K3 AGC and Apparent Noise and Signal Distortion

[Lance Wilson]

CONTENTS DELETED

The author has deleted this message.

[Grant Youngman]

Oh, good lord.  This is ridiculous.  "The noise was so bad on my K3 i had to go back to my S-38C with external Q-multiplier  just to work the contest" ... and on and on.

With no personal affronts intended, perhaps just learning to use the radio would help.

Grant/NQ5T


On Feb 2, 2010, at 7:20 PM, Lance Wilson wrote:

>
> I have started a new thread on this as there are multiple threads that are
> difficult to keep track of.
>
> Today a number of other hams have also reported "noise", signal distortion
> and other noise and signal issues that are a bit on the strange. T
______________________________________________________________
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

[Lance Wilson]

CONTENTS DELETED

The author has deleted this message.

[Julian, G4ILO]

I think you are wrong that this is not related to AGC Threshold. However, if THR = 8 is not enough to raise the level of the threshold above the noise you probably won't see a difference. The noise will still modulate the AGC which makes the radio seem noisier, and weak signals will be reduced by the AGC to the same loudness as the noise.

There seems to be a growing body of evidence including the figures quoted by W4TV that the K3's AGC Threshold cannot be set high enough at least for some radios / some user situations. If others think it is fine, that may just be due to gain variations or people being blessed with living in quiet locations.

If you cannot lift the AGC threshold above the noise then perhaps you can try reducing the noise below the threshold by turning the preamp off and using the attenuator. I don't think the RF Gain will help with this as I seem to recall reading that it is really an IF gain control.

[paulb]

Hello folks

Just to follow up on this thread...listening to quite strong CW sigs
with Slow AGC setting is not too flash. I would guess sigs over
75 uV or so which is well into the  9+ region.
It is hard to exactly describe but sounds like the Hardware AGC
loop is pumping. It is fine on SSB or weak CW. Not a big deal,
I would prefer a longer hardware AGC Hang time, but thats
personal preference.
I have an aversion to riding RF gain controls, reminds me of the bad old
days of rubbish receivers without product detectors :-]

cheers

Paul
zl1ajy
3051

[Geoffrey Mackenzie-Kennedy]

Hello Paul,

Do you hear any "raspy" sound when listening to quite strong CW sigs - not
the sound of real band or thermal noise, but more like the sound of paper
being torn?

I should not be asking as I do not own a K3, but curiousity is killing this
cat!

73,
Geoff
GM4ESD


paulb <[hidden email]> wrote on Wednesday, February 03, 2010 at 10:27
AM

> Hello folks
>
> Just to follow up on this thread...listening to quite strong CW sigs
> with Slow AGC setting is not too flash. I would guess sigs over
> 75 uV or so which is well into the  9+ region.
> It is hard to exactly describe but sounds like the Hardware AGC
> loop is pumping. It is fine on SSB or weak CW. Not a big deal,
> I would prefer a longer hardware AGC Hang time, but thats
> personal preference.
> I have an aversion to riding RF gain controls, reminds me of the bad old
> days of rubbish receivers without product detectors :-]
>
> cheers
>
> Paul
> zl1ajy
> 3051

______________________________________________________________
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

[Lance Wilson]

CONTENTS DELETED

The author has deleted this message.

[Bill W4ZV]

Grant Youngman N5QT wrote

Oh, good lord.  This is ridiculous.  "The noise was so bad on my K3 i had to go back to my S-38C with external Q-multiplier  just to work the contest" ... and on and on.

With no personal affronts intended, perhaps just learning to use the radio would help.

Spot on Grant...doesn't this remind you of Orion a few years ago?

FYI here's a note I got from Rob Sherwood NC0B regarding his experience with the K3 in the CQ 160 last weekend.  I believe everyone knows Rob is one of the most objective critics around when it comes to evaluating radios.  My point in posting this is that we need to consider the source before taking many complaints at face value.  In most cases, these are reflecting lack of knowledge of some of the posters more than performance of the K3.  Said another way, always use some caution when reading comments here or on eHam (especially if they're from folks who use screen names only and don't have the guts to post their calls).

##################################################
Hi Bill,
 
I did post a comment on how the K3 performed.  Using 40 ohm per channel stereo headphones, I was not bothered by any audio distortion issues.  I ran the AGC the whole time, and noted no problems.  I am not into the RF gain back and the audio up method.   I have the AGC setting on SOFT, an option that came out last summer I believe.
 
Due to the QRN on Saturday, I ran the attenuator most of the time.  No use having the band noise reading up scale a few S units.  I only spent about 10 to 12 hours on the contest, as I was not feeling well on Friday or Saturday.  Worked 15 different countries and 8 JAs.  Russia was a struggle for me, as I only worked one, and a few western Europe.  No VKs this time like in W1BB contest.   Used the K3 about 80% of the time.  Semi-breaking with the Alpha 89 was flawless.  Didn't have some setting right to make it full QSK, not that I normally run that way.  Will sort that out next weekend.
 
I had to drop down to 150 Hz BW several times due to adjacent QRM.   The roofing filter switches in at 400 Hz, but I did not notice any problem with the DSP between 300 Hz and 150 Hz, not that I expected that to be an issue.  Key clicks were a problem at times, but nothing fixes that, and I did hear one station in Idaho that I could copy from his phase noise 1 kHz away on either side.
 
73, Rob
#######################################################

Regarding Grant's "S-38C in the CQ 160" comment, below are the current Top Ten claimed scores (Multiop and Single op) for the CQ 160 CW (IMHO the world's toughest environment for any receiver).  I've indicated (***before the call) stations that I know for a fact were using K3s...there could be others in the list, but I'm sure of these:

http://lists.contesting.com/pipermail/3830/2010-February/191049.html

Call               QSOs StPrv    DX   hr      Score Club
USA Multi-Op HP
***KC1XX             1870    59    80    40  1,154,256 YCCC
K1TTT             1672    59    77    40    967,096 YCCC
***W2FU              1756    58    77    38    936,765 Rochester (NY) DX As
NQ4I              1573    59    82  39.5    829,362 SECC
K1LT              1559    58    75    39    776,587 MRRC
N7DD              1410    59    83    33    752,032 Arizona Outlaws Cont
K0RF              1494    59    74    33    707,693
AK9F(@WB9Z)       1542    59    75    40    700,418 SMC
***N1LN              1552    58    70  41.5    679,936 PVRC
***W8RT(@W8TOP)      1354    57    74  35.5    622,381

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
USA Single Op HP
K1DG(@N1LI)       1741    59    82    30  1,292,265 YCCC
K8PO              1702    58    76    30  1,185,900 YCCC
***K3ZM              1627    57    81    30  1,095,582
***K1UO              1441    58    77    30    941,895
***W4ZV              1542    58    68    30    829,332 PVRC
KU1CW(@N0NI)      1473    59    75    30    803,062
***AA1K              1405    58    67    30    735,750 FRC
***K5NA              1397    58    71    30    653,901 CTDXCC
W1MK              1285    55    55  18.5    637,560 YCCC
NR5M(NM5M)        1359    58    65    32    595,074 CTDXCC

I seriously doubt many of these folks considered swapping their K3 for an S-38C!  My point is, when you see ridiculous comments like Grant quoted above, always consider the source before taking them at face value.  As to learning the radio, the **current online** manual and errata should be your first stop:

http://www.elecraft.com/K2_Manual_Download_Page.htm#K3

The following links are also good learning resources with multiple links/pages in each:

http://www.elecraft.com/K3/K3FAQ.htm
http://www.elecraft.com/K3/k3_operating_tips.htm
http://n1eu.com/  (click K3 link to the left)
http://www.zerobeat.net/mediawiki/index.php/K3_Wiki_at_ZeroBeat.NET

In most cases, Elecraft's default settings are a good starting point.  The K3 is sort of like having a sports car with automatic everything.  However, when you're ready to take it to a higher level of performance, the hooks are there to do so, but you need to understand what you're doing first by educating yourself.

73,  Bill  W4ZV

[Bill W4ZV]

Bill W4ZV wrote

The following links are also good learning resources with multiple links/pages in each:

http://www.elecraft.com/K3/K3FAQ.htm
http://www.elecraft.com/K3/k3_operating_tips.htm
http://n1eu.com/  (click K3 link to the left)
http://www.zerobeat.net/mediawiki/index.php/K3_Wiki_at_ZeroBeat.NET

I should have also included K8ZOA's excellent webpage:

http://cliftonlaboratories.com/elecraft_k2_and_k3_transceivers.htm

Use some caution because some of Jack's comments/measurements were made prior to some of the latest hardware/firmware revisions.  The K3 is a moving target because Elecraft keeps improving it...and that's a good thing!  :-))

73, Bill

[Guy, K2AV]

Hi Lance,

> Indeed, activating the attenuator does improve the situation quite a bit but
> as I operate 20M and above for the most part I do not want to operate the K3
> with ATT on all of the time.

I was going to ask.

Assuming that "no ATT" is not a religious thing, and engaging ATT
works, why NOT use it, if the performance of the AGC is the end-all
kill-all aspect of the K3 for you?  All that we are talking about is
how much system gain is in use between the antenna and the analog to
digital converter.

20m ambient noise here on my C31XR at 76 feet is around 1 uV here for
SSB bandwidths, which IS engaging the AGC because it is S3.

Use ATT before you reduce RF gain, as it provides the best headroom in
the overall circuit.  "RF gain" is applied in the IF amp at Q1, as can
be determined by reading the schematics on the Elecraft web site.

The other thing, have you tried setting CONFIG:AGC DCY to SoFt?

Soapbox mode on:

I think part of the reason for the continued reemergence of this
subject is the p**s-poor performance of earlier receivers and that
fact that their AGC and S meters typically did not engage until what
was really S5 and reported it as S1.  In other words we are expecting
our new K3 to behave just as garbage on low signals as the old stuff,
just because we're used to it. And when the K3 reports 1 uV as S3 and
turns on the AGC, we think it's busted and noisy.

In olden telephone days, the accumulated noise of strung together
circuits created quite a bit of noise, and someone describing the
other person on the end of a long distance call as "distant" was
really a euphemism for noise.  People complained a lot about noise.

When the new "N" carrier systems came out and were being deployed, the
engineers were proud that they had created a quiet system (they had)
and would ease customer complaints about noise on the line.  The
telephone company was deluged with customer complaints that the lines
sometimes sounded "dead".  Telco was forced to spend a lot of money to
retrofit a noise injector which was set to a minor level of noise. One
of the maintenance tasks I had when routining "N" carrier was to
calibrate the noise injection pot. They paid me real money to do that.
To this day I think that was one of the silliest things going.

Do we REALLY want to morph the K3 back to bad old days analog just so
it sounds familiar?  Can't we just figure out how this bright new
digital radio is different, adjust and move on?  I won't even mention
RTFM for fear of having to dodge bricks.

mode off.

73, Guy.
______________________________________________________________
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

[Alan Bloom]

On Wed, 2010-02-03 at 17:41 -0500, Guy Olinger K2AV wrote:
...
> When the new "N" carrier systems came out and were being deployed, the
> engineers were proud that they had created a quiet system (they had)
> and would ease customer complaints about noise on the line.  The
> telephone company was deluged with customer complaints that the lines
> sometimes sounded "dead".  Telco was forced to spend a lot of money to
> retrofit a noise injector which was set to a minor level of noise.
...

They do the same thing with cellular phone systems.  Some systems go
completely silent when the other person is not talking, which made
people think they were disconnected.  So they had to add a little
"comfort noise" to let people know everything is still working.

Al N1AL


______________________________________________________________
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

[Grant Youngman]

The same thing was done during the transition from mechanical analog switching to electronic analog matrix and digital switching.  All the clunks and thunks that marked the progress of call connection largely disappeared with electronic analog and digital switches.   So the switch vendors added artificial progress tones/noises just to give the caller that comfort factor that a connection was still in progress ...

it's kind of silly when you think about it, but turns out to be pretty important in the overall user interface .. :-)

Grant/NQ5T


On Feb 3, 2010, at 4:52 PM, Alan Bloom wrote:

>
> people think they were disconnected.  So they had to add a little
> "comfort noise" to let people know everything is still working.

______________________________________________________________
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

[Rick Dettinger-3]

> I think you just did:-)

73,
Rick  K7MW


>   I won't even mention
> RTFM for fear of having to dodge bricks.
>
> mode off.
>
> 73, Guy.

______________________________________________________________
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

[NZ0T]

All I can say is that I love my K3 but maybe that's because I use it to operate on the bands instead of performing all kinds of nit picking tests on noise, AGC, NB and whatever else I can think of.  

But maybe that's just me.  73 Bill NZ0T

[KK7P]

> 20m ambient noise here on my C31XR at 76 feet is around 1 uV here for
> SSB bandwidths, which IS engaging the AGC because it is S3.
>
> Use ATT before you reduce RF gain, as it provides the best headroom in
> the overall circuit.

With an MDS around -136 dBm and 1 uV being -107 dBm, the noise is about
30 dB over MDS.  Adding even 20 dB of input attenuation wouldn't harm
weak signal reception and would provide 20 dB more headroom for nearby
strong signals before they started to bother you.

The K3 has a 10 dB input attenuator.  However, you can use a trick to
get a 10/20/30 dB input attenuator if you have a KXV3 or KXV3A and
aren't using a separate Rx antenna.  Just insert a 20 dB attenuator
between the RX ANT IN and RX ANT OUT jacks on the KXV3/A panel.

If you need 10 db ATT, use the ATT button. If you need 20 dB, remove the
10 dB ATT and select RX ANT.  For 30 dB select both RX ANT and ATT.  You
still can use PRE to get +10 dB, so you now have 40 dB of front end gain
control in 10 dB steps to allow you to match the front end sensitivity
of the K3 to current band/noise conditions.

The downside is that the S Meter will read 20 dB lower when you have RX
ANT engaged since the DSP doesn't know about the attenuator.

Enjoy!

73,

Lyle KK7P
______________________________________________________________
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

[Don Wilhelm-4]

Lance,

The ATTenuator and Preamp settings are remembered per band.
So you can have it on for one band and off for another - no constant
button pushing.

73,
Don W3FPR

Lance Wilson wrote:
> Indeed, activating the attenuator does improve the situation quite a bit but
> as I operate 20M and above for the most part I do not want to operate the K3
> with ATT on all of the time.
>  
>
______________________________________________________________
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

[n7ws]

Rob and I are in agreement on the AGC thing.  I believe Wayne has said much the same.

When high-end receivers had two pentode -r-f stages driving a pentagrid mixer and barn door selectivity, it made sense to reduce the r-f gain.

IMHO, riding the r-f gain control (really i-f gain in the K3) in a properly designed receiver should be totally unnecessary.

If it necessary, then the "properly-designed" nomenclature doesn't apply, except in the rarest of circumstances.

Furthermore, unless you've done the r-f gain calibration in your K3, reducing the r-f (i-f) gain with AGC on might severely degrade the SNR.

Fire at will.

Wes  N7WS


--- On Wed, 2/3/10, Bill W4ZV <[hidden email]> wrote:

FYI here's a note I got from Rob Sherwood NC0B regarding his experience with
the K3 in the CQ 160 last weekend.

I did post a comment on how the K3 performed.  Using 40 ohm per channel
stereo headphones, I was not bothered by any audio distortion issues.  I ran
the AGC the whole time, and noted no problems.  I am not into the RF gain
back and the audio up method.   I have the AGC setting on SOFT, an option
that came out last summer I believe.



     
______________________________________________________________
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

[Don Wilhelm-4]

Wes,

I agree that constant "riding the RF Gain" should not be necessary,  But
on any one band, reducing the RF Gain to the point where the atmospheric
noise is not readily apparent does make a lot of sense to me.  Set in
that way, one does not have to listen to a lot of band noise and the AGC
still works (albeit at a reduced level), and there is still S-meter
response for those who want to give reports based on the meter reading.
Once set for any one band, it does not need to be changed - AGC and the
AF Gain will take care of the rest - I would not refer to that as
"riding the RF Gain control".

Now --, if one wants to turn off the AGC altogether, yes, the only
technique that makes sense is to use the RF Gain as the 'volume' control
(in addition to preamp and attenuator controls) - the technique as I
learned it is to disconnect the antenna and turn up the AF Gain to the
point where the internal receiver noise is apparent, but not bothersome
(leave it set there), and use the RF Gain to set the desired output
audio level - in this case, changing the RF Gain is the only way to
control the volume of what is coming out of the headphones (or
speaker).  I cannot imagine anyone operating that way on a normal basis
with a modern receiver with good AGC, but I know some do that based on
habits formed long ago with old receivers using diode detectors - we had
to do it that way because the BFO would overload the AGC and kill the
receiver gain.

73,
Don W3FPR

Wes Stewart wrote:
> IMHO, riding the r-f gain control (really i-f gain in the K3) in a properly designed receiver should be totally unnecessary.
>
>  
>
______________________________________________________________
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

[Julius Fazekas n2wn]

There were at least a few of us running LP and QRP who did OK with our K3's : K1HTV, OH1VR/VP9, OK1FKD... and, at least, one future convert that I know about.

Many of us "little guns" don't have the antenna advantages of the "big boys", both the K2 and K3 have been great equalizers in the rough and tumble 160M contesting arena.

73,
Julius
n2wn

[Bill W4ZV]

Julius Fazekas n2wn wrote

There were at least a few of us running LP and QRP who did OK with our K3's

I wasn't trying to list every K3 in the contest...just those used in the Top Ten of the two most competitive categories (Multi Op High Power and Single Op High Power).  

If you want to see more detail of all K3s in the contest, I posted how to do that below:

http://www.mail-archive.com/[e-mail]/msg90719.html

73,  Bill