Accuracy re fidelity on/with the TX display while using PS

K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Accuracy re fidelity on/with the TX display while using PS

Postby K9RX » Sat Feb 17, 2018 12:58 am

I've been using Puresignal for months now ... I've had people stop in and say how narrow the TX'd waveform was ... except on 75 meters. Twice now over the last couple of weeks I've had two different stations (both W4's) break in and say I was spattering - bad. They said i was not TOO strong ... last night the guy said there was significant energy on the USB part of the spectrum. He said it was at best only 20 db down.

Here's the thing - the displayed view, using DUP, showed at least 55 db down on both sides and I had the proper green "correcting" boxes going on.

So ... what gives? I thought what I see on the panadapter while TXing using DUP is indeed my signal - what was going out the coax ...

so the main and really only question here is:

assuming I have it wired up correctly (I'm 100% sure of that - yes on the output of the amp), assuming that the components are working correctly....

is it possible that it is spattering and yet I'm not seeing it?

Gary
K9RX

[ps: I had a few days later someone else tell me he was also on frequency listening at the last time this happened and he didn't see anything wrong with my signal - BUT he was a bit unsure on how to properly evaluate for quality on his panadapter.]
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Sat Feb 17, 2018 3:40 am

Gary,

Is it possible that these people are messing with you?

Get someone you trust, that has good copy on you, S9 at least, S9+20 would be better, to grab a screen shot of their (properly adjusted) panadapter and email it to you.

Another option would be to sign into one of the many web-enabled SDRs and make your own real-time evaluation.

I bet that it looks just fine.

73,

Scott
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby K9RX » Sat Feb 17, 2018 2:08 pm

Scott,

That was one thought, that they're messing with me (and IF so possibly/probably know each other) ....

but I was still hoping to get an answer to that one question: if I am seeing a clean signal IS it indeed then a clean signal or is it possible that I could see a clean signal (DUP on, feedback from the amp output) and yet have an issue ...

that stated I hadn't thought about a Web-sdr ... I had thought about buying one but this is a great suggestion Scott - thanks.

Gary
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby K9RX » Sat Feb 17, 2018 2:19 pm

Unless you can specify a better source of Web-SDR that won't work - they have very crude displays. Doubtful it could be used for qualitative analysis... I'll see if I can borrow a 7300 from a local ham ...

Gary
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Sat Feb 17, 2018 3:16 pm

K9RX wrote:...but I was still hoping to get an answer to that one question: if I am seeing a clean signal IS it indeed then a clean signal or is it possible that I could see a clean signal (DUP on, feedback from the amp output) and yet have an issue ...
Assuming a proper coupler setup, I have never seen a case where what you saw on the DUP display was not ground truth. Unless there are some non-linear junctions in your feedline or antenna, that pretty much has to be the case!

that stated I hadn't thought about a Web-sdr ... I had thought about buying one but this is a great suggestion Scott - thanks.
If the web SDRs are not cutting it for you, then definitely consider buying one of the low cost RTL based SDR radios. I bought a Nano 3 and have been very happy with it. It works sufficiently well on 10M so that I can use it to look/listen to my off-air signal in that band. I've been using it with both SDR# and SDR Console. It's not a high performing receiver to be sure, but it's plenty good for off-air monitoring and does a creditable job of listening to local public safety radio and ADS-B. I've even decoded some P25 with it.

I'm happy enough with it that I may also buy a Ham It Up upconverter for it so I can monitor below 10M as well.
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby K9RX » Sat Feb 17, 2018 3:41 pm

Yes - I had considered that - that indeed something at the antenna or the relay box is non-linear ... and indeed that might be the case. I didn't want to mention that ahead of time as I first and foremost wanted to ascertain if indeed the signal at the output of the amp coupled back is the real signal. I had to believe it was .... so now I can focus on whether or not there is an issue at the antenna/feed.

thanks. Meanwhile I've been looking at getting a backup radio (since I've had 2 failures with the 8000, one requiring a replacement radio) ... whatever it is it will have a halfway decent panadapter built in.

Gary
WB2ZXJ
Posts: 5
Joined: Sun Apr 09, 2017 3:04 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby WB2ZXJ » Sat Feb 17, 2018 3:45 pm

.
Last edited by WB2ZXJ on Sun Nov 18, 2018 9:45 pm, edited 1 time in total.
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Sat Feb 17, 2018 3:46 pm

7300's are the best bang for the buck, bar none. Used ones are now <$1K. You might also consider a 10E, I've seen them for <$700. You won't get full power from the amp, but there's a lot to be said for commonality of operation with your 8000.
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Sat Feb 17, 2018 4:36 pm

I love my SDRplay RSP1 as a bench tool for doing audio checks and visual signal IMD checks of my ANAN. It's relatively cheap (120 bucks at HRO) for a VLF through 2.0 GHz receiver and the SDRplay hardware runs SDRuno, SDR Console, SDR# and probably a few other pieces of software I haven't discovered yet. It runs fine through its USB connection simultaneously on the same machine as it's running my ANAN with OpenHPSDR mRX PS.

73,

Rob W1AEX

SDRuno panadapter display showing my ANAN-200D driving my ALS-1306 to the 1KW level.
Image
"One thing I am certain of is that there is too much certainty in the world."
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Sat Feb 17, 2018 6:06 pm

One of the things that could cause a disparity between the DUP display and the actual RF output spectrum of the amplifier is non-linearity being introduced by the the coupler. I.e. if the output of the coupler's sampling port is not linearly proportional to the RF output of the amplifier*

Therefore, when using a separate receiver, it would be best to use a short pick-up antenna attached to the separate receiver's input, rather than connecting the separate receiver to the coupler's output port.

*Pure Signal pre-distorts the input to the transmit D/A converter to cause the input to RX1 to have minimum distortion. If there is anything in the path between the amplfier's RF output and the input to RX1 (i.e. the coupler, the RX1 input step attenuator) that has a non-linear input-output relationship... then Pure Signal will cause the actual output of the amplifier to have the inverse non-linearity.

Stu
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Sat Feb 17, 2018 8:42 pm

AB2EZ wrote:...it would be best to use a short pick-up antenna attached to the separate receiver's input...
That's what I do. I want an over-the-air look at things.

This is what I use. I just stand it up on the windowsill in my studio:

Image
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Sun Feb 18, 2018 7:19 pm

Rob

I'm thinking of buying an RSP1A from HRO

If you are transmitting with your ANAN + external AMP... and also using your RSP1 for monitoring your amplifier's RF output spectrum...

How do the 2-tone spectrum readings of IP3 compare (ANAN DUP vs. RSP1) when you have

a) Pure Signal turned off
b) Pure Signal turned on

Thanks
Stu
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Sun Feb 18, 2018 9:53 pm

Hi Stu,

I never thought to try that so I'll have to check that out to be sure. When I make my usual run of the bands tomorrow morning I'll set things up and take two screenshots so you can look for yourself. I'm betting they will be pretty close.

73,

Rob
"One thing I am certain of is that there is too much certainty in the world."
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Mon Feb 19, 2018 6:04 pm

Hi Stu,

I set the RSP1 up with a 6 inch whip antenna and positioned it about 2 feet from my dummy load which the ANAN and amplifier were connected to. After a few test runs I found that it was necessary to disable the AGC in SDRuno to prevent it from "hunting" for a level the RSP-1 was happy with. I think this is an indication that there are some proximity and overload effects going on but I'll probably need to set the RSP1 up on a different computer to test that. In the first image I ran a two-tone test with the ANAN driving the ALS-1306 to 500 watts. In the second image, the ANAN is driving the ALS-1306 to 750 watts in the SSB mode. I did not re-size the images so they are full desktop size to make it easier to read the level scales on both panadapters. I'm not sure what to conclude except that the blooming at the base of the two-tone test seems to indicate that there is some overload going on in that test. Overload does not seem to be as much of a problem when running SSB for some reason. Let me know what you think.

73,

Rob W1AEX

NOTE: If you click on the link that appears over each screenshot you can view them at full size.

http://www.w1aex.com/archive/ANAN_500W_2TONE.jpg
Image

http://www.w1aex.com/archive/ANAN_750W_SSB.jpg
Image
Last edited by W1AEX on Mon Feb 19, 2018 7:04 pm, edited 3 times in total.
"One thing I am certain of is that there is too much certainty in the world."
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Mon Feb 19, 2018 6:58 pm

Rob

The 2-tone spectrum looks reasonably clean on the RSP1 display... but it is difficult to confirm that the IMD products are more than 40dB below the principal tones. I.e. -53dm - (-93dBm).

On the SSB displays... its appears (with some interpretation to ignore the 2nd peak on the RSP1 display) that the IMD in the adjacent 3kHz frequency band is at least 50dB below the in-band modulation products.

I agree that using the RSP1 with a separate computer (perhaps a laptop running on battery power) in a more distant location (perhaps 25 feet from the amplifier and dummy load) might produce a more compelling result.

Stu
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Mon Feb 19, 2018 9:02 pm

Rob,

In your two-tone case you are demonstrating that results on the RSP1 are better than that shown by DUP. IMD products on the DUP display are down around -60dBc. On the RSP all you can see is the phase noise, probably mostly from the RSP1. The noise floor on the RSP1 appears to be -70dB from the signal peak, but I bet those IMD products are buried in the phase noise.

That's the long answer. The short answer is that the DUP display is a reasonable and accurate representation. I've had occasion to check the output of my Xtronic coupler with a "real" HP spectrum analyzer and it matched up within a dB or two to the DUP display, given equivalent settings in PowerSDR and the spectrum analyzer.

The equivalent settings issue is an important one, by the way. If you are using different FFT parameters things are going to look different.

73,

Scott
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Mon Feb 19, 2018 9:46 pm

Scott
Rob

Assuming that the software being employed with the RSP1 allows one to increase the averaging time... shouldn't one be able to significantly reduce the effect of phase noise in the panadapter display produced in the 2-tone test by increasing the averaging time to a much larger value?

Stu
Last edited by AB2EZ on Mon Feb 19, 2018 11:35 pm, edited 1 time in total.
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Mon Feb 19, 2018 10:04 pm

Stu and Scott,

Thanks for the assessment. It does seem like the RSP1 generates quite a bit of noise above and below the center frequency of the transmitted signal. As you can see in the AM screenshot below, when looking at a voice transmissions (either SSB or AM) the phase noise is well masked and the signal quality is excellent, but when a naked carrier or a two-tone test is displayed, any anomalies that might be present near F-0 are covered up.

Over the next few days I'll try the RSP1 on my gaming computer which is across the room and roughly 20 feet away from the ANAN operating bench. Perhaps that will minimize the proximity issue.

73,

Rob W1AEX

http://www.w1aex.com/archive/ANAN_100W_AM.jpg
Image
"One thing I am certain of is that there is too much certainty in the world."
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Mon Feb 19, 2018 10:11 pm

I just saw your post above Stu. The SDRuno software does allow quite a range of choices for FFT averaging in the spectrum display. I'll read up on that and give some larger values a try.

73,

Rob

Image
"One thing I am certain of is that there is too much certainty in the world."
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Mon Feb 19, 2018 10:27 pm

The FFT averaging value goes as high as 4096, which resulted in the screenshot of a 500 watt two-tone signal shown below. I keyed the two-tone test and watched the tones rise up over a span of 2 or 3 seconds to the point where they did not go higher and then took the screenshot. I think I will need to increase the coupling between the 6 inch antenna probe and the dummy load until I begin to see the third order products at the baseline. The higher FFT value seems to work pretty well Stu although the spectrum display is waaaaayyyy too slow for normal viewing!

73,

Rob

http://www.w1aex.com/archive/FFT_4096.jpg
Image
"One thing I am certain of is that there is too much certainty in the world."
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Tue Feb 20, 2018 5:32 am

Rob
Scott

Thinking about this some more....

The power leaving the transceiver is only around 13dB less than the power leaving the external amplifier.

With a dummy load attached to the external amplifier (and with the amplifier input being equivalent to a dummy load at the output of the transceiver)... it is not obvious that the power being picked up, by the small antenna, from the external amplifier's output is more than 13dB greater than the power being picked up by the small antenna from the transceiver's output.

To do this right, the output of the external amplifier should be attached to the regular transmitting antenna... and the auxiliary pickup loop should be located far enough from the transceiver so that one can verify that the signal being picked up by the auxiliary receiver

a) when the amplifier is on, is at least 30dB greater than

b) when the external amplifier is off, and the output of the external amplifier is connected to a dummy load.

This will ensure that with the external amplifier on, and the main antenna connected to the amplifier's output, the signal being picked up by the auxiliary receiver is sufficiently dominated by the signal from the main antenna (i.e. the external amplfier's output).

Stu
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby w-u-2-o » Tue Feb 20, 2018 2:32 pm

These sort of multipath problems are largely avoidable. The entire point of having and using a coupler is to avoid them. If you don't trust your coupler, then put the coupler on the test bench and test it.

If you don't trust the DUP display, then take the output of the coupler, pad it down a bit if you have to, apply it to a spectrum analyzer or other receiver of your choice, and compare that to the DUP display.

However, if you don't trust your antenna system after the amp and think there is some kind of non-linearity in your feedline or whatever, then, yes, you need to receive that off the air, and the instrumentation antenna should be located somewhere that provides a signal to noise condition that is appropriate to your test requirements.

73,

Scott
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: Accuracy re fidelity on/with the TX display while using PS

Postby K9RX » Tue Feb 20, 2018 4:08 pm

Stu,

So .... if I understand what you are suggesting you are saying that if I have another SDR radio sitting 12' away with say a wire antenna that the dominant signal isn't going to be from my main transmit antenna that is 500' away - even though it will have 1000W+ to it? Even at 1/d^2 I'd think the signal would be far stronger than any "leakage" for lack of a better term, from the equipment sitting there... or am I miss reading what you're saying?

Gary
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Tue Feb 20, 2018 4:49 pm

Stu, Scott, and Gary,

In this screenshot the combination of the ANAN (running 30 watts) + 1306 amplifier is transmitting a 700 watt two-tone signal into my station antenna while the SDRuno is receiving the signal with a 6-inch 2 meter HT antenna from about 75 feet away from the transmit antenna. The SDRuno spectrum display is using 4096 for the FFT setting. I found that the difference between the third-order peaks and the fundamental would change depending on where the receiver AGC was set which probably means the SDRplay is a little unhappy with the nearly 50-over S-9 signal it is looking at. The third order peak ranged from -40dB to -45dB down depending on the AGC setting selected.

The third-order IMD results were not as good as seen when running the transmitter into the dummy load where the signal strength into the receiver ranged between 20dB to 30dB over S-9. I think I would agree with Gary that the amount of signal that might leak out of the ANAN while running ~30 watts would be quite small compared to the power being dissipated and radiated by the dummy load. Over-the-air tests with other stations on 75 meters typically indicate the third-order products are down around -50dB with the 200D and ALS-1306.

73,

Rob W1AEX

http://www.w1aex.com/archive/sdruno_4096.jpg
Image
"One thing I am certain of is that there is too much certainty in the world."
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Tue Feb 20, 2018 7:32 pm

Gary

You appear to have misunderstood my recent post. Please read it again.

Stu
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Tue Feb 20, 2018 7:53 pm

Rob

I think your latest posted measurement results are very helpful, but:

You wrote

"The third-order IMD results were not as good as seen when running the transmitter into the dummy load where the signal strength into the receiver ranged between 20dB to 30dB over S-9. I think I would agree with Gary that the amount of signal that might leak out of the ANAN while running ~30 watts would be quite small compared to the power being dissipated and radiated by the dummy load. ..."

I don't think a dummy load, connected to an external amplifier with coaxial cable, is necessarily any more efficient a radiator than the coaxial connection from the output of the transceiver to the input of the amplifier. Maybe it is... and maybe it isn't... depending upon the details of the various cables being used, the mechanical design (shielding) of the dummy load, the mechanical design (shielding) of the external amplifier, etc.

If the amplifier is producing 700W and the transceiver is producing 30W... then there is no certainty (without confirmation via measurements) that the input power to the auxiliary receiver produced by the transceiver is more than 13.7 dB below the input power to the auxiliary receiver being produced by the amplifier.

That is not a great enough contrast for an accurate measurement of IMD at the output of the external amplifier for cases in which the external amplifier's output IMD... with Pure Signal on... is more than 13.7dB below the external amplifier's output IMD with Pure Signal off.

I have previously posted the mathematical analysis to prove this assertion... but I don't expect too many people to read it.

Stu
Last edited by AB2EZ on Wed Feb 21, 2018 3:23 pm, edited 1 time in total.
User avatar
W2PA
Posts: 166
Joined: Sun Apr 09, 2017 6:34 pm
Location: LaGrangeville, NY
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W2PA » Tue Feb 20, 2018 11:47 pm

I read it, Stu (I read nearly everything ;-). Your math makes sense.

I think your point, put another way, is that PS corrects for net IMD through the entire transmit chain, exciter plus amplifier. This means the output of the exciter alone has not been corrected optimally. One wants to make sure that those IMD products aren't superimposed on the (probably much better) net spectrum, in the presence of 13.7 dB gain in the amp. Yes?
73,
Chris, W2PA
AB2EZ
Posts: 99
Joined: Sun Apr 09, 2017 2:29 pm
Location: Princeton, NJ

Re: Accuracy re fidelity on/with the TX display while using PS

Postby AB2EZ » Tue Feb 20, 2018 11:54 pm

Chris

Yes... well stated

73
Stu
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Wed Feb 21, 2018 10:00 pm

Stu and Chris,

I have to admit that the math is well beyond my skill level (my wife looks for my handwriting whenever the checkbook doesn't balance) but between both of your explanations the light went on and I see what Stu is driving at with the verification of the (a) and (b) scenarios outlined previously in the thread. Thank you for the clarifications!

73,

Rob
"One thing I am certain of is that there is too much certainty in the world."
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: Accuracy re fidelity on/with the TX display while using PS

Postby W1AEX » Sat Feb 24, 2018 3:43 pm

Stu,

Yesterday, I had a chance to run the scenario you outlined (in italics below) and found a bit less than 20dB difference in the signal level between setup conditions a and b. It looks like running the auxiliary SDR receiver on the same computer as the ANAN adds all kinds of potential for ground loops and other RF paths that make a mess of things. Until I figure out a setup that is reliable and gives repeatable results I'm going to go back to using the K3FEF webSDR page to get a visual of my OTA signal and also to listen to what it sounds like in the voice modes. The propagation when I made the screenshots below was not great as a 500W two-tone was not enough power to make the 3rd order peaks show up, but at least it confirmed that my S-9 + 10dB signal looked clean as a whistle.

73,

Rob W1AEX

"To do this right, the output of the external amplifier should be attached to the regular transmitting antenna... and the auxiliary pickup loop should be located far enough from the transceiver so that one can verify that the signal being picked up by the auxiliary receiver

a) when the amplifier is on, is at least 30dB greater than

b) when the external amplifier is off, and the output of the external amplifier is connected to a dummy load."


http://k3fef.com/

500 watt two-tone
Image

500 watt SSB signal
Image
"One thing I am certain of is that there is too much certainty in the world."

Return to “PureSignal Operations (PowerSDR & Thetis)”