Test Method for Evaluating ESC (Antenna Diversity)

dj1yr

Once again, it’s all about antenna diversity.
I’d now like to set up a reproducible test configuration to verify antenna diversity.
I have a well-equipped lab with several RF generators capable of I/Q signals, a fading simulator, a noise source, an ARB, a multisignal generator, and more.
I also have audio analyzers, which I used to determine the improvement in signal-to-noise ratio.

Currently, I can’t detect any noticeable difference with my Flex 6700; unless the gain control is active, the phase shift doesn’t work (with the “ESC” function enabled).

When I feed the measurement signals into an Anan 200D/7000MK2/Anvelina Pro3, I can successfully reduce a synthetically generated interference signal by a good 20 dB without any problems and achieve a significant improvement in the signal-to-noise ratio.
So, under Thetis with a dual ADC configuration, the measurement setup works very well.

Under SSDR, no improvement can be achieved; although I can adjust the RF signal level between SCU0 and SCU1, the phase shift does not seem to work.

The measurement setup is as follows: 2 RF generators (SMIQ06B with ARB and noise); the signal in the ARB is 3 kHz-wide white noise with two carriers at 900 Hz and 2300 kHz; PEP signal level -63 dBm; SNR = 12 dB, one generator connected to Ant1, the second to Ant2, TX Slice A = Ant1, DIV + ESC on, Slice H = Ant2.
I also varied the SNR and signal levels of the generators; with Flex, there was no effect on the ESC function, while with Thetis, the diversity function worked very well.

How can I verify the ESC's functionality in the lab?
How does Flexradio test its functionality?

So far, I haven't found any further information on this, so I've tried to develop my own reproducible measurement methods.

Posts on other forums suggest that the ESC hasn't worked since its introduction—which, of course, would be a real indictment.


73´ René

Stan VA7NF

Thankyou for attempting to engage phase shifting on a 6700. I have been waiting 15+ years for this to work and all versions of V4 have failed. Neither noise suppression nor beam forming have worked; I could not see the phase control do anything.

This was promised on the 6700 pre-release V1 software; NOW is a good time to make it work?

Mike-VA3MW

I spent a few hours on this tonight comparing Diversity on my FLEX-6600 to my SDRPlay RSP Duo and I got about the same results on both. I don't have quite the same equipment available, so I thought I would try with what I had available.

I see you tried it with a -63 dBm signal, or 10dB over S9. You might want to try it with a weaker signa.

The test setup:

• IFR-1200S on 14.000 Mhz
• -90dBm signal
• RF Signal Split on a MiniCircuits Splitter for isolation
• RXa for Main Slice
• RXb for sample
• Check the RF-Gain (PreAmps) to make sure they are equal

Keep in mind, you cannot see a change in the panadapter/waterfall. However, on a Carrier, the S-Meter will vary allowing you to see how you are making out.

The static signal is -92 dBm or about S6.

When you vary the Phase and Gain, only the Child receiver is adjusted.

For equal signals on different SCUs I found the gain was a bit hot and once I dialed back the gain to 0.2 to 0.3, I got a very good result. I also noticed that the tuning was a bit tight and critical while the SDR Play was a bit wider.

This is with in a Peak setting and the signal is S7 or -88 dBm. There is a ~4 dB improvement over the single antenna which is to be expected.

If I flip the phase 180, the signal drops to -100 dBm or ~s4 or about an 8 dB reduction.

It may be that the gain is a bit hot if you are doing a bench test.

dj1yr

Hi Mike,

Thanks for taking the trouble to test this.

Exactly what sort of splitter did you use?
With a -90dbm signal from the generator, the RX should receive no more than -93dbm.

I’ll repeat the test using your settings; perhaps there’s a difference between the Flex 6700 and the 6600 or 8600.

I don’t think the -63 dBm PEP (please note the PEP) is a problem or is skewing the results, but you never know what Flex has done.

Unfortunately, there’s no information available on the settings Flex uses for ESC, which is very important for it to work properly.

What’s crucial for me at the moment is that everything works perfectly with Thetis, that the interference suppression sometimes exceeds 20 dB, and that the same antenna configuration has no effect on Flex.

What do you mean by ‘child receiver’?

Who is the reference here and who is the second SCU?


How can I insert images?


73´ René

Mike-VA3MW

The splitter was a minicircuits passive device. It provides isolation and sends 1/2 the signal in each direction.

When you say 20dB, can you elaborate what you are measuring? Is that Audio SNR or RF signal?

The child is the child receiver that is added for diversity. If you are using Slice A normally (Parent) and you then add DIV or ESC the second receiver is the child as it is slaved to the Parent slice.

I am able to paste in images using copy and paste. You should also be able to upload an image from one of the icons below, or can link in a video from YouTube if you wish to upload a video.

Mango

Let me start by stating I am not attempting to start a Flex vs Anan post here. I like both radios for different reasons. I will state however that I have been deeply disappointed with Flex's implementation of the ESC function.

1. I can greatly reduce local interference caused by a nearby solar inverter using Anan/Thetis using their diversity function and cancelling the noise. This does not work with a Flex 8600 using ESC.

2. I use radios beyond the ham bands and DXing on AM radio (a winter time hobby), I can use the "radar screen" using Thetis with the Anan 200D to hear two different radio stations. This does not work with Flex/ESC.

3. Someone at Flex should take a look at how Thetis implements diversity. Their software allows me to move a dot inside of the "radar" to change gain/phase simultaneously. I cannot do this with ESC. I have to use sliders and move them one at a time. (SDR Play does this too, BTW.)

Again, I'm not trying to dis Flex. I would just like someone within the company to see how it is supposed to work properly.

Bill AB7AA

Mango,

Good comments.

9 years ago I bought 2 6700's just because of the Flex promise to implement ESC. Since then, I think that everyone can witness their lack of total commitment to the concept. I hope that someone there now takes it seriously enough to make it effective and easy to use.

73,

Bill AB7AA

Mango

dj1yr's test certainly seems valid and I don't want to take away from his test, but I found the best test (if you don't have the necessary test equipment) is to tune to the medium wave band and look for two competing stations. Using two very good receive antennas with similar receive capabilities, you can null out one station vs the other. This is how diversity should work. This works both with dual receiver Anan radios using Thetis as well as the SDR Duo using SDRConnect or SDRPlay. It is amazing when it works.

Russ Ravella

Hi Mango,

I can't quite tell what you're saying here. Can you DO this with your Flex radio and ESC, or this is a good test approach in general ?

My own experience has been that after many many tries over many many months, I can only very occasionally find a signal that manipulating the ESC controls has any effect on. And when it does, it's so subtle I really have to work it to detect differences. But in general, it doesn't do anything. I have great results with DIV in general though - use it all the time. (8600 and 6700, v4.2.20, various pairs of antennas)

Very much hoping this finally gets addressed in v4.3 which if released as recently implied by Flex, will be eleven months since ESC was released and six months since the last release.

Thanks, Russ KR6W

Mango

Hi Russ,

Yeah. Sorry I wasn't clear. I have NOT gotten the Flex 8600 to work with their ESC function. Period.

With Thetis and the Anan the idea is to treat two separated antennas as two spatially different views of the radio spectrum. By changing the phase and amplitude relationship between those two views, the receiver can make signals from one direction add together while signals from another direction partially cancel. On MW, you can actually hear this happen—you can tune the diversity controls to favor one station, then shift the phase approximately 180° and bring the competing station forward.

I have found AM radio works best for demonstrating this and just made it a suggestion for anyone wanting to test phasing and nulling with dual receiver radios that offer this functionality.

My experience has shown that nulling and phasing signals works as it should using Anan radios with Thetis. I can take those two same antennas and use SDR Duo with their software and accomplish the same thing. It will not work with Flex radios and their latest version of software. It has never worked.

Russ Ravella

RIght, got it. Thanks Mango.

Yeah, there are actually three types of "phase coherent diversity reception" where the ability to sum two different signals phase coherently benefits radio reception. One is just put the output from one antenna in one ear, the other in the other ear. It's what Flex calls "DIV", and is the easiest to implement and works great in SSDR. If the antennas are of different polarity (horizontal beam or dipole for one, classic vertical antenna for the other for example) you get "polarization diversity". Super powerful. Most QSB is actually polarization rotation, not actually signal strength fading. You hear that clearly as the signal apparently moving from left to right and back as the phase rotates without disappearing. I think it's the most underrated performance improvement option in all of ham radio. If more people understood how powerful it is …. Biggest argument there is for more slices as each child slice is a second required slice.

The other two rely on combining the signals first. If you use one antenna that's especially susceptible to objectionable noise, and sum it coherently with the main antenna you can subtract some of the noise by adjusting the relative phase and amplitudes between the signals (Timewave's ANC-N does this). Or if you have two similar antennas, you can electronically point the effective beam formed by the sum of them as you describe above (DXEngineering's NCC-2 does this). Same principle used in phased array antennas for SAR, etc.

I've heard folks attempting to do both of the latter with ESC, always unsuccessfully. Something needs to be fixed in there. I know Flex is aware of an anomalous phase lag and is working on that but I'm not sure that's the whole story, especially if it's constant. I'm really hoping they're finally going to fix it in V4.3. If it's working properly, if you unplug your antennas from your Anan and plug them right into your Flex, it should work exactly the same. It's just coherent summing with relative phase and gain adjustments, no magic (though it's apparently hard to implement in practice). Fingers crossed ! It's been almost a year …

Thanks for the response, Mango ! Russ KR6W