Page 9 : Controlling Duty Cycle and the advantage of using Optical Switching versus Hall Switching

[Previous] [next]

原文: http://www.totallyamped.net/adams/page9.html

在 Robert Adams 向世界首次公開他的馬達的時候，同時也公開一個非常獨特但很簡單的開關裝置，它是一個嵌在圓盤上的星形金屬。這是一個機械開關裝置，後來他把它用來控制功率晶體。請參見下面的 Fig 26。

上面的 Fig 26 繪示 Adams 的原始星形開關圓盤。此金屬開關圓盤被裝到轉子的軸上，會和轉子一起旋轉。其會刷過銅片接點產生脈波。剛開始的時候，他直接用這些接點來控制驅動線圈，但是因為火花的問題，造成接觸不良以及換向器 (commutator) 損壞，後來他改用這開關來控制電晶體，再用電晶體來控制驅動線圈。電源經由 470 ohm的電阻接到電晶體的基極，這些接點只要控制小量的電流，電晶體則是控制主要的驅動電流，接點的火花會變得很微弱，換向器和接點的損壞可以降至最低。

此開關裝置的設計真的是古典的 KISS，並且運作良好，可以依個人需要調整或改變閉合角度(dwell angle) 及 duty cycle。設計 B 及 C 複製此種設計，只是改用厚度約2-4 mm的不透明 PVC 板來製作。原來在圓盤同一側的銅片接點，改用發射器(LED) 及接收器 (感光元件:- 感光電晶體或感光二極體或光敏電阻) 來代替。發送器裝在圓盤的一面，接收器則放在另一面。以此方式，當圓盤隨著轉子旋轉時，圓盤會遮住發送/接收配對之間的光線 (光子) 路徑，或讓光線通過圓盤上切割的槽狀或三角空隙。圓盤 B 完全模仿 Adams 的圓盤，只是它的duty cycle 不可能超過 50 %。 Disk C works just as effectively, though its controllable duty cycle cannot go as low as A or B (in the dimensions shown). It is just much easier to make than A or B, and provided the slots are not too wide or too long you will achieve a desirable duty cycle range from approx 10% to 50%. To allow lower duty cycles make the slots narrower, to allow higher duty cycles make them longer (or wider, or both). Remember anything above 50% is counterproductive.

你可以看到將接點向上移靠近圓盤的中心，或向下移靠近圓盤的外緣，可以改變 "on" 相對於 "off" 的時間的比例。也就是，你可以控制 duty cycle。藉著延著圓盤的圓周改變接點的位置，可以改變閉合角度 (dwell angle)。這類型的設計可以很精確的控制這兩個開關參數，並且在實驗時想要或需要控制這兩個參數的彈性時，建議可以使用這種設計。 The advantage of using Optical Switching over mechanical switching is there are no sparks and no physical fatigue on the control sensors. The opto-switching method offers a more precise and flexible control over switching parameters easily. A Hall IC is not as precise, and although you can vary dwell angle easily, you can not increase duty cycle easily, or lower duty cycle precisely!

In Fig 27 below we look at duty cycle and how it is affected by "Time Constants". Hopefully you read some of the material regarding "Time Constants" that I provided links to at the bottom of page 7 and are a little bit familiar now with the concept. See below Fig 27 for an explanation of the circuit differences.

在上面的 Fig 27，有 3 group A, B and C，每個group分別有 2 個線圈。每一組中的兩個線圈的中心之間的距離代表脈波的整個週期。從第一個脈波到下一個脈波，諸如此類。Between these two coils is the mid way point of travel of the rotor magnet/s moving according to the pulse.

Group A 代表 2個 "低阻抗 (low impedance)" 線圈，它們只有用在驅動模式。There is no re-circulating path for the CEMF via a diode, so there is no recycling of current. Looking at Fig 27 you will see that at the bottom of each group, there is a light green square current pulse labeled "Controlling Signal Pulses". This is the desired shape of your signal to the base / gate of your drive transistor / mosfet. Notice how above them is a series of pulse shapes labeled "Resultant Coil Pulses". These are the current pulse shapes actually formed in the coils according to the actual "Time Constants" of the circuit. They are the "Virtual Duty Cycle" of the motor.

Regarding Resultant Coil Pulses.

First and foremost, an Inductor Hates Current Change. It takes, what is known as a "rise time" before attaining its maximum voltage and current. 在 Group A，你可以看到方波的前緣已變形，方波後緣也有同樣的變形。看起來已經不像單一脈波的正正方方的形狀，看大約還是方形的。 When the pulse in the coil collapses, there is no regenerative current circuit available, so the magnetic field collapses readily, taking the same time it took to attain maximum. In the mean time the magnet has been propelled to the midway point and is ready to be attracted on to the next core.

在 Group B，同樣是 "低阻抗" 線圈，但有個二極體跨接在其兩端，所以當信號脈波 off 時，驅動線圈上的 CEMP 會建立 CEMF，其經由二極體的迴路流回驅動電感器。在 Group B，脈波前緣的形狀和 Group A 一樣，但是因為迴圈電流，使得其電流的後緣在時間上較為延伸。Think of the coil/s and diode/s together forming a unidirectional resonant "ringing bell which takes a while to subside". By the time the magnet reaches the middle passover point, however, the "ringing" current has subsided and the cores are no longer active, and the magnet is now in attraction mode to the next core.

在 Group C，線圈為 "高阻抗"。製作上使用較大的蕊心，而且繞更多圈，使得它們具有更高的感抗與阻抗。The resultant drive pulse shape is nothing like the signal shape. It takes almost the entire length of the signal pulse for the drive voltage and current in the inductor to reach maximum. Then with the collapsing edge of the signal pulse, the drive current recirculation via the diode causes the magnetic field to collapse VERY slowly. If the time it takes is too long, then the core will still be inducing a repeling force on the magnet as it passes the middle crossover point. You can see in Group C that the red magnet is ready to move forward into what should be an attraction zone, but the core it should be moving toward would still be partially active and will be in repelling mode. This will cause torque and energy loss.

在 Group C，你可以看出為何能夠精確的控制單一脈波的長度是如此的重要。特別是在利用循環電流的時候。在使用高阻抗線圈時更是如此。 The circuitry that allows a recirculating current, also creates a "Time Constant" which is dependant on the Resistive and Inductive values of the coils themselves, and / or loads, which are in series with the regenerative loop thats produced. This "Time Constant" has a dynamic feedback effect on the actual duty cycle, producing instead a "Resultant Virtual Duty Cycle". As always, theres "No Punch without Judy"!. But with careful design, you can exploit this extended pulse time, as long as your resultant duty cycle pulse is no more than 50 % of the real total cycle. See Fig 28 below for a simplified layout of a photo-controlled switching circuit.

In Fig 28 above, the pulsing signal is produced by a PVC disk with slots cut into it, rotating between a LED (Light Emiting Diode) and a Photo Dependent Resistor. When the LED shines through the slot of the PVC disk, the resistance of the Photo Dependent Resistor changes from 100 k-ohm to around 30 k-ohms. This will raise the bias voltage to the Darlington Pair of Transistors, and they will turn on and deliver current to the motor coil. Note the inclusion of a 3 V Voltage regulator to keep the signal supply voltage steady at all times. Variable resistor VR1 is included to give fine tuning to the bias of the Darlington Pair to ensure that the Darlington Pair turn off completely when they need to. The actual maximum value of VR1 is to some degree, dependent on the maximum and minimum resistances of the PDR.

現在我們要簡短的討論一下轉子的設計，請看下面的 Fig 29。

Fig 29 above shows a very simple beginners layout for an Adams motor. Its a typical Radial Design. It's a good design to start with as you learn the basics of operation. But it has a number of limitations, and can pose safety problems. In the above diagram, if the supply was 24-48 Volts, and the drive coils were low impedance coils (1-2 ohms), then the rotor would easily reach speeds between 5,000 to 10,000 RPM. At these speeds, the centrifigal force on the magnets can be so great, that there is a great risk of the magnets flying off the rotor if they are not bound to the rotor very well. 假如磁鐵飛出來的話會是非常危險的，很容易造成嚴重的傷害。

Other limitations include: limited number of coils you can place around the rotor whilst trying to keep the whole motor within respectable size limits, and only having access to one side of the magnet. In Fig 30 below we look at a Planar Design. See below for explanation.

Fig 30 above shows the rotor layout in a different manner. It shows the magnets inserted into the face of the rotor instead of the circumference or outer edge of the rotor. The magnet length and rotor thickness should be matched, to minimize wind drag and maximize safety. The rotor can be made from plastics, resins, or wood or even aluminium (O.K. but not recommended ). High temperature epoxy resin, or a thermo-hardening plastic is best if you're confident with moulding it, and lathing it smooth afterwoods. The magnets are inset from the edge of the rotor, thus providing the magnet with a solid barrier of outer rotor material, preventing it from flying off when the rotor is spinning at high speeds.

It is a slightly more difficult design to set-up, often requiring two bearings instead of just one, and two (or more) mounting plates, but it is much safer when designing high powered, high speed rotors. It also has the advantage of allowing access to both sides of the magnets, which can increase greatly the total number of coil configurations used for a particular setup. It is also easy to have multiple rotor disks on the same axle shaft, with 2 disks interacting with single drive cores between them. Why not 3 or 4 rotor disks? It's entirely up to you and how powerful a machine you want to build.

Note * Adjustment of the distance between the magnets and the cores in a planar design can be a little more difficult than in a radial design, but can be facilitated using 4 long threaded non-magnetic stainless steel or brass rods and nuts as adjustable frame guides. Soft Iron or Galvanised threaded rods will do, but will affect the rotor by proximity to it, because they are magnetically reactive metals. Same applies to your axle. Use something non-magnetic if possible. For most experiments, it will NOT BE CRITICAL to the success or failure of the experiment..

在後面的page 的 Fig 31，介紹一個非常簡單的實驗，你可以用來 "顛覆楞次定律 (Bend Lenz's Law)"，使用一個 "open magnetic system" 發電機 (這你要自己製作)，其由一個傳統的 closed magnetic system DC 馬達來驅動。(這個 DC馬達不用自己製作) (LOL).

I hope this general information and that on preceding pages has been of some help to you, the experimenter, and has not left you more confused than enlightened! 這是和 Adams Pulsed Motors 有關的最後一個 page ........ Happy motoring pulsars, I hope you discover what you think you're looking for. Keep Experimenting!