Hello all,

I am in the process of attempting to build a Cohn Crystal Filter. I will be testing every crystal in the bunch using a modified version of the Colpitts crystal oscillator circuit. I want to use my handy-dandy NanoVNA SAA2 to measure each of the crystals to find the best "match" for the filter. However to my disappointment, I read on the www.nanorfe.com website the following disclaimer:

"Note: All V2 versions switch on and off the test signal rapidly and can not measure crystals or be used as a signal generator. See FAQ. For measuring crystals we recommend using an original NanoVNA by edy555."

It sounds like all of the version 2 models can't measure a crystal alone due to the limitations of the hardware. I couldn't find out if that was still true when the crystal was placed into a Colpitts test circuit and then connected to the NanoVNA for accurate measurement. I tried to do some research on this topic but couldn't find the answer I needed.

What do you think?

73, John W5EGA

Hi John,
I don't know (yet- mine just arrived an hour ago) whether it can measure a crystal, but I would think you also need to measure the crystal's motional parameters, L1 and C1, R and the crystal holder's capacity in order to design a proper filter- at least that is the method I have always used.

Dale W4OP

Hello John,

If you are willing to install the ojisan firmware to can measure crystals, parameters are calculated:
See: https://www.rudiswiki.de/wiki9/nanoVNA-V2-ojisanFirmware#LC.2C_Crystal_series_resonant_measurement

The firmware can be downloaded at chapter :
https://www.rudiswiki.de/wiki9/nanoVNA-V2-ojisanFirmware#Screen_capture

73, Rudi DL5FA

It's a firmware limitation. There was a firmware release from a Japanese
developer
for the first V2 that can be used for measuring crystals.

On Fri, 29 Jan 2021 at 22:10, <W5EGA@outlook.com> wrote:

> Hello all,
>
> I am in the process of attempting to build a Cohn Crystal Filter. I will
> be testing every crystal in the bunch using a modified version of the
> Colpitts crystal oscillator circuit. I want to use my handy-dandy NanoVNA
> SAA2 to measure each of the crystals to find the best "match" for the
> filter. However to my disappointment, I read on the www.nanorfe.com
> website the following disclaimer:
>
> "Note: All V2 versions switch on and off the test signal rapidly and can
> not measure crystals or be used as a signal generator. See FAQ. For
> measuring crystals we recommend using an original NanoVNA by edy555."
>
> It sounds like all of the version 2 models can't measure a crystal alone
> due to the limitations of the hardware. I couldn't find out if that was
> still true when the crystal was placed into a Colpitts test circuit and
> then connected to the NanoVNA for accurate measurement. I tried to do some
> research on this topic but couldn't find the answer I needed.
>
> What do you think?
>
> 73, John W5EGA
>
>
>
>
>
>

First thank you all for your answers...

Dale: "I don't know (yet- mine just arrived an hour ago) whether it can measure a crystal, but I would think you also need to measure the crystal's motional parameters, L1 and C1, R and the crystal holder's capacity in order to design a proper filter- at least that is the method I have always used."

I am planning to use two different methods (NanoVNA and G3UUR) to analyze the batch of crystals I ordered to determine their holder capacitance, motional capacitance, motional inductance, series resonant frequency, and the parallel resonant frequency in order to notionally "match" a set for a SSB filter and a second set for a CW filter. I am putting together a modified Colpitts oscillator circuit based off the directions on Jeff's (KK9JEF) blog: https://kk9jef.wordpress.com/tag/crystal/.

My goal is/was to use the NanoVNA to double check the G3UUR method for better peace of mind with the results. I want to be as accurate as possible with my very modest (but growing) home hobby lab. Since this is the first time the new test circuit will be used I thought it was best to double check and verify my results. So for this reason I will build a thru circuit for the NanoVNA and use the Ojisan firmware. I am currently using and prefer the "blackmagic" FW and will roll back to it after these tests.

Thank you all for your suggestions, comments, and help!

***Provided it doesn't violate any of the rules of this group, I might also come back and post the results and details of my measurements for some of the more advanced users in this group to critique my methods and findings.

73, John W5EGA

Excellent- let us know your results John. The AADE site has a convenient filter synthesizer.

Dale W4OP

On Sat, Jan 30, 2021 at 07:45 PM, <W5EGA@outlook.com> wrote:
>
> you also need to measure the crystal's motional parameters, L1 and C1, R
>
Hello John,

With the ojisan firmware you can measure those parameters, see:
https://www.rudiswiki.de/wiki9/nanoVNA-V2-ojisanFirmware#LC.2C_Crystal_series_resonant_measurement

If you select the right frequency span, you get automatic the values for:
Fs
Fp
Ls
Cs
Cp
Rs
Q

73, Rudi DL5FA

would be superb is one clever person would add those fuctions to the
other fw

... dislord would be a canditate for that ... wink wink:-)

dg9bfc sigi

Am 30.01.2021 um 22:54 schrieb reuterr@web.de:

> On Sat, Jan 30, 2021 at 07:45 PM, <W5EGA@outlook.com> wrote:
>> you also need to measure the crystal's motional parameters, L1 and C1, R
>>
> Hello John,
>
> With the ojisan firmware you can measure those parameters, see:
> https://www.rudiswiki.de/wiki9/nanoVNA-V2-ojisanFirmware#LC.2C_Crystal_series_resonant_measurement
>
> If you select the right frequency span, you get automatic the values for:
> Fs
> Fp
> Ls
> Cs
> Cp
> Rs
> Q
>
> 73, Rudi DL5FA
>
>
>
>
>

Hello all,

Update and a quick question:

I got the G3UUR Colpitts Crystal Oscillator circuit working and was able to get good readings for my set of 25 crystals (9MHz). I am using a cheap Ebay frequency counter (a clone of the PLJ-8LED) and verified its accuracy with both a RTL-SDR dongle and a SI5351 board that is calibrated to 10MHz.

I also set up a test jig (with the crystal under test being analyzed in both series and shunt) for use with my NanoVNA SAA2. I flashed the Ojisan firmware without any issues, set the appropriate frequency range, and performed a SOLT calibration. However when I use the "S21 XTAL" setting I get several negative results. (Ls, Cs, Cp, Rs, and Q are all negative values). Is that normal or am I doing something wrong?

BTW the Ojisan firmware is not compatible with the 4" screens and only uses the upper left portion of the screen - which makes it tough on the eyes. NanoVNA saver does not have a XTAL setting so using it via USB Mode is not possible either.

Pictures of my test circuits and the overall results to follow in the next day or so.

73, John W5EGA

ojisan fw also exist for 4 inch

greetz sigi dg9bfc

Am 06.02.2021 um 06:33 schrieb W5EGA@outlook.com:

> Hello all,
>
> Update and a quick question:
>
> I got the G3UUR Colpitts Crystal Oscillator circuit working and was able to get good readings for my set of 25 crystals (9MHz). I am using a cheap Ebay frequency counter (a clone of the PLJ-8LED) and verified its accuracy with both a RTL-SDR dongle and a SI5351 board that is calibrated to 10MHz.
>
> I also set up a test jig (with the crystal under test being analyzed in both series and shunt) for use with my NanoVNA SAA2. I flashed the Ojisan firmware without any issues, set the appropriate frequency range, and performed a SOLT calibration. However when I use the "S21 XTAL" setting I get several negative results. (Ls, Cs, Cp, Rs, and Q are all negative values). Is that normal or am I doing something wrong?
>
> BTW the Ojisan firmware is not compatible with the 4" screens and only uses the upper left portion of the screen - which makes it tough on the eyes. NanoVNA saver does not have a XTAL setting so using it via USB Mode is not possible either.
>
> Pictures of my test circuits and the overall results to follow in the next day or so.
>
> 73, John W5EGA
>
>
>
>
>

On Sat, Feb 6, 2021 at 07:33 AM, <W5EGA@outlook.com> wrote:

> BTW the Ojisan firmware is not compatible with the 4" screens and only uses
> the upper left portion of the screen
>
Hello John,

I have attached the latest 4 inch LCD ojisan firmware (480x320 ST7796) for you.

73, Rudi DL5FA

Thank you for the firmware information! I didn't see that before. I will load it up later today and give it a go.

Here are a couple pictures of my test jig setup for the NanoVNA. Please let me know if you have any suggestions or critiques as I don't want to induce any more problems into the testing. I also attached a couple of pics of the G3UUR test jig and setup.

73, John W5EGA

Hello everyone again,

Last update for the day... I am attaching an excel spreadsheet that I setup (to simplify the "maths") and will automatically compute the motional capacitance and motional inductance based on the measured and input values. I hope that it may be of use to someone else along the way. The file is not copyrighted and is freely shared with anyone that wishes to use it. I did however protect the spreadsheet - but only to keep someone (yeah - like me) from boinking up the preconfigured equations in the spreadsheet. The legend at the bottom of the spreadsheet lays out the formulas used and includes the format of each of the measurement values.

Now I'm off to flash the 4" screen version and play around with the Ojisan Firmware's XTAL measurement and compare the two findings.

73, John W5EGA

Question for anyone that has used the XTAL measurement in the Ojisan F/W. Does the parallel resonant frequency (Fp) lack the same fine resolution as the series resonant frequency (Fs)? Am I doing something wrong? Is it something to do with a setting in the NanoVNA? Is it something to do with my test jig? (All are very possible given I am still learning here.)

For example when measuring a few crystals I will get the following reading:

Crystal # Fs Fp
1 8.996854 9.014400 ... ... ... ...
2 8.996862 9.014600 ... ... ... ...
3 8.996755 9.014800 ... ... ... ...

All of the readings for the Fp are in the hundreds of hertz resolution but the Fs is down to single hertz resolution. Splitting hairs I know... but I would like to know why I am getting the readings I am and insure that I am not to the cause for the resulting differences. If it is a limitation of the NanoVNA I am completely fine with the results. Again, I would just like to know why.

73, John W5EGA

Hans Summers, G0UPL said it best in one of his videos regarding test equipment and accuracy results, "A man with one clock always knows what time it is. But a man with two clocks is never really sure. It only gets worse when you have three or four clocks."

On Tue, Feb 9, 2021 at 06:00 PM, <W5EGA@outlook.com> wrote:
>
> Crystal # Fs Fp
> 1 8.996854 9.014400 ... ... ... ...
> 2 8.996862 9.014600 ... ... ... ...
> 3 8.996755 9.014800
>
When you have a span of 30 kHz, the resolution will be:
span = 9020 - 8990 kHz = 30 kHz = 100 = 300 Hz resolution
or
span = 9020 - 8990 kHz = 30 kHz = 200 = 150 Hz resolution

Raw Bandwidth = 300 Hz

Showing a frequency down to Hertz numbers, can only be explained
with an integer to floating point conversion, for my understanding.

See for more examples:
https://www.rudiswiki.de/wiki9/nanoVNA-V2-ojisanFirmware#LC.2C_Crystal_series_resonant_measurement

Example with a Hertz number in Fp:
ojisan_crystal_71.5MHz_analyze_series_parallel_X2.jpg

73, Rudi DL5FA

Thank you Rudi for that information! I am going to move forward and select the best set of 8 crystals to start with on my SSB filter.

On Wed, Feb 10, 2021 at 04:51 AM, <W5EGA@outlook.com> wrote:
>
> I am going to move forward and select the best set of 8 crystals to start with
> on my SSB filter.
>
Hello John,
I forgot to mention 2 parameters of the ojisan firmware for a finer frequency resolution, see menu:
1. STIMULUS
SWEEP POINTS
Input for example: 201
2. DISPLAY
RBW (Raw Band Width)
RBW 1k, 300 Hz, 100 Hz, 30 Hz, 10 Hz (select)

I have not jet evaluated the RBW respective the crystal measurement.
Maybe you can try it, when you are already working on that topic ;-)

73, Rudi DL5FA