ESG-DP冷却时工作正常但温度升高就会失去PLL锁定

我一直在研究E4436B ESG-DP，它在冷却时工作正常，但一旦温度升高，就会在整个GHz频率（1 GHz，2 GHz和3 GHz）失去PLL锁定。
PLL解锁范围仅比整个GHz频率低几MHz，即低于约980MHz，1.98GHz，2.98GHz，输出再次被锁定和清洁。
即使是1GHz和2GHz以上的最小频移（例如，10毫赫兹）也会导致立即锁定PLL并具有干净的输出。
我认为这是合成器的温度引起的问题，因为用冷冻喷雾冷却侧卡笼解决了问题，虽然是暂时的。进一步的调查，在我看来好像两个12伏冷却风扇不是
实际上，当我测量风扇两端的电压时，它只有7.5伏特，如果我断开一个或两个风扇（两个断开时都是12伏特），它会增加。
每个风扇连接器的12伏侧为12伏WRT机箱接地，但风扇连接器的负极端子从0伏特上升到大约4.5伏特，两个风扇都连接在一起。似乎12伏风扇应该得到
12伏，但这实际上可能是正常的行为吗？
是否存在某种基于温度的风扇电压调节？
如果没有，是否有一个简单的解决方案？谢谢，斯坦



     以下为原文

  I've been working on an E4436B ESG-DP that works fine when cold but loses PLL lock at the whole GHz frequencies (1 GHz, 2 GHz, and 3 GHz) once it's wARMed up a bit. The PLL unlock range extends only a few MHz lower than the whole GHz frequency, i.e., below about 980 MHz, 1.98 GHz, 2.98 GHz the output is again locked and clean. Even the smallest frequency shift (e.g., 10 millihertz) above 1 GHz and 2 GHz causes immediate PLL lock with clean output. I thought this was a temperature-induced problem with the synthesizer, since cooling off the side card cage with freeze spray got rid of the problem, albeit temporarily.

On further investigation, it seemed to me as though the two 12 volt cooling fans were not moving a lot of air, and, in fact, when I measured the voltage across the fans, it was only 7.5 volts, increasing if I disconnected one or both fans (12 volts with both disconnected). The 12 volt side of each fan connector is at 12 volts WRT chassis ground, but the negative terminal of the fan connector floats up from 0 volts to around 4.5 volts with both fans connected.

It seems as though the 12 volt fans ought to be getting 12 volts, but could this in fact be normal behavior? Is there some sort of temperature-based regulation of the fan voltage? If not, is there a simple fix?

Thanks,
Stan