php+840,哈罗CQ火腿社区 - QRP and DIY - IRF840居然可以用在7MHz的功放！？

一般使用6db参考值测试方法：(针对短波电台功放的测试)

使用频谱仪测量满cw功率输出时为参考电平(如100瓦电台在通过100瓦衰减器(如50db))为0dbm.

输入双音信号到电台，ssb模式，发射，在频谱仪观测两个信号电平，调节双关旋钮(如mic gain,pwr)使两个信号的电平为-6dbm，这是他们两边的imd产物与-6dbm的幅度差就是3阶互调，一般在20-40db左右，再往两边过去的就是5阶互调，看图：

与一般的双极发射管和线形专用发射管(如mrf150)相比，irf510的特性就差得多，可参看相关连接的介绍，在20瓦输出时，最差的情况已经只有20db了，如果输出40瓦，后果不堪设想，但看一下1000mp mark v的输出，200瓦时超过30db，75w class a时3x db,

所以厂家不会使用irf510类的管子来做线性功放，虽然它们很便宜，一个mrf150管子国内i报价400元，一个irf510才几元，差别太大了。

但是作为一种研究或者qrp应用，irf510还是可行的，它在10瓦时出时还是有不错的表现的。但如果用到20瓦以上，就似乎不太合适了。

相关连接：http://www.qsl.net/ik4auy/article_4.htm

网页里面有yaesu等成品机的imd特性图链接，好好理解一下原文吧！！里面有很好的解说！！ [表情]

hf qrp linear amplifier with low cost mosfets (here is a brief translation)

our main goal was to get an optimized low cost mosfets linear amplifier circuit for qrp, with around 10 - 15 watts pep or cw average power for ssb and cw, with lower higher order imd in the 1.8 to 30 mhz, for amateur radio use, to compare it with a similar circuit adapted around plastic case transistors with a 13.8 v power supply, both with variable bias capability, to improve home base station performances, where it is easily possible to get other than 13.8 v power supply for tx finals.

this mosfet unit operates at 28 v and the high bias current used requires a large heat sink and air blower (fan) always on while in tx to mantain these plastic devices well below their maximum power dissipation and temperature limits inside a soa safe operating area point of work. we used low cost mosfets devices that ended to show us good rf performances, but you could also change the project to try specifically rated rf mosfets, of course with some minor mods if eventually required, for even lower imd performances, with more rugged ceramic case like the mrf series, but at much higher prices, already seen in some high level commercial amateur radio transceivers (1). in general linearity at rf is seen as a) uniform gain - frequency relation; b) low intermodulation imd as monitored with a two tone test (at a small frequency separation around 1 to 2 khz or if not otherwise possible at a larger separation) with a well regulated power supply for the unit. this is of course an approximate "best case" ssb performance test any way since in ssb voice is a whole band of audio frequencies not only two (agree with w8ji tom advices, see topband reflector); c) attenuation of harmonic frequencies (the higher level is in the odd multiple of fundamental output frequency but the closer one is the second and it must be adequately filtered with output low pass filters to get values below - 40 db down as per fcc or other national laws. of course in b) and c) the higher the attenuation the better is (db with minus sign).

we and other amateurs (3) wanted to test these low cost fast switching mosfets that you can find very easily.

some advantages that mosfets have over transistors (4): higher gate impedence to dc make easier to implement a bias circuit without dc current power transistors in the bias circuit and mosfets behaves better, at raising temperature of work, with a self protection action because there is a lowering in the drain-source current; an higher output impedence together with an higher dc power supply make easier to project and built the output rf transformer; the irf510 is easy to get and inexpensive (5). in the negative side mosfets are sensitive to electrostatic discharges and it is commonly guessed that they could self-oscillate (more than transistors) but in all tests we have performed around this circuit and our pcb layout it did not happen.

of course with a 28 v power supply it is not intended for portable operation. we think that the worst of the previously negative aspects is a lower efficiency: to get a better linearity (lower imd) we need to set the point of work near class a so we have higher quiescent currents while in tx . the high (stable) gain achieved with 2 stages (2x irf510 + 2x irf510) with a lower higher imd order performances (that means lower splatters if you drive another linear amplifier to get an higher output level) and these devices are so easy to get and low priced that it is interesting to give them a try