Filters and Configuration Questions

[Bob Naumann W5OV]

I've asked about the 2.1 filter before, but here it is again.

Will the K3 configuration be changed to allow the selection of the 2.1
filter for transmission purposes? If not, aside from it being some amount of
work to allow this, is there a technical reason why not? In other words,
would my signal be "bad" if it was routed through the 2.1?

I really want to have the following 5 filters in my K3:  6k, 2.1k, 1k, 500,
200. The 2.7 wastes a filter slot since all I use it for is to transmit.

Regarding the 6 KHz filter, why can't it be selected to receive CW?

I think it was permitted earlier firmware- wasn't it? This one really
puzzles me why this would be restricted. I like listening to a wide passband
and narrowing it down only as much as necessary - sometimes 6 kHz is OK.

Bob Naumann W5OV


______________________________________________________________
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]

Bob Naumann wrote

I've asked about the 2.1 filter before, but here it is again.

Will the K3 configuration be changed to allow the selection of the 2.1
filter for transmission purposes?

Only Elecraft can answer this question.  In early 2008, I tried using a 1 kHz filter for CW TX (telling the configuration it was a 2.7) for CW TX and got some very bad signal quality reports.  I think Elecraft has since changed the firmware to prevent one from doing this...apparently for good reasons, which I don't understand either.

73,  Bill

[P.B. Christensen]

> Only Elecraft can answer this question.  In early 2008, I tried using a 1
> kHz filter for CW TX (telling the configuration it was a 2.7) for CW TX
> and
> got some very bad signal quality reports.

If the CW Tx frequency is not reasonably centered in the filter passband,
poor keying quality reports will result, especially if placed on, or near
the upper or lower skirt of a crystal filter.  This asymmetrical group delay
issue plagued the Omni VI+ as the CW carrier frequency was placed right on
the skirt of the 2.4 kHz @ 9 MHz filter.  The placement of the Tx BFO
frequency near the filter skirt was done in order to better accommodate
variable CW Tx offset in the Omni 6+ design.  In the Omni V, the CW BFO was
fixed and located about 400 Hz above where Ten Tec placed it in the Omni 6+.
That shift of a few hundred Hz made all the difference in the world.  Among
several other contributions to poor keying in the Omni 6+, including a poor
ALC design that mutated from the Omni 5 and 6 (non plus), the primary
culprit was the BFO placement near the skirt of the filter.

The overall solution included filter replacement with a custom 2.8 kHz
filter from INRAD and re-working the ALC with a handful of components.  I
ended up bypassing the filter with a hard wire in CW Tx.  Passing the BFO
through the filter simply isn't required when the rig is capable of
generating a clean waveform.  Plenty of old homebrew boatanchor rigs were
capable of producing a high-quality RF envelope and didn't rely upon ALC nor
a filter to pass the CW waveform.  True, just as many homebrew rigs produced
horrendous signals on the air.

Since keying bandwidth is a function of the RF envelope shape, I'm not sure
why anyone would want to narrow the K3 filter on CW transmit.   As long as
the ALC is stabilized and the slope is reasonably smooth and absent sharp
discontinuities, all a narrower filter will accomplish is to produce
envelope distortion if not placed squarely in the center of the filter
passband.

Paul, W9AC





 

______________________________________________________________
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]

P.B. Christensen wrote

Since keying bandwidth is a function of the RF envelope shape, I'm not sure
why anyone would want to narrow the K3 filter on CW transmit.

Because most rigs have spurs and I wanted to make my signal as clean as possible.  Typically these may be <80 dBc but still readable if the transmitter is using high power and the receiving station has good antennas (e.g. on 160m in particular).  I often hear +/- 1 kHz -80 dBc spurs from FT1000 series rigs and of course the 2.4 kHz -60 dB spur from the Flex 5k (recently fixed thanks to W5ZN and others' persistence).  

If you look carefully at Figure 2 in ARRL's original K3 product review, you can see a few suspicious blips which I wanted to clean up.  Unfortunately using 1 kHz to TX in the K3 caused other problems so I abandoned the idea.  

I'm not aware of any spur reports on the K3 since the VCO stiffener and KPA3 mods were implemented, so I agree there's probably no need now.  

http://www.elecraft.com/manual/ELECRAFT%20KSYN3%20Stiffener%20Mod%20Rev%20D.pdf
http://www.elecraft.com/K3/mods/K3%20Technical%20Alert%201.pdf

For any interested in assuring clean signals you should also consider the chirp mod (if applicable):

http://www.elecraft.com/K3/mods/KSYN3%20DDS%20Gain%20Modification.pdf

73,  Bill

[Guy, K2AV]

I'll vote with Bill here.

Since the CW envelope in the digital number stream is pristine, the
filters and any envelope delay can only distort it. In the case of the
2.7, the delay is far enough away frequency-wise from the signal to be
insignificant.

Given that the filters are subject to manufacturing variance, and a
super clean filter would be prohibitively expensive, to use a narrow
one would involve some kind of routine to "reverse modify" the CW
envelope number stream to the transformative opposite of the
distortion, so that the delay undoes its distortion by distorting it,
kind of like two amps in series that together have less IMD than
either alone, common enough.

While that is attractive in a geeky way, it would have to be done on a
bench for the exact filter.

Given the results we have, aside from unfortunate and now-fixed issues
such as the microphonic sensitivity of the VCO boards, CW is
remarkably clean. It's the only rig available to me I would consider
using to park on 7000.25 as a run frequency.

It MAY be that Wayne et al have some of that reverse magic in effect,
but I haven't heard of it.

Trying to fix something that ain't broke?

73, Guy.

On Wed, Jan 5, 2011 at 11:19 AM, Bill W4ZV <[hidden email]> wrote:

>
>
> P.B. Christensen wrote:
>>
>> Since keying bandwidth is a function of the RF envelope shape, I'm not
>> sure
>> why anyone would want to narrow the K3 filter on CW transmit.
>>
>
> Because most rigs have spurs and I wanted to make my signal as clean as
> possible.  Typically these may be <80 dBc but still readable if the
> transmitter is using high power and the receiving station has good antennas
> (e.g. on 160m in particular).  I often hear +/- 1 kHz -80 dBc spurs from
> FT1000 series rigs and of course the 2.4 kHz -60 dB spur from the Flex 5k
> (recently fixed thanks to W5ZN and others' persistence).
>
> If you look carefully at Figure 2 in ARRL's original K3 product review, you
> can see a few suspicious blips which I wanted to clean up.  Unfortunately
> using 1 kHz to TX in the K3 caused other problems so I abandoned the idea.
>
> I'm not aware of any spur reports on the K3 since the VCO stiffener and KPA3
> mods were implemented, so I agree there's probably no need now.
>
> http://www.elecraft.com/manual/ELECRAFT%20KSYN3%20Stiffener%20Mod%20Rev%20D.pdf
> http://www.elecraft.com/K3/mods/K3%20Technical%20Alert%201.pdf
>
> For any interested in assuring clean signals you should also consider the
> chirp mod (if applicable):
>
> http://www.elecraft.com/K3/mods/KSYN3%20DDS%20Gain%20Modification.pdf
>
> 73,  Bill
>
> --
> View this message in context: http://elecraft.365791.n2.nabble.com/Filters-and-Configuration-Questions-tp5891986p5892454.html
> Sent from the Elecraft mailing list archive at Nabble.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
>

______________________________________________________________
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

[Joe Planisky]

I do not know the details, but there is some sort of limitation which prevents the synthesizer from reliably acquiring lock with RX bandwidths greater than 2.8 kHz.  

I ran into a firmware bug about a year and a half ago (since fixed) that was allowing my 13 kHz FM filter to be selected in CW mode.  The result was the RX would go dead intermittently when returning from TX.  At that time Wayne explained that the permanent fix for the problem was to correct the firmware so CW mode never allowed a filter > 2.8 kHz to be used.

73
--
Joe KB8AP



On Jan 5, 2011, at 5:47 AM, Bob Naumann wrote:

> ...

> Regarding the 6 KHz filter, why can't it be selected to receive CW?
>
> I think it was permitted earlier firmware- wasn't it? This one really
> puzzles me why this would be restricted. I like listening to a wide passband
> and narrowing it down only as much as necessary - sometimes 6 kHz is OK.
>
> Bob Naumann W5OV

______________________________________________________________
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