85032F 9 GHz N套件的系数混乱

嗨Dave，我们的标准实验室人员在过去15年左右对85032F的系数进行了一些调整/优化，并且当他们这样做时，并非所有支持站点上显示的表都经过更新以反映这些变化，
所以最终成为你所指出的所有差异的原因。
我想象的8753D表中显示的滑动负载是复制和粘贴错字的结果;
你会注意到那里显示的类定义没有引用那个标准数字。
今天表示它是PNA的85032F系数的网页表实际上对应于您在现有* ENA * *上看到的85032F系数，但正如您所指出的那样，这些系数与实际存储在该套件的* PNA * *上的系数不同。
该表中显示的系数实际上仅在高达9 GHz时进行了优化，因此高于该频率时，它们无法准确地对标准建模，因此您不希望使用那些频率以上的系数。
虽然该套件的PNA中存储的系数实际上确实可以准确地为高于9 GHz的标准建模，即使该套件官方仅指定高达9 GHz。
高达9 GHz这两组系数实际上同样准确地对标准进行建模（我们标准实验室的工程师最近验证了这一点）。
客户支持应该很快纠正该网页，以指示最新PNA中真正使用的系数，我将检查是否要求他们删除8753D表中指示的滑动负载。
我记得你之前提到的PNA，它是一个较旧的型号。
在那个年代的PNA中，该套件中两个开口的偏移延迟存储为41.19 ps。
几年后，实验室决定更新该定义，使两个打开之间的建模偏移延迟略有不同。
我认为更新是为了更好地匹配由于女性接触而导致的实际身体反应。
我相信该支持网页上的8753D表是根据该更新进行调整的，这就是为什么你在那里看到两个不同的延迟值的原因。
旧版PNA的固件没有得到更新;
较新的PNA基本上得到了更新，但两个开放延迟值进一步细化了一点。
顺便说一下，该套件的FieldFox'.xkt'文件中显示的所有系数编号与后面的PNA固件版本中的定义完全匹配。
希望这有助于减轻混乱和最好的问候，布拉德



     以下为原文

  Hi Dave,

Our standards lab folks have made a few tweaks/optimizations to the coefficients for the 85032F over the past 15 years or so, and as they did so, not all the tables shown on that support site were updated to reflect those changes, so that ended up being the reason for just about all of the differences you pointed out.  The sliding load indicated in the 8753D table I imagine was the result of a copy-and-paste typo; you'll note that the class definitions shown there don't reference that standard number.

The webpage table of 85032F coefficients that today indicates it is for PNA does actually correspond to the 85032F coefficients you see on existing *ENAs*, but as you pointed out, those differ from the coefficients actually stored on *PNAs* for that kit.  The coefficients shown in that table are actually optimized specifically just up to 9 GHz, such that above that frequency they don't accurately model the standards so you wouldn't want to use those above that frequency.  Whereas the coefficients stored in the PNA for that kit actually do accurately model it's standards for well above 9 GHz, even though that kit is officially only specified for up to 9 GHz.  Up to 9 GHz both of those sets of coefficients actually model the standards about equally accurately (an engineer in our standards lab recently verified that).  Customer support is supposed to soon correct that webpage to indicate the coefficients that are truly being used in latest PNA, and I'll check to see about asking them to remove the sliding load indicated in the 8753D table.

I recall you mentioning before which PNA you have, that it is an older model.  In that vintage of PNA, the offset delay for both opens in that kit was stored as 41.19 ps.  A couple years later the lab decided to update that definition to just have the modeled offset delay differ slightly between the two opens.  I think that update was done to better-match the actual physical response due to the female contact.  I believe the 8753D table on that support webpage was adjusted per that update so that's why you see the two different delay values there.  The firmware for the older PNAs didn't get that update; the newer PNAs essentially got that update but with the two open delay values further refined a bit.  By the way, all the coefficient numbers shown in the FieldFox '.xkt' file for that kit exactly match the definition in the later PNA firmware revisions.

Hope this helps alleviate the confusion and best regards,
Brad

> {quote：title = bhokkan写道：} {quote}> Hi Dave，嗨布拉德>我们的标准实验室人员在过去的15年左右对85032F的系数进行了一些调整/优化，并且他们这样做了
，并非所有支持网站上显示的表都已更新以反映这些更改，因此最终成为您指出的所有差异的原因。
好的我明白了。
>今天表示它适用于PNA的85032F系数的网页表确实对应于您在现有* ENA * *上看到的85032F系数，但正如您所指出的那样，这些系数与实际存储在该套件的* PNA * *上的系数不同。
该表中显示的系数实际上仅在高达9 GHz时进行了优化，因此高于该频率时，它们无法准确地对标准建模，因此您不希望使用那些频率以上的系数。
虽然该套件的PNA中存储的系数实际上确实可以准确地为高于9 GHz的标准建模，即使该套件官方仅指定高达9 GHz。
它可用于18 GHz吗？
>高达9 GHz这两组系数实际上同样准确地对标准进行建模（我们标准实验室的工程师最近验证了这一点）。
我认识的人需要低成本的VNWA，这是一个500 MHz（动态范围降低为1.3 GHz）的VNA。
对他而言，不必担心连接器性别的便利性将超过使系数工作在9 GHz以上的任何优势。
>客户支持应该很快纠正该网页，以指示最新PNA中真正使用的系数，并且我将检查是否要求他们删除8753D表中指示的滑动负载。
>我记得你提到过PNA之前，它是一个较旧的型号。
是的，它的9 GHz 3端口N3383A - 信不信由你，我还没用过它。
我注意到比较了一些仪器，其中一个组件有两个略有不同的系数（我认为偏移Zo是49.99对49.992欧姆），虽然我不确定这是在N3383A，8720D还是8753ES上
。
但是从类中查看组件实际上没有使用它会出现。
至少在8720D和8753ES上，它不要求连接器的性别。
正如我所说，我还没有使用PNA，虽然我需要很快。
>在那个年代的PNA中，该套件中两个打开的偏移延迟存储为41.19 ps。
几年后，实验室决定更新该定义，使两个打开之间的建模偏移延迟略有不同。
总的来说，这似乎是一个不错的工具包，不必担心不同的性别 - 除了9 GHz以上。
这是一个明显优于85054B的优势，虽然当然套件应该更准确的滑动负载，它还有连接器量具和漂亮的适配器。
>我认为更新是为了更好地匹配由于女性接触而导致的实际身体反应。
我相信该支持网页上的8753D表是根据该更新进行调整的，这就是为什么你在那里看到两个不同的延迟值的原因。
旧版PNA的固件没有得到更新;
较新的PNA基本上得到了更新，但两个开放延迟值进一步细化了一点。
顺便说一下，该套件的FieldFox'.xkt'文件中显示的所有系数编号与后面的PNA固件版本中的定义完全匹配。
谢谢。
所以为FieldFox列出的那些基本上是你拥有的最好的。
>希望这有助于缓解混乱和最好的问候，是的，确实如此。
有兴趣的话，你有没有看过eBay上85033F的价格？
有一个售价2600美元，没有任何适配器，还有一个售价3000美元的适配器。
Keysight套件的新成本列为2203美元*，因此它们在eBay *上的价格高出400-800美元！
>布拉德戴夫



     以下为原文

  > {quote:title=bhokkan wrote:}{quote}
> Hi Dave,

Hi Brad

> Our standards lab folks have made a few tweaks/optimizations to the coefficients for the 85032F over the past 15 years or so, and as they did so, not all the tables shown on that support site were updated to reflect those changes, so that ended up being the reason for just about all of the differences you pointed out.  

OK, I see. 

> The webpage table of 85032F coefficients that today indicates it is for PNA does actually correspond to the 85032F coefficients you see on existing *ENAs*, but as you pointed out, those differ from the coefficients actually stored on *PNAs* for that kit.  The coefficients shown in that table are actually optimized specifically just up to 9 GHz, such that above that frequency they don't accurately model the standards so you wouldn't want to use those above that frequency.  Whereas the coefficients stored in the PNA for that kit actually do accurately model it's standards for well above 9 GHz, even though that kit is officially only specified for up to 9 GHz.  

Is it usable to 18 GHz? 
 
> Up to 9 GHz both of those sets of coefficients actually model the standards about equally accurately (an engineer in our standards lab recently verified that). 

The person I know who wanted them, needed for the low-cost VNWA, which is a 500 MHz (1.3 GHz at reduced dynamic range) VNA. For him, the convenience of having not to worry about the sex of the connector will outweigh any advantage of having coefficients that work above 9 GHz. 

> Customer support is supposed to soon correct that webpage to indicate the coefficients that are truly being used in latest PNA, and I'll check to see about asking them to remove the sliding load indicated in the 8753D table.

> I recall you mentioning before which PNA you have, that it is an older model.  

Yes, its a 9 GHz 3-port N3383A - believe it or not, I have not used it yet. 

I did notice comparing some of the instruments, that there were two slightly different coefficients for one of the components (I think the offset Zo was 49.99 vs 49.992 Ohms), although I'm not sure if this was on the N3383A, 8720D or 8753ES. But it would appear from looking at the classes that the component was not actually used. At least on the 8720D and 8753ES, it does not ask for the sex of the connector. As I say, I've not used the PNA yet, although I will need to soon. 

> In that vintage of PNA, the offset delay for both opens in that kit was stored as 41.19 ps.  A couple years later the lab decided to update that definition to just have the modeled offset delay differ slightly between the two opens. 

Overall, that seems a nice kit, with one not having to worry about the different sexes - except above 9 GHz. That's a distinct advantage over the 85054B, although of course that kit should be more accurate with the sliding loads, and it also has connector gages and nice adapters. 

> I think that update was done to better-match the actual physical response due to the female contact.  I believe the 8753D table on that support webpage was adjusted per that update so that's why you see the two different delay values there.  The firmware for the older PNAs didn't get that update; the newer PNAs essentially got that update but with the two open delay values further refined a bit.  By the way, all the coefficient numbers shown in the FieldFox '.xkt' file for that kit exactly match the definition in the later PNA firmware revisions.

Thank you. So those listed for the FieldFox are basically the best you have. 

> Hope this helps alleviate the confusion and best regards,

Yes, it does. 

As a matter of interest, have you ever looked at the prices of 85033F's on eBay? There is one at $2600 without any adapters, and a couple at $3000 with an adapter. The new cost from Keysight for the kit is listed as $2203, *so they are $400-$800 more expensive on eBay*! 

> Brad

Dave

> {quote：title = drkirkby写道：} {quote}>它可用于18 GHz吗？
是的，我确实得到了最新PNA和FieldFox目前用于该套件的那些系数被选择为具有宽带足够的配合，以便它们跟踪标准的实际响应至少18 GHz。
> {quote：title = drkirkby写道：} {quote}>谢谢。
别客气。
> {quote：title = drkirkby写道：} {quote}>感兴趣的是，你有没有看过eBay上85033F的价格？
有一个售价2600美元，没有任何适配器，还有一个售价3000美元的适配器。
Keysight套件的新成本列为2203美元*，因此它们在eBay *上的价格高出400-800美元！
不，我没有，嗯，这对我来说确实很有趣/令人惊讶。



     以下为原文

  > {quote:title=drkirkby wrote:}{quote}
> Is it usable to 18 GHz? 

Yes I did get confirmation that those coefficients currently being used by latest PNA and FieldFox for that kit were chosen to have a broadband-enough fit such that they track the standards' actual response out to at least 18 GHz.

> {quote:title=drkirkby wrote:}{quote}
> Thank you.
You're welcome.

> {quote:title=drkirkby wrote:}{quote}
> As a matter of interest, have you ever looked at the prices of 85033F's on eBay? There is one at $2600 without any adapters, and a couple at $3000 with an adapter. The new cost from Keysight for the kit is listed as $2203, *so they are $400-$800 more expensive on eBay*! 

No I hadn't, hmm that's indeed interesting/surprising to me.