|
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for low-energy Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a resonance frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j. Can I take the value of this impedance at this frequency for the matching process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
|
What matters is the impedance of the antenna at the frequency where you will use it. Impedances at other frequencies, resonant or not, don't really matter.
SherpaDoug
|
Hello Douglas,
Thanks for your response.
Best regards
Le ven. 22 juil. 2022 ¨¤ 12:55, Douglas Butler <sherpadoug@...> a
¨¦crit :
toggle quoted message
Show quoted text
What matters is the impedance of the antenna at the frequency where you
will use it. Impedances at other frequencies, resonant or not, don't
really matter.
SherpaDoug
|
You should measure the impedance at the frequency you plan on using. I have found fairly large variations over a few MHz in this range before. Zack W9SZ On Fri, Jul 22, 2022 at 6:34 AM Diane BONKOUNGOU <dianebonk2@...> wrote: Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a
resonance frequency of 2.28GHz and at this frequency, I have an impedance
of 47.8-7.06j. Can I take the value of this impedance at this frequency
for the matching process or do I have to take the value of impedance at
2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
<> Virus-free. www.avg.com <> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
|
On 7/22/22 8:36 AM, Zack Widup wrote: You should measure the impedance at the frequency you plan on using. I have
found fairly large variations over a few MHz in this range before.
Zack W9SZ
Most simple antennas, near their design frequency, have an impedance curve that looks a lot like a RLC resonance. The real part changes fairly slowly, but the imaginary part changes quickly. On the attached graph (impedance of a dipole), you can see that X changes from about -40 to +40 going from 90 to 100% of a half wavelength (resonance is at about 95% half wavelength), while R goes from 53 to 75 ohms over the same frequency range. A 5% variation in frequency isn't huge (at 2.45 GHz, 5% is 120 MHz) - If you think of the antenna as a filter, it's not even down 3dB at 5%. Zmag is the absolute value of the Z. So if you put a L or C in to "cancel" the reactive part,
|
On Fri, Jul 22, 2022 at 04:34 AM, Diane BONKOUNGOU wrote:
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a resonance
frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j.
Can I take the value of this impedance at this frequency for the matching
process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
You never posted in the other topic what you finally did to correctly measure the antenna with the NanoVNA. If you don't establish the "reference plane" rightat the antenna your impedance measurements will be way off and you won't be able to design a matching network. Even if you get a decent "reference plane" the antenna must be measured in circuit so that you have a the same ground plane as when the antenna will be used. Also the outer shield of your test cable may form part of the antenna and this will affect the measurements. At these GHz frequencies you will have better results if you tune the antenna for resonance at the desired frequency instead of building a matching network. The reason I say this is that VSWR and Return Loss will only change slightly if your reference plane is not right on but the impedance will be off a fair amount. The reason is that the magnitude of the reflection coefficient stays the same and only the phase angle changes if the cable is slightly long or short. However the phase angle rotation results in a different impedance calculation. Even if you get a reasonable estimate of impedance you then have the issue of designing with components and PCB design at Bluetooth frequencies. Roger
|
Hello Roger, Thanks for reminding me, I will post on the other topic how I went about correcting the electronic delay introduced by the SMA cable. My cable was too long to correct the electronic delay at its end after doing the calibration at the NanoVNA end that's why I had a lot of circles in the Smith chart. I cut the cable further and have a quarter circle in the Smith chart and then use the Edealy parameter in the NanoVNA to correct. Best regards Le ven. 22 juil. 2022 ¨¤ 17:03, Roger Need via groups.io <sailtamarack= [email protected]> a ¨¦crit : On Fri, Jul 22, 2022 at 04:34 AM, Diane BONKOUNGOU wrote:
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a
resonance
frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j.
Can I take the value of this impedance at this frequency for the matching
process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
You never posted in the other topic what you finally did to correctly
measure the antenna with the NanoVNA. If you don't establish the
"reference plane" rightat the antenna your impedance measurements will be
way off and you won't be able to design a matching network. Even if you
get a decent "reference plane" the antenna must be measured in circuit so
that you have a the same ground plane as when the antenna will be used.
Also the outer shield of your test cable may form part of the antenna and
this will affect the measurements.
At these GHz frequencies you will have better results if you tune the
antenna for resonance at the desired frequency instead of building a
matching network. The reason I say this is that VSWR and Return Loss will
only change slightly if your reference plane is not right on but the
impedance will be off a fair amount. The reason is that the magnitude of
the reflection coefficient stays the same and only the phase angle changes
if the cable is slightly long or short. However the phase angle rotation
results in a different impedance calculation. Even if you get a
reasonable estimate of impedance you then have the issue of designing with
components and PCB design at Bluetooth frequencies.
Roger
|
You should better calibrate at the cable end and NOT on the vna..
No circles anymore!! Nothing to compensate via edelay..
I only use edelay to compensate say an sma to n adaptor.. But not to compensate a cable length
Dg9bfc sigi
Am 01.08.2022 12:01 schrieb Diane BONKOUNGOU <dianebonk2@...>:
toggle quoted message
Show quoted text
Hello Roger,
Thanks for reminding me, I will post on the other topic how I went about
correcting the electronic delay introduced by the SMA cable. My cable was
too long to correct the electronic delay at its end after doing the
calibration at the NanoVNA end that's why I had a lot of circles in the
Smith chart. I cut the cable further and have a quarter circle in the
Smith
chart and then use the Edealy parameter in the NanoVNA to correct.
Best regards
Le ven. 22 juil. 2022 ¨¤ 17:03, Roger Need via groups.io <sailtamarack=
[email protected]> a ¨¦crit :
On Fri, Jul 22, 2022 at 04:34 AM, Diane BONKOUNGOU wrote:
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for
low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a resonance
frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j.
? Can I take the value of this impedance at this frequency for the matching
process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
You never posted in the other topic what you finally did to correctly
measure the antenna with the NanoVNA.? If you don't establish the
"reference plane" rightat the antenna your impedance measurements will
be
way off and you won't be able to design a matching network.?? Even if you
get a decent "reference plane" the antenna must be measured in circuit so
that you have a the same ground plane as when the antenna will be used.
? Also the outer shield of your test cable may form part of the antenna and
this will affect the measurements.
At these GHz frequencies you will have better results if you tune the
antenna for resonance at the desired frequency instead of building a
matching network.?? The reason I say this is that VSWR and Return Loss will
only change slightly if your reference plane is not right on but the
impedance will be off a fair amount. The reason is that the magnitude of the reflection coefficient stays the same and only the phase angle changes
if the cable is slightly long or short.? However the phase angle rotation
results in a different impedance calculation.?? Even if you get? a
reasonable estimate of impedance you then have the issue of designing with
components and PCB design at Bluetooth frequencies.
Roger
|
Hello,
Thanks for your reply.
I am using a quarter wavelength antenna.
Best regards
toggle quoted message
Show quoted text
Le ven. 22 juil. 2022 ¨¤ 16:59, Jim Lux <jimlux@...> a ¨¦crit : On 7/22/22 8:36 AM, Zack Widup wrote:
You should measure the impedance at the frequency you plan on using. I have
found fairly large variations over a few MHz in this range before.
Zack W9SZ
Most simple antennas, near their design frequency, have an impedance
curve that looks a lot like a RLC resonance. The real part changes
fairly slowly, but the imaginary part changes quickly.
On the attached graph (impedance of a dipole), you can see that X
changes from about -40 to +40 going from 90 to 100% of a half wavelength
(resonance is at about 95% half wavelength), while R goes from 53 to 75
ohms over the same frequency range.
A 5% variation in frequency isn't huge (at 2.45 GHz, 5% is 120 MHz) - If
you think of the antenna as a filter, it's not even down 3dB at 5%.
Zmag is the absolute value of the Z.
So if you put a L or C in to "cancel" the reactive part,
|
On Mon, Aug 1, 2022 at 03:15 AM, Siegfried Jackstien wrote:
You should better calibrate at the cable end and NOT on the vna..
No circles anymore!! Nothing to compensate via edelay..
I only use edelay to compensate say an sma to n adaptor.. But not to
compensate a cable length
I agree with you. Using edelay to de-embed a coaxial cable is not a good idea if impedance measurements are the goal. The attenuation of the cable affects the measurement. Things get worse the higher the frequency as cable loss increases. Roger
|
And please remember resonance is not necessarily at 1:1 SWR. It may or may not be. Resonance is *defined* by (consult the complex portion of the measured impedance): +jX = -jX. The resistance term is purely resistive (consisting "mostly" of the radiation resistance) while the reactive component is strictly reactive: capacitive or inductive. Dave - W?LEV On Mon, Aug 1, 2022 at 10:01 AM Diane BONKOUNGOU <dianebonk2@...> wrote: Hello Roger,
Thanks for reminding me, I will post on the other topic how I went about
correcting the electronic delay introduced by the SMA cable. My cable was
too long to correct the electronic delay at its end after doing the
calibration at the NanoVNA end that's why I had a lot of circles in the
Smith chart. I cut the cable further and have a quarter circle in the Smith
chart and then use the Edealy parameter in the NanoVNA to correct.
Best regards
Le ven. 22 juil. 2022 ¨¤ 17:03, Roger Need via groups.io <sailtamarack=
[email protected]> a ¨¦crit :
On Fri, Jul 22, 2022 at 04:34 AM, Diane BONKOUNGOU wrote:
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for
low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a resonance
frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j.
Can I take the value of this impedance at this frequency for the matching
process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
You never posted in the other topic what you finally did to correctly
measure the antenna with the NanoVNA. If you don't establish the
"reference plane" rightat the antenna your impedance measurements will be
way off and you won't be able to design a matching network. Even if you
get a decent "reference plane" the antenna must be measured in circuit so
that you have a the same ground plane as when the antenna will be used.
Also the outer shield of your test cable may form part of the antenna
and
this will affect the measurements.
At these GHz frequencies you will have better results if you tune the
antenna for resonance at the desired frequency instead of building a
matching network. The reason I say this is that VSWR and Return Loss will
only change slightly if your reference plane is not right on but the
impedance will be off a fair amount. The reason is that the magnitude of
the reflection coefficient stays the same and only the phase angle changes
if the cable is slightly long or short. However the phase angle rotation
results in a different impedance calculation. Even if you get a
reasonable estimate of impedance you then have the issue of designing with
components and PCB design at Bluetooth frequencies.
Roger
-- *Dave - W?LEV* *Just Let Darwin Work* -- Dave - W?LEV
|
Note that when the load looks resistive the VSWR versus frequency is at a minimum.
toggle quoted message
Show quoted text
On August 2, 2022 4:38:51 PM MDT, W0LEV <davearea51a@...> wrote: And please remember resonance is not necessarily at 1:1 SWR. It may or may
not be. Resonance is *defined* by (consult the complex portion of the
measured impedance): +jX = -jX. The resistance term is purely resistive
(consisting "mostly" of the radiation resistance) while the reactive
component is strictly reactive: capacitive or inductive.
Dave - W?LEV
On Mon, Aug 1, 2022 at 10:01 AM Diane BONKOUNGOU <dianebonk2@...>
wrote:
Hello Roger,
Thanks for reminding me, I will post on the other topic how I went about
correcting the electronic delay introduced by the SMA cable. My cable was
too long to correct the electronic delay at its end after doing the
calibration at the NanoVNA end that's why I had a lot of circles in the
Smith chart. I cut the cable further and have a quarter circle in the Smith
chart and then use the Edealy parameter in the NanoVNA to correct.
Best regards
Le ven. 22 juil. 2022 ¨¤ 17:03, Roger Need via groups.io <sailtamarack=
[email protected]> a ¨¦crit :
On Fri, Jul 22, 2022 at 04:34 AM, Diane BONKOUNGOU wrote:
Hello,
Hi, I am a beginner in antenna design and tuning.
I want my antenna to have a resonant frequency of 2.45GHz for
low-energy
Bluetooth communication.
And I want to know what to consider for the matching process.
I have made some measurements with Nanovna of my antenna. I have a resonance
frequency of 2.28GHz and at this frequency, I have an impedance of 47.8-7.06j.
Can I take the value of this impedance at this frequency for the matching
process or do I have to take the value of impedance at 2.45GHz?
Where to take the value of impedance for tuning?
Thanks.
You never posted in the other topic what you finally did to correctly
measure the antenna with the NanoVNA. If you don't establish the
"reference plane" rightat the antenna your impedance measurements will be
way off and you won't be able to design a matching network. Even if you
get a decent "reference plane" the antenna must be measured in circuit so
that you have a the same ground plane as when the antenna will be used.
Also the outer shield of your test cable may form part of the antenna
and
this will affect the measurements.
At these GHz frequencies you will have better results if you tune the
antenna for resonance at the desired frequency instead of building a
matching network. The reason I say this is that VSWR and Return Loss will
only change slightly if your reference plane is not right on but the
impedance will be off a fair amount. The reason is that the magnitude of
the reflection coefficient stays the same and only the phase angle changes
if the cable is slightly long or short. However the phase angle rotation
results in a different impedance calculation. Even if you get a
reasonable estimate of impedance you then have the issue of designing with
components and PCB design at Bluetooth frequencies.
Roger
--
*Dave - W?LEV*
*Just Let Darwin Work*
--
Dave - W?LEV
|
Incorrect. The VSWR is based on impedance - which is a combination of resistance and reactance. A vertical at resonance will have an impedance of about 35 + 0j ohms or an impedance of 35 ohms and a 50/35 or 1.43:1 SWR. However, slightly off resonance it could have an impedance of say 40 + 30j ohms, which would have a total impedance of 50 ohms (just an example for ease of calculations) and an SWR of 1:1. The antenna is non-resonant at the lowest SWR reading.
|
40+j30 is a Z of 50 ohms. In this case, this is an SWR 1:2 !!! Not 1:1 .
Only swr 1:1 is 50 ohms resistive load (and then again, using a 50 ohm cable ;-) ).
73
Arie PA3A
Op 3-8-2022 om 02:26 schreef Jerry Stuckle:
toggle quoted message
Show quoted text
Incorrect. The VSWR is based on impedance - which is a combination of resistance and reactance. A vertical at resonance will have an impedance of about 35 + 0j ohms or an impedance of 35 ohms and a 50/35 or 1.43:1 SWR. However, slightly off resonance it could have an impedance of say 40 + 30j ohms, which would have a total impedance of 50 ohms (just an example for ease of calculations) and an SWR of 1:1. The antenna is non-resonant at the lowest SWR reading.
|
On Tue, Aug 2, 2022 at 05:26 PM, Jerry Stuckle wrote:
The VSWR is based on impedance - which is a combination of resistance and
reactance.
Can you explain this in more detail. I thought VSWR was based on voltage.
|
SWR = Zt / Zo applies only when both impedances are resistive. When Zt = 40 + j30 ohms, the load is highly reactive. The coefficient of reflection rho is: rho = (Zt - Zo) / (Zt + Zo) = 0 + j0.3333 ... SWR = (1 + abs(rho)) / (1 - abs(rho)) = 2.0 73, Maynard W6PAP On Tuesday, August 02, 2022 05:26:38 PM Jerry Stuckle wrote: Incorrect. The VSWR is based on impedance - which is a combination of
resistance and reactance. A vertical at resonance will have an impedance
of about 35 + 0j ohms or an impedance of 35 ohms and a 50/35 or 1.43:1 SWR.
However, slightly off resonance it could have an impedance of say 40 + 30j
ohms, which would have a total impedance of 50 ohms (just an example for
ease of calculations) and an SWR of 1:1. The antenna is non-resonant at
the lowest SWR reading.
-----------------------------------------
|
Russ,
If the load is not perfectly matched to the source, there will be a voltage standing wave ratio (VSWR) because of the mismatch.
Learn the Smith chart.
toggle quoted message
Show quoted text
On Wednesday, August 3, 2022, 11:52:31 AM EDT, Russ <u.rusty@...> wrote: On Tue, Aug? 2, 2022 at 05:26 PM, Jerry Stuckle wrote:
The VSWR is based on impedance - which is a combination of resistance and
reactance.
Can you explain this in more detail. I thought VSWR was based on voltage.
|
Helpful document:
toggle quoted message
Show quoted text
On 8/3/2022 9:10 AM, Kenneth Hendrickson via groups.io wrote: Russ,
If the load is not perfectly matched to the source, there will be a voltage standing wave ratio (VSWR) because of the mismatch.
Learn the Smith chart.
On Wednesday, August 3, 2022, 11:52:31 AM EDT, Russ <u.rusty@...> wrote:
On Tue, Aug? 2, 2022 at 05:26 PM, Jerry Stuckle wrote:
The VSWR is based on impedance - which is a combination of resistance and
reactance.
Can you explain this in more detail. I thought VSWR was based on voltage.
|
Yes!!! Learn the Smith Chart!!!! Also, learn and internalize Ohm's Law!!!! Both will serve you well with all things RF. SWR can be measured as current, or voltage standing wave ratios. A good test for all this confusion is to establish a system which measures 1:1 SWR at a known frequency. Then add an electrical 1/8 to 1/4-wavelength of transmission line to the system. If the SWR remains 1:1 with the added transmission line in place, indeed, the SWR is 1:1. If the measurement with the additional line length measures something other than 1:1, the original 1:1 measurement is in error. This is the primary reason S-parameters deal strictly with power, the product of voltage and current for a sine wave. With SWR on the transmission line, voltage and current along the line changes as measured at single points along the line. However, power remains the same along the line - it's just Ohm's law with a little very simple algebra. Dave - W?LEV On Wed, Aug 3, 2022 at 4:10 PM Kenneth Hendrickson via groups.io <dsp_stap= [email protected]> wrote: Russ,
If the load is not perfectly matched to the source, there will be a
voltage standing wave ratio (VSWR) because of the mismatch.
Learn the Smith chart.
On Wednesday, August 3, 2022, 11:52:31 AM EDT, Russ <
u.rusty@...> wrote:
On Tue, Aug 2, 2022 at 05:26 PM, Jerry Stuckle wrote:
The VSWR is based on impedance - which is a combination of resistance and
reactance.
Can you explain this in more detail. I thought VSWR was based on voltage.
-- *Dave - W?LEV* *Just Let Darwin Work* -- Dave - W?LEV
|
Time is well spent learning a bit about the Smith Chart.
You can plot a complex impedance and then draw a circle that intersects the point and get the SWR. You can also measure the distance and get the reflection coefficient and gamma.
A picture is worth a thousand words so here is a plot on a Smith Chart of several complex impedances that all result in a SWR of 2 and a reflection coefficient of 0.333 ( a Reflection Loss of 9.54 dB). The 40 + j30 from the earlier discussion is one of these.
Roger
|
Douglas Butler
Roger Need
Arie Kleingeld PA3A
