学习中。。。。。。。。。
本帖最后由 vodka 于 2009-6-10 19:46 编辑
50-Ω Helix Feed
=============
Joe Cadwallader, K6ZMW, presented this feed method in June 1981 QST. Terminate the helix in an N connector mounted on the ground screen at the periphery of the helix. See Fig 17. Connect the helix conductor to the N connector as close to the ground screen as possible (Fig 18). Then adjust the first quarter turn of the helix to a close spacing from the reflector.
This modification goes a long way toward curing a deficiency of the helix—the 140-Ω nominal feed-point impedance. The traditional λ/4 matching section has proved difficult to fabricate and maintain. But if the helix is fed at the periphery, the first quarter turn of the helix conductor (leaving the N connector) acts much like a transmission line—a single conductor over a perfectly conducting ground plane. The impedance of such a transmission line is:
Z0=138 log (4h/d)
where:
Z0 = line impedance in ohms
h = height of the center of the conductor above the ground plane
d = conductor diameter (in the same units as h).
The impedance of the helix is 140 Ω a turn or two away from the feed point. But as the helix conductor swoops down toward the feed connector (and the ground plane), h gets smaller, so the impedance decreases. The 140-Ω nominal impedance of the helix is transformed to a lower value. For any particular conductor diameter, an optimum height can be found that will produce a feedpoint impedance equal to 50 Ω. The height should be kept very small, and the diameter should be large. Apply power to the helix and measure the SWR at the operating frequency.Adjust the height for an optimum match.
Typically, the conductor diameter may not be large enough to yield a 50-Ω match at practical (small) values of h. In this case, a strip of thin brass shim stock or flashing copper can be soldered to the first quarter turn of the helix conductor (Fig 19). This effectively increases the conductor diameter, which causes the impedance to decrease further yet. The edges of this strip can be slit every 1/2 inch or so, and the strip bent up or down (toward or away from the ground plane) to tune the line for an optimum match.
This approach yields a perfect match to nearly any coax. The usually wide bandwidth of the helix (70% for less than 2:1 SWR) will be reduced slightly (to about 40%) for the same conditions. This reduction is not enough to be of any consequence for most amateur work.
The improvements in performance, ease of assembly and adjustment are well worth the effort in making the helix more practical to build and tune.
我在4NEC2X里仿真了缩小1/4space的，晚上空的时候让小奴隶帮我优化。这样就能知道到底情况如何了。
VODKA太赞了。
果然，1/4的缩小后，阻抗马上降到20欧了，哈哈，晚上优化一下应该没问题的。
上图让大家看看
http://www.anywlan.com/data/attachment/album/200906/10/56374_12446231214hAh.jpg
都是E文的,有谁能翻译下.
都是E文的,有谁能翻译下.
wawj605 发表于 2009-6-10 17:00 http://www.anywlan.com/images/common/back.gif可能性太小了，
我都不愿意翻译，还有谁愿意？:lol
本帖最后由 vodka 于 2009-6-10 19:36 编辑
让俺翻译, 一个单词要收 10 美分,还是交给 google 翻译便宜一些:
50 Ω螺旋饲料
=============
乔 Cadwallader ， K6ZMW ，提出此供稿方法1981年6月QST 。终止螺旋在一个N连接器安装在地面上屏幕边缘的螺旋。参见图17 。螺旋导线连接到N连接器作为接近地面尽可能屏幕（图18 ） 。然后调整第一季度之交的螺旋结束间距从反射。 这一修改不用了漫长的道路走向固化A缺乏症的螺旋140 Ω名义饲料点阻抗。传统的λ / 4 匹配部分已被证明难以制造和维护。但是，如果螺旋是美联储在外围，第一季度之交的螺旋导体（离开了N连接器）的行为就像输电线路，一个单一的指挥一个理想导体接地平面。的阻抗这样的输电线路是：
Z0 = 138 log（ 4h/d）
其中：
Z0 =行阻抗欧姆
h =高度的中心导体以上接地平面
d =导体直径（在同一单位为H ） 。
阻抗螺旋140 Ω转弯或两个远离饲料点。但是，随着螺旋导体swoops下跌对饲料连接器（和地面飞机）水量会较小，所以阻抗减小。 140 Ω标称阻抗螺旋转化为较低的价值。任何特定导体直径，最佳高度可以发现，将产生一个馈阻抗等于50 Ω 。高度应保持非常小，直径应大。适用电源对螺旋和衡量水资源小组委员会在运作的高度frequency.Adjust的最佳匹配。
通常情况下，导体直径可能不足够大，收益率为50 Ω符合实际（小）值的H.在这种情况下，薄带黄铜沉股票或闪烁的铜可以焊接到今年第一季度之交的螺旋导体（图19 ） 。这实际上增加了导体直径，造成的阻抗，以进一步减少尚未。边缘的这个地带可以狭缝每1 / 2英寸左右，并带一心向上或向下（朝向或远离地面飞机） ，以调整生产线的最佳匹配。
这种做法产生一个完美的比赛几乎任何同轴电缆。通常宽带的螺旋（ 70 ％不到2:1驻波比）将减少略（大约为40 ％ ）的相同的条件。这一减少是不够的，是任何后果最业余工作。 改进性能，易于装配和调整是非常值得的努力，使螺旋更实际，建立和调整
VODKA太赞了。
sanking 发表于 2009-6-10 16:28 http://www.anywlan.com/images/common/back.gif
Sanking 兄过誉了,俺大部分是现炒现卖,况且真正做实事的是您,俺只是吹吹水
果然，1/4的缩小后，阻抗马上降到20欧了，哈哈，晚上优化一下应该没问题的。
上图让大家看看
http://www.anywlan.com/data/attachment/album/200906/10/56374_12446231214hAh.jpg
sanking 发表于 2009-6-10 16:39 http://www.anywlan.com/images/common/back.gif
你把 segments 搞多 10 倍看看.
另外, 尾端(接接头的地方)要近可能接近 Ground plane,不知道是俺的视觉误差还是怎样? 俺觉得你的模型的尾巴悬空多了一点.
看了楼上俩位的介绍很有收获，看来仿真也要一定的天线基础啊。
你把 segments 搞多 10 倍看看.
另外, 尾端(接接头的地方)要近可能接近 Ground plane,不知道是俺的视觉误差还是怎样? 俺觉得你的模型的尾巴悬空多了一点.
vodka 发表于 2009-6-10 19:54 http://www.anywlan.com/images/common/back.gifsegs增多没有用，试过了，而且segs多了计算起来太慢太慢了。
尾端的地方，这个是没有经过优化的，优化后具体怎么样现在还不清楚。