DIY wideband antenna VSWR - beginner

[flowwolf]

TL;DR Summary

How much VSWR is good in a DIY wideband antenna

Dear forumers,

I would like to do some RF measurements by using a spectrum analyzer or an SDR in the ~ 100mhz-6ghz range.

I'm new to this and to the concepts in general and I would like to hear a few recommendations.

I primarily interested here in how much it is acceptable to use a DIY wideband or ultra-wideband directional antenna (log periodic or spiral etc.). I saw that a few people came up with or sell such antennas cheaply:

https://www.rtl-sdr.com/a-homemade-pcb-log-periodic-antenna/

but after comparing VSWR with professional antennas I see that the VSWR of pro antennas are nearly flat.I am not sure if I understand the concepts correctly, so

how bad VSWR is in the following image (from above) (EDIT: it's return loss but)?: https://www.rtl-sdr.com/wp-content/uploads/2017/01/lpad-RL.jpg

If I used this antenna to measure RF signals, would that mean that the instrument would measure a 1ghz signal 10dB weaker than a signal with the same power at 1.5ghz (refer to the above pic)?

If the antenna's impedance is high and a balun is used (e.g. from 300 -> 50ohm), how much does that change this curve?Another question, I see that there are coax and dipole antennas, I read that for spiral antennas a balanced signal is required so does it make a difference if I feed a pcb log periodic antenna with a balanced signal?Regards,
Akos

 

[berkeman]

Summary:: How much VSWR is good in a DIY wideband antenna

I would like to do some RF measurements by using a spectrum analyzer or an SDR in the ~ 100mhz-6ghz range.

I'm new to this and to the concepts in general and I would like to hear a few recommendations.

What spectrum analyzer will you be using? Will you be using a vector impedance S-parameter test instrument to measure the VSWR and other s-parameters?

https://en.wikipedia.org/wiki/Network_analyzer_(electrical)

 

[jasonRF]

If I used this antenna to measure RF signals, would that mean that the instrument would measure a 1ghz signal 10dB weaker than a signal with the same power at 1.5ghz (refer to the above pic)?

No. The return loss measurement is done as follows: hook up an analyzer to the connector on the antenna, the analyzer then transmits a signal and measures what fraction is reflected from the antenna and sent back to the analyzer. If $\Gamma$ is the reflection coefficient, then the return loss is $|\Gamma|^2$; in dB that is $20*\log_{10}|\Gamma|$. The fraction of the power transmitted is then $1-|\Gamma|^2$. So a return loss of -10 dB means $|\Gamma|^2 = \frac{1}{10}$, so 90% of the power is transmitted. Of course, the same is true on receive; only 10% of the power is lost. The antenna you showed has a worst-case return loss of -7 dB; I will let you calculate what fraction of the power you lose in that instance.

If the antenna's impedance is high and a balun is used (e.g. from 300 -> 50ohm), how much does that change this curve?

I suspect it was measured with whatever balun is recommended to match to whatever the system impedance is (usually 50 Ohms). Otherwise the worst-case return loss would be much worse.

 

[flowwolf]

What spectrum analyzer will you be using? Will you be using a vector impedance S-parameter test instrument to measure the VSWR and other s-parameters?

I think I will start with an SDR like rtl-sdr.
I read that it's possible to do that with an SDR and a configurable signal generator and a directional coupler.

Is this possible?

jasonRF, so this means in the case of -7dB return loss, it is 20% while -28dB would be 0.16% power loss?

I suspect it was measured with whatever balun is recommended to match to whatever the system impedance is (usually 50 Ohms). Otherwise the worst-case return loss would be much worse.

OK

 

[Baluncore]

VSWR is a measure of impedance mismatch based on reflected energy. If you have a more lossy antenna or feedline you will reduce the VSWR. Don't confuse low SWR due to bad resistive losses with a low SWR due to good matching. You must understand the difference. Cheap coax has lower SWR than expensive coax because it is more lossy. A resistive dummy load has a perfect 1:1 VSWR, but it does not work well as an antenna.

If the antenna's impedance is high and a balun is used (e.g. from 300 -> 50ohm), how much does that change this curve?

Use a balun only when it is needed to match the impedance. Use the correct ratio. If you do not match impedance, the flat response of the system across a wide band will be lost because the electrical lengths of the feed line segments will become critical.

 

[Merlin3189]

I definitely agree with your first para, but I always thought a balun was used to match an unbalanced cable like coax to a symmetric antenna like a dipole, even if there was no impedance mismatch. (Or to join balanced feeder like twin line to unbalanced cable or socket.) Balun and matching transformer functions are often combined, and called, say, a 4:1 balun used to join 300 Ω twin feeder to 75 Ω TV coax input socket.

 

[Baluncore]

I definitely agree with your first para,

@Merlin3189 Are you addressing me?
Strictly speaking, a balun is used to convert between balanced and unbalanced lines.
But as you say, a balun will usually have an impedance ratio, typically 4:1 for commercial TV.
That impedance ratio is often more important than the balance.

 

[flowwolf]

Thanks for the help, all of them was useful.

I re-read that SDR and VSWR measurement article, it says that an RF bridge or a directional coupler is needed to do that.
So I checked RF bridges and directional couplers, the cheaper ones don't seem to go higher than 3ghz.

Also cheaper VNAs don't go higher than that.

If you have a more lossy antenna or feedline you will reduce the VSWR. Don't confuse low SWR due to bad resistive losses with a low SWR due to good matching. You must understand the difference. Cheap coax has lower SWR than expensive coax because it is more lossy.

You mean a lossy antenna can have a low (good) VSWR value and still not receiving as well?

Another question, I see that there are coax and dipole antennas, I read that for spiral antennas a balanced signal is required so does it make a difference if I feed a pcb log periodic antenna with a balanced signal?

but I always thought a balun was used to match an unbalanced cable like coax to a symmetric antenna like a dipole

Just to be sure, there are both coax and symmetric antennas, e.g. log-periodic antennas.
Unbalanced (coax) signal cannot be used to feed a symmetric antenna.

Is a symmetric log-periodic antenna better in any way than an unbalanced one btw.?

 

[Baluncore]

You mean a lossy antenna can have a low (good) VSWR value and still not receiving as well?

Exactly.