【導(dǎo)讀】Abstract. I.INTRODUCTION

　　

【正文】 d during the 1 .5 kV test at 400 Hz is shown in Fig. 15. From the figure it can be seen that the PFN charges up。 after 2 ms the DUT is fired and the voltage collapses. Also it can be seen that the waveform has a droop of 25%. The first 10pswide anode current pulse of the fivepulse burst has a peak amplitude of kA and a di/dt of approximately 10 000 A/ps (Fig. 16).The 100A gate pulse, which initiates the turnon of the DUT, is also shown in Fig. 16. The amplitude of the fifth current pulse is 25% lower than the first pulse. These devices were tested under these conditions more than ten times and they successfully worked under these repetitive conditions. Both the devices were then characterized by measuring the foward and reverse leakage currents at 25 and 125176。C. and the gatecathode VI characteristics. The devices showed no degradation from the switching stresses. The experimental results presented above verify the numerical analysis and results presented in [10], by confirming that these devices do have the potential to operate reliably under repetitive high dildt conditions. Fig. 15. Typical anodecathode voltage across the thyristor under test. CH1500 V/div. Time base2 ms/div. Fig. 16. Gate pulse to the DUT (CH1) and anode current (CH2). CH1 20 A/div, CH21350 A/div. Time base2 pddiv. VII. CONCLUSION The differences in the switching characteristics of the shorted and unshorted devices were explained using a physical model. The role of the amplifying gate structure in the turn on process and its detrimental effect on the performance of the device for switching highcurrent pulses with risetimes on the order of 100 to 200 ns was explained. The phase I1 interval seen in the anode current waveform of the unshorted device was theoretically shown to be the time taken for the lateral and vertical transit of the carriers in the pbase. The experimentally observed burntspot area was correlated with theory based on the temperature for failure of the device. Using this information the current density for failure was obtained. The safe operating frequencies of the devices under high dildt conditions were estimated from the thermal analysis. Two types of thyristors, an SCR and a GTO, were tested as closing switches for switching a 400 Hz, 5pulse repetitive burst from a PFN. Both the devices successfully switched a 5pulse burst of 8 kA, 10 ps wide current pulses having a difdt of 10000 A/ps from a state of forward blocking at kV. The high difdt repetitive switching was made possible by driving the shorted device using 100 A, 800 ns wide gate pulses, thus ensuring that sufficient area was initially tumed on to keep the localized heating to acceptable levels. These results help verify the numerical safe operating frequency estimations. These results indicate that thyristors have the potential to replace some conventional gas switches currently being used in various pulsed power systems (in the 110kV 1100 kA range). REFERENCES (1) N. Mapham, “The rating of siliconcontrolled rectifiers when switching into high currents,” IEEE Trans. Commun. Electron., vol. 83, pp. 515519, Sept. 1964. (2) J. L. Hudgins and W. M. Portnoy, “Gating effects on thyristor anode current di/dt,” IEEE Trans. Power Electron., vol. PE2, pp. 149153, 1987. (3) V. A. Sankaran, J. L. Hudgins, and W. M. Portnoy, “Role of the amplifying gate structure in the tumon process of involute structure thyristors,” IEEE Trans. Power. Electron.. vol. 5, pp. 125132, Apr. 1990. (4) C. E. Kennedy, J. L. Hudgins, V. A. Sankaran, and W. M. Portnoy, “ Comparison of GTOs and SCRs for high di/dt switching,” Conf. Rec. /AS Ann. Meeting, Oct. 1990, pp. 16431647. (5) J. L. Hudgins and W. M. Portnoy, “High di/dt pulse switching of thyristors,” IEEE Trans. Power Elertron., vol. PE2, pp. 143148, 1987. (6) K. Okamura, Y. Watanabe, I. Ohshima, and S. Yanabu, “Highspeed,highpower switching of semiconductor devices,” 7thlEEE Pulse Power Conf., June 1989, pp. 836839. (7) J. Vitins, J. L. Steiner, and A. Schweizer, “Reverse conducting thyristors replace thyratrons in submicrosecond pulse generation,” Pror. 6th IEEE Pulse Power Conf.. June 1987, pp. 591594. (8) W. E. Newell, “Transient thermal analysis of solidstate power de vicesmaking a dreaded process easy,” IEEE Trans. Ind. Appl., vol. IA12, pp. 405420, July/Aug. 1976. (9) S. Ikeda, S. Tsuda, and Y. Waki, “The current pulse ratings of thyristors.” IEEE Trans. Electron Devices, vol. ED17, pp. 69M93, Sept. 1970. (10) V. A. Sankaran, J. L. Hudgins. C. A. Rhodes, and W. M. Portnoy, “A numerical approach based on transient thermal analysis to estimate the safe operating frequencies of thyristors,” to be published in the IEEE Trans. Power Electron..vol. 6, July 1991. (11) G. D. Bergman, “The gate triggered tumon process in thyristors,” SolidState Electron.. vol. 8, pp. 157765, 1965. 晶閘管的 高能脈沖式開(kāi)關(guān)特性 摘要 —— 通過(guò)實(shí)驗(yàn)對(duì)有和沒(méi)有放大門的 可控硅 的高能、 高 di/dt 的 脈沖式開(kāi)關(guān)特性 進(jìn)行了研究。 發(fā)現(xiàn)沒(méi)有放大門的裝置比有放大門的裝置性能好的多。一種 物理模型描述 了 放大 門在開(kāi)關(guān)接通過(guò)程中放大門的作用，可以用來(lái)解釋開(kāi)關(guān)特性的差異 。 器件接通部分的失誤都來(lái)自于理論上的計(jì)算和相關(guān)的觀察。 這使計(jì)算的電流密度 存在失誤的可能 .在高 di/dt 條件 下這些器件出現(xiàn)的都是發(fā)熱問(wèn)題，可以 利用 模擬 有限元方法 對(duì)器件的溫升進(jìn)行估算 。 這個(gè)模擬試驗(yàn)表明 有放大門器件的 溫升明顯高于 沒(méi)有放大門的器件 。 從這些 結(jié)果中可以知道 , 在高 di/dt 條件 下所有設(shè)備的 安全運(yùn)行頻率 是可以進(jìn)行估算的 。 這些估 算通過(guò)在高 di/dt 條件下對(duì)器件進(jìn)行 重復(fù) 試驗(yàn)而得到證實(shí)。 導(dǎo)言 近來(lái)隨著 半導(dǎo)體器件設(shè)計(jì) 的改革及其 制造技術(shù)的 發(fā)展， 大功率晶閘管 得到了廣泛的應(yīng)用 。 這些裝置的設(shè)計(jì) 都是應(yīng)用連續(xù)的模式實(shí)現(xiàn)運(yùn)行， 如 AC/DC 電源轉(zhuǎn)換 以及電動(dòng) 機(jī) 。 直到最近 才發(fā)現(xiàn) 其 在 高功率脈沖開(kāi)關(guān) 的應(yīng)用中還是 未知數(shù) . 主要原因之一 是在 低 di/dt 下 阻礙了高速 可控硅 , 高能 開(kāi)關(guān)的使用 。 限值的 di/dt 值在故障發(fā)生 之前 與剛接通部分的面積和傳播速度有關(guān)。 最近的實(shí)驗(yàn)結(jié)果 [2][4]表明 ,隨著門裝置 的發(fā)展， 具有高度 interdigitated 門 陰極結(jié)構(gòu) 的 可控硅及 GTO 可以可靠地在高 di/dt 條件 下運(yùn)行 . 此前 ,可控硅也被重復(fù) 10 小時(shí)用于 具有 di/dt 的 1KA, 10ps、 55HZ 及 800HZ 的 寬脈沖 中。 據(jù)報(bào)道 [6]GTO 模塊 (五器件系列 )在 頻率為100 赫茲 時(shí) 可以 阻礙 115 千伏和 轉(zhuǎn)換 具有 2500/ks 的 di/dt 的脈沖。 不對(duì)稱的裝置 ,如 AS