选择8510C的放大器要注意什么？

> {quote：title = samtsitrin写道：} {quote}>大家好。
我是一个完整的新手！
>我们使用旧的8510C通过光子晶体发送微波，当我们改用同轴天线作为接收器时，它们的信号非常低。
此VNA输出的最大功率为25dbm（与316mW相同）我不知道8510C。
真的可以输出25 dBm吗？
这在微波频率上是相当大的功率。
但我认为8510系统是由几个不同的组件组成的，所以我想这很有可能。
它不仅仅是大多数现代VNA所能提供的。
在不知道你在做什么的情况下，很难知道你想要建议什么。
您检测到的功率是否太低而无法测量VNA？
如果您发布了系统的框图，以及对您正在做什么，您希望衡量的内容等的某种描述，那么人们可能会有一些建议。
但是你的帖子中没有任何东西可以表明前进的方向。
>一些合作者建议放大器放大信号。
>浏览放大器，我发现例如这两个型号：> 83018A，功率为125mW，增益为27db（便宜！）>和> 83020A，输出功率为1W，增益为30db（昂贵）>所以我的问题是：
看来我们正在支付瓦特的输出，所以这是要注意的参数？
>如果我们选择更便宜的型号，这是否意味着我们的316mW最大VNA信号将被放大125mW，这就是它？
功率只是放大器的一个方面。
增益和带宽可能是功率放大器的另外两个最重要的特性。
噪声系数对放大低信号电平的放大器很重要，但对于高达1 uW的功率电平则不重要。
但是放大器还有许多其他参数，特别是它们的线性度。
如果放大器指定为125 mW，那就是最大功率。
它不增加125 mW。
有很多方法可以组合放大器，但它不是在公园散步。
如果您需要1 W来弥补所有损耗，请购买1 W放大器。
>请帮助，因为我是一名物理系学生，而不是工程师，任何智慧的话都会受到赞赏......你需要确定你需要什么样的力量。
如果你有25 dBm，这还不够，为什么还不够？
而不是看一个放大器的catalgoue，憎恶你需要什么样的规格。
也许您可以在探测器上使用前置放大器。
真的没有足够的信息让任何人帮助你。
我们至少需要一个框图和你想要实现的目标的描述。
戴夫



     以下为原文

  > {quote:title=samtsitrin wrote:}{quote}
> Hello everyone. I am a complete newbie here!
> We are using an old 8510C to send microwaves through photonic crystals and are getting a really low signal when we switch to using a coaxial antenna as receiver. The max power at output for this VNA is 25dbm (is that the same as 316mW) 

I don't know the 8510C. Can it really output as much as 25 dBm? That is quite a bit of power at microwave frequencies. But I think an 8510 system is made up from several different components, so I guess that is quite possible. It is more than most modern VNAs would provide. 

Without knowing what you are doing, it is hard to know what you want to suggest. Is the power you are detecting too low for the VNA to measure? 

If you posted a block diagram of your system, and some sort of description of what you doing, what you hope to measure etc, people might have some suggestions. But there is not really anything in your post to indicate a way forward. 

> Some collaborators suggested an amplifier to amplify the signal. 
> Browsing the amplifiers , I found for example these two models:
> 83018A which gives 125mW and a gain of 27db (cheap!)
>  and
> 83020A which gives 1W output and a gain of 30db (Expensive)
> 
> So my questions are: It seems we are paying for Watts of output so that's the parameter to pay attention to?
> If we go with the cheaper model , does that mean that our max VNA signal of 316mW will be amplified by an additional 125mW and that's it?

Power is only one aspect of an amplifier. Gain and bandwidth are probably the two other most important properties for power amplifiers. Noise figure is important for amplifiers amplifying low signal levels, but not a power levels as high as even 1 uW. But there are many other parameters of amplifiers, specifically how linear they are. 

If an amplifier is specified as 125 mW, that is the maximum power. It does not add 125 mW. There are ways to combine amplifiers, but it is not a walk in the park. If you need 1 W to overall all your losses, buy a 1 W amplifier. 

> Please help, since I am a physics student and not an engineer and any words of wisdom are appreciated...

You need to determine what power you need. If you have 25 dBm, and that is not enough, why is it not enough? Rather than look at a catalgoue of ampliiers, deteremine what specification you need. Perhaps you can use a pre-amplifier on the detector. 

There really is insufficient information for anyone to help you much. We would need at least a block diagram and a description of what you are trying to achieve. 

Dave

戴夫 - 谢谢你的回应。
我们的实验如下：VNA连接到S参数测试装置。
我们通过在端口1使用波导喇叭通过光子晶体结构发送微波（7Ghz-15Ghz）。我们用来收集S21传输，另一个喇叭位于我们晶体的另一侧并连接到端口2并获得了良好的数据。
现在实验正在改变，我们想要使用同轴偶极天线从顶部扫描结构并收集结构不同部分的透射S21数据，以绘制晶体内部的电场强度。
在一些初步试验中，我们发现信号非常低，另一个使用类似设置的研究小组提到他们使用1W放大器来提升信号输出端口1。
8510C的25dbm值是我在显示器上读取的输出功率...我猜这还不够，因为天线收集辐射的效率远低于波导喇叭？
根据你所写的，125mW输出放大器似乎根本没用，因为它的输出比我们目前的VNA本身要少？
谢谢你的建议。



     以下为原文

  Dave-
thanks for responding. Our experiment is the following:
the VNA is connected to the S-parameter test set. We are sending microwaves (7Ghz-15Ghz) through a photonic crystal structure by using a waveguide horn at port 1. We used to collect S21 Transmission with another horn positioned on the other side of our crystal and connected to port 2 and got good data.

Now the experiment is changing, where we want to use a coaxial dipole antenna to scan the structure from the top and collect transmission S21 data in different parts of the structure to map the Electric field intensity inside the crystal. On some preliminary trials we found that the signal is quite low and another research group that uses a similar set-up mentioned they used a 1W amplifier to boost the signal out of port1.

The 25dbm value for the 8510C is what I read on the display for output power...I am guessing it's not enough because the antenna is much less efficient in collecting the radiation than the waveguide horn?

From what you wrote it seems the 125mW output amplifier will not be useful at all, because it's less output than our current VNA itself?

Thanks for your advice.

> {quote：title = samtsitrin写道：} {quote}> Dave->感谢您的回复。
我们的实验如下：> VNA连接到S参数测试装置。
我们通过在端口1使用波导喇叭通过光子晶体结构发送微波（7Ghz-15Ghz）。我们用来收集S21传输，另一个喇叭位于我们晶体的另一侧并连接到端口2并获得了良好的数据。
>>现在实验正在改变，我们想要使用同轴偶极天线从顶部扫描结构并收集结构不同部分的透射S21数据，以映射晶体内部的电场强度。
在一些初步试验中，我们发现信号非常低，另一个使用类似设置的研究小组提到他们使用1W放大器来提升信号输出端口1。
这个水晶有多大？
我不禁感觉你的空间分辨率与你的囊肿大小相比会很差，这让我想知道你会通过扫描看到什么样的优势。
你使用什么尺寸的喇叭天线？
根据尺寸，我们可以计算出增益本来就是多少，因此使用偶极天线可以大致减少多少信号。
> 8510C的25dbm值是我在显示器上读取的输出功率...我猜这还不够，因为天线收集辐射的效率远低于波导喇叭？
是的，天线的增益会低得多。
对于大多数喇叭，喇叭的增益在10到25 dBi之间。
我不认为在喇叭上超过25 dBi是非常实用的。
但是一个人会大约2 dBi，所以会有相当大的损失，特别是如果你用两个偶极子替换了两个角。
如果每一侧最初有25 dBi增益，你可能会损失大约46 dB，现在减少到2 dBi。
>根据你所写的，125mW输出放大器似乎根本没用，因为它的输出比我们目前的VNA本身要少？
是的，就是这样。
使用125 mW放大器是最不明智的。
显然增加功率将提高您的S / N，但您是否尝试过其他技术，如降低IF带宽和/或平均值？
1 W放大器只能为您带来5 dB的改善，但是您可以通过将IF带宽降低10倍来改善问题。这样做的代价是扫描速度较慢。
即使你的喇叭只有10 dBi的增益，你也可能会失去16 dB变成一对偶极子，因为现在每个天线的增益都低8 dB。
在这种情况下，增加5 dB的功率将有所帮助，但它不会恢复到以前的位置。
我想知道你是否可以在端口2（接收器）上使用前置放大器，而不是端口1（源）？
如果您使用的是低噪声放大器而不是功率放大器，那么您可以以更低的成本获得显着的收益。
>感谢您的建议。
我必须补充一点，我从未使用过8510C，但我确信它有能力降低IF带宽，我也希望能够进行平均。
您能否提供其他研究小组正在做的事情的参考？
我将看看论文，看看我是否还有其他想法。
DaveEdited：drkirkby于2013年7月11日上午12:27



     以下为原文

  > {quote:title=samtsitrin wrote:}{quote}
> Dave-
> thanks for responding. Our experiment is the following:
> the VNA is connected to the S-parameter test set. We are sending microwaves (7Ghz-15Ghz) through a photonic crystal structure by using a waveguide horn at port 1. We used to collect S21 Transmission with another horn positioned on the other side of our crystal and connected to port 2 and got good data.
> 
> Now the experiment is changing, where we want to use a coaxial dipole antenna to scan the structure from the top and collect transmission S21 data in different parts of the structure to map the Electric field intensity inside the crystal. On some preliminary trials we found that the signal is quite low and another research group that uses a similar set-up mentioned they used a 1W amplifier to boost the signal out of port1.

How big is this crystal? I can't help feeling your spatial resolution is going to be quite poor in comparision to the size of your cystal, which makes me wonder what advantage you will see by scanning. 

What size horn antenna(s) were you using? From the size we can work out what the gain would have been and so roughly how much the signal would be weaker with dipole antennas. 

> The 25dbm value for the 8510C is what I read on the display for output power...I am guessing it's not enough because the antenna is much less efficient in collecting the radiation than the waveguide horn?

Yes, the gain of the antenna would be much lower. A horn would have a gain of between 10 and 25 dBi for most horns. I don't think its very practical to go beyond 25 dBi on a horn. But a diople would be about 2 dBi, so there would be quite a loss, especially if you have replaced two horns with two dipoles. You could have lost around 46 dB, if each side originally had 25 dBi gain, and is now reduced to 2 dBi. 

> From what you wrote it seems the 125mW output amplifier will not be useful at all, because it's less output than our current VNA itself?

Yes, that is so. You would be most unwise to use a 125 mW amplifier. 

Clearly increasing the power will improve your S/N, but have you tried other techniques like reducing the IF bandwidth and/or averaging? A 1 W amplifier is only going to gain you a 5 dB improvement, but and you can improve matters more than that by reducing your IF bandwidth by a factor of 10. The penalty for doing that is a slower sweep speed. 

Even if your horns were only 10 dBi gain, you could have lost 16 dB changing to a pair of dipoles, since each antenna now has 8 dB less gain. In that case, getting 5 dB more power is going to help, but it wont recover to where you were before. 

I wonder if you could use a pre-amp on port 2 (receiver), rather than port 1 (source)? You could get a significant gain for far less money if you used a low-noise amplifier, rather than a power amplfier. 

> Thanks for your advice.

I must add I have never used an 8510C, but I'm sure it has the ability to reduce the IF bandwidth, and I expect averaging too. 

Can you provide a reference to what the other research group were doing? I'll take a look at the paper and see if I have any other ideas. 

Dave

Edited by: drkirkby on Jul 11, 2013 12:27 AM

戴夫 - 晶体直径约10-15厘米。
根据频率，我们应该能够确定内部的电场模式。
它是一个类似于本文所述的设置：http：//www.lci.kent.edu/muri/nim_papers/negative_index/2006Smithmapping.pdf最新的喇叭天线是17db MA86551 http://www.advancedreceiver.com
/page34.html我们仍计划使用1个喇叭天线和1个偶极扫描天线。
+显然增加功率将提高您的S / N，但您是否尝试过其他技术，如降低IF带宽和/或平均值？
一个1 W的放大器只能让你获得5 dB的改善，但你可以通过将你的IF带宽减少10倍来改善问题。这样做的代价是扫描速度较慢。+你能不能请
告诉我你对减少IF带宽的意思？
你的意思是我们在频率扫描中采取的点数？
我们已经使用相当高的平均值。
+我想知道你是否可以在端口2（接收器）上使用前置放大器，而不是端口1（源）？
如果使用低噪声放大器而不是功率放大器，你可以以更低的成本获得显着的增益。+因此，端口2上的前置放大器可能比端口1上的1W放大器更好吗？
你有任何我应该抬头的模特吗？
我们是否必须以某种方式过滤噪音？
+你能提供其他研究小组的工作参考吗？
我会看一下这篇论文，看看我是否还有其他的想法。+不幸的是，其他小组的论文并没有多谈他们的设置，而是在个人交流中，他们告诉我他们使用了放大器。
谢谢你的帮助。



     以下为原文

  Dave-
The crystals are about 10-15cm in diameter. Depending on the frequency we should be able to determine E-field pattern inside. It is a similar set up as described in this paper:
http://www.lci.kent.edu/muri/nim_papers/negative_index/2006Smithmapping.pdf

The most recent horn antennas were 17db MA86551 http://www.advancedreceiver.com/page34.html
and we still plan on using 1 Horn antenna and 1 dipole scanning antenna. 

+Clearly increasing the power will improve your S/N, but have you tried other techniques like reducing the IF bandwidth and/or averaging? A 1 W amplifier is only going to gain you a 5 dB improvement, but and you can improve matters more than that by reducing your IF bandwidth by a factor of 10. The penalty for doing that is a slower sweep speed.+ 

Could you please tell me what you mean about reducing the IF bandwidth? Do you mean the number of points we take in the frequency sweep? We do use pretty high averaging already.

+I wonder if you could use a pre-amp on port 2 (receiver), rather than port 1 (source)? You could get a significant gain for far less money if you used a low-noise amplifier, rather than a power amplfier.+ 

So a pre-amp on port 2, might work better than a 1W amp on port 1? Do you have any models I should look up? Would we have to filter the noise somehow?

+Can you provide a reference to what the other research group were doing? I'll take a look at the paper and see if I have any other ideas.+ 

Unfortunately the other groups papers don't talk much about their set-up, it's in personal communication that they told me they used the amp.

Thanks for your help.

> {quote:title=samtsitrin wrote:}{quote}
> Dave-
> The crystals are about 10-15cm in diameter. Depending on the frequency we should be able to determine E-field pattern inside. It is a similar set up as described in this paper:
> http://www.lci.kent.edu/muri/nim_papers/negative_index/2006Smithmapping.pdf

I had a quick look at that paper, and was amuzed how they uses feet, mm and cm in the same paper. I guess being from the UK, we  never use imperial in any scientific paper. Anyway, that aside, *IF* I'm not mistaken, there is a pretty fundamental error in their reasoning. The authors seem to think they have achieved a plane wave from their coax to waveguide adapter, but I very much doubt that is the case.  For 10.5 GHz, the wavelength is 28.6 mm. WR90/WG16 waveguide has an internal width of 22.86 mm. To get a planar wave, generally defined in RF terms to have a phase varition of no more than 22.5 degrees, you have to be in the far-field of an antenna. That occurs at a distance of at least 2*D^2/lambda where D is the largest dimension of an antenna, and lambda is the wavelength. So for a wavelgnth of 28.6 mm, with an antenna dimension of 22.86, you need to be at a distance of at least 57 mm. But as far as I can tell, the spacing between the source and detector is only 11 mm. Perhaps I have mis-understood their work - I did not read it in any depth, but there will not be a TEM wave 11 mm away from a coax to waveguide transition. It would be hard to know what the field is then. I can't help but feel this is a near-field scanning technique, with all the computational problems that brings. 

It's also unclear to me how they calibrated the VNA. The paper has a distinct lack of information about how the VNA was set up. 

> The most recent horn antennas were 17db MA86551 http://www.advancedreceiver.com/page34.html
> and we still plan on using 1 Horn antenna and 1 dipole scanning antenna. 

So for each dipole you horn you change for a horn, you will be losing around 15 dB *assuming far-field conditions*, which is not the case here. 

> Could you please tell me what you mean about reducing the IF bandwidth? Do you mean the number of points we take in the frequency sweep? We do use pretty high averaging already.

Many radio frequency instruments, including something as simple as a domestic AM/FM radio, will convert the incoming signal to one or more fixed "intermediate frequencies" (IF). Narrow band filtering is performed at the IF. The narrower you make the IF filter bandwidth, the lower the noise will be - think of noise power being K*T*B, although it will be higher due to the noise figure of the instrument. But the basic point is the noise power will increase with bandwidth. So to reduce the noise, you can reduce the IF bandwidth. In the case of my HP 8720D VNA, the default IF bandwidth for measurements of the sort of frequency ranges used in that paper, would probably be 3 kHz, but I have the option of reducing that to as low as 10 Hz, so reducing the noise by a factor of 3000/10=300 or 24 dB. I've no idea what your 8510C is set to, or what it can be changed to, but I suspect you could get a significant reduction in noise by reducing the IF bandwidth. Given the whole lot is computer controlled, and in a reseach environment, measurement time might be less of an issue than it would be for purely manual measurements or in a production environment, where mesurrement speed is critical. 

(As a matter of interest, is this a Ph.D project, are you a post-doc working on it? I assume this is a university research project, but perhaps not.)

The downside of reducing the IF bandwidth is the time-constant of the filter is increased, and so the sweep must be reduced. To a first order approximation, by the same ratio, so if you reduce the bandwith by a factor of 300, the time to acquire the data would go up by a factor of 300. But even your expensive 1 W amplifier is only going to give you about 5 dB improvement in S/N, whereas it is possilble to achieve better than that by reducing the IF bandwidth a modest amount. 

> So a pre-amp on port 2, might work better than a 1W amp on port 1? Do you have any models I should look up? Would we have to filter the noise somehow?

It would be cheaper. One could get much higher gain, although without know much about an 8510C, you should seek the advice of people that know them. You would also need to state your configuration, as I believe the 8510 series is built from sepparate modules, and how it works would depend very much on what you have. 

But personally I would look to reduce the IF bandwidth first. Find the Agilent application note "Understanding and Improving Network Analyzer Dynamic Range"

> Unfortunately the other groups papers don't talk much about their set-up, it's in personal communication that they told me they used the amp.

It's a shame there is no details about the measurement system. Yet more papers where it would be improssible to reproduce results. 

> Thanks for your help.

No problem. 

Dave

嗨戴夫 - 在论文中他们形成了距离光源30厘米的“波导”长度，因此他们确保在接收器处得到平面波。
为了回答你的另一个问题，这是一个大学研究项目，如果事情锻炼，希望将成为我硕士论文的一部分。
我查看了IF带宽调整，但是它不是8510C的选项，IF带宽是预设的。
尽管平均值是由我们控制的，但我们确实需要将信号提升到噪声之上。
我想我会转发我的问题，并特别询问你对8510C端口2上的前置放大器的想法，因为1W放大器只有5db增益才真的很贵。
谢谢，山姆



     以下为原文

  Hi Dave -
in the paper they formed a "waveguide" length of 30cm from the source so they made sure they got plane waves at the receiver. To answer your other question, this is a university research project, and hopefully will be a part of my Master's thesis if things workout. 

I looked into the IF bandwidth adjustment, but alas it's not an option on the 8510C, the IF bandwidth is preset. Averaging though is controlled by us, but we really need to boost the signal above the noise. 

I think I'll repost my question and ask specifically about your idea of a pre-amp on port 2 with 8510C, because the 1W amp is really expensive for only a 5db gain.

Thanks,
Sam

> {quote：title = samtsitrin写道：} {quote}>嗨Dave - >在论文中他们形成了距离光源30厘米的“波导”长度，因此他们确保在接收器处得到平面波。
我没有看到，但如果波导距探测器太近，那么波浪就不会是平坦的。
>为了回答你的另一个问题，这是一个大学研究项目，如果事情锻炼，希望将成为我硕士论文的一部分。
听起来不错。
我在计算机模拟上完成了我的硕士学位 - 没有你的那么有趣。
我的博士学位更有趣。
>我调查了IF带宽调整，但是它不是8510C的选项，IF带宽是预设的。
你是100％肯定的吗？
我觉得很难相信。
>平均虽然是由我们控制的，但我们确实需要将信号提升到噪声之上。
为什么？
什么阻止你增加平均值？
在计算机控制的大学研究项目中，测量时间不太可能成为问题。
>我想我会转发我的问题，并特别询问你对8510C端口2上的前置放大器的想法，因为1W放大器只有5db增益才真的很贵。
是的好主意。
我认为棘手的一点是如何使用前置放大器进行校准。
但我确信它可以做到，但我个人不知道该怎么做。
我认为完整的2端口校准可能是不可能的，因为前置放大器的处理能力远低于VNA中的源。
>谢谢，> Sam Dave



     以下为原文

  > {quote:title=samtsitrin wrote:}{quote}
> Hi Dave -
> in the paper they formed a "waveguide" length of 30cm from the source so they made sure they got plane waves at the receiver. 

I did not see that, but if the waveguide is spaced too close to the detector, the waves will not be plain. 

> To answer your other question, this is a university research project, and hopefully will be a part of my Master's thesis if things workout. 


Sounds good. I did my MSc on a computer simulation - less interesting than yours. My Ph.D was more interesting. 


> I looked into the IF bandwidth adjustment, but alas it's not an option on the 8510C, the IF bandwidth is preset. 

Are you 100% sure of that? I find that hard to believe. 

> Averaging though is controlled by us, but we really need to boost the signal above the noise. 

Why? What stops you increasing the averaging? Measurement time is not likely to be a problem in a computer controlled university research project. 

> I think I'll repost my question and ask specifically about your idea of a pre-amp on port 2 with 8510C, because the 1W amp is really expensive for only a 5db gain.

Yes, good idea. 

I think the tricky bit would be how to get a calibration with a pre-amp. But I'm sure it can be done, but I don't personally know how to do it. I think a full 2-port cal might be impossible, as the pre-amp would have far less handling power than the source in the VNA. 

> Thanks,
> Sam

Dave

在8510上：功率是源功率设置，但源和测试端口之间的损耗约为12-15 dB。
而且，源功率设置只是一个设置;
你必须查看源，看看源是否能够创建25 dBm;
大多数不是。
我认为sw工程师只是在UI中放置了一个非常大的数字（当时），以确保您可以询问任何可用源（当时）的最大源功率。
如果您正在进行简单的响应校准或无校准，则在接收天线和8510的端口2之间使用LNA（低噪声放大器）是非常合理的。
将放大器尽可能靠近天线放置。
我猜测源的最大输出大约是+20 dBm，并且在丢失的情况下，天线的输出大约为+10 dBm，因此使用125 mW放大器（大约+20 dBm）可以提供大约10 dB的改善。
但是，在这种情况下，除非进行参考校准，否则必须非常小心，不要将放大器驱动到压缩状态，否则测量结果不会很好。
放大器中的任何漂移都会在测量中产生很大的误差。
我认为最好使用接收器放大器。



     以下为原文

  On the 8510: the power is the source power setting, but there is about 12-15 dB loss between the source and the test port.  And, the source power setting is just that, a setting; you have to look at the source to see if the source is capable of creating 25 dBm; most are not. I think that the sw engineers just put a very big number (for the time) in the UI to ensure that you can ask for the max source power of any source that is available (at the time).

If you are doing simple response calibrations, or no calibrations, using an LNA (low noise amplifier) between the receive antenna and port 2 of the 8510 is very reasonable. Put the amplifier as close to the antenna as possible.  

I'm guessing the source max output is about +20 dBm, and with loss, you are getting about +10 dBm out to the antenna, so using a 125 mW amplifier (about +20 dBm) will give about 10 dB improvement.  BUT, in this case you have to be very careful not to drive the amplifier into compression or you measurements will not be good, unless you do a reference calibration.  And any drift in the amplifer will be big error in your measuerments.  Best to use a receiver amplifier I think.

感谢您的投入。
所以你说接收器天线和VNA之间的低噪声放大器（这与前置放大器Dave建议的相同吗？）可以帮助增强噪声信号吗？
您建议使用8510的任何LNA型号？



     以下为原文

  Thanks for the input.
So you're saying a Low Noise Amp (is this the same as the preamp Dave was suggesting?) between the receiver antenna and the VNA can help boost the signal over the noise? Any LNA models you suggest to try with 8510?

> {quote：title = samtsitrin写道：} {quote}>感谢您的输入。
>所以你说接收器天线和VNA之间的低噪声放大器（这与前置放大器Dave建议的相同吗？）可以帮助增强噪声信号吗？
您建议使用8510的任何LNA型号？
乔尔和我说同样的话。
前置放大器必须是低噪声才能在此应用中使用，因此LNA ==前置放大器。
我没有仔细查看它们，但是在Minicircuits网站上快速浏览显示ZX60-183 +和170美元。
噪声系数略低于7 dB，覆盖6至18 GHz。
我会尝试找一些噪音较低的东西，但我认为这比你裸露的8510要好一些。
基本上找到尽可能低的东西。
你真的需要一个放大器在这里无条件稳定，因为通过同轴到波导的过渡将它连接到波导将在低于约6.5GHz的波导被切断的情况下呈现非常差的回波损耗。
在这种情况下，有些放大器会变得不稳定，所以我想找到一个无条件稳定的放大器。
看看你是否能找到噪音较低的数字，但那将是一个起点。
戴夫



     以下为原文

  > {quote:title=samtsitrin wrote:}{quote}
> Thanks for the input.
> So you're saying a Low Noise Amp (is this the same as the preamp Dave was suggesting?) between the receiver antenna and the VNA can help boost the signal over the noise? Any LNA models you suggest to try with 8510?

Joel is saying the same thing as me. A pre-amplifier would have to be low-noise to be any use in this application, so LNA==pre-amp. 

I've not looked in detail for them, but a quick look on the Minicircuits site shows an ZX60-183+ and $170. That has a little under 7 dB noise figure and covers 6 to 18 GHz. I would try to find something with a lower noise figure, but that would be an improvement over your bare 8510 I think. Basically find something as low as possible. You really need an amplifier to be unconditionally stable here, as connecting it to waveguide via a coax to waveguide transition will present a very poor return loss below about 6.5 GHz where the waveguide is cut off. Some amps would go unstable in such a case, so I'd want to find an unconditionally stable amp. See if you can find something of lower noise figure, but that would be a starting point. 

Dave

这对我们来说可能是对的。
因为我们感兴趣的带宽是7-15Ghz。
还有一个问题，可以解释一下声明“7db噪声系数”，我解释说噪声最多可以放大7db而不是理想的接收器，而LNA的总增益为24db，因此信号将被放大更多
比噪音？
我在这里走在正确的轨道上吗？
对不起，如果这些是愚蠢的问题
我只想了解LNA如何区分信号和噪声。
噪音主要来自我们系统的微波元件，这是LNA最小化的吗？
再次感谢你的帮助。



     以下为原文

  This may be the right thing for us. Because our bandwidth of interest is 7-15Ghz.
One more question, could please explain the statement "a little under 7db noise figure" I interpret this that noise will be amplified at most 7db over what an ideal receiver would while the LNA has an overall gain of 24db so the signal will be amplified more than the noise?

Am I on the right track here? Sorry if these are dumb questions. I am just trying to understand how the LNA will distinguish between the signal and the noise. Is the noise mostly from the microwave components of our system and that's what the LNA minimizes?
Thanks again for your help.

> {quote：title = samtsitrin写道：} {quote}>这对我们来说可能是对的。
因为我们感兴趣的带宽是7-15Ghz。
这就是为什么我选择那个，虽然我不太难找到你的放大器。
我决定查看Minicircuits网站，因为他们的产品价格合理。
这是我看到的覆盖X波段的任何放大器的最低噪声系数。
（8.2至12.4 GHz）。
>还有一个问题，请解释一下“7db噪声系统下的声音”这句话我解释说，噪声最多会被理想接收器放大7db，而LNA的总增益为24db，因此信号将被放大
噪音多吗？
>>我在这里走在正确的轨道上？
对不起，如果这些是愚蠢的问题
我只想了解LNA如何区分信号和噪声。
噪音主要来自我们系统的微波元件，这是LNA最小化的吗？
>再次感谢您的帮助。
不幸的是，你没有走上正轨。
我给出的噪声系数是数据表中的值。
它们将由制造商测量。
事实上，一些VNA可以测量噪声系数，但不能测量8510.放大器输入端的信号会有一些信噪比。
放大器输出端的信号将具有*较低的*信噪比。
因此任何实用的放大器都会降低信噪比。
较低噪声系数的放大器会降低信噪比，使其低于较高噪声系数的放大器。
您可能认为放大器是您需要的最后一件事，因为我已经声明它会降低S / N比。
但是，由于放大器到达VNA的信号会更高，因此它将大于网络分析仪中的噪声。
因此，放大器的目的是将信号增加到显着高于VNA中噪声电平的水平。
噪声系数定义为输入端的信噪比除以输出端的信噪比。
它通常以dB表示。
低噪声放大器的噪声系数小于0.1 dB，但这些放大器往往是非常窄的放大器。
我不知道有什么低的覆盖你想要的频率范围这么低的噪音数字。
我不知道这个频率范围内的最新技术水平，或者具有较低噪声系数的设备是否可以商业化。
当您将放大器与VNA级联时，整体噪声系数将略高于放大器。
如果放大器具有无限增益，则级联噪声系数将是放大器的噪声系数。
但是没有实用的放大器有无限的增益，这真的是一件好事！
但这确实意味着整体噪声系数将比放大器的噪声系数差。
我认为我找到的放大器会改善你的情况，但可能会有更好的选择。
戴夫



     以下为原文

  > {quote:title=samtsitrin wrote:}{quote}
> This may be the right thing for us. Because our bandwidth of interest is 7-15Ghz.

That is why I picked that one, though I did not look hard to find you an amplifier. I deceided to look on the Minicircuits web site, as their products tend to be reasonably priced. That was the lowest noise figure of any amplifier I see that covered X-band.  (8.2 to 12.4 GHz). 

> One more question, could please explain the statement "a little under 7db noise figure" I interpret this that noise will be amplified at most 7db over what an ideal receiver would while the LNA has an overall gain of 24db so the signal will be amplified more than the noise?
> 
> Am I on the right track here? Sorry if these are dumb questions. I am just trying to understand how the LNA will distinguish between the signal and the noise. Is the noise mostly from the microwave components of our system and that's what the LNA minimizes?
> Thanks again for your help.

Unfortunatley you are not on the right track. 

The noise figure I gave was the values on the data sheet. They would be measured by the manufacturer. In fact, some VNA can measure noise figure, but not the 8510. 

The signal at the input of the amplifier will have some signal to noise ratio. The signal at the output of the amplifier will have a *lower* signal to noise ratio. So any practical amplifier will degrade the signal to noise ratio. A lower noise figure amplifier will degrade the signal to noise ratio less than a higher noise figure amplifier. 

You might think an amplifier is the last thing you need, given I've already stated it will degrade the S/N ratio. However, since the signal reaching the VNA will be higher with the amplifier, it will be larger than the noise in the network analyzer. So the purpose of the amplifier is to increase the signal to a level which is significantly above the noise level in the VNA. 

The noise figure is defined as the signal to noise ratio at the input divided by the signal to noise ratio at the output. It is normally expressed in dB. A low noise amplifier can have a noise figure of less than 0.1 dB, but these tend to be very narrow band amplifiers. I'm not aware of anything that low which will cover the frequency range you want with such low noise figures. I don't know what the state of the art is over this frequency range, or what devices with a lower noise figure are commericallly available. 

When you cascade the amplifier with the VNA, the overall noise figure will be somewhat higher than that of the amplifier. If the amplifier had infinite gain, then the cascaded noise figure would be that of the noise figure of the amplifier. But no practical amplifiers have infinite gain, which is quite a good thing really! But this does mean the overall noise figure will be poorer than that of the amplifier. 

I think the amp I found for you will improve your situation, but there may well be a better choice. 

Dave