Noise Blanker DSP Values?

gkendric
Posts: 4
Joined: Sun Feb 26, 2023 3:25 am

Noise Blanker DSP Values?

Postby gkendric » Mon May 29, 2023 12:14 pm

I've looked through the topics and the forum responses and there is much talk about the Noise Blanker and ineffective noise blanker settings however I don't see anyone that has shared the values that they found most useful. I am new to the ANAN community and have had my ANAN 7000DLE Mk II for a short time but still am learning. I thought this might be a useful source of info that would help me set up my program. I have Protocol 2 and Thetis 2.9.0.

DSP (NB/SNB) setting values are currently as follows:

Noise Blanker
Threshold = 25

SNB
Threshold 1 = 8.0
Threshold 2 = 32.0

Times (ms):
Slew - 0.55
Lead - 0.18
Lag - 0.71

NB2 Mode
Linear Interpolate

I wish I could say I understand what all the Thresholds and Times were and how they were achieved but I'm not and know i have yet a lot of study and research to do. I've always heard about how special the receiver of the ANAN is and how it cuts interference and has such quiet performance. I do believe this and I would love to know what's best to have for the readings. I'm also using the ANAN 7000 for mostly SSB but also enjoy CW but mostly on the lower bands 80 - 30 Meter Bands.

Much appreciation to those who respond.
User avatar
Tony EI7BMB
Posts: 651
Joined: Sun Apr 09, 2017 2:31 pm
Location: Dublin
Contact:

Re: Noise Blanker DSP Values?

Postby Tony EI7BMB » Mon May 29, 2023 1:50 pm

I would be very interested in this too . Terrible noise here on 6M in USA direction
K1LSB
Posts: 639
Joined: Wed Feb 05, 2020 5:25 pm

Re: Noise Blanker DSP Values?

Postby K1LSB » Mon May 29, 2023 2:17 pm

Here are my settings, I can't remember how or why I arrived at these particular values, it was basically trial and error but I have no idea how optimal the values are. Noise Blanker has never really behaved as I would wish but that may just be the nature of how it works.

Capture-2.jpg
Capture-2.jpg (19.58 KiB) Viewed 3345 times

Mark
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: Noise Blanker DSP Values?

Postby w-u-2-o » Mon May 29, 2023 2:21 pm

Remember that noise blankers only work on impulsive noise. They will be pretty much ineffective against normal band noise. For band noise we have the NR modes.

If none of the noise blanker modes (NB, NB2, SNB) are having any substantial effect then it's likely the noise in question is not impulsive.

Perhaps a discussion of the types of noise you are experiencing might be worthwhile.

I have no input into modifying the various NB settings. However, I will note that the NR2 gain method should always be set to Gamma for maximum effectiveness.
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: Noise Blanker DSP Values?

Postby K9RX » Tue May 30, 2023 11:45 am

I have some high impulse noise issues on 6M... and fortunately, most of the time, don't have overly strong signals that mess with the NB action like it does on HF. On HF I set the Threshold at 24. That works well as a compromise. On 6M I will run it there or maybe down to 20. But on 6M EME where I am assured of no high signal levels I run it at 13. By going to 13 I gain an additional 1.5-2 db of noise floor reduction depending of course where I am pointed (relative to the noise sources).

Gary
K9RX
User avatar
KA5KKT
Posts: 129
Joined: Thu Aug 06, 2020 6:51 pm

Re: Noise Blanker DSP Values?

Postby KA5KKT » Sat Oct 28, 2023 7:05 pm

For the past couple of months I've had high noise levels. Based on experience I surmise the noise is Line Noise. I seem to be getting a lot of Noise Blanker overload.

Is the perceived lack of performance of the noise blankers in the Anan predominantly tied to hardware or software?


Kind regards,
Dick - KA5KKT
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: Noise Blanker DSP Values?

Postby w-u-2-o » Sun Oct 29, 2023 3:08 am

KA5KKT wrote:For the past couple of months I've had high noise levels. Based on experience I surmise the noise is Line Noise. I seem to be getting a lot of Noise Blanker overload.

Is the perceived lack of performance of the noise blankers in the Anan predominantly tied to hardware or software?
Neither. Any perceived lack of performance is tied to unrealistic expectations.

Noise blankers are designed to help suppress impulsive noise. Impulsive noise is relatively rare, unless perhaps you are operating in a moving vehicle with an ignition system. On the other hand random noise is ubiquitous and endemic. If there is some uncertainty with respect to what kinds of noise are present and activating any of the noise blanking modes makes little if any difference, then the noise that is present is undoubtedly random noise. This is especially true if SNB is ineffective, as SNB is substantially more capable than the legacy NB modes, the latter being what one would find in any common radio from the Big Three makers. When faced with random noise, and this is always the case, random noise should be mitigated using NR or NR2. FWIW, I never turn NR2 off unless running digi modes. And if you do have impulsive noise you can certainly run NR and NB simultaneously.

Finally, be sure to set the step attenuator to a value that is commensurate with band conditions. To do this simply keep adding attenuation and watch the panadapter until you see the SNR of the largest signals start to drop (this will be because system noise figure starts to exceed the natural noise floor). Then back off a couple or three dB. This will optimize gain for best ADC dynamic range under those conditions.

If you haven't already seen it, or haven't seen it in a while, take a look at the first video referenced in this topic: viewtopic.php?f=9&t=4018
User avatar
KA5KKT
Posts: 129
Joined: Thu Aug 06, 2020 6:51 pm

Re: Noise Blanker DSP Values?

Postby KA5KKT » Sun Oct 29, 2023 3:32 am

A) Is so call line noise not impulse noise?

B) SNB seems to neutralize quite a bit of the low end of the audio frequency...eliminating much of the presense.


As ever,
Dick - KA5KKT
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: Noise Blanker DSP Values?

Postby w-u-2-o » Sun Oct 29, 2023 10:34 am

It's not clear what you mean by "line noise". Is this 60 cycle hum?

If SNB is helping then by all means use it.
User avatar
KA5KKT
Posts: 129
Joined: Thu Aug 06, 2020 6:51 pm

Re: Noise Blanker DSP Values?

Postby KA5KKT » Sun Oct 29, 2023 2:46 pm

Power Line Noise -
Virtually all power-line noise, originating from utility company equipment, is caused by a spark or arcing across some power-line related hardware. A breakdown and ionization of air occurs, and current flows between two conductors in a gap.
per ARRL

https://ncetest.com/buzz-kill-heres-why ... ake-noise/
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: Noise Blanker DSP Values?

Postby w-u-2-o » Sun Oct 29, 2023 5:59 pm

Is this what your noise sounds and looks like? If so, then SNB is already doing as good or better job than any other NB algorithm.

Consider also that you might be dealing with more than one type of noise.

Trucker
Posts: 304
Joined: Wed Nov 03, 2021 5:16 pm

Re: Noise Blanker DSP Values?

Postby Trucker » Sun Oct 29, 2023 7:21 pm

I have found with Noise Blankers, there is no, "one size fits all", noise blanker. Thetis gives a lot of options in how the noise can be handled. But, even then, it may not be enough. If it is something similar to Scott's example, the answer might be to try and track down the source of the noise. I have found that some noise blankers work better than others. If I cannot locate the source of the noise, I turn to the tools in Thetis and my other radios. Sometimes, if the source is a neighbor, you "might " be able to get them to let you help find and hopefully, fix, the problem if it is coming from something on their property. Otherwise, you then have to research other alternatives.
Good luck. I know from experience how frustrating noise from outside your home can be to find and correct.
James
WD5GWY
User avatar
KA5KKT
Posts: 129
Joined: Thu Aug 06, 2020 6:51 pm

Re: Noise Blanker DSP Values?

Postby KA5KKT » Mon Oct 30, 2023 5:52 am

...some clarification.

"A) Is so call line noise not impulse noise?

B) SNB seems to neutralize quite a bit of the low end of the audio frequency...eliminating much of the presense."

To Scott... Yes, my noise is quite a bit like that in the link you posted. I will have to say that I haven't 'looked at it' over as broad a range that was displayed in that link, but I don't suppose that makes a lot of difference.

* Again, is so-called <power> line noise not impulse noise?

* Yes, I know that some noise blankers are not particularly effective against some types of noise. And regarding my B) statement, SNB seems to neutralize quite a bit of the low end of the audio frequency (of the signal I'm trying to listen to) and thus diminishing the 'presense' (of what I'm trying to hear through the noise)...not so much the noise itself be it impulse or however it may be classified. So while SNB has an effect in this situation, I can't say the effect does much to clarify what I'm trying to hear.

Yes, your suggestion and my experience of the past couple of months cause me to consider that my expectations might ought to be downgraded.


As ever,
Dick - KA5KKT
rdwing
Posts: 92
Joined: Sat May 21, 2022 2:05 am

Re: Noise Blanker DSP Values?

Postby rdwing » Thu Nov 02, 2023 7:03 am

I unfortunately suffer from a ton of line noise, specifically 60hz and 120hz fundamental broadband hash. Actually even worse than the video Scott linked, and even worse there are similar noises that I suspect are caused by neighboring SMPS...

Regular NB1 works great for that, sometimes, because it's a WNB (wideband noise blanker). For the right noise can cancel out 20 dB of hash garbage from the panadapter and waterfall. I have found it works best at the default settings, Threshold 25-30, and Slew/Lead/Lag at 0.01.

My only bone to pick is that like a lot of noise blankers, a strong signal will tend to distort both that strong signal itself, and raise the noise floor and distort other nearby signals of interest. This effect can wipe out weaker signals nearby. It is possible to tune the slew control between 0.01 and 0.05 and try to contain this effect.

NB2 is an interpolating noise blanker, essentially NB1 but can be configured to linearly interpolate the previously removed impulse region of signal. It also has other modes to estimate the original signal. Generally I have found NB2 to not work as well as NB1, and to be more sensitive to the raising of the noise floor when used in the presence of a strong signal nearby.

Additionally, the SNB (Spectral Noise Blanker) behaves like an AF noise blanker. It is frankly just not very effective here.
I think what Dick is saying is that currently in Thetis, the SNB has a forced low cut of 300 Hz, in that ballpark. And thus it tends to wipe out what audio is recovered anyways.

It is capable of having its output bandwidth adjusted, but it doesn't currently seem to be functioning. I will open a bugreport for that.

I took a video illustrating the noise and how the noise blankers function here, as well as an older ones showing one of the noises previously.
https://youtu.be/HCKnaHiKU2s
https://youtu.be/zoQbKFPbsSc?si=WCy6bmv2RRRLpIoG&t=30
W7GES
Posts: 76
Joined: Sat Mar 13, 2021 6:56 pm
Location: Phoenix AZ

Re: Noise Blanker DSP Values?

Postby W7GES » Fri Nov 03, 2023 4:53 am

I had terrible power line noise to my NE that the NB circuit in Thetis (or my TS-890) could make a dent with.

Sometimes it was easier to aim my beam SW to work Europe since I could hear them better off the back of the beam (and no this wasn’t during times when Long Path was open).

Using Thetis I recorded the noise floor values of 10-80 meters with my beams pointed NE and SW and recorded the values, times, etc in a spreadsheet.

Since I live in Phoenix, AZ it is rare that we get rain. After a rainy day I noticed no line noise (and recorded this in my spreadsheet).

I called the electric company and they created a problem ticket. After a few weeks (and an email for help on my contest club email reflector) a local ham mentioned he knew the electric company troubleshooter and he would give him a call.

A few days later I was contacted by the power company trouble shooter. He said my trouble ticket had been closed. Evidently a normal crew drove by, didn’t see anything obvious and closed the ticket.

He reopened the ticket and said he was doing noise abatement procedures for the Super Bowl preparations but would try to fit me in when he was finished.

A few weeks later he called and said he would be at my house in 20 minutes.

He drove by my house on the N/S street and the E/W street and did pick up line noise.

When he stopped at my house I took him to my shack and showed him my spreadsheet. I also showed him a screen grab video I had taken when the line noise was starting up again the day after a rain storm.

Then I started Thetis and showed him the waterfall / spectrum display as I rotated the antenna from SE > NE > N > NW.

He said I could walk along with him while he used his noise locating equipment.

3 Powerpoles later he said the noise was so bad it pegged his meter. He set up the bucket truck, extended it near the transformer and told me my noise would be gone when we went back to my ham shack. He said a wire was routed incorrectly and was arcing.

He was correct. This was on Dec 1, 2022 and although I get a little interference from neighbor’s pool motor’s, etc once in a while the horrible line noise hasn’t returned.

Below are screen shots from when he arrived and when he left. The top display is both beam antennas after the problem was fixed, the bottom display is from the antenna pointed at the power line. Unfortunately 15 meters didn’t have any activity when I took this screen shot but trust me my noise level is now low…

The screen shots are from N1MM as I could show the lineman both radios at once. I run Thetis on a stand alone computer so the noise floor readings are not visible on this computer but it’s easy enough to see there’s a giant difference.

Sorry for the long email, I just wanted to point out that it is worthwhile to document your noise and call your power company.

I hope you are able to get resolution for your noise. I don’t think there is going to be a noise blanker algorithm that will be able to fix it.

73,

George / W7GES
Attachments
2022_1201_1836z_w7ges_after.png
2022_1201_1836z_w7ges_after.png (1.11 MiB) Viewed 2402 times
2022_1201_1716z_w7ges_before.png
2022_1201_1716z_w7ges_before.png (1.25 MiB) Viewed 2402 times

Return to “Thetis”