Re: K3 AGC Adjustment Parameters

[Bill W4ZV]

N6KR:

>2. Some combinations of settings can end up being very confusing,
except for the advanced and/or experimentally inclined operator who can
recognize the side effects. This has been endlessly played out on some
of our competitors' rigs. I had sought to avoid such difficulties once
it became clear that there was such a small range in use by our
testers.

        Just to reinforce what Wayne said, I used an Orion for about
4 years.  It probably has the most adjustable AGC of any rig on the market,
with Threshold, Decay and Hang separately adjustable over a wide range.
After some initial user problems, Ten-Tec integrated Threshold with the
Gain control which effectively made a separate Threshold AGC adjustment
redundant.  Decay (e.g. separately adjustable "Fast", "Medium" or "Slow")
was also not really necessary beyond a couple of settings.  Hang was the
only parameter I used much and that was mainly to defeat Orion's Flat AGC
response on weak signals near the noise floor (I would set Hang very high
to effectively defeat the AGC).  With the variable AGC SLP setting, I
suspect
that the need for a separate Hang control is unnecessary.

        Just a comment on the widely publicized DSP "AGC problem" Rob
Sherwood gave in a 2007 Dayton paper (below).  I've listened to his
recordings
and my Orion **never** had any AGC recovery problems like that even
though I operate 160m a lot where there is plenty of impulse (lightning)
QRN.
In his recording, it sounds like the RX desenses for a long time following
an impulse.  If my Orion had done that, it would have been sold in a
heartbeat!  To my ears, it sounded like someone who did not know
how to adjust a rig in high QRN, i.e. like you were using Slow AGC with
a long Hang time and then wondering why the RX desensed.  I know
Rob knows better and I'm sure the rig mentioned (IC-7000 specifically)
does have that problem, but my Orion definitely did not.  Given that
Elecraft
says they have addressed this problem in the K3, I am going to take
them at their word until I hear otherwise on my K3.

http://www.sherweng.com/documents/Dayton2007w.pdf
(On slide #8 click on the speaker at the bottom to hear the sound clip)

        Bottom line:  I'm not too conceerned about AGC issues as long
as both Sloped AGC and the Impulse issues are implemented.  I would
be curious to know how the AGC "Off" is implemented however, since
the Analog AGC can never be truly turned off due to the need to protect
the DSP's ADC from overload.  In Orion, "AGC Off" was emulated by
a very high Decay setting with zero Hang.  I suspect the same is true in
the K3.

                                                73,  Bill  W4ZV

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Brendan Minish]

I Don't know about the orion but i certainly can hear the problem Rob
talks of on my Icom 7800.

Basically the problem is due to very short (but quite large) pulses
causing excessive AGC action.

certan types of Electric fence noise, Precipitation static and at a
guess Dry air static (not a problem in EI!) all create very short
pulses, only a sample or two long at the DSP IF frequency.


The AGC acts fully on these pulses (& presumably has to to protect the
DAC from an 'over' condition)  .
With a more conventional receiver architecture (& with the K3) the AGC
is detected AFTER the conventional filters (in DSP terms these would be
the roofing filters) narrow filters will stretch the pulse out and this
integrates the energy of the pulse over a longer time period and reduces
the amount of AGC action.
One of the 'defects' of traditional filters (filter ringing) is a real
advantage in this instance.
The other effect is of course the rise time and characteristics of the
AGC

Looking at the k3 architecture, I have high hopes that the k3 will be at
least as good as the k2 in this area  

Noise such as 160m QRN is usually quite time stretched by propagation
and the 7800 performs well here (although the flat, 'perfect'
 AGC response is irritating.)


Wayne, Please count my vote in favor of full access to AGC parameters, I
would very much like to explore this issue further.

73
Brendan EI6IZ    


On Thu, 2007-10-11 at 22:02 -0400, Bill Tippett wrote:

>
>         Just a comment on the widely publicized DSP "AGC problem" Rob
> Sherwood gave in a 2007 Dayton paper (below).  I've listened to his
> recordings
> and my Orion **never** had any AGC recovery problems like that even
> though I operate 160m a lot where there is plenty of impulse (lightning)
> QRN.
> In his recording, it sounds like the RX desenses for a long time following
> an impulse.  If my Orion had done that, it would have been sold in a
> heartbeat!  To my ears, it sounded like someone who did not know
> how to adjust a rig in high QRN, i.e. like you were using Slow AGC with
> a long Hang time and then wondering why the RX desensed.  I know
> Rob knows better and I'm sure the rig mentioned (IC-7000 specifically)
> does have that problem, but my Orion definitely did not.  Given that
> Elecraft
> says they have addressed this problem in the K3, I am going to take
> them at their word until I hear otherwise on my K3.
>
> http://www.sherweng.com/documents/Dayton2007w.pdf
> (On slide #8 click on the speaker at the bottom to hear the sound clip)
>
>         Bottom line:  I'm not too conceerned about AGC issues as long
> as both Sloped AGC and the Impulse issues are implemented.  I would
> be curious to know how the AGC "Off" is implemented however, since
> the Analog AGC can never be truly turned off due to the need to protect
> the DSP's ADC from overload.  In Orion, "AGC Off" was emulated by
> a very high Decay setting with zero Hang.  I suspect the same is true in
> the K3.
>
>                                                 73,  Bill  W4ZV
>
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
>  http://mailman.qth.net/mailman/listinfo/elecraft   
>
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com