开云体育

Easy to do in a long thread; I've done it myself.

Nice to see you posting; this group has been quiet for far too long.

Kerry VK2TIL.

Ooops, sorry, missed your earlier comment.

73, Vladimir

G'day Vladimir.

I mentioned this earlier.

I also have a VNWA but I don't like to smooth traces, I prefer the raw data.

I like to use Dave's myVNA with the N2PK; it's my tribute to Dave.

Kerry VK2TIL.

Hi,

You might want to check the smoothing function, just one of the beauties in DG8SAQ's VNWA software.
It also runs with N2PK hardware.

Cheers,
Vladimir

The attached photos show the setup and calibration standards.

Note that Load calibration is always 50 ohms; the 200-ohm device is used to check that setup and calibration are OK before beginning a measurement. It is NOT part of the calibration.

If I don't "see" 200 ohms, as per the attachment, for a Reflection setup or 9.54 dB loss for a Transmission setup, I know that something is wrong.

Kerry VK2TIL.

Where did you calibrate the load? What resistance?

Attached are;

Plot of raw data and

Plot of same data with Trendlines.

My engineer friend would be horrified!

Kerry VK2TIL.

Some VNAs, eg the VNWA, have a trace-smoothing function.

myVNA does not have this. If you want a nice straight trace, set the vertical axis scale to a high value per division.

I think that an Excel plot of the data can be smoothed with a Trendline; I've never done it, though.

I prefer to show the raw data; as a friend who is a high-level microwaves engineer says, "Smoothing is cheating".

Kerry VK2TIL.

things are working a little better, or were, up until I swapped the power and
ground rebuilding the cable I can't find. Glad I had a current limited supply
feeding it! Things are less noisy now that the cable it tighter in the socket.

Another question I have is if it is possible to get an average only line? Like
on a scope when you hit average in acquisition. I can do the average
accumulate and put the average line up, but I want something that runs for
like 10 minutes averaging all along the way. I'll keep looking.

Jerry

I posted recently in the VNWA group about building my N2PK;

/g/VNWA/message/22644

I'd been learning about vector measurements, first with a home-made noise bridge and then with a lucky purchase of a HP8405A vector voltmeter, when I heard about the N2PK; I didn't now what a VNA was but I thought I should get one.

Like you, it was a memorable experience; for you, it's a wonderful way to remember your father.

Although I never met him, I still remember Art Allen who sent me the parts; he was so pleasant and patient with a mug like me, and even sent some more parts when I found that one lot, of resistors I think, was the wrong value.

The N2PK was indeed a unique experience; Jeff AC0C said it well in a recent post;

"... every N2PK that comes into existence should be celebrated. They are increasingly rare beasties".


Kerry VK2TIL.

I forgot to label the frequency axis of my chart; see attached.

I should also have said, in relation to the power supply, that there is a 5v linear regulator on the rear panel of the VNA; the 12v and 5v lines are protected by "crowbar" over-voltage circuits; see attached.

Kerry VK2TIL.

Hi Kerry,

Very nicely built set-up! The supply internals look great.
I started building the VNA and RF-IV adapter just over 10 years ago, and I was a bit naive.
Knowing what I know now, I too would have gone with a linear supply.
I will modify at some point, and post the results.

It was the largest home-built project I'd attempted at the time (and might still be!). I couldn't have done it without the help from everyone on the N2PK mailer. My father was around then, and even though he didn't know what it was, he loved seeing it being assembled, with the careful coax plumbing inside and all, so the VNA project is a whole nice memory for me.

________________________________
From: [email protected] <[email protected]> on behalf of Kerry via groups.io <planningpower@...>
Sent: 29 January 2024 02:24
To: [email protected] <[email protected]>
Subject: Re: [N2PK-VNA] RFIV vs bridge

G'day Shabaz.

I did some measurements a few days ago on a 10k metal-film leaded resistor; this topic energised me this morning to process the S11 file.

See attached.

It's clear that the RF I-V fixture does not do well at low frequencies; its function is to provide measurements of low and high impedances.

Mine works from 0.5 - 100 MHz.

I haven't used it for low impedances; I have a project that requires measurements down to milliohms so I might use it there.

When I built my N2PK, I built a separate linear power supply; I'm old-fashioned and I don't feel entirely comfortable with switching supplies, although I admit that they have improved greatly over the years.

Photos of my home-made "wall-wart" :) are attached.

Regards, Kerry VK2TIL.

G'day Shabaz.

I did some measurements a few days ago on a 10k metal-film leaded resistor; this topic energised me this morning to process the S11 file.

See attached.

It's clear that the RF I-V fixture does not do well at low frequencies; its function is to provide measurements of low and high impedances.

Mine works from 0.5 - 100 MHz.

I haven't used it for low impedances; I have a project that requires measurements down to milliohms so I might use it there.

When I built my N2PK, I built a separate linear power supply; I'm old-fashioned and I don't feel entirely comfortable with switching supplies, although I admit that they have improved greatly over the years.

Photos of my home-made "wall-wart" :) are attached.

Regards, Kerry VK2TIL.

Hi Kerry,
Impressive!
I don't achieve that level. I tried a 5.1k resistor just now too, the charts are set to 10 ohms per division (same as yours), with 1-60MHz range total. I am using ADC speed:9, and ADC Avg:5.
The blue trace is a single sweep, and the red trace is average of 36 sweeps.
There's a chance it could be improved by changing my power supply (I'm using just an average AC/DC converter 5V barrel-plug output brick (Meanwell GST25B05-P1J<>), and inside the N2PK enclosure I created the 10V rail with an ADM660 switched-capacitor voltage doubler chip. That can't be too wise, I should try swapping it out one day. I might also try with a linear 5V PSU at some point, but nevertheless the results seem acceptable for now, just a few tens of ohms noise at such a high DUT resistance.


[cid:6e197815-ba88-496f-92e2-5f27e245fc02]
[cid:48575797-959d-4f89-9a17-76fe560a4bda]


________________________________
From: [email protected] <[email protected]> on behalf of Kerry via groups.io <planningpower@...>
Sent: 28 January 2024 21:07
To: [email protected] <[email protected]>
Subject: Re: [N2PK-VNA] RFIV vs bridge

Check that Instrument Mode is set to Reflection and RF I-V; Transmission mode should not be used with RF I-V mode.

Also check that calibration mode is set to Response Calibration.

These settings should be applied automatically when RF I-V mode is selected but they may have been inadvertently changed.

Also check that the two co-axial cables are not transposed; I once did this when I first used the fixture with my VNWA.

As I recall, the results were incorrect but, surprisingly, "in the ballpark", probably due to the awesome power of calibration.

There was a discussion of the RF I-V fixture some years ago on the old N2PK Yahoo group; Bill Carver was having a problem with his, I think.

To assist, I did some measurements of a 5k1 0805 resistor and obtained results as seen in the attachment; Paul opined that this was very good, the kind of result that might be expected.

Kerry VK2TIL.

Check that Instrument Mode is set to Reflection and RF I-V; Transmission mode should not be used with RF I-V mode.

Also check that calibration mode is set to Response Calibration.

These settings should be applied automatically when RF I-V mode is selected but they may have been inadvertently changed.

Also check that the two co-axial cables are not transposed; I once did this when I first used the fixture with my VNWA.

As I recall, the results were incorrect but, surprisingly, "in the ballpark", probably due to the awesome power of calibration.

There was a discussion of the RF I-V fixture some years ago on the old N2PK Yahoo group; Bill Carver was having a problem with his, I think.

To assist, I did some measurements of a 5k1 0805 resistor and obtained results as seen in the attachment; Paul opined that this was very good, the kind of result that might be expected.

Kerry VK2TIL.

I found my RFIV box but not the cable, so I wired it up and its working. It seems noisy. Just measuring my load references I get about +/- 1db noise. If I use a bridge, I get a ruler flat line. Also, changing references dont change the noise on the RFIV. The noise peaks are still there so its in the cables or RFIV unit. Granted I'm not using the top of on the box as I haven't found it yet. I have some cheap reference loads and when I use them the noise shifts with their readings. I have to play with it more, of course.

Any config errors that can cause the noise to be higher in RFIV mode?

thanks

Jerry

I found my RFIV box, but not the top or the connecting cable. I'll post a picture in the morning.

I also found an unpopulated dual detector board. I think it is the latest, but would have to verify it. If someone promises to build it, you can have the board and a decent oscillator, I think they are Connor-Winfield 25ppm, I can't remember, for the cost of shipping from California. In the US I would ship it flat rate, I think I have an extra DDS chip too, and if so, I would throw that in.

I have an oscillator that takes a 10Mhz reference input and outputs something above 126Mhz. It's low phase noise, something around -163dBc @100Khz but I never tested it. I wanted to build that 2nd VNA and I have it about 75% populated. Instead of a premium oscillator (one of the Connor-Winfields) and a better DDS, I think I would use LT3045 and LT3094 ldo regulators, and this oscillator that takes a reference. Those regulators are incredible. They will clean-up somewhat dirty supplies. I'm thinking about taking my VNA apart and putting them in. I can just jumper around the regulators that are in there. Back when I built my VNA, we were testing them to see how low the noise floor was and I think using the LT regulators, a better oscillator and a new DDS would be about the limit.

Anyway, I'm going to test my RFIV board in the morning if I can find the cable. I put a connector on the back and would hate have gone thru that trouble and not use it. I wish I could find the top as it had a cutout shield machined into it.

Speaking of home made test equipment, if you want to really go upscale, go here and build Andrew's Phase Noise test system. I did and it is amazing. He's added other features to it now.
In case you are wondering, if you get the Spartan board for around $500, and that's possible, then the rest of the setup will run another $500 by the time you get the ADC board and a pair of low phase noise reference oscillators. So for around $1000USD, you get a board that can outperform many PN test setups that are 10X expensive if not more.

Jerry

yes of course, I know what it is. I just have to find my original. I had two of everything. The plan was to build a second one using a lower phase noise oscillator and a better DDS. I got the 2nd about half build a while back and stalled for some reason. I think that was around when I broke my arm and couldn't solder for quite a while.

The RF-IV is one of the most frequent things I use with the N2PK-VNA. It's really excellent.
Here's a photo of my setup:

<!AlgR3ZgXGd0MwBRu2dqu9KfOE--y?e=SkxVmb>
[]
[]20240127_015755.jpg

I used RJ45 connectors to attach the RF-IV control pins to the N2PK VNA.

I use the RF-IV feature for exploring what components look like, since I don't have any other general-purpose impedance meter!
My N2PK VNA and the RF-IV sensor were created with the most recent boards (V5.x) that Ivan Makarov offered, with the built-in USB interface.
I'm quite interested in the VNWA, I might buy that one day, but I really? like the N2PK VNA. It was great value-for-money, even though it took me about a year to assemble it, acquiring the enclosure bits, cables, etc.

My N2PK VNA has engraved connector legends, they might be faintly visible in the photo. I got those done at such a low cost; in the UK, there are stores nationwide called Timpson; they do key cutting, and also wristwatch and baby spoon engraving etc., so I just asked the kid working there if he wouldn't mind engraving onto my alu box, and he did a fantastic job.


________________________________
From: [email protected] <[email protected]> on behalf of Kerry via groups.io <planningpower@...>
Sent: 27 January 2024 01:51
To: [email protected] <[email protected]>
Subject: Re: [N2PK-VNA] photo

As Kurt said; the RF I-V PCB and parts (some common parts such as resistors are not included).

I still use the one I built years ago;

[img][/img]

(Hope the photo comes-through; I haven't used PostImage for years).

Its virtue lies in the measurement of low & high impedances, well out of any VNA's "comfort zone"; here is a measurement of a 10k 1/4-watt leaded resistor;

RF I-V fixture;

[img][/img]

Normal reflection bridge;

[img][/img]

Its frequency range is limited by the ferrite transformers; mine has a range from 0.5MHz to 100MHz.

The supplied ferrites are #61 material; I have an unbuilt kit like yours and may, one day, build it with a higher-permeability material to extend the lower limit.

Building the transformers requires some patience;

[img][/img]

and housing requires some thought as the connectors are irregularly-placed.

Kerry VK2TIL.

As Kurt said; the RF I-V PCB and parts (some common parts such as resistors are not included).

I still use the one I built years ago;

[img][/img]

(Hope the photo comes-through; I haven't used PostImage for years).

Its virtue lies in the measurement of low & high impedances, well out of any VNA's "comfort zone"; here is a measurement of a 10k 1/4-watt leaded resistor;

RF I-V fixture;

[img][/img]

Normal reflection bridge;

[img][/img]

Its frequency range is limited by the ferrite transformers; mine has a range from 0.5MHz to 100MHz.

The supplied ferrites are #61 material; I have an unbuilt kit like yours and may, one day, build it with a higher-permeability material to extend the lower limit.

Building the transformers requires some patience;

[img][/img]

and housing requires some thought as the connectors are irregularly-placed.

Kerry VK2TIL.