关闭后8753A响应校正丢失

> {quote：title = imett写道：} {quote}>嗨，我用它的85046a保持8753a但是4线不是原装的，我用的是短的RG223。
>网络运行良好，但响应不是很平坦，例如对于低频率的S11测量（dB），测量值非常大，但3GHz大约为-1.5dB。
没问题，校准（SOL和85032套件）响应变得平坦。
>当我关闭仪器后退时，任何时候我都要连接校准套件。
>有一种方法可以存储这种校准，如电缆补偿或校正吗？
我需要一个特定的软件o校正工具？
>我阅读了服务/用户/现场手册，但没有关于这个问题。我想知道手册没有谈论真正的仪器校准/校正，我认为这是正常的有相同的组件（指令耦合器，电缆
，连接器..）>差异很小。
>>提前谢谢。
> matteo只要这些电缆是相稳定的，就没关系了。
它们不*必须相位匹配。
查看此线程从8753A到S参数测试装置的电缆是否需要相位匹配？
现在已经很晚了，我累了，所以没有时间仔细看你写的东西。
但VNA校准程序的重点是补偿耦合器，内部电缆等的非理想行为。只要您的电缆稳定，就可以了。
我个人认为我会使用半刚性，但这只是我的看法。
如果校准和测量之间的事情发生变化，校准不能做的是正确的，例如温度变化，连接器脏污，电缆因弯曲和温度引起的相变等。但是电缆损耗，只要它保持固定
并不过分，不会造成问题。
一个简单的问题。
你的85032B中的女性N标准是否有中心导体？
戴夫



     以下为原文

  > {quote:title=imett wrote:}{quote}
> Hi, i've an hold 8753a with its 85046a but the 4 cable aren't original, i've used a short RG223 instead.
> The network works well but the response are not soo flat, for example for the S11 measure (dB) at low frequency the measure is quite 0dB but at 3GHz is about -1.5dB. No problem, with the calibration (SOL with 85032 kit) the response became flat.
> When i turn off the instrument everthing going back and any time i've to connect the calibration kit.
> There is a way to store this calibration like a cable compensation or correction? I need a particular software o correction tool? 
> I read the service/user/on site manual but nothing about this problem.I'm wondering about that the manuals doesn't talk about a real instrument calibration/correction, i think that is normal have same components (directive coupler, cable, connector..)
> with small difference.
> 
> thanks in advance.
> matteo

As long as those cables are phase stable, it will not matter. They do *not* have to be phase matched. See this thread
Do cables from 8753A to S-parameter test set need to be phase matched?
 

It's late here and I'm tired, so don't have time to look carefully at what you wrote. 

But the whole point of the VNA calibration routine is to compensate for the non-ideal behavior of couplers, internal cables etc. So as long as your cable is stable, it will be ok. Personally I think I would have used semi-rigid, but that is just my view. 

What the calibration can't do is correct if things change between calibration and measurement, such as happens with temperature changes, dirty connectors, phase change in cables due to bending and temperature etc. But losses in your cables, as long as it stays fixed and is not excessive, will not cause a problem. 

A quick question. Does your female N standard in your 85032B have the center conductor? 



Dave

谢谢你的快速反应，我的校准套件是85032e，由909F + 85032-60011开/短制成。
使用原始电缆，响应变得平坦？
否则我将使用哪个校准程序来校准vna以便在关闭后校准不会消失？我看到有一个gpib工具包（KE5FX gpib工具包2）可以将校准保存在
仪器，但我认为，每次你必须回忆这个记忆......比连接溶胶套件和直通连接要容易得多。
对于我的工作，我不需要高精度，它只适用于我的家庭实验室，在我的工作中我们有一个运行良好的E5071！
再次感谢



     以下为原文

  Thank you dave for the quick response,my calibration kit is 85032e and it is made of 909F+ 85032-60011 open/short.

With the original cables the response became flat? Otherwise which calibration routine i've to use to calibrate the vna in order to have a calibration that doesn't disappears after the switch off?I see that there is a gpib tool kit (KE5FX gpib tool kit2)  which save the calibration inside the instrument but i think that everytime you have to recall this memory...is much easier than connect the sol kit and a thru connection.

For my job i don't need high precision , it is only for my home lab, at my work we have an E5071 that works well!
Again thanks

请记住，8753A现在已经有28年了，它的记忆可能正在消失（就像我自己的一样）。
事实上，我认为你必须使用保存/调用键存储仪器状态来存储寄存器。
它将存储仪器状态和校准。
但如果关闭它，对于8753，如果仪器不到1年，则保存时间为3天。
它有一个“超级上限”（请记住，这是在nicad电池随时可用，甚至闪存之前），它可以在静态撞锤上充电约3天。
但随着仪器老化，超级电容器寿命逐渐降低。
还要意识到，您的洗衣机lkely比8753A具有更强大的微处理器。
它是在1979年到1986年之间设计的（是的，7年制作，当时也没有设计工具）。
这是在任何PC之前，在电子邮件之前，我们（HP）在Palo Alto有一台计算机，并且如果我们想要进行诸如过滤器模拟之类的计算，则调制解调器与它通信。
采样器的模拟花了20个小时来写回数据文件（50,000点时间响应）。



     以下为原文

  remember, the 8753A is now about 28 years old, and it's memory may be fading (like my own).

In fact, I think you must store the instrument state using the save/recall keys to store a register. It will store the instrument state and calibration.  But if you turn it off, for the 8753, the save time was specified at 3 days if the instrument is less than 1 year old.  It has a "super cap" (remember, this was before nicad batteries were readily available, or even flash memory), that could hold a charge on the static ram for about 3 days. But as the instrument got older, the supercap life degrades.

Also realize, your washing machine lkely has a more powerful microprocessor than the 8753A.

It was designed between 1979 and 1986 (yup, 7 years in the making, no design tools back then either).  This was before any PCs, before Email, we (HP) had one computer, in Palo Alto, and a modem to talk to it if we wanted to make calculations such as a filter simulation.  The simulation for the sampler took 20 hours to write back the data file (50,000 point time response).

事实上，从这些图中，分析仪和测试装置仍然完美无缺！
这么多年后非常值得注意。
R和A通道的规格为+ -1 dB，A / R的比例为+ - 1/2 dB至1.3 GHz和+ -1至3 GHz。
那就是分析仪本身。
关于R和A通道的测试设置丢失的规范也是+ -1 dB（如果我没记错的话），因此如果没有用户校准，打开的A / R测量值应在+ -1.5到+之间读取 -
它在该范围内为2 dB。
为了改善这一点，您必须使用纠错，8753A提供85046A测试装置的完整2端口纠错。
在线查找旧手册并阅读Calibraiton章节和误差修正操作理论，了解如何应用用户纠错。



     以下为原文

  In fact, from those plots, the analyzer and test set are still perfectly in spec!  Remarkable after so many years.

The specification on R and A channels were +-1 dB and on the ration of A/R it is +- 1/2 dB up to 1.3 GHz and +-1 to 3 GHz.  That's on the analyser by itself.

The specification on the loss of the test set to the R and A channels is also +-1 dB (if I remember correctly) so without a user calibraiton, the A/R measurement of an open should read within +-1.5 t o +-2 dB over the range, which it does.

To improve this, you must use error correction and the 8753A provides for full 2 port error correction wtih the 85046A test set.  Find an old manual online and read the chapter on Calibraiton and the Theory of Operation for Error Correction to learn how to apply a user error correction.

感谢大家，我读了校准章，校正错误我需要这些部分：校准套件85029B带软盘（20dB衰减器50dB衰减器和不匹配适配器）-HP 9122磁盘驱动器问题是磁盘驱动器，这种
部分很老，在ebay上找工作单位并不容易。
有一种方法可以在gpib上使用磁盘驱动器模拟器吗？
对于校准套件软盘很容易找到副本？
我不需要高精度，我可以使用虚拟衰减器。
在您看来，校准实验室有这些旧零件吗？
谢谢matteo



     以下为原文

  Thanks to everyone, 
i read the calibration chapter and for the  correction error i need these parts:

-calibration kit 85029B with floppy disk (20dB attenuator 50dB attenuator and mismatch adapter)
-HP 9122 disk drive

The problem is the disk drive, this kind of part is very old and find a working unit over ebay is not easy.
There is a way to use a disk drive emulator over gpib?
For the calibration kit floppy is easy can find a copy? i don't need high precision,  i can use a dummy attenuator.

In  your opinion a calibration lab has this old parts?
thanks matteo

> {quote：title = Dr_joel写道：} {quote}>它有一个“超级上限”（请记住，这是在nicad电池随时可用，甚至是闪存之前），它可以在静态ram上保持电量约为
3天。
但随着仪器老化，超级电容器寿命逐渐降低。
我想我们可以感谢Nicad在设计8753A时没有*，因为我预计它们现在都会泄漏，并且会有更多损坏的8753A和更少可用的。
而8753A仍然很强劲。
我有一些人在eBay上看到了一些低端PNA现在用于少于后来的8753系列乐器。
戴夫



     以下为原文

  > {quote:title=Dr_joel wrote:}{quote}
>  It has a "super cap" (remember, this was before nicad batteries were readily available, or even flash memory), that could hold a charge on the static ram for about 3 days. But as the instrument got older, the supercap life degrades.

I think we can be thankful Nicads were *not* available when the 8753A was designed, as I expect they would all be leaking by now, and there would be more damaged 8753As and less usable ones. Whereas the 8753As are still going strong. 

I have some seen some low-end PNAs on eBay now for less than some of the later 8753 series instruments. 

Dave

NiCads当时可用，但我想我们决定不使用它们。
我在1972年以395美元的价格购买的HP35计算器有NiCads。
同样的电池组也在8350A中用于内存备份。
我们可能认为超级电容是一个更好的主意，因为不需要复杂的充电电路......并且它们不会因电池泄漏而损坏。



     以下为原文

  NiCads were available back then, but I guess we decided not to use them.  My HP35 calculator that I bought in 1972 for $395 had NiCads.  That same battery pack was also used in the 8350A for memory backup.  We probably thought the supercap was a better idea since no complicated charging circuitry was needed...and they was no chance a damage due to battery leakage.

> {quote：title = jvall写道：} {quote}> NiCads当时可用，但我想我们决定不使用它们。
我觉得这是一个明智的选择
>我在1972年以395美元的价格购买的HP35计算器有NiCads。
我想当前的消耗不会使电容非常实用。
> 8350A中也使用了相同的电池组进行内存备份。
我们可能认为超级电容是一个更好的主意，因为不需要复杂的充电电路......并且它们不会因电池泄漏而损坏。
随着老化，许多测试设备上的电池泄漏确实是一个问题。
我想我应该在我的8720D中更换电池 - 鉴于安捷伦仅在一两个月前校准过，我现在不愿意打破校准密封件，但我怀疑我会在明年进行校准之前做到这一点
。
另一个特别令人讨厌的电池问题是HP 70700A数字转换器模块。
当电池耗尽时，RAM内容丢失，设备无法正常运行。
仅仅更换电池是不够的，因为它需要重新编程，为此你需要一些非常古老的惠普电脑。
当然，很多HP测试设备的使用时间远远超出其设计寿命。
我相信有些人在8753中安装了液晶显示器来更换破旧的CRT。
戴夫



     以下为原文

  > {quote:title=jvall wrote:}{quote}
> NiCads were available back then, but I guess we decided not to use them.  

A wise choice I feel. 

> My HP35 calculator that I bought in 1972 for $395 had NiCads.  

I guess the current consumption of that would not have made a capacitor very practical. 

> That same battery pack was also used in the 8350A for memory backup.  We probably thought the supercap was a better idea since no complicated charging circuitry was needed...and they was no chance a damage due to battery leakage.

Battery leakage does seem to be a problem on a lot of test equipment as it ages. I guess I should get the battery replaced in my 8720D - given it was calibrated by Agilent only a month or two ago, I'm reluctant to break the calibration seals now, but I suspect I will do it before it goes for calibration next year. 

Another particularly annoying battery problem is the HP 70700A digitizer module. When that battery dies, the contents of the RAM are lost and the device is inopperable. Simply replacing the battery is not enough, as it needs to be reprogrammed, and for that you need some pretty ancient HP computer. 

Of course, a lot of HP test equipment is used well beyond its design lifetime. 

I believe some people have fitted LCD displays in the 8753 to replace worn out CRTs. 

Dave

事实上，这是一个“超级帽子”，用于保持公羊活着。
我们需要能够持续至少10年的东西，1985年的电池不够可靠。
最终它会降级，几年后记忆会丢失（现在已经超过了25年）。



     以下为原文

  In fact it was a "super-cap" that was used to keep the ram alive.  We needed something that would last a minimum of 10 years and batteries in 1985 weren't reliable enough.  Eventually it would degrade and memory would be lost after several years (and now it has been more the 25 years).

我几个月来第一次打开我的8753A，尝试进行一些测量，并且在我没有足够的内存错误之前它将不会进行很长时间的校准。
如果我将扫描点数设置为1601，则根本不会让我进入校准过程。
所有内部系统测试都通过......你认为这是一个失败的“超级上限”或失败的RAM的症状吗？
如果是这样，8753A的更换内存/“超级上限”部件是否仍然可用？



     以下为原文

  I just turned my 8753A on for the first time in months, tried to make some measurements and it will not hold a calibration for very long before I get an insufficient memory error. If I set the number of sweep points to 1601, it will not let me into the calibration process at all. All internal system tests pass...  Do you think this is symptomatic of a failing "super cap" or failing RAM? If so, are replacement memory/"super cap" parts for the 8753A still readily available?

可能不是超级上限问题（这与内存是否能在电源循环中存活有关）。
RAM可能会失败。
这些部件可能与1975款福特Maverick的部件一样可用（这是我在CPU的设计进行时所驾驶的）。
CPU设计已有30年历史，因此您可以寻找所有16K内存。
哦，不要忘记它是通孔部件。



     以下为原文

  Probably not a super-cap issue (that has to do with whether the memory will survive a power cycle).

Might be failing RAM.  The parts are probably just as available as parts to a 1975 Ford Maverick (which is what I was driving when the design of the CPU was going on).  The CPU design is a full 30 years old so look for all the 16K ram you can.  Oh, and don't forget it's thru-hole parts.