Copper Mountain Technologies VNA

[TELCO]

Hello to all,

Is anybody here using or have experience with the Copper Mountain Technologies Vector Network Analyzers? Specially for use in RF applications such as filters, duplexers, antennas, coax feedlines, attenuators. I am currently using the CMT TR1300/1 two port VNA and it has performed better than we expected (I am in no way associated with them other than just another happy customer).

Looking forward to comments.

Peter 

[Colt45]

They seem to work at least as well as 80's HP boatanchors IME, though my use of them has been pretty limited... sweeping amps, cables, etc.

[paul@yahrprobert.com]

We used those analyzers in an RF tester product and got them do do what we wanted.  As I recall there were a few "learning curves" with the software.  Ours were running in an NI PXIe chassis with a windows 10 controller.  The VNA was controlled by a piece of Copper Mountain software running on Windows, and we sent SCPI commands to the Windows software to control the VNA. The tester was to operate in a corporate environment where keeping the windows version up to date was mandatory, so the chance of that chain of software staying viable for 10 years was virtually nil.
   As far as the VNA performance was concerned it was good enough for us.  I'm pretty sure the dynamic range of that unit doesn't compare with real, expensive, standalone VNA's, so caveat emptor there.

[virtualparticles]

The dynamic range of the TR1300 is 130 dB which is actually better than a certain "E" series VNA. The performance of the CMT VNAs are every bit as good as those made by the other guys. The ISO-17025 lab in Indianapolis, certified to 50 GHz, can attest to the performance of the gear and provide uncertainties traceable to NIST. The "expensive standalone" VNAs are simply more expensive and the embedded OS is an IT nightmare. These days everyone has a laptop computer. Why not run sweeps on your own laptop and save the data? Running back and forth between the VNA and your office with a flash drive is just silly.

[joeqsmith]

Good to see you back Brian. 

[pdenisowski]

Why not run sweeps on your own laptop and save the data? Running back and forth between the VNA and your office with a flash drive is just silly.

I agree that moving files with a flash drive is ... suboptimal

However, most modern standalone VNAs (and all R&S VNAs) allow you to remotely control them and transfer data/files/settings over a LAN connection using standard web browser.

[joeqsmith]

His point of "IT nightmare" is valid for many companies.   That old Agilent PNA I have is Windows based and supports Ethernet.  It allows me to do pretty much anything I want over the LAN.   At work, the IT department would have a fit over connecting it to the LAN.   Still, I could run it from a laptop and bridge to the corporate network like he mentions.   

[G0HZU]

Every so often I backup all my s-parameter data files from my old Agilent E5071B. Last time, I counted the number of SnP files and it was over 3000. Many of these were semi automated measurements of semiconductors (eg BJTs or FETs) at various bias conditions so 50 files can be generated in one session for example.
However, it is a major pain to have to transfer the data across to my main PC via a USB stick.

In that sense, I'd like to have a USB based VNA. However, I'd be concerned about the long term stability of some of the USB based VNAs from various manufacturers. A component measurement session might take me quite a while as I have to change the bias conditions maybe 50 times. I need to rely on a VNA that can hold its calibration for the whole time. The E5071B (and the C version at work) seems to be really good at this. It has fan controlled thermal management to try and keep the internal temperature stable for example. There's also lots of thermal mass in the way of any sudden changes in room temperature. So in this sense I'm quite happy to have a huge and heavy VNA on my bench.

[joeqsmith]

Why wouldn't you just link it to your LAN with Ethernet and store the data direct to the PC?   In my case, I just setup a shared area on my PC and mount it when the PNA boots.  I've not had any problems using Windows 2000 or XP to connect to Windows 10.

[KE5FX]

Yeah, storing data on the VNA itself is the mother of all bad ideas.  Nobody ever thinks to back up those drives, if/when it is even possible to do so.   When it gets upgraded, you now have to think about how to get the data over to the new instrument.  Finally, when it is scrapped or sent to surplus, hopefully somebody remembers to zeroize it.

Usually best to treat all instruments like DAQ peripherals.

[G0HZU]

I agree it would make sense to adopt the LAN method but this is banned at work for various reasons and I prefer to adopt the same rules here at home.

With this in mind, it does actually make sense to store to the VNA internal HDD as well as to a USB stick because I get a fairly safe (temporary?) copy for free on the VNA HDD and if something happens to the integrity of the USB stick  (data corruption?) it means that I don't have to repeat the measurements again.

I might take a dozen s2p measurements in maybe 30 minutes at various operating points or drive levels and saving each one to the internal HDD first is the best thing to do as I know (within reason) that the data has been saved reliably without having to verify on my desktop PC.

In other words, a typical session for me would involve me making 10-50 measurements and I'd save each one to the HDD as an SnP file as I go along. At the end of the session I transfer all of these files to the desktop PC via a USB stick. Obviously, this is a far from ideal way to do it but it's the way I've done it for many years and I've not lost any data yet. I used to do this stuff via GPIB on my first VNA but I now also need to save other things apart from s-parameter files and I want to preserve the original Agilent file format. Also, my main desktop PC is in the opposite corner to the VNA in my workroom now. So any cable routing would be a pain.

Networking it all is the obvious answer, but this is banned at work for various security reasons and I've adopted the same rules here.

Nobody ever thinks to back up those drives

I do back up the VNA drive data quite regularly including the original system restore (ghost?) files put on there by Agilent and the options *.lic file.

[joeqsmith]

If I saddled my home hobbies with work procedures,  it would strip away the fun and I would not get much done.   It's better for me to keep the two separate. 

In the case of the LAN, for me that's the natural way to run tests anyway so all of my equipment is on it.  It's not just how I save data but how I automate any experiments I run.   For GPIB, the interface is on the LAN.   I can walk outside of the house and have access to the entire lab at my finger tips.  I could even open up to the WAN and let you run it from your home.   

For backup, I could just save the data twice from the VNA like you mention, but once it is on the PC, I back that all up frequently anyway to multiple devices.

[virtualparticles]

In that sense, I'd like to have a USB based VNA. However, I'd be concerned about the long term stability of some of the USB based VNAs from various manufacturers. A component measurement session might take me quite a while as I have to change the bias conditions maybe 50 times. I need to rely on a VNA that can hold its calibration for the whole time.

This is a very valid concern. A poorly made VNA may drift over time and the calibration would become invalid. All VNAs do drift a tiny bit of course. I've tested this a number of times on a CMT VNA and measured around 0.03 to 0.05 dB drift on the S21 measurement of a 15 dB attenuator over a 24 hour period.

[virtualparticles]

Why wouldn't you just link it to your LAN with Ethernet and store the data direct to the PC?

Hi Joe!

Mounting the VNA over the LAN is definitely better than running around with a flash drive stick. My only concern would be that if the OS in the VNA is older, it will be vulnerable to all the old exploits. We had a VNA that got hacked (at a prior company) and it was the base for getting into everything else in the building.

[tomud]

Mounting the VNA over the LAN is definitely better than running around with a flash drive stick. My only concern would be that if the OS in the VNA is older, it will be vulnerable to all the old exploits.

It's more of a company security policy issue. What is the difference whether a remote or local vulnerability will be used, e.g. using a flash drive.
With a good security policy implemented, a dedicated VLAN, such a thing can be minimized. Although it is known that it will never be eliminated, because here sometimes there is user stupidity (in general, the most common problem of hacking into the network).

As for the VNAs described here, I'd rather see sample measurements than opinions. With a description of what calibrators, test leads, etc. were used.

[virtualparticles]

As for the VNAs described here, I'd rather see sample measurements than opinions. With a description of what calibrators, test leads, etc. were used.

I can do that! I'll set up a drift test now, and post the results tomorrow morning. I'll include all the details of the measurement as well.

[abeyer]

It's more of a company security policy issue. What is the difference whether a remote or local vulnerability will be used, e.g. using a flash drive.
With a good security policy implemented, a dedicated VLAN, such a thing can be minimized. Although it is known that it will never be eliminated, because here sometimes there is user stupidity (in general, the most common problem of hacking into the network).

The whole reason behind those security policies is the lack of attention to that kind of thing. Sure, you can isolate things in a properly secured VLAN... but who's gonna do that when you can just plug it into your wifi router? (and then ignore the fact that your toaster is now part of a DDoS botnet because every other Internet of Shit device in your home was on the same network.) And while you're correct that there can be attacks via USB or other things, too... at least you seldom pick up random USB drives from people on the internet and plug them into your test equipment. (I hope)

This isn't even just a concern for old devices. There is(was?) a recent RCE against fully updated current model Rigol scopes.

[G0HZU]

In my case, my Agilent VNA is quite old now. It's running Win 2k. There will be ways to network it to my main Win 11 PC but I don't want to do it for various reasons including security.
I'm going to keep doing things the way I always have done with this old VNA. I'll just keep on trotting back and forth from PC to VNA, slowly wearing out the carpet

I've got various bits of test gear here that run old versions of Windows as the OS. The oldest is a scope that runs Win98. I've got a couple of really old Tek RTSAs here that run WinXP although one of them (RSA6114A) runs WinXP embedded whatever that means. I don't intend to network any of them via LAN. I do have them linked via GPIB for screenshots to another netbook PC that is not on any network.

I remember there being a big panic when one of the company VNAs came back from repair/calibration and it had been to two places to do the repair. Along the way something got onto it that caused a virus scan to trip an alert at the calibration house. That caused a lot of confusion and delay. It turned out to be a false alarm but sending old Win 2k and Win XP VNAs and analysers for calibration has its risks.

The company also has to be careful to not store any sensitive data on the equipment before sending it for calibration. It has to go to approved places only that also do virus scanning on arrival and dispatch. I guess that's another reason to move towards USB based analysers. Much easier to manage the annual calibration or any repairs.

[joeqsmith]

As for the VNAs described here, I'd rather see sample measurements than opinions. With a description of what calibrators, test leads, etc. were used.

I can do that! I'll set up a drift test now, and post the results tomorrow morning. I'll include all the details of the measurement as well.

Look forward to seeing how you test it.  I tried looking at a few of the low cost VNAs by running a cal directly on their connectors and then changing the temperature by 10degC.   

https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg3890843/#msg3890843

[rfclown]

Several years ago I purchased a Copper Mountain R60 for the company I was working for to test antenna assemblies in production. My past experience had mainly been with HP boxes (HP8753, 5071), and I was pleased with the R60. When at that company, I had a FieldFox for a VNA at my bench. With the HP VNAs, I didn't have the TDR option, so I had to do my own math for that function. For antenna assembly testing (connector, cable, antenna), I needed to have a TDR function to find issues with manufacturing. The R60 software threw in the TDR option as free, so that was a real bonus. In fact, you can download their software, use whatever VNA you have, save in Touchstone, and read the files in and convert to time. Just make sure you do the sweep correctly (evenly spaced points from 0) because the software doesn't check that and you'll get bad results (at least it did this 7 or so years ago). I wouldn't hesitate to consider Copper Mountain again.

In my case, my Agilent VNA is quite old now. It's running Win 2k. There will be ways to network it to my main Win 11 PC but I don't want to do it for various reasons including security.
I'm going to keep doing things the way I always have done with this old VNA. I'll just keep on trotting back and forth from PC to VNA, slowly wearing out the carpet
...

I completely don't understand this. Whatever VNA I use, I always read the data into a PC, usually using VISA/SCPI (GPIB, Ethernet, Whatever interface the device has). I never use sneakernet; I don't understand why anyone does. If you have a virus issue on your Windows based VNA, plug in a USB flash drive, and sneakernet to a PC, you can also transfer problems that way. If I connect GPIB or Ethernet and get my data via SCPI, how is that a security issue? If you connect with Ethernet, Firewalls work. I have a specific subnet which I only use for instrument control. I don't try the do the log into the instrument from the other side of the world thing. If I need to work remotely, I'll remote into my lab computer, and run my automation programs.

Copper Mountain is completely different in that you run their application and communicate with that to get the data.

[virtualparticles]

I ran an SC5090 compact VNA, measuring a 10 dB attenuator over the full span. The cables were 2 ft N to 3.5mm and the attenuator was a male to female 3.5mm with a female to female 3.5 mm adapter. After allowing the machine to warm up for one hour, I saved the S21 results to memory at 4PM yesterday. This morning at 11AM I did a data/mem plot to show the drift over the 19 hours. I also saved the two Touchstone files. As expected, I see about 0.05 dB drift, mostly at 9 GHz.

Not that it really matters, but I calibrated using a 20 GHz Automatic Calibration Module.

[joeqsmith]

Copper Mountain is completely different in that you run their application and communicate with that to get the data.

If I understand your statement, you can not directly control their VNA with custom software but have to use theirs?    The LibreVNA also works like this.  Personally, I would rather have direct control.   

It sounds like the sneaker net is what he is used to and he may not have any experience with networking.  I'm far too lazy to work like that.  I would rather the PC run the test and collect the data for me. 

[joeqsmith]

I ran an SC5090 compact VNA, measuring a 10 dB attenuator over the full span.  ..

Any idea how much the rooms air temp varied between those two data points?   

***
Looking at the temperature test I ran with the LiteVNA, I was concerned about the cables adding error and this is why I did not use them.   Still looks like the majority of the error was the phase.   I only ran a narrow range but can try to repeat your test.  For the $120, don't expect much out of it. 
***
Attached looking at your normalized mag&phase data.   I wonder how much of this is due to a mechanical change.   Interesting.  And thanks for running it.

[G0HZU]

I don't want to do the file management via GPIB. I've done that in the past and it's not for me. Whatever method I use I can't avoid the short walk from VNA to PC at the end of a session...

The simulation SW I use is well suited to this system. I don't need to do any browsing or launch any GPIB based file management programs via GPIB at the PC end because the simulator does it for me within its GUI. It's been written to make the SnP file import process easier. The PC pops up the files for me when the USB stick is plugged in. Just double click any SnP module already in the simulator and right click and paste. If there is only one file the simulator automatically loads it into the SnP block for me as well as saving it to the desktop PC in the same folder as the main project file. It saves all the files if there is more than one file being transferred.

It isn't a perfect system, LAN would be slightly more efficient but please don't tell me to use LAN (again) as I don't want to use LAN for various reasons including security.

I also don't want the main PC on my RF workbench. It's nice to have it in the opposite corner by the window (for lots of reasons).

Sometimes it's nice to just use a nanovna with the main PC and that's why I would also consider a fairly decent USB based 2 port VNA that can be used away from the main RF workbench for casual stuff. But it isn't going to replace the main VNA any time soon for various reasons.

[joeqsmith]

I don't believe anyone has told you what to do.  More offering different perspectives.  You lost me with that last post.   I assume the simulator is some sort of post processing software. 

In my case, the desktop PC just runs normally.   I setup the equipment's file save and print buttons to point to the shared area of the desktop.   So when I press print or save on the equipment, it shows up on the PC.   The software I use on the PC side points to the shared area so I just run it.   I setup the Windows equipment to mount the shared area automatically as part of the boot process.  I don't even really think about it.  All of my test equipment plugs into a single switch.   There's just one cable then that routes back to another switch located at my desk.   

Running the equipment is all done over that same bus.   The PCs connect to that second switch along with a hacked (WinRT) wireless router.    The printers are also on Ethernet, so I can have the PNA point to the printer if I want (like any other PC).   For equipment that only supports the Centronix parallel port for printing, I designed an adapter that emulates a printer and sends the data over Ethernet to the printers.  Wrote that stack in assembler for a 6811 back when I thought I was smart.