Simen,
Some of the other replies you have received are quite informative. If you want an in depth technical explanation on series and parallel resonant circuits and how the Q of the LC components affects the resonant frequency it was well covered by the Terman in his Radio Engineers handbook in 1943. You can download his book (out of copyright now) and find what you are looking for starting on page 135.
Roger
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
Why the big concern over the cal standards supplied with the NANOs? To 1.5 GHz, I have compared those to (expensive) HP cal standards using the 8753C and found good agreement. Above 1 GHz, your mileage may vary. I have and evaluated sets from the original NANO through the NANO-V2N. Dave - W?LEV On Fri, Mar 26, 2021 at 7:23 PM Roger Need via groups.io <sailtamarack= [email protected]> wrote: On Fri, Mar 26, 2021 at 08:30 AM, jmr wrote:
I'll try testing a few networks over the weekend with it. I also had a read of
the whole of this thread and the test fixture setups described at the start of
the thread aren't going to be good enough for critical measurements like this.
They are absolutely fine for measuring components or circuits for capacitance
or resistance or inductance but to give a VNA a fighting chance you absolutely
need to have a good calibration kit and fixture. In my case I absorb the
fixture and calibration hardware together so I don't have to do any
de-embedding. This avoids some measurement uncertainties. My fixture/cal kit
uses parts rated to 18GHz (18GHz for normal use rather than metrology) and it
is almost self correcting. I do still have to include a tiny delay correction
which I do via my PC tools.
jmr,
Manfred and I have both posted jigs that can be easily built and operated
with software that is available to everyone. Yes there are limitations to
what these jigs can do but it helps others who want to get started
measuring components like inductors, capacitors and filters.
In your posts I don't recall that you ever posted a photo of your test jig
and cal loads. Your software is something you wrote for your own use and
is not available to others.. So your posts may show what the NanoVNA is
capable of doing with custom hardware/software but there is no way anyone
can duplicate your results or know how to improve their own setup.
Roger
-- *Dave - W?LEV* *Just Let Darwin Work*
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
On Fri, Mar 26, 2021 at 08:30 AM, jmr wrote: I'll try testing a few networks over the weekend with it. I also had a read of
the whole of this thread and the test fixture setups described at the start of
the thread aren't going to be good enough for critical measurements like this.
They are absolutely fine for measuring components or circuits for capacitance
or resistance or inductance but to give a VNA a fighting chance you absolutely
need to have a good calibration kit and fixture. In my case I absorb the
fixture and calibration hardware together so I don't have to do any
de-embedding. This avoids some measurement uncertainties. My fixture/cal kit
uses parts rated to 18GHz (18GHz for normal use rather than metrology) and it
is almost self correcting. I do still have to include a tiny delay correction
which I do via my PC tools.
jmr, Manfred and I have both posted jigs that can be easily built and operated with software that is available to everyone. Yes there are limitations to what these jigs can do but it helps others who want to get started measuring components like inductors, capacitors and filters. In your posts I don't recall that you ever posted a photo of your test jig and cal loads. Your software is something you wrote for your own use and is not available to others.. So your posts may show what the NanoVNA is capable of doing with custom hardware/software but there is no way anyone can duplicate your results or know how to improve their own setup. Roger
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
I had another play this afternoon and it appears the latest FW doesn't support the edelay command in the same way as the older FW. Therefore, I can't calibrate the nanoVNA to the same level of accuracy as I can with the old FW when using my PC tools. There is actually a slight mathematical error in the previous graph even though it looks impressive. The ESR traces in the previous graph are ever so slightly low in value because of this error. I did some more testing and I've had to go back to the 060 FW because I can't calibrate out the fine errors in my cal kit with the latest FW. This is at its most significant when I try and measure 2 port networks for s11 s12 s21 s22 because I can't calibrate as accurately any more.
|
This may have been stated in slightly different terms, but: The circuit Q is the culprit. Lower Q results in dragging the resonant frequency lower. Maybe get yourself some LN2 and immerse the parallel tuned circuit and measure resonant frequency. Then progressively measeure resonant frequency and temperature as the circuit warms. Of course, you can take it the other way and increase the temperature while noting the resonant frequency and temperature. You can also simulate this by adding / subtracting loss - resistance - in series with each component. Dave - W?LEV On Fri, Mar 26, 2021 at 11:52 AM Simen Tobiassen <simen@...> wrote: Simulations show Currents do NOT cancel at the Parallel LC-resonance
frequency, contrary to what we learn is one of the main characteristics of
parallel resonance circuits. Why?
Testing various LC combinations, the two Currents always meet and cancel
at a frequency lower than the XL = XC frequency.
?
Micro-Cap 12 (Transient Analysis; Oscilloscope) shows LC-resonans
frequency / 1.034 ¡Ö frequency of current cancellation. Moving the input
frequency up and down from this point you can find the perfect frequency
where the two currents have the same amplitude = cancelling. Down in
frequency makes the current through the inductor stronger, up makes the
current through the capacitor stronger.
I mostly use RS = 10m Ohm in both the Capacitor and in the low induction
Coil. Same is true with RS = 0 Ohm.
LC wise they are perfect(ly fake ;) 0 parasitic L and C. Anyway, the
simulation makes no difference; Capacitance added to the inductor gives the
same result as if the C was in a separate capacitor; It just moves the
LC-resonance to a lower frequency. And the current-cancelling frequency
also thereby moves down.
This also correlate to the readings on my nanoVNA; two different
frequencies, and I do not understand how to combine them into an ¡°official¡±
Parallel LC-resonance circuit.
--
Simen Tobiassen
-- *Dave - W?LEV* *Just Let Darwin Work*
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
Here's a quick preview. This is a repeat of the ESR test of the 1111 560pF cap + 0.333R series resistor. Also included is an attempt to measure the raw ESR of the 1111 560pF cap and the datasheet suggests it will be somewhere around 0.03R at VHF.
This was taken with 100Hz bandwidth (and no averaging) with DiSlord's 1.0.50 FW and the results are really impressive across 10Mhz to 50MHz. It's obviously still not going to be possible to measure the ESR of caps like this down at 100kHz but I think the nanoVNA is going to at least be capable of doing comparative tests between different types of 470pF cap across 5MHz to 50MHz as long as you don't expect metrology grade accuracy.
Things will get tougher for ceramic caps with smaller value so these would have to be tested at higher frequencies or (preferably) some other measurement technique should be used for high Q caps with low capacitance values.
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
I managed to find Dislord's 1.0.50 firmware for the H. It was posted up onto this user group in the files section. I had previously been looking on Dislord's github page without success. The good news is that this firmware seems to be at least as good (maybe better?) than all previous firmware versions I've tried. It also works with my PC tools without freezing.
I'm hoping to keep this 1.0.50 firmware on my nanoVNA all the time now as it seems much nicer all round. Big thanks to DiSlord for this!
I'll try testing a few networks over the weekend with it. I also had a read of the whole of this thread and the test fixture setups described at the start of the thread aren't going to be good enough for critical measurements like this. They are absolutely fine for measuring components or circuits for capacitance or resistance or inductance but to give a VNA a fighting chance you absolutely need to have a good calibration kit and fixture. In my case I absorb the fixture and calibration hardware together so I don't have to do any de-embedding. This avoids some measurement uncertainties. My fixture/cal kit uses parts rated to 18GHz (18GHz for normal use rather than metrology) and it is almost self correcting. I do still have to include a tiny delay correction which I do via my PC tools.
If I used the fixtures described earlier in this thread I'd expect to have poor results when doing critical testing like this. It might be possible to correct the limitations of those jigs electronically but why start from there?
|
He already gave this info, below:?? Micro-Cap 12
Google it - it's free windows s/w
toggle quoted message
Show quoted text
On Friday, March 26, 2021, 10:08:55 a.m. EDT, Mark A. Wilkinson <markwilkinson805@...> wrote: Hello Simen, Please let me know the name of the simulation software that you are use. Thanks in advance. Mark Wilkinson ________________________________ From: [email protected] < [email protected]> on behalf of Simen Tobiassen <simen@...> Sent: Friday, March 26, 2021 7:52 AM To: [email protected] < [email protected]> Subject: [nanovna-users] Why does the two currents in a parallel LC-resonance circuit cancel on a lower frequency than the LC-resonance? #coils #charts #test #teaching #traps #simulation #rant #problem #measurement #learning #circuit Simulations show Currents do NOT cancel at the Parallel LC-resonance frequency, contrary to what we learn is one of the main characteristics of parallel resonance circuits. Why? Testing various LC combinations, the two Currents always meet and cancel at a frequency lower than the XL = XC frequency. ? Micro-Cap 12 (Transient Analysis; Oscilloscope) shows LC-resonans frequency / 1.034 ¡Ö frequency of current cancellation. Moving the input frequency up and down from this point you can find the perfect frequency where the two currents have the same amplitude = cancelling. Down in frequency makes the current through the inductor stronger, up makes the current through the capacitor stronger. I mostly use RS = 10m Ohm in both the Capacitor and in the low induction Coil. Same is true with RS = 0 Ohm. LC wise they are perfect(ly fake ;) 0 parasitic L and C. Anyway, the simulation makes no difference; Capacitance added to the inductor gives the same result as if the C was in a separate capacitor; It just moves the LC-resonance to a lower frequency. And the current-cancelling frequency also thereby moves down. This also correlate to the readings on my nanoVNA; two different frequencies, and I do not understand how to combine them into an ¡°official¡± Parallel LC-resonance circuit. -- Simen Tobiassen
|
Re: Saving .s1p, .s2p, and Screenshots Directly from the nanoVNA-H4
Please read the Console Command document in the forum files section - it has all the commands.
Also, use the Nanovna-App application - it has a clock set command in setup.
Search the forum messages for more info - it's all there.
toggle quoted message
Show quoted text
On Friday, March 26, 2021, 9:56:31 a.m. EDT, Rui Valente <ct2ggx@...> wrote: Hello i install to a Xtal and now i can see the clock is running in to menu version...but how i can set the hour and date? A sexta, 26/03/2021, 13:35, Mario Vano <mvano@...> escreveu: I got my crystal from a cheap clock in my junk box that had died from
mechanical problems - the crystal worked great!
M
|
Hello Simen, Please let me know the name of the simulation software that you are use. Thanks in advance. Mark Wilkinson ________________________________ From: [email protected] < [email protected]> on behalf of Simen Tobiassen <simen@...> Sent: Friday, March 26, 2021 7:52 AM To: [email protected] < [email protected]> Subject: [nanovna-users] Why does the two currents in a parallel LC-resonance circuit cancel on a lower frequency than the LC-resonance? #coils #charts #test #teaching #traps #simulation #rant #problem #measurement #learning #circuit Simulations show Currents do NOT cancel at the Parallel LC-resonance frequency, contrary to what we learn is one of the main characteristics of parallel resonance circuits. Why? Testing various LC combinations, the two Currents always meet and cancel at a frequency lower than the XL = XC frequency. ? Micro-Cap 12 (Transient Analysis; Oscilloscope) shows LC-resonans frequency / 1.034 ¡Ö frequency of current cancellation. Moving the input frequency up and down from this point you can find the perfect frequency where the two currents have the same amplitude = cancelling. Down in frequency makes the current through the inductor stronger, up makes the current through the capacitor stronger. I mostly use RS = 10m Ohm in both the Capacitor and in the low induction Coil. Same is true with RS = 0 Ohm. LC wise they are perfect(ly fake ;) 0 parasitic L and C. Anyway, the simulation makes no difference; Capacitance added to the inductor gives the same result as if the C was in a separate capacitor; It just moves the LC-resonance to a lower frequency. And the current-cancelling frequency also thereby moves down. This also correlate to the readings on my nanoVNA; two different frequencies, and I do not understand how to combine them into an ¡°official¡± Parallel LC-resonance circuit. -- Simen Tobiassen
|
Many simulators apply default series R to L and perhaps C to prevent creating a singularity matrix. In some cases you can over ride the default.
|
Re: Saving .s1p, .s2p, and Screenshots Directly from the nanoVNA-H4
Hello i install to a Xtal and now i can see the clock is running in to menu
version...but how i can set the hour and date?
A sexta, 26/03/2021, 13:35, Mario Vano <mvano@...> escreveu:
toggle quoted message
Show quoted text
I got my crystal from a cheap clock in my junk box that had died from
mechanical problems - the crystal worked great!
M
|
If you write the node equations for a parallel LC with an R in series with the L
wo= 1/¡ÌLC ¡Á ¡Ì1-R^2 ¡Á C/L
Yes, the series R will perturb the resonance point from the expected value.
|
Thanks, interesting! How is that computed?
But the result however, is the same with Serial Resistance = 0 Ohm in capacitor and in coil; Except I get beats on top of the 888630 Hz waves, probably modulated with the 918881 LC-resonance.
--
Simen Tobiassen
|
Re: Saving .s1p, .s2p, and Screenshots Directly from the nanoVNA-H4
I got my crystal from a cheap clock in my junk box that had died from mechanical problems - the crystal worked great!
M
|
root (LC) is only true if R=0
If any R is present either in series with L or C, The root (LC) is modified... downward in freq...
|
Simulations show Currents do NOT cancel at the Parallel LC-resonance frequency, contrary to what we learn is one of the main characteristics of parallel resonance circuits. Why?
Testing various LC combinations, the two Currents always meet and cancel at a frequency lower than the XL = XC frequency.
?
Micro-Cap 12 (Transient Analysis; Oscilloscope) shows LC-resonans frequency / 1.034 ¡Ö frequency of current cancellation. Moving the input frequency up and down from this point you can find the perfect frequency where the two currents have the same amplitude = cancelling. Down in frequency makes the current through the inductor stronger, up makes the current through the capacitor stronger.
I mostly use RS = 10m Ohm in both the Capacitor and in the low induction Coil. Same is true with RS = 0 Ohm.
LC wise they are perfect(ly fake ;) 0 parasitic L and C. Anyway, the simulation makes no difference; Capacitance added to the inductor gives the same result as if the C was in a separate capacitor; It just moves the LC-resonance to a lower frequency. And the current-cancelling frequency also thereby moves down.
This also correlate to the readings on my nanoVNA; two different frequencies, and I do not understand how to combine them into an ¡°official¡± Parallel LC-resonance circuit.
--
Simen Tobiassen
|
Re: Saving .s1p, .s2p, and Screenshots Directly from the nanoVNA-H4
Larry,
Thanks for the info on how to use the SD card. I have not used the time command yet as thought I should install the crystal first.
Ken, NU4I
Sent from my iPod
toggle quoted message
Show quoted text
On Mar 25, 2021, at 19:42, Larry Rothman <nlroth@...> wrote:
?The 32KHz xtal is only needed if you want a real time of day clock on your unit for accurate dates on data saved to the SD card.
Without the xtal, you'll need to turn on the Nano, set the time using console commands and then use the sdcard.
Many older PC motherboards use the narrow 32KHz xtal and that's where I got mine from for my units. They work fine and old junked motherboards are easy to come by and usually free.
On Thu., 25 Mar. 2021 at 4:02 p.m., Fred, W9KEY<fred.schwierske@...> wrote: So just to clarify - are the new H4 units from R&L Electronics presently shipping with all required hardware to use a (user supplied) microSD card?
1) Or must the above discussed 32KHz crystal first be installed?
2) And do units currently shipping from R&L have cases which include the card slot, or does the user have to cut an access slot?
|
Re: Saving .s1p, .s2p, and Screenshots Directly from the nanoVNA-H4
The 32KHz xtal is only needed if you want a real time of day clock on your unit for accurate dates on data saved to the SD card.?
Without the xtal, you'll need to turn on the Nano, set the time using console commands and then use the sdcard.?
Many older PC motherboards use the narrow 32KHz xtal and that's where I got mine from for my units. They work fine and old junked motherboards are easy to come by and usually free.?
On Thu., 25 Mar. 2021 at 4:02 p.m., Fred, W9KEY<fred.schwierske@...> wrote: So just to clarify - are the new H4 units from R&L Electronics presently shipping with all required hardware to use a (user supplied) microSD card?
? 1)? Or must the above discussed 32KHz crystal first be installed?
? 2)? And do units currently shipping from R&L have cases which include the card slot, or does the user have to cut an access slot?
|
Re: Pitfalls of measuring components with the NanoVNA
#measurement
So does mine but it still was a clone, nothing major, few resistors where there should be none and some wrong component values. Only way to be sure is to buy from Hugen store. On Thu, 25 Mar 2021 at 19:15, jmr via groups.io <jmrhzu= [email protected]> wrote: It was a long time ago but I did try and buy the Hugen version with the
blue PCB. I just checked and it does say nanoVNA-H v3.3 on the blue PCB and
the case and lettering are the same as the Hugen version. I'm not sure if
the Hugen version is a clone but at the time this seemed to be the one to
buy.
|
Roger Need
