NanoVNA Custom Software

[jspencerg]

Re: Transco relays
Joe's relay has different activation wiring.  Unit I ordered has only two wires.  No sign of Transco, absorbed by another company?  Dow-Key microwave appears to make similar units.  Identified likely schematic for my unit. There are several units with 3+ control wires. Attached catalogue of coaxial relays.
  2-dpdt-transfer-coaxial-switch.pdf

[joeqsmith]

The manual provided the part number for the relay I used but again the software will not care. 

[jspencerg]

The relay activation methods seem to be only difference between models.  I don't know why or when one switching activation would be preferred.  My two-wire input is simplest of them all.

Creating this port- switching arrangement is a rabbit hole I had not planned on.

[joeqsmith]

Your relay will require a constant supply for one state vs my latching type that only requires a small pulse to change the state.   My average current draw is basically 0 off the USB port.  You could use a simple RC diode circuit to reduce the hold current after it switches if that is a concern for you.  They may not spec the hold current but easy enough to swag.   

Do you have a clear goal in mind or just wanting to play?  Just curious.  Obviously you are adding variables and as we saw from realfrans, it can cause some problems.  If the goal is to measure active circuits, keep in mind that square wave drive.   If you just want to play with it, maybe build a T-Check to go with it.   In the last video I made, I show the home made T-Check and transfer relay giving some decent results to a GHz or so. 

[realfran]

The transfer relay on my set-up is one design from HP there is two really switches on H the 141T" semirigid from the relays to the N connectors hare the original length. The only thing  I change is the relay drives, the connection from LiteVNA port and the relays are connected directly.
The LiteVNA for is limited specs don't notice a big problem of crosstalk or isolation on the end of all is hobby instrument, not metrology grade device.
The calibration on the software helps a lot to minimize the difference.

[jspencerg]

Re: My goals
I'm interested in the science. I'm studying the theories and applications of EE. Waves were always central to my engineering and science.  Learning about, measuring and controlling EM wave behavior fascinates me. As a hands-on person I've always considered that being able to measure something is a core endeavor.
I'm having fun with it.

[joeqsmith]

The HP8711A seems to have had an upper end of 1.3G.  This makes sense that the data you posted for your Mini-Circuits attenuator was so poor at 6GHz.  I imagine the extra stubs are a problem.   I had no luck finding a clear photo of what it looks like inside. 

I'm studying the theories and applications of EE.

I've been in that same boat for many decades and still know very little about electronics.   There's just not enough time.   

[jspencerg]

Re: Learning EE et. al.
The amount to learn just keeps growing too. Plasmonics... Where will this field lead to?
The great thing for me is the availability of learning resources and hands- on experiments I can do while learning.
Having so many experts,  or at least very competent people, on YouTube,  etc sharing and teaching is a huge boost for me.
I'm going to recreate your power circuit for relay using Microcap and then modify for my relay.  It'll be a big learning process all the way.
Do you use RG58 with stranded core for your instrumental connections?

[joeqsmith]

Mostly I use  RG223, RG400/U, RG188

https://www.pasternack.com/flexible-0.122-rg188-ds-50-ohm-coax-cable-ptfe-jacket-rg188-ds-p.aspx

Link shows a photo of the custom blue cables I use in a fair number of videos. 
https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg3954551/#msg3954551

[jspencerg]

Re: Meaning of vat-15 scan?
I'm thinking the v2plus4 did good job of measuring the gain.  Shows loss of functionality at +5GHz because attenuator data sheet says attenuation should be going down as frequency increases.
Unknown is how dependable the units are to follow spec's. Have to verify attenuator on another VNA.
How's that for reading results?

Looking at cable specs:  I see the double- shielded cables have higher operating frequencies.

Thanks.

[joeqsmith]

I wouldn't expect to see that much ripple, even when using the ideal models. 

Yes, the cable I use will normally outperform RG58.   

[jspencerg]

Re: ripple in attenuation level
Would the ripple decrease if measurement taken over narrower frequency range?
Is there good rule for samples/Mhz where not expecting filter or resonant effects?

[joeqsmith]

That's a good question for you and realfran to sort out. 

[jspencerg]

I set up v2plus4 as shown in screen shot. Calibrated it. I'm getting nothing when I sweep. Seems locked up.
Restarted all. Ran without calibration.  Normalized with a thru. Connected 2W 20db attenuator from aliexp. Obtained screen shot.
v2plus4 clearly breaks down about 4.5GHz. Results lower than that look really good.

This locking up after calibration always occurs.  My nanovna-h does not have that problem. Is there something to check or fix for this issue? 

[joeqsmith]

Re: lockup

You would need to provide more details beyond suggesting it locked up.  I suggest you look at the specs for your V2Plus4:
https://nanorfe.com/nanovna-v2.html

Someone else once told me how my software was locking up after calibration.  They too provided no details.  In the end, after they provided screen shots, I could see they had entered their own coefficients for the standards.   They had zero'ed out C0 which then caused a divide by zero and the software was showing NAN (not a number).   I talked about that specific problem in the last video I made.   

I don't typically limit the software to try and protect the user from themselves.   If you want to program your V2+4 for 20GHz, the software will gladly attempt it.   As I said during that last video, I now add a small offset to C0 to avoid diving by zero.   I also talked about someone who was attempting to set the number of data points to something very small  (as this apparently improves the distance resolution in TDR mode), which would put a very high demand on the CPU to the point where the OS may become non-responsive.   The new software attempts to throttle the system.  So there are a few things that I added in an attempt to prevent users from shooting themselves in the foot.  We also now handle the regional settings for much the same reason. 

[joeqsmith]

Maybe this will help:

https://nanorfe.com/forum/Preview-of-my-latest-software-updates.html

scroll down to:
Guido ON7CH 2021/11/19 13:57

Guido,

If you want to try and solve it, as I have wrote many times, you need to provide enough detail for me to replicate the problem.  I don't have enough time to guess at what you are doing.   Post a screen shot of the Calibration terms you have entered and AFTER you calibrate, post a picture of the coefficients.  Are you saving the settings, if so, upload the defaults file you have created.    I wonder what you are actually seeing when you suggest CH0_Z freezes.   Again, a screen shot would have helped.

I have not actually used more than the first entry in the standards with the V2Plus4 but I have used more than 10 with the other VNAs and they pull from the same code base.  CH0_Z should never freeze and there is nothing magic about the tenth entry.  Maybe post a photo of the data supplied with your calibration standards.   I think with all of this information, we should see the problem rather quickly.

[jspencerg]

I have not adjusted any values you mentioned prior to or after calibration.  I'll study on this, see if I can find pattern of failure or success. 

[joeqsmith]

That's one clue.   I have not tried to run the V2Plus4 outside of it's operating range like you are and I am not sure how my firmware responds compared with yours.  You could provide screen shots like the other person did.  That would certainly help me understand what is happening.

[joeqsmith]

This is not something I would normally do but again, the software will certainly allow you to. Don't expect it to somehow magically increase the upper range of your equipment. 

[jspencerg]

Re: vat-15+ results and documenting programs
I now divine you ran that test with the litevna rather than v2plus4. That is why you ran it to 6.4Ghz. Yes,  I know that is outside the v2plus4 range. My assumption based on program version you ran,  would have thought you'd run litevna on the new solver64 software. Did not understand hardware ID at top of page.
I'm reading suggested posts.  Realizing the depth and breadth of responses you've made.  For me,  repeating myself and continually referring back to past answers would be tiresome and frustrating.  Would compiling and sorting question/answers into a searchable FAQ style page create a user reference akin to a manual? I understand you're not writing another manual. Your postings have created an unapproachable(big) archive of advice and details for us newbies.
Thanks for your help.

[joeqsmith]

Yes, realfran had ran their VAT15 on their LiteVNA.  I assume they used their transfer relay to collect that data.  I suspect they ran it at 6.4GHz to show how well it worked but they show a lot of ripple.   There was a reason I suggested you work with them to answer your question about the ripple. 

Because the newer LiteVNA is somewhat backwards compatible with the V2Plus4, the software I wrote for the V2Plus4 does indeed support it to some extent.   

Answering questions, especially repeated ones can also be tiresome and frustrating.   You certainly could create a FAQ if you want to just like nctnico could also write an installation manual.   Don't expect me to create any new documentation.   People will not read it.   In the future I will most likely stop attempting to answer the more basic questions. 

[jspencerg]

Re: answering questions
"In the future I will most likely stop attempting to answer the more basic questions."

I expect anyone would grow tired of repeating themselves.  Regrettable you don't want to organize the help you've already provided. I don't have to understanding necessary to transform your conversations into a categorized FAQ.

[jspencerg]

Re: v2plus4 locking up
Attempted calibration after first reading of postings with Guido. Same behavior. Program not displaying same errors as his.  Don't know where to go from here. Attaching screen shots.
PS Reset all. Attempted calibration between 10M and 300M with 401 pts. Same results.  Calibration terms identical.

[jspencerg]

Another pic

[jspencerg]

One more.  Attaching all three to same post was not working.