Difference in SWR readings between nanoVNA and SWR meter

[vinlove]

I just did SWR reading on my HF antenna which is a EFLW with a balun.

At 7.15 Mhz, my SWR meter off ATU is 1:1.1 SWR.
On nanoVNA reading for the antenna at the same freq. is 1:2.87 SWR.

That I would say is, a huge difference.

Not sure which is right at this point and time.

Obviously one is wrong, or both wrong.
But it can't be both right?

[vinlove]

I have tried with 2x other SWR meters (high quality ones), and they all read SWR as 1:1.1 at the freq. 7.15Mhz.

nanoVNA is reading 1:2.87 SWR at 7.15 Mhz seems incorrect reading? I was using about 40W FM carrier for the SWR readings on the freq.

[vk6zgo]

If you have a "dummy load", which is (or should be) a known 50 , you can check its VSWR with, first, the SWR meter, then with the VNA,-----both should give you the same result.
If you have no dummy load, you should be able to check the VNA alone* with two 100 metal film (not wirewound) resistors in parallel.
*Fairly obviously, normal 1/4 & 1/2 watt resistors would not be suitable for transmitter output power.

At 7.15MHz, any stray inductance or capacitance should have minimal effect.

It is only of recent years that inexpensive VNAs have been available.
Back when I did most of my RF testing, they were "rare as hen's teeth" & cost as much as a fairly nice car, so we usually used Scalar Network Analysers.

[Kalvin]

Check also that you have calibrated your NanoVNA using the open/short/50ohm.

[radiolistener]

In short, if you're using NanoVNA properly (performed calibration before measurement), it shows correct value.
And your SWR meter shows wrong value.

Scalar SWR meters have huge measurement error. For example, if you're using low power TX for measurement, your scalar SWR meter will show smaller value than real. Try to use at least 100-200 W power for SWR measurement with scalar meter, in such case it will show you more realistic value. But note, it may be dangerous for your rig, because SWR=2.9 is not good.

As you can see, I was right when I told you that there is possible issue with your antenna in the topic about PSU. 

[vk6zgo]

This guy has experienced errors with a NANO VNA compared to another VNA, so that makes the Vector -vs- Scalar argument a bit redundant.

[Kalvin]

You can always compare your SWR-meter and NanoVNA with a known load. Just get a known resistor 25 ohms - 100 ohms and a capacitor or an inductor, and measure the SWR.  Then compare the measurement values to the calculated value. You will have much less guesswork.

[vinlove]

Great info !! Thanks.

I just checked with a dummy load, and the nanoVNA shows flat 1:1 SWR across the band, as expected.

But why the HF antenna which reads 1:1.02 on 3x different SWR meters (all high quality ones) is reading SWR 1:2.8 in the nanoVNA is a mystery to me.

My rig has also inbuilt SWR meter, and it reads 1:1 on the 7Mhz with the HF antenna, so I have been using the HF antenna straight into the rig with just PWR/SWR meter all the time for ages with full 100W most of the time, and making loads of good contacts with good reports.

[Kalvin]

Antenna may be picking a strong signal, and NanoVNA is quite sensitive device with small output power, so that may one cause that NanoVNA is showing high SWR.

[Kalvin]

The effect of the received signal from the antenna might be reduced (although not completely eliminated) if NanoVNA could sweep by alternating TX on and TX off with some averaging, so that the background noise coming from the antenna could be canceled out, and the device should be able to provide somewhat improved SWR readings. I do not know how much improvement is possible, though.

[radiolistener]

But why the HF antenna which reads 1:1.02 on 3x different SWR meters (all high quality ones) is reading SWR 1:2.8 in the nanoVNA is a mystery to me.

Most of all because you're using too small power for SWR measurement with your SWR meter. Most of scalar SWR meters using diode as amplitude detector for a coupler output. The diode has limited working dynamic range and non-linear response at low voltage.

This is why it will show you SWR=1.1 while in reality it is SWR=2.9. Just because amplitude taken from a coupler is too small for diode. It is small, because output power is too small.

If you don't believe me, you can perform simple experiment to make sure that this is the truth. Just connect 75-100 Ohm resistor as a load on SWR meter and try to measure SWR at different power level. You will see that your SWR meter will show you SWR=1.0 at low power.

75 Ohm resistor means SWR=1.5. But your SWR meter will show you SWR=1.0. Just because not enough power output was used for measurement.

Almost all scalar SWR meters shows wrong SWR value at low power output.

Another issue with scalar SWR meter is that it shows SWR value for entire bandwidth, it don't take into account that your TX has harmonics and out of band emissions. It show you average SWR value for all components in the signal.

In opposite, VNA meters show you SWR value at specified frequency only. Also, VNA meters don't have diode issue (at low power output). This is why VNA measurement is more correct and more precise than scalar SWR meters.

[radiolistener]

My rig has also inbuilt SWR meter, and it reads 1:1

inbuild SWR meters have very bad measurement performance. They are not intended for precise measurement. The goal of inbuild SWR meters is just to protect your rig from burning out at very high SWR.

[vk6zgo]

Great info !! Thanks.

I just checked with a dummy load, and the nanoVNA shows flat 1:1 SWR across the band, as expected.

But why the HF antenna which reads 1:1.02 on 3x different SWR meters (all high quality ones) is reading SWR 1:2.8 in the nanoVNA is a mystery to me.

My rig has also inbuilt SWR meter, and it reads 1:1 on the 7Mhz with the HF antenna, so I have been using the HF antenna straight into the rig with just PWR/SWR meter all the time for ages with full 100W most of the time, and making loads of good contacts with good reports.

Have you done the full check, which entails also checking the transmitter plus SWR meters into the dummy load?

Also, try checking the SWR of your antenna at full power, if your SWR meters are rated for it.(do it fast, as other hams may look askance at a massive carrier up for too long )
Write that down, then check at the lowest power level you can reduce the radio to.
Write that down, also.

Compare the two------ does it confirm the point made by radiolistener?

Another test you can try, this time on the nanoVNA, is to use it to check the SWR with a 75 load, then a 150 load
They should measure 1.5:1 & 3.0:1 respectively.

I tend towards  Kalvin's suggestion of interference from a strong external carrier.

Back in the day, we occasionally needed to measure the return loss of the antenna bays of our TV site, over their designed range of frequencies.

To do this, we used a Polyskop SWOB, which produced a swept RF signal, which was applied to the item under test, then detected the reflected signal & compared the two, to show return loss on a CRT screen.

We had to wait for the TV Station next door to close down for the night, (stations did that back then) as the Polyskop was "as broad as a bull's bum", & would display their signal as errors.

[vinlove]

Will do more tests suggested over the weekend, and get back with the results.

[vinlove]

I don't quite agree or believe with Radiolistners simple suggestion that my rigs SWR meter is bad, and all 3x of my SWR meters are c%ap.  Because the rigs is quite happy when it is transmitting via 1:1 SWR antenna.

If SWR goes up to more than 1:2 on the inbuilt SWR meter, then it is not happy, and reduce the output power drastically.  I always have to ensure the SWR is less than 1:1.5, and ideally 1:1  using the SWR meters and also inbuilt SWR meter.  They all worked as the rig wanted when TXing for years.

I have worked thousands of QSO with my HF antenna using full power 100W.  And suddenly you get told all these SWR meters are useless and not working and only nanoVNA is the one to trust is a bit hard to take in.

There must be some other factors involved in this situation.

[radiolistener]

I don't quite agree or believe with Radiolistners simple suggestion that my rigs SWR meter is bad, and all 3x of my SWR meters are c%ap. 

It doesn't matter believe it or not, it is really crap if you compare it with VNA.
The same you may not believe in Ohm's law, but it still works despite the fact that you don't believe it.

Scalar SWR meters is not intended for precise measurement. Especially at low transmission power. They are used just to estimate relative SWR level.

Inbuild SWR meters are even worse than external SWR meters. Because the goal of inbuild SWR meter is not to measure SWR. Their goal is to protect your rig when SWR is too high. So they don't need to measure low SWR precisely. Their goal is just to estimate if SWR is acceptable or not. That's it.

[radiolistener]

I have worked thousands of QSO with my HF antenna using full power 100W.

The thing that you didn't realized is that high SWR don't prevent QSO.
High SWR just reduces efficiency of your transmitter-antenna system.
With low efficiency you're needs to spend more power to get the same QSO.
It is also may be dangerous for your rig, because it needs to dissipate more power into heat and needs to keep higher voltage on the amplifier output.

And suddenly you get told all these SWR meters are useless and not working and only nanoVNA is the one to trust is a bit hard to take in.

I didn't told you that your SWR meters "are useless and not working". You can use it for SWR estimation.
They are turns useless when you try to use it for precise measurement of SWR. They are not intended for that purpose.

It is possible to make scalar SWR meter more precise, in order to do that you're needs to replace diode detector with detectors on logarithmic amplifier, such as AD8310.

With logarithmic detectors you can make your SWR more sensitive for low power transmission. But precision of such scalar SWR meter is still limited, due to linearity issues.

[vk6zgo]

I have worked thousands of QSO with my HF antenna using full power 100W.

The thing that you didn't realized is that high SWR don't prevent QSO.
High SWR just reduces efficiency of your transmitter-antenna system.
With low efficiency you're needs to spend more power to get the same QSO.
It is also may be dangerous for your rig, because it needs to dissipate more power into heat and needs to keep higher voltage on the amplifier output.

And suddenly you get told all these SWR meters are useless and not working and only nanoVNA is the one to trust is a bit hard to take in.

I didn't told you that your SWR meters "are useless and not working". You can use it for SWR estimation.
They are turns useless when you try to use it for precise measurement of SWR. They are not intended for that purpose.

It is possible to make scalar SWR meter more precise, in order to do that you're needs to replace diode detector with detectors on logarithmic amplifier, such as AD8310.

With logarithmic detectors you can make your SWR more sensitive for low power transmission. But precision of such scalar SWR meter is still limited, due to linearity issues.

I don't believe the OP is using an SWR meter for "precise measurements".
The difference between an SWR of 1.1:1 & one of 2.87:1 is not a small change.

If the dispute was between a VNA in a large box with "HP", "Anritsu", or "Rohde & Schwarz" on the front panel & a cheap SWR meter, I might be more inclined to agree with you, but a nanoVNA bought on the Internet is a cheap piece of equipment, of often unknown build quality or history.

The "Scalar" term is normally used for "Scalar Analysers" like the Polyskop or similar, which can determine return loss to a high degree of accuracy.
SWR can then be calculated if required.

[Bud]

I have worked thousands of QSO with my HF antenna using full power 100W.  And suddenly you get told all these SWR meters are useless and not working and only nanoVNA is the one to trust is a bit hard to take in.

There must be some other factors involved in this situation.

All these cheap VNAs use wideband detectors, some have the bandwidth extending into GHz region. Any signals within the bandwidth of the detector will skew the measurements when their combined power is comparable with the dicky weak excitation signal of those VNAs. This is the factor you are looking for.

The proper instruments have narrow band detectors and as such measure only at the frequency of interest.  In order to use a half bees dick VNA to measure external aerials you have to connect it via a filter. Ideally via a bandpass filter that has a wider bandwidth than the antenna being measured, but you can try an LPF with say 10MHz cut off (for measurements at 7MHz) and see if that helps.

[radiolistener]

All these cheap VNAs use wideband detectors, some have the bandwidth extending into GHz region. Any signals within the bandwidth of the detector will skew the measurements when their combined power is comparable with the dicky weak excitation signal of those VNAs. This is the factor you are looking for.

SWR meter has wideband detectors (just simple diode). VNA measures SWR value for specific frequency with very narrow bandwidth.

But I think this is not the root of cause for his case. The reason is that he uses low power transmitter and since SWR is too small, the amplitude from reverse wave sensor is very low. It is not enough to open diode detector, so SWR meter just don't see reflected wave and shows low SWR. Just because power is too small and SWR value is not high enough to open diode.

[radiolistener]

In order to use a half bees dick VNA to measure external aerials you have to connect it via a filter.

your words are correct with one exception, it is relevant for SWR meter, not for VNA.
Yes, in order to get better reading from SWR meter, it's better to use bandpass filter on the SWR meter input.

For VNA there is no needs band pass filter, because it already measure SWR for specific frequency with very narrow bandwidth. NanoVNA uses PLL synthesizer for frequency sweep and SA602 mixers to shift frequency down and digitize it with 48 kHz ADC, further processing is performed in digital domain.

But again, filter cannot solve low SWR reading on SWR meter. Because this issue happens because voltage from refltected wave sensor is too low to open diode detector and SWR meter just cannot see reflected wave due to it's small amplitude.

Since SWR meter cannot see reflected wave it shows low SWR. But things will be changed when transmitter power is very high or SWR is very high. In such case amplitude from reflected wave sensor will be high enough to open diode detector and SWR meter starts to see reflected wave and starts to show more correct SWR value. But it happens for high SWR.

When SWR is low, classic SWR meter cannot measure it precisely.

But NanoVNA can do it precisely, because it uses SA602 mixers and ADC with DSP for measurement instead of simple diode detector which is used in SWR meters

[radiolistener]

If the dispute was between a VNA in a large box with "HP", "Anritsu", or "Rohde & Schwarz" on the front panel & a cheap SWR meter, I might be more inclined to agree with you, but a nanoVNA bought on the Internet is a cheap piece of equipment, of often unknown build quality or history.

I'm agree NanoVNA quality may depends on seller, and if topic starter has defective unit, it can show incorrect values.

But if he uses good NanoVNA and performed proper calibration, the difference with "HP", "Anritsu", or "Rohde & Schwarz" will be noticed for SWR < 1.02 (40 dB dynamic range for S11). But if NanoVNA shows SWR=2.9, that's definitely not the case.

I didn't noticed significant difference for S11 measurement with professional equipment. And cannot imagine the reason why it may have significant difference. I assume that NanoVNA works ok and properly calibrated. If we talking about defective unit, this is different story.

The "Scalar" term is normally used for "Scalar Analysers" like the Polyskop or similar, which can determine return loss to a high degree of accuracy.

I'm agree "Scalar" term may be applied also for precise equipment. It just means that equipment measures value in scalar form instead of vector form.

I mean classic SWR meters with diode detectors when I used word "scalar".

[nali]

There's probably some other factor OP hasn't mentioned. Like, I doubt that his 100W rig is using SMA connectors so there must be a patch lead or adapter to the antenna lead. Is that performing OK? How was the nanoVNA calibrated - with patch lead / adapter or without?

Would be interesting to see a S11 sweep or maybe a Smith plot over the band of interest.

[radiolistener]

yes, I'm agree, first we needs to make sure that his NanoVNA works properly and has proper calibration.

We needs screenshots taken with O, S and L loads on CH0 (S11) with enabled Smith chart trace.

[vinlove]

My rig is  using a PL259 patch lead to the SWR meter, and to the antenna from the meter.
I did calibration on the nanoVNA using all three Terminators for Open, Short and Load, and also Thru using the SMA pact cable.

The result? = Same as before, nothing changed.

So if I trim the antenna to bring down the SWR to 1:1 on  the nanaVNA, making the SWr reading on the analogue SWR meter to shoot up, and if the rig does not function, then what?