Re: K3S issues - audio hiss

[Roger-3]

Deviating from the original topic but somehow related to the noise gate I thought this may be of general interest as at least one other K3S user observed the same thing –  a HB9 station has the same phenomenon with his K3S (see below).

 

Bob:

 

I tried different microphones (MH4, SHURE 444, HEIL HC-5) with mic gain and compressor settings down to inacceptable and with all three of them I get the same effect. That is with PTT pressed and the microphone level other than “0” and the noise gate OFF on an SDR receiver I see a tx noise floor block according to the filter bandwidth. However at approx. 2,5 kHz I see a “needle” that reaches 17 dB above the ambient noise floor. If I turn on the noise gate the tx noise floor is suppressed and so is the needle. Speaking into the microphone in order to open the noise gate the needle is masked by the speech energy however in breaks and before the gate closes again the level drops to the ambient noise level and the needle becomes visible again. It is always there but just covered by the speech energy. That is what makes the modulation ringing on top and at the end of each word. Reducing the compression has no effect other than lowering the overall signal – the tone is still there.  I even tried to lower the 2,4 kHz level in the TX EQ but it has no effect as the noise is right on 2,5 kHz. The only thing that would help I guess is a software notch filter in the DSP that eliminates this spur or artefact or best avoid it at all. Disconnecting the microphone doesn´t change anything either. As soon as the Mic Gain level is other than “0” the tone is there microphone connected or not. So it seems it is an artefact that´s created in the radio but absolutely no noise pickup from the microphone.

 

I tried 4 different radios – legacy and SDR and none of them shows a single tone on top of the tx noise floor but a smooth and more or less straight top of the tx noise caused by ambient noise. If no SDR is on hand a local remote SDR will do.

 

73s Roger, DL5RBW

 

From: Bob McGraw K4TAX [mailto:[hidden email]]
Sent: Monday, September 16, 2019 8:05 PM
To: Roger
Subject: Re: [Elecraft] K3S issues

 

Roger:

Radio wise, I set my K3S just as I normally use it for SSB QSO's.  The mike is selected at FP.H and the mike is a Shure SM-58 dynamic mike.  The MIC gain value is 35 and the CMP value is 15.   When monitoring the radio using another receiver, I do hear some noise in the audio spectrum.  This noise is displayed on the SDR receiver using HDSDR as being in the 1500 Hz to 2500 Hz range with a peak value about 2200 Hz.   Setting the MIC gain value to 0,  I find the noise goes away.  Nothing observed or heard.  The setting the MIC gain back to 35 but disconnecting the mike from the front panel and then keying the radio as before,  there is no noise.   My conclusion is that the mike is just picking up ambient sound in the room.  The use of the Speech Processor does bring the noise level up as opposed to when the CMP is set to a value of 0.   I normally run the CMP at a value of 15.

My conclusion is there is nothing I can see or hear coming out of the radio, unless it it going into the radio via the connected microphone.  The measurements show the level to be some -70 dB below 100 watts PEP.   With 100 watts PEP being 50 dBm re 50 ohms, then - 70 dB below that is 0.00001 watts.   Now if your friend is seeing your 10M signal at -43 dBm or S-9 +30 dB, then a signal that is -70 dB down is still and S-2 signal of -115 dBm.  In a quiet band condition, that could easily be heard.

My measurement were with the noise gate set to OFF.

73

Bob, K4TAX

 

 

On 9/16/2019 12:30 PM, Roger wrote:

Hi Bob,

 

it is strange with that noise as nobody ever told me but gave me unsolicited reports on the good audio which I believe is true when it comes to frequency response. The first time I encountered this problem is when F5VNE and I4LEC told me about it on 40 m SSB 4 weeks ago. We checked and rechecked and finally it appeared that it was gone and the TL-922 ´s fan was the noise source. Signals went down by that time and so I guess we came to a wrong conclusion.

 

Then a week ago a local friend (1 mile away) immediately told me about this noise. As there was no QSB and the signals well over S9 I tried all the things I described including battery power to make sure the switching mode powersupply is not causing any trouble. No way I could get rid of the noise except cranking down the mic gain to the level where the audio became drastically weaker and the noise gate took care of it. As soon as the noise gate opened when talking the “ringing”  could be heard in the background. However it is best heard with the noise gate disabled and just the PTT pushed.

 

I wonder what you find out.

 

 

73s Roger, DL5RBW

 

From: Bob McGraw K4TAX [mailto:[hidden email]]
Sent: Monday, September 16, 2019 3:16 PM
To: Roger
Subject: Re: [Elecraft] K3S issues

 

Roger:

Thanks for the detailed response.   As my friend says "there's a lot of buckshot in that message".  On the other side, as is often said, "if it is working, don't try to fix it".  I agree with that.

Perhaps time permitting today, I'll make an effort to look closer for the noise you describe as a distant cell phone ringing.  I've not had any reports of noise on my signal in this regard.  Some of the fellows I chat with are serious audio critics and I'm sure if something was amiss I'd be told about it.

As the weather is supposed to be very hot today with temps in the mid to upper 90's F, likely this will be and "indoor under the air conditioning" day for me.  Time to play radio.   That is after I get the things done which my wife dictates.

73

Bob, K4TAX

 

On 9/16/2019 6:40 AM, Roger wrote:

Hi Bob,

 

I was briefly thinking about a connector problem as especially the older versions of the K3 (K3S?) had issues with the tin plated pins. Starting to disassemble the radio and puliing/pushing boards is something I had lower down on the priority list especially as it gives the false impression of a successful “repair attempt” as long as the fluctuation is on an irregular basis and doesn´t show up after pulling the boards.

 

So I wanted to check out if someone else had a similar issue. As it didn´t show up anymore I leave it with that until it becomes a bit more steady.

 

Concerning the audiochecks I only use a 2nd station receiver for a basic first check if something is really off track or at least acceptable to start with because of exactly the anatomic reasons you lined out. The same is true if listening thru headphones on a remote SDR – HOWEVER the latency of the signal until it is processed in the SDR takes some time and there is yet another delay in getting the signal thru the internet to your computer and finally to the headphones. With the local SDR (5 miles away) the signal on an empty 10 m band is strong enough to even hear the details yet there is a gap of 2 to 3 and sometimes 5 seconds before the audio arrives at my headphones/loudspeakers. So there is plenty of time to talk and transmit for a couple of seconds then stop and listen to your signal like any other listener would. The latency on some remote SDRs can in fact be controlled to extend the latency. No impair due to “anatomic feedback” whatsoever. Recording and playing it back can be done with this method as well as the SDRs offer a record and playback function. So the only difference is that you record your audio thru an artificial communication path with absorbers or attenuators while the other way is via a real path. Once you use a local SDR, the signal level is strong enough and the noise level is low (like the closed 10 m  band) it is an absolutely valid method proved by many others with a good audio.

 

ALC indication on my setup is up to 7 bars which is just fine. The compressor level has no impact on when the noise gate starts to open. If it is easy to accomplish (which I don´t know) a selectable level up to the point where even speech close to the microphone is almost cut off would be fine. Everyone could pick the level to suit his individual situation from 0 to max. In OPENHPSDR you can almost choke off your speech by selecting the maximum noise gate level. But you can go down 0 and everywhere in between. The noise situation here is as it is. It is not a particularly quiet environment with the amplifier running 20 inches away from the radio nor is it a particular noisy environment – from what I have seen the last 40+ years of ham radio it is below average so nothing to worry about. In fact with OPENHSDR and the possibility to select a higher Noise Gate level the noise gate doesn´t open with the background noise yet is still sensitive enough to open reliably while speaking into the microphone. And there is still plenty of headroom to lift the level without being not sensitive enough for speech.

 

By the way testing around with the noise gate level somebody told me that a high frequent hiss is coming along with the signal especially if it goes beyond approx. – 43 dBm. Noise Gate on or off didn´t make any difference. Switching to other radios the noise was gone. Swapping powersupplies and even going for an emergency battery the noise was always present although the ambient was completely quiet (amplifier off, computer off etc.). I listened to the remote SDR and in fact it is there. Yesterday listening to a HB9 station on 80 m he had exactly the same noise coming along and after having a few words with him he confirmed that he is using a K3S. It sounded like a distant cellphone ringing – one of those typical old analog phones where the rings merge to an almost constant tone. With the SDR I may be able to determine the frequency of that tone.

 

Anyway it sounds like I don´t like the radio which is not the case. It´s fun to experience the phenomenal receiver and the possibilities it offers especially with a second receiver and some other useful accessories. But if something odd shows up it is always worthwhile to stop and investigate…………………………………

 

73 Roger, DL5RBW

 

From: Bob McGraw K4TAX [mailto:[hidden email]]
Sent: Saturday, September 14, 2019 1:54 AM
To: Roger
Subject: Re: [Elecraft] K3S issues

 

Roger:

I'd say the issue with the current indication is a front panel issue.  There are several connections that affect the display.   Might do well to check those, even pulling the front panel and cleaning the connectors.

I don't agree using a remote SDR for the purpose of checking ones signal or audio.  It is much better to use a local receiver with termination on the antenna input.  Record the output so one can listen to the signal.  Record a bit of transmission, and play it back for evaluation.   Due to sound conduction through the human head, it is not realistic to use headphones and talk at the same time.  The sounds of speech as conducted through the bones in the head to the inner ear,  while one is talking, is not realistic of what the transmitter is producing nor what the receiver is hearing.   Recording and playback is the much preferred method.

You are correct in that the monitor function only samples analog audio.  It does not represent the RF waveform.    If you find that TX Gate value of 25 is not adequate, that indicates you have a lot of mike gain which would be considered excessive and thus a large amount of noise ambient or otherwise in the room.     I suggest you carefully review your mike-to-mouth technique and your MIC gain settings.  If you are running the MIC gain at a value of 25 or less, I recommend you check to see if High Gain is activated.   Ideally the MIC gain should run in the 35 to 45 numerical range.    This will be somewhat dependent on mike technique and the type of mike being used.

Values typical of my satisfactory operation for TX Gate is 06 to 07.  The MIC SEL is set to FP.H with a MIC gain of 35 using a Shure SM-58 dynamic mike.  Also the TX EQ is set as follows:  50 Hz -16dB, 100 Hz -16dB, 200 Hz - 9dB, 400 Hz -3dB, 800 Hz -0dB, 1.6 kHz -0dB, 2.4 kHz +3dB, 3.2 kHz +6dB.    If you haven't attenuated the low end significantly, as shown above, this could be the source for some very low frequency energy which consumes lots of power in the spectrum but provides little communications intelligibility.   This could also be the source for the TX Gate which appears as not having adequate range.

Just trying to be of assistance with an audio issue.

73

Bob, K4TAX

 

On 9/13/2019 4:55 PM, Roger wrote:

-----Original Message-----
From: Bob McGraw K4TAX [mailto:[hidden email]]
Sent: Friday, September 13, 2019 11:04 PM
To: Roger
Subject: Re: [Elecraft] K3S issues

Roger:

The current value of 1.4 A is about correct for receive mode. Then in transmit mode, SSB no signal, the current is about 2.8 A. If the last digit of the current is moving around, I see the same thing. It is minor and just a toggle between bits or resolution.

Maybe it was a bit unclear - the radio sat in RX mode and the display was jumping from 1,4 A to 2,7 A in RX mode without doing anything. It was not the actual current consumption but only the display that was erratic. Idle current on TX is higher in my case but that´s another story.

I think first one should set the MIC Gain and CMP values for normal expected operation with the TX Gate at OFF. Then set the TX Gate to ON and start increasing the value from 00, while listening to the MON via headphones. The correct point is the value +1 where the noise is muted. In my station a Noise Gate value of 06 to 07 is ideal.

If you change mikes, you will have to reset the Noise Gate value and likely the Mic Gain values as well. This is due to different mikes having different output levels and different pick-up patterns. Be sure to start with the CMP off or 0 and then adjust the MIC gain such with normal voice and the mike of your choice you will observe no more than 5 to 7 bars on the ALC scale. If you change to a different mike, you will likely need to adjust the MIC gain value for that mike. Once this is done, set the CMP value to indicate about 3 to 5 bars on the CMP scale. This is the correct method for setting MIC Gain, and CMP value,

based on the ALC indication.   I don't consider listening on another

receiver to be a valid means of adjusting the MIC Gain or CMP value.  

All of this is covered on page 30 of the manual.

I don´t know whether you ever used a remote SDR but that´s the best way you can check your signal in many aspects. Listening to a second receiver next to the transmitter with the headphones on is a means to get a rough estimate at best. However using an SDR which is "decoupled" also in terms of time (latency is the keyword here) is an ideal means. Don´t forget to set the bandwidth wide and flat! The monitor function can be used for setting the noise gate but other than that it only gives you the unprocessed signal. Most of the radios with monitor function will NOT sample or tap the actual signal on the antenna port.

I don't find any issues where the system does not perform as designed and expected.

Well an erratic display can hardly be considered a design goal but I believe it is more likely a part problem. The noise gate however could easily offer a bit more headroom or a higher “release” level. If you get away with a lower level no problem if it is adjustable down to 0. If one needs a higher level because of a higher ambient noise the maximum setting as it is right now couldn´t be enough.

73

Bob, K4TAX

73/Roger

On 9/13/2019 3:37 PM, Roger wrote:

> Good evening,

>

> 1.

> when I turned on the K3S tonight I noticed that the current display

> was jumping between 1,4 and 2,7 Amperes on idle. Never saw the rig

> doing it before. The actual current consumption didnt change but the

> readout of the K3S only. That approx. +1,3 A also was noticed on

> transmit. Checking the circuit diagram I see a 0,005 ohm current sense

> resistor (R3). The voltage drop goes into a current sense monitor

> ((U6) and finally into a multi/demultiplexer (U6.1).

> As I already had a bad (=oscillating) op amp in the temperature

> readout circuit I wonder if bad luck struck twice and I got a bad

> current sensor or multi/demultiplexer?

> Now as the radio ran for a while and warmed up I dont see the jump in

> the current display anymore.

>

> 2. not really an issue but maybe a software fix is possible the Noise

> Gate level is very much on the edge. With the handmicrophone and in

> particular with a desk microphone the mic gain level has to be set

> very carefully on the conservative side. That is the mic gain has to

> be set just at the point where the signal becomes a little bit weaker

> compared to the asymptotic setting where a maximum is reached and no

> increase possible anymore even though there is still headroom in the

> mic gain range (received on a remote SDR). However even with mic gain

> just set under the maximum permissible value and the Noise Gate level

> set to max (25) still an occasional popping can be heard by pushing

> and holding the ptt but no ambient noise. If the mic gain level is set

> to the permissible limit (ALC

> deflection) the noise gate doesnt work anymore and an high frequent

> noise is coming along with the signal. No it is not the fan of the

> powersupply as I ran the station on a convection cooled linear

> powersupply and everything else in the shack muted. I even switched

> off the fan in the K3S which I normally leave running on "1". The

> compressor setting has NO effect at all on the point where the Noise Gate releases.

> The question is whether it is possible to have the firmware changed

> for more headroom in setting the Noise Gate level as the range from 0

> to 15 is almost useless.

>

>

> 73/Roger - DL5RBW

>

>

>

>

> ______________________________________________________________

> Elecraft mailing list

> Home:  <http://mailman.qth.net/mailman/listinfo/elecraft> http://mailman.qth.net/mailman/listinfo/elecraft

> Help:  <http://mailman.qth.net/mmfaq.htm> http://mailman.qth.net/mmfaq.htm

> Post:  <mailto:[hidden email]> mailto:[hidden email]

>

> This list hosted by:  <http://www.qsl.net> http://www.qsl.net Please help support this email

> list:  <http://www.qsl.net/donate.html> http://www.qsl.net/donate.html Message delivered to

>  <mailto:[hidden email]> [hidden email]

>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]