- 积分
- 3799
- 在线时间
- 7930 小时
- 最后登录
- 2024-4-22
- 阅读权限
- 150
- 精华
- 6
- UID
- 11979
- 帖子
- 3951
- 精华
- 6
- 经验
- 3799 点
- 金钱
- 3381 ￥
- 注册时间
- 2006-9-3
|
楼主 |
发表于 2008-12-28 14:59
|
显示全部楼层
DM的回复，先把英文贴出来再慢慢翻译～～翻译水平很差～
Hi Yu Thanks for your e-mail. You raise a number of good questions. It is certainly true that phase relationships will change as you move off axis vertically, and to a lesser extent horizontally. It is also true that there is a fundamental difference between electrical phase and acoustic phase. Electrical phase is pretty much irrelevant unless the acoustic centers of the various drivers (tweeter, mid, woofer) are exactly aligned. This is not true for any of my designs. The acoustic center of the woofer will be behind the mid, which will be behind the tweeter. But to back up one step. All of my designs use Linkwitz-Riley 2nd order or 4th order acoustic slopes. In both cases, the drivers being crossed should be precisely in acoustic phase around the crossover region on the design axis. This is not true of other crossover topologies. In order for the drivers to be in acoustic phase, the roll-offs will have to be adjusted to compensate for the differing acoustic centers of the drivers (which cause different arrival times at the listening position, which affects phase). One way to do this is to use a slightly lower crossover order for the woofer on the low pass circuit, and a slightly lower order on the mid for its low pass filter. But how do you know when you have it right? Well, if the drivers are supposed to be in phase in the crossover region, then reversing the phase will produce the deepest possible null when the drivers are in phase using the right polarity. That's a big advantage of the Linkwitz-Riley topology--you know when it's right on the design axis. But what's the right design axis? On-axis, or a little off axis horizontally? On-axis is right if the speakers will be angled in toward the listener. A little off-axis is right if the speakers are pointed straight ahead with the listener in the middle. In reality, it doesn't make much difference. If you achieve a deep reverse null on-aixs, it will still be quite deep 15-20 degrees off axis horizontally unless the drivers are highly directional (and I don't use those kinds of drivers). As for vertical off-axis response--well, you can never avoid a suck-out at the crossover region unless you're using a acoustically aligned coaxial driver or a single driver with no crossover. That's because the relative distance of the tweeter and woofer to the listener will change as you move up. This will only affect the reflected response field, and the practical effect of that on perceived sound is quite controversial. I, and most other speaker designers, really haven't come to terms with that issue. Cheers, dennis.
中文大意：
谢谢你的邮件，你提了一些很好的问题，确实，当你在垂直方向上偏轴测量的话，相位将会变化，但是当你水平移动的时候，相位的变化程度要比垂直时要小，电相位和声学相位确实是有着根本的差别，其实电相位并没有太大的作用，除非你可以把你的音箱里高中低音的声学中心精确的对齐，我的任何一个设计里面都做不到这一点，低音单元的声学中心落后于中音单元，而中音又会落后于高音，但是退一步，我所有的设计都是使用 Linkwitz-Riley2阶或者4阶分频，无论是2阶还是4阶，在设计轴向位置，单元在分频点的声相位必须要精确对齐，但是对其他的分频结构来说并不一定是这样，为了保证单元的声相位准确，滚降曲线必须要进行调整来补偿不同单元的声学中心（是不是指发声的中心位置？--lakerblue）这个不同将会影响不同频率声音到达聆听位置的时间差异，也就是会影响相位，一个解决办法就是在低音和中音的低通线路中采用较低阶的分频（知道为什么有很多箱子低通都比高通低一阶--lakerblue），怎样才能知道你做的对不对呢？如果单元在分频点附近相位准确的话，反接就会产生一个最深的谷，这是Linkwitz-Riley分频结构非常大的一个优势，但是你要知道只有在设计轴向是才会正确，但是什么是正确的设计轴向？正对或者水平稍微偏移一些？如果你听音的时候喇叭直接对着你，那正对就是准确的，如果喇叭平放，而你在中间听音，那水平偏移一点就是正确的，但是实际上这两种情况并没有太大区别，如果你正对测量会产生一个很深的谷，那么你水平偏移15到20度，反接的谷还是会很深（除非你使用指向性很强的单元），对垂直的偏轴响应来说，你永远也避免不了一个深谷，除非你使用同轴单元或者全频单元，这是因为高音和低音到聆听者的距离会发生变化，这个只会影响到反射的频响区域，而实际有效的所谓影响现在还有很大的争议，我和其他的音箱设计者还没有研究到那个地步
|
|
发表于 2008-12-28 14:58
）
