Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On Wed, Oct 28, 2020 at 10:22 AM, Erik Kaashoek wrote:
The calibration load
For accurate S11 measurements of that microstrip at 1 GHz. you need a good reference plane and termination load. You can't put an SMA connector at the end of the microstrip and then use a SMA male terminator and get accurate results. I suggest a quality SMA connector for the input and a small 50 ohm SMD resistor soldered to the end of the microstrip. For calibrating the NanoVNA some female SMA SOL can be made to go on the end of the cable from the VNA. You may need to adjust the offset delay in NanoVNA Saver. Roger
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On 10/28/20 8:14 AM, Jose Mihotek via groups.io wrote: Roger & the group,
I stand corrected.The Z0 does change with freq.For coax, the change is small (<5%).For Ham purposes (HF), it is negligible.For VHF & UHF, if you use quality coax (Heliax), it is very negligible.For Microstip and Coplanar Waveguide, FR4 is a very poor choice. it is batch dependent. There is FR4 and there is? FR4...You should use Rogers4350B, it is much more stable and guaranteed.Alumina would be even better...
FR4 or XXXP is significantly cheaper than Rogers, though. There are plenty of applications where you just don't care. A microwave motion detector at 10GHz has short traces, and doesn't care what the loss happens to be.
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On Wed, Oct 28, 2020 at 08:45 AM, Erik Kaashoek wrote: Far end was terminated with nominal load. This was 20cm coax!
What did you use as a load for close to 50 +j0 at 1 GHz. ?? Roger
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On Wed, Oct 28, 2020 at 07:24 AM, Jim Lux wrote:
reflecting back the unterminated far end.
Far end was terminated with nominal load. This was 20cm coax! -- NanoVNA Wiki: /g/nanovna-users/wiki/homeNanoVNA Files: /g/nanovna-users/filesErik, PD0EK
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
Good point on FR4 Er vs frequency, I forget what reasonably affordable dielectrics do! Here are a couple of simulations of microstrip:
Er vs freq from (4.65 to 4.35 from 10MHz to 10GHz)
Modeled with Er = 4.65 - (log(freq)-1)*0.1
tan(d) = 0.016
55 mil wide microstrip, 30mil dielectric, 1.4mil trace thickness, 1um roughness
Series transmission line terminated in 50 ohms
S11 is FR4 with constant Er vs freq, S33 is (s11 of ) FR4 with the Er formula above. That's more difference than I expected, maybe you can in fact measure that on a NanoVNA.
The second plot is a 75mil wide microstrip line, a bit lower than 50 ohms, for another view at one one might see.
If I were to extract the impedance of a line using S11, I'd use some simulator that I could build a transmission line in and twiddle the Zo and loss vs frequency to get a good fit to measured data.
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
Roger & the group,
I stand corrected.The Z0 does change with freq.For coax, the change is small (<5%).For Ham purposes (HF), it is negligible.For VHF & UHF, if you use quality coax (Heliax), it is very negligible.For Microstip and Coplanar Waveguide, FR4 is a very poor choice. it is batch dependent. There is FR4 and there is? FR4...You should use Rogers4350B, it is much more stable and guaranteed.Alumina would be even better...
73
JoseN0GU
toggle quoted message
Show quoted text
On Wednesday, October 28, 2020, 10:34:24 AM EDT, Jim Lux <jimlux@...> wrote: On 10/28/20 5:09 AM, KENT BRITAIN wrote: ? Except for the high dollar RF materials, yes the Er of common fiberglass
PCB material does change with frequency!!
The Er = 4.4 measurement is made at 1000 Hz!?? Audio range.
By the time you get to 2.4 GHz is has dropped to the 3.8-3.9 range
I have made thousands of antennas on FR4 type materials. However, changing from 4.4 to 3.8 is going to show a Z change (ignoring the effects of loss, which are substantial) of sqrt(4.4/3.8) or about 7-8%. A 50 ohm line at 1 GHz would be a 54 ohm line at audio frequencies. 3.8 is sort of on the low side - but in any case, it depends on your specific brand and what the glass and resin contents are. It could easily vary +/- 1. Atmel's app note on designing transmission lines gives epsilon (@1GHz) above 4 for all 4 of their FR4. with dissipation factors (tan d) of 0.013 to 0.030
Kent WA5VJB
? ? ? On Wednesday, October 28, 2020, 2:47:03 AM CDT, Erik Kaashoek <erik@...> wrote:
?
? And here is the 0-2GHz scan of a 3cm coplanar stripline on FR4 that seems to confirm the impedance is changing with frequency
The marker is at 1GHz
|
Re: Selecting proper firmware
@Christian, thanks, I'll do that.
toggle quoted message
Show quoted text
On 10/28/20, Christian Zietz <czietz@...> wrote: Perhaps it's best to ask NanoVNA *V2* specific questions in the NanoVNA *V2*
group: /g/NanoVNAV2? They make it very clear over there
that "you will only get support on [that] forum". However, albeit without
any personal knowledge w.r.t. to the NanoVNA V2, it would say it's very
clear: if you have V2.2 hardware, you need the V2.2 firmware build.
Regards
Christian
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On 10/28/20 5:09 AM, KENT BRITAIN wrote: Except for the high dollar RF materials, yes the Er of common fiberglass
PCB material does change with frequency!!
The Er = 4.4 measurement is made at 1000 Hz!?? Audio range.
By the time you get to 2.4 GHz is has dropped to the 3.8-3.9 range
I have made thousands of antennas on FR4 type materials. However, changing from 4.4 to 3.8 is going to show a Z change (ignoring the effects of loss, which are substantial) of sqrt(4.4/3.8) or about 7-8%. A 50 ohm line at 1 GHz would be a 54 ohm line at audio frequencies. 3.8 is sort of on the low side - but in any case, it depends on your specific brand and what the glass and resin contents are. It could easily vary +/- 1. Atmel's app note on designing transmission lines gives epsilon (@1GHz) above 4 for all 4 of their FR4. with dissipation factors (tan d) of 0.013 to 0.030 Kent WA5VJB
On Wednesday, October 28, 2020, 2:47:03 AM CDT, Erik Kaashoek <erik@...> wrote:
And here is the 0-2GHz scan of a 3cm coplanar stripline on FR4 that seems to confirm the impedance is changing with frequency
The marker is at 1GHz
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On 10/28/20 3:43 AM, Erik Kaashoek wrote: Forgot the S11 plot of the bad coax
Yes, that's pretty typical for lossy coax - but you're not seeing the coax's impedance changing, you're seeing the terminal impedance changing as as the coax is different numbers of wavelengths long, reflecting back the unterminated far end. And since the loss increases with frequency, the "mismatch" is attenuated more and more with frequency. As they say, 100 feet of cheap coax in a bucket of water makes a fine dummy load at UHF.
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
On 10/28/20 12:46 AM, Erik Kaashoek wrote: And here is the 0-2GHz scan of a 3cm coplanar stripline on FR4 that seems to confirm the impedance is changing with frequency
The marker is at 1GHz
Is the line terminated in the nominal impedance? or are we just seeing the reflection from the end, with loss?
|
Re: Selecting proper firmware
Perhaps it's best to ask NanoVNA *V2* specific questions in the NanoVNA *V2* group: /g/NanoVNAV2? They make it very clear over there that "you will only get support on [that] forum". However, albeit without any personal knowledge w.r.t. to the NanoVNA V2, it would say it's very clear: if you have V2.2 hardware, you need the V2.2 firmware build. Regards Christian
|
Re: Selecting proper firmware
Mike4U's comment on github:
Indicates with the new firmware, there is a workaround by doing a
little dance using the on-board set-up.
IAE, still would be good to have some confidence in which firmware to
use. I am still at the "getting my feet wet" stage of using NanoVNA
and there seems to be so many variants out there.
And any useful information regarding the saver crashing with V2 issue
would be helpful as well. (It doesn't crash on my V1 nanovna.)
toggle quoted message
Show quoted text
On 10/28/20, OneOfEleven <cmoss296@...> wrote: I can't help with the firmware (yet), but if NanoVNA-saver is crashing it's
not the firmware causing it, it's NanoVNA-save itself (needs fixing).
|
Re: Selecting proper firmware
I can't help with the firmware (yet), but if NanoVNA-saver is crashing it's not the firmware causing it, it's NanoVNA-save itself (needs fixing).
|
Selecting proper firmware
I'd like to update my firmware in hopes of correcting a problem with
nanovna-saver crashing with my device.
In the "Version" display of my vna, it says "NanoVNA V2_2"
The current firmware is git-20200617-1a9a11d
When I check the versions page:
while all selections show 20201013, there appear to be variants as to
the files when I hover over the download links, eg,
nanovna-v2-20201013-v2_2.bin, or nanovna-v2-20201013-v2plus4.bin, the
operatives being, v2_2.bin or v2plus4.bin.
Since mine says "NanoVNA V2_2" I would guess that I should use the
v2_2.bin variant, which is the 3rd selection in the "Official"
variants section, which also says "No longer sold".
I would like to verify this with the group, as I am queezy about doing
an improper firmware update. I would also welcome any other input you
may have on the matter.
Thanks, K...
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
Hi Kent, nice to see you here. I have a couple of your pcb logper antennas (850-6500 MHz) that I use in my lab. Great antennas, real bargain!
Tip for buyers: buy 2 of them, then you can measure the gain of these antennas with your NanoVNA. No need for an expensive calibrated antenna if you need to do some serious measurements.
Reinier
Op 28-10-2020 om 13:09 schreef KENT BRITAIN:
toggle quoted message
Show quoted text
Except for the high dollar RF materials, yes the Er of common fiberglass
PCB material does change with frequency!!
The Er = 4.4 measurement is made at 1000 Hz!?? Audio range.
By the time you get to 2.4 GHz is has dropped to the 3.8-3.9 range
I have made thousands of antennas on FR4 type materials.
Kent WA5VJB
On Wednesday, October 28, 2020, 2:47:03 AM CDT, Erik Kaashoek <erik@...> wrote:
And here is the 0-2GHz scan of a 3cm coplanar stripline on FR4 that seems to confirm the impedance is changing with frequency
The marker is at 1GHz
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
Except for the high dollar RF materials, yes the Er of common fiberglass
PCB material does change with frequency!!
The Er = 4.4 measurement is made at 1000 Hz!?? Audio range.
By the time you get to 2.4 GHz is has dropped to the 3.8-3.9 range
I have made thousands of antennas on FR4 type materials.
Kent WA5VJB
toggle quoted message
Show quoted text
On Wednesday, October 28, 2020, 2:47:03 AM CDT, Erik Kaashoek <erik@...> wrote: And here is the 0-2GHz scan of a 3cm coplanar stripline on FR4 that seems to confirm the impedance is changing with frequency The marker is at 1GHz -- NanoVNA Wiki: /g/nanovna-users/wiki/homeNanoVNA Files: /g/nanovna-users/filesErik, PD0EK
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
|
Re: Measuring characteristic impedance of a stripline with a nanoVNA-H
|
Roger Need
