E类的PLL SSTC


A small Tesla coil will have a high resonant frequency, making conventional driving techniques difficult.一个小型的特斯拉线圈将具有较高的共振频率，使传统的驾驶技术困难。 What were once low capacitances suddenly seem huge, and insignificant switching times become a considerable portion of the switching period.什么是一次低电容似乎突然巨大的，微不足道的切换时间成为相当一部分转换时期。 A whole array of new problems arise as frequency increases.一系列新的问题出现的频率增加。 However small coils only need small amounts of power to make a decent spark-show, allowing for simpler topologies. Originally I had intended to run this coil off-line with a half-bridge but I was unable to achieve decent gate drive through the GDT.然而小线圈只需要少量的电力，使体面火花出现，使简单的拓扑结构。本来我打算运行此线圈离线与半桥但我无法实现体面栅极驱动通过医疗设备生产商。 So instead I switched to class E, which only requires one switching device and no GDT.所以不是我转向E级，只需要一个开关装置，也没有医疗设备生产商。


Update: This circuit doesn't do real PLL. 更新：这条赛道不会真正的PLL 。 It will try to adjust the frequency until a 90 degree phase shift occurs between it's own output and the CT feedback, which is impossible at resonance because the CT feedback and 4046 output are supposed to be in phase. 它会尝试调整频率 ， 直至90度之间发生相移它自己的输出和反馈的CT ，这是不可能的 ， 因为在共振的CT反馈和4046输出都应该是在第一阶段。 Therefor you will need to turn the pot all the way in one direction to bias the VCO enough to outweigh this. 为此 ， 您需要把锅所有的方式 ， 一个方向偏压控振荡器足以超过这一点。 Using an antenna instead of CT will fix the problem, and I'm working on a improved circuit. 而不是使用一个天线的CT将修复这个问题，我工作的改进电路。 Bottom line, wait for the upgrade. 底线，等待升级。




Click on the schematic for a larger image. 点击示意图大图。

Circuit Function: The phase locking is preformed by the 4046 chip. 电路功能：锁相是预制的4046芯片。 First some basics.第一部分基础知识。 The 4046 has an internal VCO, or voltage controlled oscillator.在4046有一个内部振荡器，或电压控制振荡器。 It's frequency is controlled by the voltage at pin 9.它的频率是控制的电压引脚9 。 The 4046 also has internal phase comparators, in this circuit only phase comparator 1 is used, which is just a XOR logic gate.在4046也有内部相比较，在这条赛道上只有1相比较使用，这仅仅是一个异或逻辑门。 First one sets the frequency range of the internal VCO with the 100pF capacitor, and resistors on pins 11 and 12.第一台的频率范围内的内部振荡器与100pF电容和电阻器的引脚11和12 。 The pin 11 resistor sets the upper frequency, while pin 12 the lower frequency. 11引脚上的电阻套频率，而12针的频率较低。 By adjusting the voltage at pin 9 the VCO frequency can be moved up or down between the set frequency points.通过调整电压引脚9振荡器频率可以向上或向下移动集之间的频率点。 Pin 4 is the VCO output, which will oscillate regardless of input.引脚4是压控振荡器的输出，这将振荡无论投入。 This allows it to start the SSTC, which in turn provides feedback through the secondary base current transformer.这使得它开始SSTC ，这反过来又提供反馈通过电流互感器二次基地。 The base current signal and the output from the VCO itself are compared by a XOR gate.该基地目前的信号和输出压控振荡器本身比较的异或门。 (The VCO output is actually sampled at the IRF630 drain, due to delays in the driver transistors and IRF630 itself.) The output from the XOR gate is a PWMed signal, which represents the phase difference between the VCO output and base current itself. （压控振荡器输出实际上是取样在IRF630流失，由于延误司机晶体管和IRF630本身。 ）的输出异或门是一个PWMed信号，它代表了阶段之间的差异和基础的VCO输出电流本身。 Since the VCO is controlled linearly by the voltage at pin 9, pulsed DC would simply send it to max frequency and back down again.由于压控振荡器的线性控制电压引脚9脉冲直流只会发送到最高频率和回了。 What is needed is a steady DC voltage proportional to the PWMed signal, which is created by the 120k resistor and 1nF cap. The VCO can be biased by changing the constant voltage at pin 9 with the potentiometer.现在需要的是稳定的直流电压成正比PWMed信号，这是所造成的12万电阻1nF帽。压控振荡器可偏见通过改变电压引脚9上的电位器。 This effectively allows one to adjust the phase angle.这实际上允许一个调整相位角。 Music can be modulated into the output signal by inserting it into pin 9.音乐可调制到输出信号的插入引脚9 。 Unfortunately the corona hisses and distorts the music.不幸的是日冕hisses和歪曲了音乐。

Class E is almost as simple as it looks, basically one switches in resonance with the series resonant circuit formed by the primary inductance and matching capacitor. E类几乎是那样简单，因为它看起来，基本上一个开关谐振与串联谐振电路所形成的初级电感和匹配电容。 The whole point is to tune the primary and resonant capacitor until the circuit is critically damped (doesn't ring below zero), and turn on the mosfet just as the voltage reaches zero. This allows the mosfet to turn on with no voltage across it, ZVS, which eases switching and decreases switching losses. Damping of the resonant circuit is done by adjusting the primary coupling.整个的一点是要调整小学和谐振电容器，直到电路极其阻尼（不低于零环） ，并打开的MOSFET正如电压达到零。这使得MOSFET的开启电压，没有它，开关，开关，从而简化和减少开关损耗。阻尼谐振电路进行调整的主要耦合。 Tighter coupling causes energy to be drawn from the primary faster, which causes more damping.严格的耦合原因是能源的主要选自更快，从而导致更多的阻尼。 Unlike conventional SSTCs coupling should be fairly loose, almost like a SGTC.与传统的SSTCs耦合应该是相当松散的，几乎像一个SGTC 。 Since this has been elaborated much better before, I'll point you to some other pages which describe class E switching much better. Richie's page and Steve Ward's page are great resources. Steve Ward's page has simulated waveforms which greatly help the tuning process.由于这已经制定更前，我就点你的一些其他网页描述E类开关好得多。 里奇的网页和史蒂夫沃德的网页是巨大的资源。史蒂夫沃德的页面已经模拟波形，这大大有助于调整进程。 See if you can recognize these:看看你是否可以承认这些：


  

Bottom trace is gate voltage, top is drain. 底部Trace是栅极电压，顶端是流失。 The gate signal sure looks nasty when out of tune. 门信号肯定期待讨厌当格格不入。


The phasing of the CT and primary are very important.分阶段的CT和初级是非常重要的。 While experimenting with an antenna I found that I could only achieve breakout with the primary phased oppositely as the secondary.虽然试验天线我发现我只能实现突破的主要分阶段相反的是次要的。 The CT phasing was also critical for a phase lock, but I was unable to determine which direction is required.的CT分期也是很关键的一个阶段锁，但我无法确定哪个方向是必要的。 The symptoms of improper phasing are no breakout, breakout only after "coaxing" one out with an arc or sudden loss of phase lock while tuning.症状逐步不当没有突破，突破后，才“哄”一出一弧或突然失去锁相而调整。

My coil draws about 80W from 50V, and runs at 1.38MHz.我国线圈提请有关80W的50V ，并运行在1.38MHz 。 The secondary former is 5 diameter * 8 cm tall.中学前5直径8厘米。 The entire coil with control electronics fit in the palm of my hand, hence the nickname palm-top SSTC.整个线圈与控制电路配合手掌的我的手，因此昵称掌上SSTC 。


  
   这个是用google自动翻译的


这个是原版

Class E PLL SSTC


A small Tesla coil will have a high resonant frequency, making conventional driving techniques difficult. What were once low capacitances suddenly seem huge, and insignificant switching times become a considerable portion of the switching period. A whole array of new problems arise as frequency increases. However small coils only need small amounts of power to make a decent spark-show, allowing for simpler topologies. Originally I had intended to run this coil off-line with a half-bridge but I was unable to achieve decent gate drive through the GDT. So instead I switched to class E, which only requires one switching device and no GDT.


Update: This circuit doesn't do real PLL. It will try to adjust the frequency until a 90 degree phase shift occurs between it's own output and the CT feedback, which is impossible at resonance because the CT feedback and 4046 output are supposed to be in phase. Therefor you will need to turn the pot all the way in one direction to bias the VCO enough to outweigh this. Using an antenna instead of CT will fix the problem, and I'm working on a improved circuit. Bottom line, wait for the upgrade.




Click on the schematic for a larger image.

Circuit Function: The phase locking is preformed by the 4046 chip. First some basics. The 4046 has an internal VCO, or voltage controlled oscillator. It's frequency is controlled by the voltage at pin 9. The 4046 also has internal phase comparators, in this circuit only phase comparator 1 is used, which is just a XOR logic gate. First one sets the frequency range of the internal VCO with the 100pF capacitor, and resistors on pins 11 and 12. The pin 11 resistor sets the upper frequency, while pin 12 the lower frequency. By adjusting the voltage at pin 9 the VCO frequency can be moved up or down between the set frequency points. Pin 4 is the VCO output, which will oscillate regardless of input. This allows it to start the SSTC, which in turn provides feedback through the secondary base current transformer. The base current signal and the output from the VCO itself are compared by a XOR gate. (The VCO output is actually sampled at the IRF630 drain, due to delays in the driver transistors and IRF630 itself.) The output from the XOR gate is a PWMed signal, which represents the phase difference between the VCO output and base current itself. Since the VCO is controlled linearly by the voltage at pin 9, pulsed DC would simply send it to max frequency and back down again. What is needed is a steady DC voltage proportional to the PWMed signal, which is created by the 120k resistor and 1nF cap. The VCO can be biased by changing the constant voltage at pin 9 with the potentiometer. This effectively allows one to adjust the phase angle. Music can be modulated into the output signal by inserting it into pin 9. Unfortunately the corona hisses and distorts the music.

Class E is almost as simple as it looks, basically one switches in resonance with the series resonant circuit formed by the primary inductance and matching capacitor. The whole point is to tune the primary and resonant capacitor until the circuit is critically damped (doesn't ring below zero), and turn on the mosfet just as the voltage reaches zero. This allows the mosfet to turn on with no voltage across it, ZVS, which eases switching and decreases switching losses. Damping of the resonant circuit is done by adjusting the primary coupling. Tighter coupling causes energy to be drawn from the primary faster, which causes more damping. Unlike conventional SSTCs coupling should be fairly loose, almost like a SGTC. Since this has been elaborated much better before, I'll point you to some other pages which describe class E switching much better. Richie's page and Steve Ward's page are great resources. Steve Ward's page has simulated waveforms which greatly help the tuning process. See if you can recognize these:




Bottom trace is gate voltage, top is drain. The gate signal sure looks nasty when out of tune.


The phasing of the CT and primary are very important. While experimenting with an antenna I found that I could only achieve breakout with the primary phased oppositely as the secondary. The CT phasing was also critical for a phase lock, but I was unable to determine which direction is required. The symptoms of improper phasing are no breakout, breakout only after "coaxing" one out with an arc or sudden loss of phase lock while tuning.

My coil draws about 80W from 50V, and runs at 1.38MHz. The secondary former is 5 diameter * 8 cm tall. The entire coil with control electronics fit in the palm of my hand, hence the nickname palm-top SSTC.