HP8753A网络分析仪输出功率有问题

你有什么（如果有的话）测试集？
你在哪里测量功率？
我相信85046A＆amp;
85047A都有一个由VNA控制的步进衰减器，虽然我希望VNA可以进行微调。
上周我购买了一个8503A S参数测试，希望能用于我的8753ES。
它只在今天到达，我还没有打开它。
但我不希望8753ES + 8503A组合具有很大的功率控制，因为8503A S参数测试装置没有85046A或85047A测试装置所具有的步进衰减器。
假设我可以让8753ES至少部分地与8503A配合使用，我可以忍受测试装置的所有其他限制，因为它是85046A的价格的10％。
我有一个20 GHz 8720D，所以只希望8753ES在50 MHz以下偶尔使用。
Dr_Joel无疑是回答你问题的最佳人选，因为我相信他在8753A中设计了源模块。



     以下为原文

  What (if any) test set do you have? Where are you measuring the power?

I believe that the 85046A & 85047A both have a step attenuator which is controlled by the VNA, although I would expect fine adjustments to be done by the VNA. 

Last week I purchased an 8503A S-parameter test that I hope to use with my 8753ES. It only arrived today and I have yet to unwrap it. But I don't expect to have much power control with the 8753ES+8503A combination, as the 8503A S-parameter test set doesn't have the step attenuator that an 85046A or 85047A test set would have.  Assuming I can get the 8753ES to at least partially work with the 8503A, I can live with all the other limitations of the test set, as it was 10% the price of an 85046A & I have a 20 GHz 8720D, so only want the 8753ES for occasional use below 50 MHz. 

Dr_Joel is undoubtedly the best person to answer your questions,  as I believe he designed the source module in the 8753A.

我有HP8546A测试装置，但我使用双向分配器进行测试。
所以我测量的功率是源的功率 - 3dB（因为分路器）。



     以下为原文

  I have HP8546A test set but I am using a two way splitter to conduct my tests. So the power that I measure is power of the source – 3dB (because of the splitter).

> {quote：title = ffranulo写道：} {quote}>我有HP8546A测试集我认为你的意思是85046A>但我使用的是双向分配器来进行我的测试。
所以我测量的功率是源的功率 - 3dB（因为分路器）。
这听起来好像输出水平低。
您可能会发现可以使用测试装置中的衰减器将其进一步降低，但是如果源功率低于应有的25 dB，您当然会失去很多动态范围。
HP / Agilent Yahoo小组是另一个寻求建议的地方。
我相信这些VNA上的源模块在二手市场上非常昂贵，尽管我自己从未真正寻找过。
我刚打开了我的8503A（500 kHz到1.3 GHz测试装置）。
我会把它介绍给我的8753ES，希望他们能成为朋友！
戴夫



     以下为原文

  > {quote:title=ffranulo wrote:}{quote}
> I have HP8546A test set 

I assume you mean 85046A

> but I am using a two way splitter to conduct my tests. So the power that I measure is power of the source – 3dB (because of the splitter).

That does sound as if the output level is low then. You might find you can drop it down further with the attenuator in the test set, but of course you are going to lose a lot of dynamic range if the source power is 25 dB below what it should be. 

The HP/Agilent Yahoo group is another place to ask for advice. I believe the source modules on those VNAs are very expensive on the used market, although I have never actually looked for one myself. 

I just unwraped my 8503A (500 kHz to 1.3 GHz test set). I will introduce it to my 8753ES and hope they can be friends!

Dave

你是对的，正确的型号是85046A。
根据我的测试，输出功率水平可能比应该低30 dB（全功率）。



     以下为原文

  You are right the correct model number is 85046A.  

According to my test the output power level is probably 30 dB lower than it should be (at full power).

这个问题很可能是功率放大器，它对ESD有点敏感，如果不使用适当的ESD控制，很容易损坏。
常见的故障模式是用户测试射频电缆。
通常，电缆测试中的第一件事是测试直流泄漏，因此在电缆上放置一个大电压以寻找泄漏。
如果没问题则将其连接到VNA进行RF测试。
良好的电缆可以长时间存储电压，来自电缆的直流脉冲会使输出状态变为静电。
该放大器稍后在8753B及更高版本中进行了重新设置，以增加更多的ESD保护，并使输出放大器更加稳定，并且兼容。
因此，如果你从一个不能以其他方式运行的单元获得一个源模块，那么很容易更换放大器，实际上它们可能是现场可替换的，但该程序从未实现过。



     以下为原文

  This problem is most likely the power amplifier, which is a little sensitive to ESD and is easily damaged if proper ESD control is not used.  A common mode of failure was for users testing RF cables. Often the first thing in the cable test is to test the DC leakage, so a large voltage is placed on the cable to look for leakage. If it is OK then it is hooked up the the VNA for RF test.  Good cables can store voltage for a long time and the DC pulse from the cable will blow the output state. 

The amplifier was redsigned later in the 8753B and beyond to add substantially more ESD protection and to chagne the output amplifier for more power, and it is compatible. So if you get a source module from a unit that is not functioning in some other way, it is easy to swap out the amplifiers, in fact they were orginally intended to be field replacable but that program was never implemented.

谢谢Dr_joel的回复。
我已经确认放大器确实不好。
我建立了自己的测试放大器，我把它放在原来的放大器模块的位置。
我这样做仅用于测试，放大器的工作范围为100到1500 MHz。
我担心源模块很难获得，所以我可以构建自己的放大器模块，适合源外壳。
因为我没有完整的原理图，所以我会很感激一些信息。
我想知道放大器的最佳增益是多少。
我在混频器输出上测得-10 dBm，这意味着我应该获得至少35 dB的增益。
那是对的吗？
此外，还有一个E3（偏置板上的参考号）引脚连接到RF放大器模块，看起来某些DC值是从放大器板采样的，用于某些反馈或测量目的。
如果可能的话，我会很感激这方面的一些信息。



     以下为原文

  Thank you Dr_joel for your reply. I have confirmed that is indeed bad amplifier. I built my own test amplifier and I put it in place of original amplifier module. I did this only for test and the amplifier worked in range from 100 to 1500 MHz. I am afraid that the source module is very difficult to get so I may build my own amplifier module that fits in the source enclosure. Because I don’t have complete schematic, I would appreciate some information. I would like to know what the optimal gain of the amplifier.  I have measured -10 dBm on the mixer output, so that means that I should have gain of the least 35 dB. Is that correct? Also, there is a E3 (ref. designator on the bias board) pin that is connected to the RF amplifier module and it looks like some DC value is sampled from the amplifier board and used for some feedback or measure purpose. If possible I would appreciate a little info about this.

回到我设计源模块时，它是第一个将多个功能组合在一个封装中的集成微电路。
很多人担心“如果一个模块出现故障，客户更换整个电源的成本会有多高”，所以我试图设计自我测试。当然，我当时非常喜欢和一些
这些想法并没有那么顺利。
E3测量来自二极管检测器的信号，该二极管检测器通过1000欧姆电阻器从主线耦合。
我的想法是查看是否有信号，这可能会说是混合器已经死了，或者放大器，例如。
但是我们决定采用“蓝色条纹”程序，用一个经过修复的源代替一个破损的源，以获得平均修复成本（约占总源成本的1/10）。
大多数故障是放大器和混频器，但偶尔会出现PCB故障，EYO（yig振荡器）故障或腔体振荡器故障。
放大器的增益为300 kHz至3 GHz，37 dB增益+ - 2 dBm，输出功率为+23.5 dBm（如果存储器可用）。
有三个直流输入，一个用于22伏，一个用于15伏，另一个用于-5伏。
我记得我们使用了一种键合采样方案来设置栅极上的负电压来控制偏置以获得最佳的谐波和功率。
如果它的输出级是熔断的，并且你可以使用一个键合器，那么你可以在输出状态下进行键合（它会短暂地死掉）并且你应该可以从其余部分获得大约+12 dBm左右的距离
放大器
输出是FET，NEC9002。
你可以谷歌数据表。



     以下为原文

  Back when I designed the source module, it was the very first integrated microcircuit that combined multiple functions in a single package. There was a lot of concern about "what if a module fails, how expensive will it be for a customer to replace the whole source" so I tried to design self-test in.  Of course I was quite younge at the time and some of the ideas didn't pan out all that well.  E3 measured a signal from a diode detector coupled off the main line through a 1000 ohm resistor. The idea was to look and see if there was signal or not, and that could say if it was the mixer that was dead, or the amplifier, for example.  But we decided to have a "blue-stripe" program where we replaced a broken source with a refurbed source for the average repair cost (about 1/10th at total source cost).  Most failures were the amplifier and the mixer, but occasionally we would get a PCB failure, an EYO (yig oscillator) failure or a Cavity Oscillator failure.  

The gain of the amplifier was 300 kHz to 3 GHz, 37 dB gain +- 2 dBm, with +23.5 dBm output power, if memory serves.  There are three DC inputs, one for 22 volts and one for 15 volts, and another for -5 volts.  As I recall we used a bond plucking scheme to set the negative voltage on the gates to control the bias to get optimum harmonics and power.

If its the output stage that is blown, and you have access to a bonder, you can just bond across the output state (it goes dead in a short ) and you should be able to get about +12 dBm or so out of the rest of the amplifier.  The output is a FET,  NEC9002.  you can google the data sheet.

谢谢Dr_joel这个答案。
因此，如果我理解正确，仪器不关心E3引脚上的信号。
放大器覆盖有陶瓷盖。
如果我找到了如何在不打破盖子的情况下打开盖子的方法，我可能会尝试修复它，或者只是像你建议的那样将跳线粘贴在放大器的最后一级。
此外，我正在考虑布置适合源组件的放大器。
感谢您的时间和帮助。



     以下为原文

  Thank you Dr_joel for this answer. So if I understand correctly, the instrument does not care about the signal on the E3 pin. 

The amplifier is covered with the ceramic lid. If I find the way how to open the lid without breaking it I may attempt to fix it or just bond a jumper over the last stage in the amplifier like you suggested. Also, I am thinking about laying out amplifier that would fit in the source assembly. 

I appreciate your time and help.

> {quote：title = ffranulo写道：} {quote}>谢谢Dr_joel的回答。
因此，如果我理解正确，仪器不关心E3引脚上的信号。
>正确，E3不用于正常操作。
在服务情况下可以在模拟总线上查看电压，但由于灵敏度非常低，因此不太有用。
>放大器覆盖有陶瓷盖。
如果我找到了如何在不打破盖子的情况下打开盖子的方法，我可能会尝试修复它，或者只是像你建议的那样将跳线粘贴在放大器的最后一级。
此外，我正在考虑布置适合源组件的放大器。
>该线路采用非导电环氧树脂连接。
我们通过在125-150℃的加热板上设置放大器插座来将其移除，然后加热环氧树脂软化后，我们使用镊子将其弹出。
它只是一个“防尘罩”，用于保持手指脱离粘合。
最后阶段的NEC 9002位于一个小型基座上，通过电路作为散热器进行环氧树脂处理，您也可以将其弹出，然后取下芯片并换上另一个芯片。
芯片与基座共晶连接（基本上焊接），因此可能更难以移除。



     以下为原文

  > {quote:title=ffranulo wrote:}{quote}
> Thank you Dr_joel for this answer. So if I understand correctly, the instrument does not care about the signal on the E3 pin. 
> 
Correct, E3 is not used in normal operation.  The voltage can be viewed on the analog bus in a service situation, but not too useful due to very low sensitivity.
> The amplifier is covered with the ceramic lid. If I find the way how to open the lid without breaking it I may attempt to fix it or just bond a jumper over the last stage in the amplifier like you suggested. Also, I am thinking about laying out amplifier that would fit in the source assembly. 
> 
The line is attached with non-conductive epoxy. We useally remove it by setting the amplifier slug on a 125-150 C hotplate and after it heats the epoxy softens and we use a tweezer to pop it off.  It is just a "dust cover" used to keep fingers off the bonds.  The NEC 9002 in the last stage sits on a little pedestal which is alos epoxied through the circuit as a heat spreader, you can pop it out as well, then take off the chip and put on another.  The chip is eutectic attached (soldered, essentially) to the pedestal, so it might be a bit more difficult to remove.

感谢Dr_Joel的详细解释。
我已经取下了盖子，现在我可以使用微电路了。
我不得不施加大量的热量（200摄氏度）才能取下盖子。
我没有太多时间详细查看放大器。
但似乎我已经损坏了输出FET。
今晚我会花一些时间确认输出FET是坏的。
如果我有这个放大器的原理图，那么对此进行故障排除要容易得多。
过去两天我对NE9002进行了一些研究，似乎很难找到晶体管。
我将尝试找到输出晶体管的替代品。
如果您有任何更换建议，请告诉我。
我确实可以使用引线接合器，所以我应该可以更换晶体管。
谢谢。



     以下为原文

  Thank you Dr_Joel for your detailed explanation. I have removed the cover, and now I have access to the microcircuit. I had to apply a lot of heat (200 deg. C) in order to remove that lid.  I did not have much time to look the amplifier in detail. But it seems that I have damaged the output FET. I will spend some time tonight to confirm that the output FET is bad. If I have a schematic diagram for this amplifier, it would be much easier to troubleshoot this. 
I did a little research about the NE9002 past two days, and it seems that is very difficult to find that transistor. I will try to find replacement for the output transistor. If you have any replacement suggestion please let me know.  I do have access to a wire bonder so I should be able to replace the transistor.  Thank you.