Re: Antenna simulation result different from experiment with NanoVNA
sonnet works superb and is quite easy to use ;-)
see picture (circular polarized patch ... a square with truncated corners)
dg9bfc sigi
many thanks for making that list of the free eda soft ... maybe i will try one or two of them (BESIDES SONNET)
Am 01.09.2022 um 16:42 schrieb Jim Lux:
toggle quoted message
Show quoted text
On 9/1/22 6:21 AM, Gary W9TD wrote:
EZNEC is now free.
EZNEC as a Method of Moments code that is derived from NEC is really a "wire" modeling code and doesn't do patches very well.
One can "grid" a patch, and put it over ground with soil properties set to match the substrate, but it doesn't have a way to do the ground plane on the other side of the substrate. (well, in NEC4 you could bury a grid of wires)).
NEC based codes will not do the matching sections or transmission line.
What the OP needs is a low cost tool that is designed for microstrip patches.
Ancient, but maybe PUFF (an early microwave circuit modeling package) would work?
At work we use HFSS, but that's sort of the antithesis of free.
I'd start with the list here:
Sonnet has a student version
|
Re: Antenna simulation result different from experiment with NanoVNA
On Thu, Sep 1, 2022 at 08:59 AM, Brian Howie GM4DIJ wrote:
The dielectric constant you chose might differ from the real one depending on
the material.
Many microwave materials publish TWO dielectric constant values; one which is used in standardized QC tests, and another one, the "design" value, which should be used for simluation. Be certain that you are using the correct one, the simulation tool will happily spit out beautiful, inaccurate plots. Also, how are you decoupling the antenna from the measurement instruments? Do they change as you move you hand along the feedline? If so, the feedline is not adequately decoupled. How big is your ground plane, if any is used? Is it adequate? An antenna's environment can mage a big difference. 73, Don N2VGU
|
Re: stand for the NanoVNA???
AT LEAST USE A PORT EXTENSION!!
a male female adaptor (port saver) ... if worn out just replace the adaptor and NOT the socket in the vna
any sma connector that you use often should have such an extender ... be it the vna or an sdr (funcube dongle, airpsy hf+, whatever)
so ... go any buy "a dozen" and put them on ALL of your sma sockets that you use more then 5 times a year :-)
dg9bfc sigi
Am 01.09.2022 um 17:27 schrieb F1AMM:
toggle quoted message
Show quoted text
You may have forgotten that you can put an extension cord of the length you want between the nanaoVNA box and the location where you are going to do the calibration. Even the imperfections of the extension will be compensated by the calibration. I did a test with a 93 ¦¸ cable.
Take advantage of this extension to change the type of connector. Switch to a female BNC socket for example, firmly fixed on a support. SMA are too fragile to be screwed and unscrewed too often.
|
Re: stand for the NanoVNA???
you shoud measeure a ht antenna like you would use them .. means hold the vna in your hand
the coupling to the body is needed on some if not all ht antennas (the transmiter needs something "to feed against" ... a virtual ground so to speak)
you see big changes if you hold the vna or not?? no wonder!! ... the case of the vna is to small for a ground on 70cm and even more on 2m ...
think about ....
dg9bfc sigi
Am 01.09.2022 um 20:20 schrieb n1evh via groups.io:
toggle quoted message
Show quoted text
I don't think I'm going to need any of those BNC to SMA
connector holders... as I have a few of these that I could
put onto the NanoVNA...
<>
or if you want them gold plated...
<>
These would sit right against the body of the NanoVNA.
I'm looking for a stand... that I can put the other end into
so that the antenna connectors (SMA, BNC, whatever) would
be facing up... to keep the HT antennas vertical. an not
touching anything when testing them. I was wondering if
anyone has made one or just designed one so it could be
printed out on a DC printer.
If you just stand the NanoVNA on its end and you have it
connected to a laptop or tablet with a USB cable... morning
the laptop or tablet around... will make the USB cable move
and most likely make the NanoVNA fall over. A plastic stand
to put the NanoVNA in... would keep it from falling over.
Mike - N1EVH
On 9/1/2022 12:46 PM, Dave Merrill wrote:
If you have access to a 3-D printer, there are a ton of designs for NanoVNA
cases, holders, connection extensions, etc. Here is but one:
Many public libraries have Maker Labs. You upload the design file to their
server and they notify you when it's done. My library charges a nominal fee
for the amount of materials used.
Dave N9ZC
On Thu, Sep 1, 2022, 9:28 AM F1AMM<18471@...>? wrote:
You may have forgotten that you can put an extension cord of the length
you want between the nanaoVNA box and the location where you are going to
do the calibration. Even the imperfections of the extension will be
compensated by the calibration. I did a test with a 93 ¦¸ cable.
Take advantage of this extension to change the type of connector. Switch
to a female BNC socket for example, firmly fixed on a support. SMA are too
fragile to be screwed and unscrewed too often.
--
F1AMM Fran?ois
De la part de n1evh via
Envoy¨¦ : jeudi 1 septembre 2022 16:10
|
Re: stand for the NanoVNA???
I don't think I'm going to need any of those BNC to SMA
connector holders... as I have a few of these that I could
put onto the NanoVNA...
<>
or if you want them gold plated...
<>
These would sit right against the body of the NanoVNA.
I'm looking for a stand... that I can put the other end into
so that the antenna connectors (SMA, BNC, whatever) would
be facing up... to keep the HT antennas vertical. an not
touching anything when testing them. I was wondering if
anyone has made one or just designed one so it could be
printed out on a DC printer.
If you just stand the NanoVNA on its end and you have it
connected to a laptop or tablet with a USB cable... morning
the laptop or tablet around... will make the USB cable move
and most likely make the NanoVNA fall over. A plastic stand
to put the NanoVNA in... would keep it from falling over.
Mike - N1EVH
toggle quoted message
Show quoted text
On 9/1/2022 12:46 PM, Dave Merrill wrote: If you have access to a 3-D printer, there are a ton of designs for NanoVNA
cases, holders, connection extensions, etc. Here is but one:
Many public libraries have Maker Labs. You upload the design file to their
server and they notify you when it's done. My library charges a nominal fee
for the amount of materials used.
Dave N9ZC
On Thu, Sep 1, 2022, 9:28 AM F1AMM<18471@...> wrote:
You may have forgotten that you can put an extension cord of the length
you want between the nanaoVNA box and the location where you are going to
do the calibration. Even the imperfections of the extension will be
compensated by the calibration. I did a test with a 93 ¦¸ cable.
Take advantage of this extension to change the type of connector. Switch
to a female BNC socket for example, firmly fixed on a support. SMA are too
fragile to be screwed and unscrewed too often.
--
F1AMM Fran?ois
De la part de n1evh via
Envoy¨¦ : jeudi 1 septembre 2022 16:10
|
Re: Antenna simulation result different from experiment with NanoVNA
You have to look closely at their data sheets, but they come up with i.d.? numbers that are not FR4.? That was a specific material no longer made..
But this is much like making a Xerox copy on your Cannon copier.? ? FR-4 has become a General term for any Epoxy/Fiberglass board material..??
Measuring the Er at a specific frequency can be fun.? ?When the material sample is big enough, I like to make a patch antenna that is expected to come out near the frequency of interest.? ?Usually I cut the patch with a razor knife, then measure it's actually frequency.? ?Trim if necessary.
Next I use Roger Cox's patch antenna program.? I play with the Er value in the program until it predicts the frequency I measured.? ? ? Next you can take the bandwidth you got from the sweep, and again backing into Roger's patch program, get a good number for the? Loss Tangent.? ? ?Kent
Here is Rogers website.? ?It says Patch 2.0, but it's really Patch 20
toggle quoted message
Show quoted text
On Thursday, September 1, 2022 at 11:32:32 AM CDT, W0LEV <davearea51a@...> wrote: Yes, I spaced the Rogers products.? Thank you for mentioning the Rogers 4350 substrate. FR-4 is available everywhere and is relatively inexpensive. What's more and being critical, both the dielectric constant and dielectric thickness of FR-4 board may vary across a large sheet.? Been there, done that being paid to force Fibre Channel "square waves" across 15-inches of FR-4 with no dispersion.? Bad things happen for EMC/RFI when the pre-emphasis is jacked (by the designers to make it function) in an attempt to accomplish that stated goal.? The GM of the division dictated absolutely bargain basement prices on every component of the system.? Ultimately, he got egg on his face and I was banned from working further with that division.? Standard practice with the present-day Corporate America:? Slay the messenger!! Dave - W?LEV On Thu, Sep 1, 2022 at 3:51 PM KENT BRITAIN <WA5VJB@...> wrote: ? Hi Dave
Very very difficult to get Duroid processed these days.I don't know of any
US shops that will touch that stuff any more.(Rumors of one in California
with a Plasma etcher)
If you want plated though holes, you have to etch the raw material to make
it porous enough to take the Copper.? Do you know what kind of chemicals
are necessary to etch Teflon!!!? ? Stuff the EPA does not like!
Rogers 4350 is a ceramic filled material that uses FR4 processing.? Not
quite as low Er and loss tangent as Duroid, but it's not bad.? And regular
PCB houses can process it.? Kent
? ? On Thursday, September 1, 2022 at 10:43:29 AM CDT, W0LEV <
davearea51a@...> wrote:
? I was also going to suggest you actually measure the ¦År of the PCB you
actually used at the frequency you intend to use it.? The ¦År of FR-4 board
falls between an upper an lower limit, but it is not controlled.? If you
want something more controlled and predictable, use RT/Duroid, but it is
expensive.? Also, Teflon board is also better controlled than FR-4, but
also expensive.? Both are far better substrates for microwave applications.
Dave - W?LEV
On Thu, Sep 1, 2022 at 10:31 AM Diane BONKOUNGOU <dianebonk2@...>
wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results
that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it?? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be
good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The
others are professional and paid tools for PCB trace antenna simulation.
Best regards.
--
*Dave - W?LEV*
*Just Let Darwin Work*
--
Dave - W?LEV
-- *Dave - W?LEV* *Just Let Darwin Work* -- Dave - W?LEV
|
Re: Antenna simulation result different from experiment with NanoVNA
On 01/09/2022 11:31, Diane BONKOUNGOU wrote: Hello everyone,
I used OpenEMS software for my simulation and I got simulation results that are very different from my experimental result, I have a PCB antenna whose resonant frequency is around 2.4GHz. I design the same thing on OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz. Such a difference of 0.2GHz is important in RF, isn't it? Knowing that the frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be good. I know that because I tested the Texas instrument PCB IFA antenna and I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The others are professional and paid tools for PCB trace antenna simulation.
Best regards. On your 2nd point I get good correlation between 4NEC2, EZNEC models and VNA measurements for conventional non PCB antennas in most cases. Brian -- Brian
|
Re: Antenna simulation result different from experiment with NanoVNA
On 01/09/2022 14:46, Diane BONKOUNGOU wrote: Hi Brian,
Thanks for your reply, attached is my code. I put the dielectric at 4.8
it's FR4. I started with 4.29, my PCB was made by PCBway company. I remark
that when I increase permittivity the frequency decrease.
I also try to put a thickness to my copper traces, I finally forgot this
Idea.
After I changed the metal trace with the copper having losses. *%CSX =
AddMetal( CSX, 'ifa' );* change this with the second line
*CSX = AddConductingSheet( CSX, 'ifa', 59e6,
70e-6); %* 59 copper conductivity, 70e-6 is the losses
Also, I try to put a layer of copper in the PCB because it is a 4-layer PCB
and it gives me 2.66GHz, otherwise, I have 2.8GHz.
My PCB is 0.8mm thicker, with 4 layers of elements. And yes, you are right,
the mesh size can be a big factor. I spend a lot of time on the mesh part.
I tried using Detect edge but I get errors all the time. I ended up doing
the meshing by hand.
-I also check the simulation box position my PCB is inside, but I am
wondering if the position is good.
Best regards I tried the file and get 2.8GHz. The S11 plot seems to be upside down. There must be a negative sign somewhere I've missed; the f_res is calculated for min |s11|. If I change it to max|s11| it makes sense. I have to change the dielectric by 1.5 times to get it down to 2.4GHz, which is away out ,so I don't think it's the choice of dielectric constant. Thickness of PCB didn't change it much. The resonant frequency will go down if the elements are embedded. You have to make a 4 layer stack as you've pointed out. It's probably this causing the error. The PCB antennas I have are double sided PCB yagis . Brian
|
Re: stand for the NanoVNA???
If you have access to a 3-D printer, there are a ton of designs for NanoVNA
cases, holders, connection extensions, etc. Here is but one:
Many public libraries have Maker Labs. You upload the design file to their
server and they notify you when it's done. My library charges a nominal fee
for the amount of materials used.
Dave N9ZC
toggle quoted message
Show quoted text
On Thu, Sep 1, 2022, 9:28 AM F1AMM <18471@...> wrote: You may have forgotten that you can put an extension cord of the length
you want between the nanaoVNA box and the location where you are going to
do the calibration. Even the imperfections of the extension will be
compensated by the calibration. I did a test with a 93 ¦¸ cable.
Take advantage of this extension to change the type of connector. Switch
to a female BNC socket for example, firmly fixed on a support. SMA are too
fragile to be screwed and unscrewed too often.
--
F1AMM Fran?ois
De la part de n1evh via
Envoy¨¦ : jeudi 1 septembre 2022 16:10
|
Re: Antenna simulation result different from experiment with NanoVNA
Yes, I spaced the Rogers products. Thank you for mentioning the Rogers
4350 substrate.
FR-4 is available everywhere and is relatively inexpensive.
What's more and being critical, both the dielectric constant and dielectric
thickness of FR-4 board may vary across a large sheet. Been there, done
that being paid to force Fibre Channel "square waves" across 15-inches of
FR-4 with no dispersion. Bad things happen for EMC/RFI when the
pre-emphasis is jacked (by the designers to make it function) in an attempt
to accomplish that stated goal. The GM of the division dictated absolutely
bargain basement prices on every component of the system. Ultimately, he
got egg on his face and I was banned from working further with that
division. Standard practice with the present-day Corporate America: Slay
the messenger!!
Dave - W?LEV
toggle quoted message
Show quoted text
On Thu, Sep 1, 2022 at 3:51 PM KENT BRITAIN <WA5VJB@...> wrote: Hi Dave
Very very difficult to get Duroid processed these days.I don't know of any
US shops that will touch that stuff any more.(Rumors of one in California
with a Plasma etcher)
If you want plated though holes, you have to etch the raw material to make
it porous enough to take the Copper. Do you know what kind of chemicals
are necessary to etch Teflon!!! Stuff the EPA does not like!
Rogers 4350 is a ceramic filled material that uses FR4 processing. Not
quite as low Er and loss tangent as Duroid, but it's not bad. And regular
PCB houses can process it. Kent
On Thursday, September 1, 2022 at 10:43:29 AM CDT, W0LEV <
davearea51a@...> wrote:
I was also going to suggest you actually measure the ¦År of the PCB you
actually used at the frequency you intend to use it. The ¦År of FR-4 board
falls between an upper an lower limit, but it is not controlled. If you
want something more controlled and predictable, use RT/Duroid, but it is
expensive. Also, Teflon board is also better controlled than FR-4, but
also expensive. Both are far better substrates for microwave applications.
Dave - W?LEV
On Thu, Sep 1, 2022 at 10:31 AM Diane BONKOUNGOU <dianebonk2@...>
wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results
that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be
good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The
others are professional and paid tools for PCB trace antenna simulation.
Best regards.
--
*Dave - W?LEV*
*Just Let Darwin Work*
--
Dave - W?LEV
-- *Dave - W?LEV* *Just Let Darwin Work*
--
Dave - W?LEV
|
Re: Antenna simulation result different from experiment with NanoVNA
Hi Dave
Very very difficult to get Duroid processed these days.I don't know of any US shops that will touch that stuff any more.(Rumors of one in California with a Plasma etcher)
If you want plated though holes, you have to etch the raw material to make it porous enough to take the Copper.? Do you know what kind of chemicals are necessary to etch Teflon!!!? ? ?Stuff the EPA does not like!
Rogers 4350 is a ceramic filled material that uses FR4 processing.? Not quite as low Er and loss tangent as Duroid, but it's not bad.? And regular PCB houses can process it.? ?Kent
toggle quoted message
Show quoted text
On Thursday, September 1, 2022 at 10:43:29 AM CDT, W0LEV <davearea51a@...> wrote: I was also going to suggest you actually measure the ¦År of the PCB you actually used at the frequency you intend to use it.? The ¦År of FR-4 board falls between an upper an lower limit, but it is not controlled.? If you want something more controlled and predictable, use RT/Duroid, but it is expensive.? Also, Teflon board is also better controlled than FR-4, but also expensive.? Both are far better substrates for microwave applications. Dave - W?LEV On Thu, Sep 1, 2022 at 10:31 AM Diane BONKOUNGOU <dianebonk2@...> wrote: Hello everyone,
I used OpenEMS software for my simulation and I got simulation results
that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it?? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be
good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The
others are professional and paid tools for PCB trace antenna simulation.
Best regards.
-- *Dave - W?LEV* *Just Let Darwin Work* -- Dave - W?LEV
|
Re: Antenna simulation result different from experiment with NanoVNA
I was also going to suggest you actually measure the ¦År of the PCB you actually used at the frequency you intend to use it. The ¦År of FR-4 board falls between an upper an lower limit, but it is not controlled. If you want something more controlled and predictable, use RT/Duroid, but it is expensive. Also, Teflon board is also better controlled than FR-4, but also expensive. Both are far better substrates for microwave applications. Dave - W?LEV On Thu, Sep 1, 2022 at 10:31 AM Diane BONKOUNGOU <dianebonk2@...> wrote: Hello everyone,
I used OpenEMS software for my simulation and I got simulation results
that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be
good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The
others are professional and paid tools for PCB trace antenna simulation.
Best regards.
-- *Dave - W?LEV* *Just Let Darwin Work* -- Dave - W?LEV
|
Re: stand for the NanoVNA???
You may have forgotten that you can put an extension cord of the length you want between the nanaoVNA box and the location where you are going to do the calibration. Even the imperfections of the extension will be compensated by the calibration. I did a test with a 93 ¦¸ cable.
Take advantage of this extension to change the type of connector. Switch to a female BNC socket for example, firmly fixed on a support. SMA are too fragile to be screwed and unscrewed too often.
--
F1AMM Fran?ois
De la part de n1evh via
Envoy¨¦ : jeudi 1 septembre 2022 16:10
|
Re: Antenna simulation result different from experiment with NanoVNA
Hi Zack?The big inventory is Arlon 25N which is also space qualified!Currently have 6 antenna assemblies in orbit and working on 3 more!I do a lot of antennas were the Er of the PCB material is not that important, like Vivaldi, LP, and Planar Disk antennas.? ?Also putting the antenna on VERY thin material makes the Er less important.? i.e. .008" multi-layer material normally used for the interlayers in multilayer PCB's.? KentPS? The Nano's are cute, I've got 3.? (Different work areas).
toggle quoted message
Show quoted text
On Thursday, September 1, 2022 at 10:03:46 AM CDT, Zack Widup <w9sz.zack@...> wrote: Kent Britain WA5VJB may chime in, but he found the variation in dielectric constant for various samples of G10/FR4 so large that he bought a large stock of board with the same dielectric constant and makes his pc board antennas from that. Using a piece of G10 or FR4 at random may not produce the expected results. Zack W9SZ On Thu, Sep 1, 2022 at 7:59 AM Brian Howie GM4DIJ <brian@...> wrote: On 01/09/2022 11:31, Diane BONKOUNGOU wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it?? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The others are professional and paid tools for PCB trace antenna simulation.
Best regards. Can you upload the file and I'll have a look The dielectric constant you
chose might differ from the real one depending on the material. There
are also issues in getting the mesh riggt.
I have a couple of PCB antennas for 1.3, 2.3 and 3.4GHz which are ok on
the nanoVNA , I was planning to model. them . They seem to have a
stripline matching section which might be tricky to model accurately.
Brian
--
Brian
|
Re: Antenna simulation result different from experiment with NanoVNA
Kent Britain WA5VJB may chime in, but he found the variation in dielectric constant for various samples of G10/FR4 so large that he bought a large stock of board with the same dielectric constant and makes his pc board antennas from that. Using a piece of G10 or FR4 at random may not produce the expected results. Zack W9SZ On Thu, Sep 1, 2022 at 7:59 AM Brian Howie GM4DIJ <brian@...> wrote: On 01/09/2022 11:31, Diane BONKOUNGOU wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The others are professional and paid tools for PCB trace antenna simulation.
Best regards. Can you upload the file and I'll have a look The dielectric constant you
chose might differ from the real one depending on the material. There
are also issues in getting the mesh riggt.
I have a couple of PCB antennas for 1.3, 2.3 and 3.4GHz which are ok on
the nanoVNA , I was planning to model. them . They seem to have a
stripline matching section which might be tricky to model accurately.
Brian
--
Brian
|
I recently purchased a NanoVNA (with a 2.8 inch screen). I was
wondering if anyone has designed/printed a stand for the
NanoVNA???
I'm looking for something that I can put the end of it into (so that
the SMA jacks are facing up) so I can attach an SMA antenna
for an HT to it.
With a stand... I won't have to worry about the NanoVNA falling
over if the USB cable between my laptop or tablet gets hit or
moves (more of a problem if holding a tablet) and since it will
keep the NanoVNA from falling over... I won't have worry about
anything touching the antenna to affect the measurements.
Thanks,
Mike - N1EVH
|
Re: Antenna simulation result different from experiment with NanoVNA
On 9/1/22 6:21 AM, Gary W9TD wrote: EZNEC is now free.
EZNEC as a Method of Moments code that is derived from NEC is really a "wire" modeling code and doesn't do patches very well. One can "grid" a patch, and put it over ground with soil properties set to match the substrate, but it doesn't have a way to do the ground plane on the other side of the substrate. (well, in NEC4 you could bury a grid of wires)). NEC based codes will not do the matching sections or transmission line. What the OP needs is a low cost tool that is designed for microstrip patches. Ancient, but maybe PUFF (an early microwave circuit modeling package) would work? At work we use HFSS, but that's sort of the antithesis of free. I'd start with the list here: Sonnet has a student version
|
Re: Antenna simulation result different from experiment with NanoVNA
Hi Brian,
Thanks for your reply, attached is my code. I put the dielectric at 4.8
it's FR4. I started with 4.29, my PCB was made by PCBway company. I remark
that when I increase permittivity the frequency decrease.
I also try to put a thickness to my copper traces, I finally forgot this
Idea.
After I changed the metal trace with the copper having losses. *%CSX =
AddMetal( CSX, 'ifa' );* change this with the second line
*CSX = AddConductingSheet( CSX, 'ifa', 59e6,
70e-6); %* 59 copper conductivity, 70e-6 is the losses
Also, I try to put a layer of copper in the PCB because it is a 4-layer PCB
and it gives me 2.66GHz, otherwise, I have 2.8GHz.
My PCB is 0.8mm thicker, with 4 layers of elements. And yes, you are right,
the mesh size can be a big factor. I spend a lot of time on the mesh part.
I tried using Detect edge but I get errors all the time. I ended up doing
the meshing by hand.
-I also check the simulation box position my PCB is inside, but I am
wondering if the position is good.
Best regards
Le jeu. 1 sept. 2022 ¨¤ 13:59, Brian Howie GM4DIJ <brian@...> a
¨¦crit :
toggle quoted message
Show quoted text
On 01/09/2022 11:31, Diane BONKOUNGOU wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results that are very different from my experimental result, I have a PCB antenna
whose resonant frequency is around 2.4GHz. I design the same thing on
OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz.
Such a difference of 0.2GHz is important in RF, isn't it? Knowing that the
frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be good. I know that because I tested the Texas instrument PCB IFA antenna and
I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The others are professional and paid tools for PCB trace antenna simulation.
Best regards. Can you upload the file and I'll have a look The dielectric constant you
chose might differ from the real one depending on the material. There
are also issues in getting the mesh riggt.
I have a couple of PCB antennas for 1.3, 2.3 and 3.4GHz which are ok on
the nanoVNA , I was planning to model. them . They seem to have a
stripline matching section which might be tricky to model accurately.
Brian
--
Brian
|
Re: Antenna simulation result different from experiment with NanoVNA
Yes, and I have known Roy for nearly 30 years, but EZNEC does not do PCB antennas.? ? ?When you download his Pro version, be sure to get the manual as well.? He has RETIRED!? ?Kent
toggle quoted message
Show quoted text
On Thursday, September 1, 2022 at 08:21:44 AM CDT, Gary W9TD <w9td@...> wrote:
EZNEC is now free.
|
Re: Antenna simulation result different from experiment with NanoVNA
What Er did you use for the PCB Material?Unless it is very expensive material where the plot the Er by frequency, you really don't know the Er.
It is industry standard to measure the Er at 1 kHz.Typically they list the Er for fiberglass PCB material.(Yea, the call it FR4, but you have not been able to buy real FR4 for the last 10 years.? The spec for FR4 calls for a Bromide anti-flammability that has been banned)Er is listed as 4.4, but as you go up in frequency Er drops.? At 2.4 GHz it is typically in the 3.8-3.9 range.? ? Personal experience in making thousands of 2.4 GHz patch antennas.? Kent WA5VJB
toggle quoted message
Show quoted text
On Thursday, September 1, 2022 at 05:31:26 AM CDT, Diane BONKOUNGOU <dianebonk2@...> wrote:
Hello everyone,
I used OpenEMS software for my simulation and I got simulation results that are very different from my experimental result, I have a PCB antenna whose resonant frequency is around 2.4GHz. I design the same thing on OpenEMS but the simulation gives me a resonant frequency of around 2.66GHz. Such a difference of 0.2GHz is important in RF, isn't it?? Knowing that the frequency range of Bluetooth Low Energy is 2.4-2.5GHz.
I have calibrated very well my NanoVNA and my experiment result must be good. I know that because I tested the Texas instrument PCB IFA antenna and I get the same result as in their Datasheet.
Can you recommend a free tool for antenna simulation?
I have searched but I can only find OpenEMS based on the FDTD method. The others are professional and paid tools for PCB trace antenna simulation.
Best regards.
|
Donald S Brant Jr
Brian Howie GM4DIJ
Dave Merrill