component. This offset component turns on the transistor T2, which is degraded by the offset voltage due to the transistor T2. Therefore, it is proposed to provide another capacitor C10 which receives the above-mentioned offset component-voltage as a coupling electric power. The output of the control circuit AS has a pure AC voltage component falling across the resistor R 1 0 so that a sufficient negative voltage can be immediately applied to control the transistor T2. Figures 4 through 7 show the voltage across the starting capacitor C5, the two half-bridge transistor voltages Um, the half-bridge voltage UHB, and the flow through the lamp choke L2 at different bits in different embodiments of the present invention. The time curve of the flow U. Figure 4 shows the voltage and current characteristics at the beginning of the oscillation. It can be clearly seen from the first that the control voltage UCS of the transistors T1, T2 gradually increases. The voltage UC5 on the capacitor C5 starts to decrease from the time point t1, and is discharged via a potential C5, which is the source potential type. The output from this can prevent the voltage from being formed in the container and the voltage is formed into an operational phase. The load is added to the load, although the -14-200913789 half-bridge-voltage Uhb can reach the positive supply potential. Thus, the starting capacitor C5 is discharged via the diode D 1 and the transistor T 1 . The gradient of the rise of the control voltage Ucs is greater than the gradient of the rise of the load current U. When the resistor R3 is bypassed by a series circuit composed of the capacitor C4 and the resistor R4, the rise of the voltage jump pattern on the secondary winding HW1 can be differentially transmitted to the parallel resonant circuits C3, L3. Figure 5 shows a successful oscillation process with different time resolutions, which is done after the failed test. As can be seen from Fig. 5, in a failed oscillation test, the voltage on the starting capacitor C5 drops because the starting capacitor C5 is discharged at least once via the diode D1. Since the half bridge-voltage UHB is again guided into the direction of the negative power supply potential by the pull-down resistor R2, an original voltage state existing before the application of the main power supply voltage can be reached and a new one can be automatically performed. Start the test. This feature ensures a high operational safety of the circuit configuration. Figure 6 shows the failed oscillation test of Figure 5 with a higher time resolution. As is apparent from Fig. 6, the control voltage Ucs is sufficient to turn on the transistor T1, so that the half bridge-voltage Uhb reaches the height of the positive power supply potential and the starting capacitor C5 can pass through the diode D1. Perform the first discharge. Of course, the negative oscillation of the control voltage Ucs is insufficient to turn on the transistor T2. The entire oscillation must be subjected to a progressive oscillation stop process. The subsequent oscillation caused by the load current It of the lamp choke coil L2 can periodically drive the half bridge voltage Uhb to the positive power supply potential, so that the starting capacitor C 5 can achieve a multiple discharge action. Figure 7 shows the oscillation of the circuit configuration in a manner similar to Figure 4, -15-200913789 which uses the circuit configuration shown in Figure 2. As is apparent from Fig. 7, when the half bridge-voltage Uhb has reached the height of the positive power supply potential, the starting capacitor C5 is no longer discharged in a sudden manner. When the embodiments of Figures 1 to 3 simultaneously show complementary transistors τι, T2, the invention can also be used to implement a half-bridge circuit having two independent control circuits and transistors of the same polarity. Of course, this requires more cost than the various circuit configurations described above. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a diagram showing a first embodiment of a circuit configuration of the present invention. Figure 2 is an illustration of a second embodiment of the circuit arrangement of the present invention. Figure 3 is an illustration of a third embodiment of the circuit configuration of the present invention. Figure 4 is a graph showing the time at which the voltage across the movable capacitor, the control voltage of the two half-bridge transistors, the half-bridge-voltage, and the load current flowing through the lamp's anti-flow current in the embodiment of Figure 1. Figure 5 shows the voltage on the moving capacitor, the control voltage of the two half-bridge transistors, the half-bridge voltage, and the load current flowing through the lamp choke in the embodiment of Figure 1 different from Figure 4 The time plot of the time resolution' is especially a time plot of a rare oscillation process that occurs after a failed test. Fig. 6 is a graph showing the time curves of various flaws in the range of the failed oscillation process at a higher time resolution in Fig. 5. Fig. 7 is a timing chart showing the corresponding number of circuit configurations of Fig. 2 in a manner similar to Fig. 4. -16- 200913789 Symbols of main components Description Cl~ C9 Capacitor D 1 Diode LI, L3 Winding L2 Lamp choke R1~ R5 Resistor GL Rectifier SI Fuse AS Control circuit HB Half bridge 1 point Hw 1 Inductance T1~ T2 Electronics Switch II Load Current EL Discharge Lamp KL Positive Temperature Coefficient Thermistor U Η B Half Bridge - Voltage Ugs Control Voltage UC5 Voltage-17-