Elecraft › [K3]

K3 on-channel strong signal overload?

Classic

List

Threaded

42 messages Options

123

Leigh L. Klotz Jr WA5ZNU

Re: K3 on-channel strong signal overload?

Administrator

Jim,
I have some questions.
> DUH! Why are you surprised by this? You had the K3 set for maximum
> sensitivity and then hit it with a 100 watt transmitter. OF COURSE
> it's going to overload!
I don't think this is follows naturally. Doesn't the automatic gain
control ("AGC") do the same thing as the RF gain control? As Oliver
pointed out, he didn't have the preamp on.

It seems to me Oliver is describing a failure of the AGC to properly
engage the HAGC , staying only in the DSP gain control area.
I've asked Oliver to report the firmware version, as I believe there
have been updates in the firmware since the "official" release that are
in the beta releases only.

>
> These are fundamental concepts of radio. It's too bad that they are seldom
> taught today. Most of us old farts learned them as Novices when we put our
> first rigs on the air. In those days, there were no transceivers. You had a
> transmitter and you had a receiver, and when you transmitted your receiver
> overloaded.
>
A few minutes of search would have shown that Oliver is the person who
originally reported this problem with the K2 in 2003, which resulted in
the back-to-back diode clamp. You may be right that Elecraft doesn't
design for strong in-passband signals, but it's not clear that that's an
fundamental limitation of RF technology, especially since the RF gain
control can handle it manually. It doesn't seem to me an unreasonable
request that the firmware be able to detect this same situation and
properly handle the gain stages (hardware and DSP) at least as well as
manual RF gain control. You might want to re-read this message:
http://www.qsl.net/ve3mcf/elecraft_reflect/K2_Howling_Fix.txt

Leigh/WA5ZNU

______________________________________________________________
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

Geoffrey Mackenzie-Kennedy

Re: K3 on-channel strong signal overload?

In reply to this post by Mike Harris-9

Hi Mike,

There is a penalty to be paid if diodes are used to clamp the level of
strong signals in a receiver's signal path, and that is the generation of
intermodulation products. In a "no-split" pile-up situation where two or
more closely spaced strong signals from other callers are in the passband,
but not zero beat with the DX, chances are that an intermodulation product
will land on top of the DX's signal if he returns to somebody before the
others stop calling. You may ask how a weak DX station could be heard at all
in this situation even if intermodulation products did not exist, to which
my reply must be that it is possible.

>From reading the posts, my impression is that this case of in-passband
strong signal overload could be viewed in the same fashion as a classic
example of 3rd Order Dynamic Range droop or "bottoming". There is a
similarity in the causes even though only a single signal is involved in the
overload case. Among the usual causes for the dynamic range droop problem
are (1) too much gain in the front end before the roofer to overcome the
insertion loss of the roofer, and to maintain a useful overall noise figure,
or (2) The Input IP3 of the roofer itself is too low, or (3) the 3rd Order
Dynamic Range of the IF cascade and what follows is inadequate, or (4) a
combination of these three. All of these are relevant when considering
overload.

According to my back of envelope calculations, the 3rd Order Dynamic Range
of the K3's 8.215 MHz IF amplifier and the second mixer is in the region of
80db before the onset of hardware AGC, the second mixer being the dominant
element. This calculation does not include the stages after the second mixer
nor the roofer nor LO phase noise, so I believe that the actual in-passband
dynamic range looking into the roofer would be less than 80db. I can only
speculate on the effect that the application of hardware AGC might have on
the dynamic range of this part of the receiver.

What this means of course is that while the overall 3rd Order Dynamic Range
of the K3 is very good in an enviroment where strong unwanted signals are
outside of the roofer's passband, one should expect some decrease within the
roofer's passband, likewise possible overload. The cure could be difficult
to implement.

73,
Geoff
GM4ESD

Mike Harris" <[hidden email]> wrote on Monday, May 04, 2009 at
7:20 PM

> Just been looking through my archives. I remembered that there was a
> simple mod to the K2 to overcome audio distortion in an extreme overload
> environment, I installed it on my K2 and it was totally immune to
> co-channel audio distortion when my very near neighbour VP8LP transmitted:
>
> "K2 - Improved Handling of Extremely strong (and nearby) Signals".
>
> Maybe the K3 architecture offers similar opportunities.

______________________________________________________________
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

orbarrett

Re: K3 on-channel strong signal overload?

In reply to this post by orbarrett

Hello all,

Thank you for all the responses and information, and I may look into the diode option that Wayne suggested.

I like operating HF mobile a lot (just started using the K3 mobile last weekend), and very often my friends and I will both be using our HF mobile setups while driving together (in one vehicle following the other) to a hamfest. We'll both be working the same distant station on a 60M channel for example, and I just turn down the audio volume when my friend transmits and turn it back up again for the distant station. I didn't want to make too much of this overload issue, just was unaccustomed to it occurring in my K3 as my friend and I have been used to not hearing it in the other rigs we've used when operating in close proximity like this. Not a big problem.

73,
Oliver

pd0psb

Re: K3 on-channel strong signal overload?

In reply to this post by Don Wilhelm-4

Quite a number of radios I know disable pre-amp and enable ATT when the AGC voltage reaches a certain threshold. A handy way of extending the range of total AGC with 30/40dB without having to reach for the knobs. Automatically switched by firmware to catch the first unexpected RF hits.

73'
Paul
PD0PSB

It is also likely that the K3 Hardware AGC is maxed out and can
lower the gain no more with the resulting of overload of the ADC.
the K3 dynamic range (as good as
it is) was exceeded - everything has its limits.

73,
Don W3FPR

David Gilbert

Re: K3 on-channel strong signal overload?

That sounds like a way to effectively turn the K3 into an AGC-pumping
monster like my old 756Pro ;)

73,
Dave AB7E

pd0psb wrote:

> Quite a number of radios I know disable pre-amp and enable ATT when the AGC
> voltage reaches a certain threshold. A handy way of extending the range of
> total AGC with 30/40dB without having to reach for the knobs. Automatically
> switched by firmware to catch the first unexpected RF hits.
>
> 73'
> Paul
> PD0PSB
>
>
> It is also likely that the K3 Hardware AGC is maxed out and can
> lower the gain no more with the resulting of overload of the ADC.
> the K3 dynamic range (as good as
> it is) was exceeded - everything has its limits.
>
> 73,
> Don W3FPR
>
>

______________________________________________________________
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

pd0psb

Re: K3 on-channel strong signal overload?

The "pumping" is up to the FW programmer, and his/her choises of threshhold and delay-times..

73'
Paul
PD0PSB

That sounds like a way to effectively turn the K3 into an AGC-pumping
monster like my old 756Pro ;)

73,
Dave AB7E

David Gilbert

Re: K3 on-channel strong signal overload?

Yes, but then you'd mask the weaker signals underneath. Anything that
requires computation, reaction, and delay (variable or otherwise) to
stretch the dynamic range is a non-linearity of one type or other (time
or level or both) that negatively affects the signal handling ability
(either in terms of sensitivity or distortion) somewhere within its range.

The thing that puzzles me about this entire thread is ... why do people
want to take a world-class communication rig like the K3 and make it
function like an FT-817 or a BC-455 or a Hammarlund Super-Pro or a Radio
Shack Scanner or a 1940's radio for listening to NPR? Did people buy
the K3 simply because it was the "in" rig without considering what they
really needed? That's like buying a race horse to plow a field. If
someone wants to do that, that's fine with me ... as long as they don't
mess with the breed to turn that race horse into a mule.

I'm honestly puzzled by some of the comments I've seen here lately ...

73,
Dave AB7E

pd0psb wrote:

> The "pumping" is up to the FW programmer, and his/her choises of threshhold
> and delay-times..
>
> 73'
> Paul
> PD0PSB
>
>
> That sounds like a way to effectively turn the K3 into an AGC-pumping
> monster like my old 756Pro ;)
>
> 73,
> Dave AB7E
>
>

n7ws

Re: K3 on-channel strong signal overload?

In reply to this post by pd0psb

I started composing a message yesterday about this and got called away. Paul has now sort of beaten me to it, but what I had written was:

"Seems to me that the radio ought to be smart enough to correct (at least partially) these issues.

If the preamp needs disabling and/or the attenuator needs engaging then the firmware should do it."

--- On Tue, 5/5/09, pd0psb <[hidden email]> wrote:

> From: pd0psb <[hidden email]>
> Subject: Re: [Elecraft] K3 on-channel strong signal overload?
> To: [hidden email]
> Date: Tuesday, May 5, 2009, 10:20 AM
> Quite a number of radios I know disable pre-amp and enable
> ATT when the AGC
> voltage reaches a certain threshold. A handy way of
> extending the range of
> total AGC with 30/40dB without having to reach for the
> knobs. Automatically
> switched by firmware to catch the first unexpected RF hits.
>
> 73'
> Paul
> PD0PSB
>
>

wayne burdick

Re: K3 on-channel strong signal overload?

Administrator

So noted on the firmware list. I'll be doing some controlled tests with
very strong signals.

tnx
Wayne
N6KR

On May 5, 2009, at 10:33 AM, Wes Stewart wrote:

>
> I started composing a message yesterday about this and got called
> away. Paul has now sort of beaten me to it, but what I had written
> was:
>
> "Seems to me that the radio ought to be smart enough to correct (at
> least partially) these issues.
>
> If the preamp needs disabling and/or the attenuator needs engaging
> then the firmware should do it."

---

http://www.elecraft.com

______________________________________________________________
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

pd0psb

Re: K3 on-channel strong signal overload?

Thanks Wayne,Wes,

Nice to see that "thinking practical" about ways to make K3 even better is not always interpreted as "attacking the K3 concept" ;-)

Best 73'
Paul
PD0PSB

wayne burdick wrote

So noted on the firmware list. I'll be doing some controlled tests with
very strong signals.

tnx
Wayne
N6KR

On May 5, 2009, at 10:33 AM, Wes Stewart wrote:

>
> I started composing a message yesterday about this and got called
> away. Paul has now sort of beaten me to it, but what I had written
> was:
>
> "Seems to me that the radio ought to be smart enough to correct (at
> least partially) these issues.
>
> If the preamp needs disabling and/or the attenuator needs engaging
> then the firmware should do it."

---

http://www.elecraft.com

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:Elecraft@mailman.qth.net

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

AC7AC

Re: K3 on-channel strong signal overload?

In reply to this post by wayne burdick

CONTENTS DELETED

The author has deleted this message.

Mike Harris-9

Re: K3 on-channel strong signal overload?

In reply to this post by Geoffrey Mackenzie-Kennedy

Hi Geoffrey,

I guess there are no simple answers to this apart from proper operator
intervention. I guess that is why the buttons are there. I've no
intention of adding any diode clipping, I don't have a problem even with
my S-9+60++ neighbour. One thing to note is that #1400 doesn't have the
hardware AGC mod installed so it is cutting in approx 20dB before the
latest versions. I was on the verge of doing it before this thread
started but I've put it away again.

So many hooks into the firmware, many possibilities to end up flying with
crossed controls.

Regards,

Mike VP8NO

______________________________________________________________
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

wayne burdick

Re: K3 on-channel strong signal overload?

Administrator

Mike,

The AGC mod is definitely preferred. The old AGC circuit allows in-band
signals as low as S9 to activate hardware AGC. This would cause desense
if you were using a wide crystal filter but a narrow DSP filter. The
new AGC circuit has an in-band desense threshold of more like S9+20 to
+30.

Of course we also recommend that you have at least one narrow crystal
filter (e.g., 400 or 500 Hz) if you're using CW or narrow data modes.
This changes the definition of "in-band" to such a small value that
desense due to adjacent signals becomes very rare.

Those really pushing the envelope should consider our 200 Hz filter.
Ask a few 160 m contest operators about this :)

73,
Wayne
N6KR

On May 5, 2009, at 1:29 PM, Mike Harris wrote:

> Hi Geoffrey,
>
> I guess there are no simple answers to this apart from proper operator
> intervention. I guess that is why the buttons are there. I've no
> intention of adding any diode clipping, I don't have a problem even
> with
> my S-9+60++ neighbour. One thing to note is that #1400 doesn't have
> the
> hardware AGC mod installed so it is cutting in approx 20dB before the
> latest versions. I was on the verge of doing it before this thread
> started but I've put it away again....

wayne burdick

Re: K3 on-channel strong signal overload?

Administrator

In reply to this post by AC7AC

If I add anything like this at all, it will default to OFF in a menu
entry.

First I have to do some tests to see if in fact there's a problem. Then
I have to consider hardware vs. firmware changes, keeping in mind that
only rarely will anyone encounter such circumstances.

73,
Wayne
N6KR

On May 5, 2009, at 12:46 PM, Ron D'Eau Claire wrote:

> It would be great if any such automatic changes can be turned off.
> ...

> ---

http://www.elecraft.com

______________________________________________________________
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

Mike Harris-9

Re: K3 on-channel strong signal overload?

In reply to this post by wayne burdick

Hi Wayne,

Thanks for your comment. I have the stock 2700Hz for SSB and five pole
500Hz for CW which works for me.

Encouraged by some recent comments I've been playing with width and shift
and find that 1800Hz BW centred on 1200Hz on SSB works very well and have
loaded it into profile II.

One nasty winters day I might just drag out the mod kit again. How do you
know which end is which on the SMD LED?

Regards,

Mike VP8NO

----- Original Message -----
From: "wayne burdick" <[hidden email]>
To: "Mike Harris" <[hidden email]>
Cc: "Elecraft Discussion List" <[hidden email]>; "Geoffrey
Mackenzie-Kennedy" <[hidden email]>
Sent: Tuesday, May 05, 2009 5:02 PM
Subject: Re: K3 on-channel strong signal overload?

| Mike,
|
| The AGC mod is definitely preferred. The old AGC circuit allows in-band
| signals as low as S9 to activate hardware AGC. This would cause desense
| if you were using a wide crystal filter but a narrow DSP filter. The
| new AGC circuit has an in-band desense threshold of more like S9+20 to
| +30.
|
| Of course we also recommend that you have at least one narrow crystal
| filter (e.g., 400 or 500 Hz) if you're using CW or narrow data modes.
| This changes the definition of "in-band" to such a small value that
| desense due to adjacent signals becomes very rare.
|
| Those really pushing the envelope should consider our 200 Hz filter.
| Ask a few 160 m contest operators about this :)
|
| 73,
| Wayne
| N6KR
|
| On May 5, 2009, at 1:29 PM, Mike Harris wrote:
|
| > Hi Geoffrey,
| >
| > I guess there are no simple answers to this apart from proper operator
| > intervention. I guess that is why the buttons are there. I've no
| > intention of adding any diode clipping, I don't have a problem even
| > with
| > my S-9+60++ neighbour. One thing to note is that #1400 doesn't have
| > the
| > hardware AGC mod installed so it is cutting in approx 20dB before the
| > latest versions. I was on the verge of doing it before this thread
| > started but I've put it away again....

______________________________________________________________
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

wayne burdick

Re: K3 on-channel strong signal overload?

Administrator

Mike Harris wrote:

> One nasty winter's day I might just drag out the mod kit again. How
> do you
> know which end is which on the SMD LED?

DMM :)

Actually, I think the instructions cover this.

73,
Wayne

---

http://www.elecraft.com

______________________________________________________________
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

JIM DAVIS-11

Re: K3 on-channel strong signal overload?

On Tue, 5 May 2009 14:50:53 -0700
wayne burdick <[hidden email]> wrote:

> Mike Harris wrote:
>
>> One nasty winter's day I might just drag out the mod kit again. How
>> do you
>> know which end is which on the SMD LED?
>
> DMM :)
>
> Actually, I think the instructions cover this.
>
> 73,
> Wayne
>
> ---
>
> http://www.elecraft.com

*******************************************************************************************
Dear Wayne,

Your quote: "> Actually, I think the instructions cover this."

Not every nuance is covered by the "MANUAL" and there's been a few times when I've had to call
"APTOS" for guidance just to muddle our way thru the "Bell-Curve" learning process!!!

Not all of the users of the "K3" are "IT-SAVVY!!!!!!!

Jim/nn6ee

______________________________________________________________
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

Merv Schweigert

Re: K3 on-channel strong signal overload?

In reply to this post by wayne burdick

wayne burdick wrote:

> If I add anything like this at all, it will default to OFF in a menu
> entry.
>
> First I have to do some tests to see if in fact there's a problem. Then
> I have to consider hardware vs. firmware changes, keeping in mind that
> only rarely will anyone encounter such circumstances.
>
> 73,
> Wayne
> N6KR
>

How many K3s are mobile, how many are used instead of a 2 meter radio
talking car to car bumper to bumper, or sitting under someones antenna
in the driveway. I sure do not want my K3 to act as a 897 or any other
peanut radio, I would have bought one of those if I did.
As the K3 sits now it can handle stronger signals than any other radio while
hearing a weak signal with no problems, using diodes has been
nothing but IMD and distortion products all over the band in the old
radios I
have tried in the past. No diodes please. Remember all those upgrades to
radios that changed out the diodes in the front end, filter switching
etc. due
to IMD products etc. ?
I think Elecraft makes a product for that called an antenuator,
buy one and stick it on the front end if the signal is too strong.
Or use the gift of manual dexterity to turn down the volume/gain.
I think the thread mutated as usual into one report and many replies
that are not the same problem.
Sure more pressing issues to deal with on the K3. Rare is good on a
steak,
rare is not so good when dealing with a customer base. Gain one, loose
many.
nuff said. Merv KH7C
______________________________________________________________
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

wayne burdick

Re: K3 on-channel strong signal overload?

Administrator

Don't worry, Merv. Any hardware mod we might come up with for handling
extremely strong in-band signals (i.e., from adjacent transmitters)
would have absolutely no impact on K3 performance otherwise. Or we
wouldn't do it.

73,
Wayne
N6KR

On May 5, 2009, at 3:22 PM, Merv Schweigert wrote:

> I sure do not want my K3 to act as a 897 or any other peanut radio..

---

http://www.elecraft.com

______________________________________________________________
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

Bill W4ZV

Re: K3 on-channel strong signal overload?

In reply to this post by Mike Harris-9

Mike Harris-9 wrote

One thing to note is that #1400 doesn't have the
hardware AGC mod installed

Are you sure? The hardware AGC mod was implemented around S/N 280 in January 2008. The blue (?) dot on the SMD diode faces toward the front panel.

73, Bill

123