K3 SSB - Is the 5 pole filter good enough

[Stewart Baker]

Having run a quick test switching between USB and LSB on a quiet
band 10m, I noticed an appreciable difference in the tone of the
noise spectrum.

I dug out my noise generator and examined the audio response using
Spectrum Lab. I set the DSP B/W to the maximum as what I wanted to
see was the 2.7KHz SSB roofing filter response.

On USB the trace was substantially flat with the -3dB points at
430Hz and 2.77KHz. Not too bad and could probably be improved by
moving the carrier offset frequency.

On LSB, however the trace was far from flat having a slope from
audio LF to HF of about 6dB. Not very good at all...

I tried adjusting the FL FRQ, but all this did was shift the
overall position of the sloping trace.

Although the crystal filters are spoken about as being roofing
filters designed to keep strong signals away from the DSP
filtering they do affect the wanted passband.

On the K2 there used to be a big difference between the USB and
LSB response until a modification to the output of the filter was
added. The KSB2 uses a 7 pole filter, giving a more or less
symmetrical response. I wonder if 5 poles is just too few ?

Or maybe I have a duff (defective) 2.7KHz filter ?

I wonder if others have plotted their 5 pole filters.

73
Stewart G3RXQ




_______________________________________________
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

[John A. McCabe]

Hi Stewart,

Although I have not ran all the tests you did, I also noticed that there
is a significant difference in tone response between LSB and USB. I also
tried adjusting  FL FRQ with about the same results as you got. My 5
pole 2.7  filter does not seem to have a very flat response. I have
thought about switching to the 8 pole filter because of that.

73,

John, KD8K

Stewart Baker wrote:

> Having run a quick test switching between USB and LSB on a quiet
> band 10m, I noticed an appreciable difference in the tone of the
> noise spectrum.
>
> I dug out my noise generator and examined the audio response using
> Spectrum Lab. I set the DSP B/W to the maximum as what I wanted to
> see was the 2.7KHz SSB roofing filter response.
>
> On USB the trace was substantially flat with the -3dB points at
> 430Hz and 2.77KHz. Not too bad and could probably be improved by
> moving the carrier offset frequency.
>
> On LSB, however the trace was far from flat having a slope from
> audio LF to HF of about 6dB. Not very good at all...
>
> I tried adjusting the FL FRQ, but all this did was shift the
> overall position of the sloping trace.
>
> Although the crystal filters are spoken about as being roofing
> filters designed to keep strong signals away from the DSP
> filtering they do affect the wanted passband.
>
> On the K2 there used to be a big difference between the USB and
> LSB response until a modification to the output of the filter was
> added. The KSB2 uses a 7 pole filter, giving a more or less
> symmetrical response. I wonder if 5 poles is just too few ?
>
> Or maybe I have a duff (defective) 2.7KHz filter ?
>
> I wonder if others have plotted their 5 pole filters.
>
> 73
> Stewart G3RXQ
>
>
>
>
> _______________________________________________
> 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

[John A. McCabe]

 >On USB the trace was substantially flat with the -3dB points at 430Hz
and 2.77KHz. Not too bad >and could probably be improved by moving the
carrier offset frequency.
 >On LSB, however the trace was far from flat having a slope from audio
LF to HF of about 6dB. >Not very good at all...
 >I wonder if others have plotted their 5 pole filters.

Stewart,

I just ran a similar test using spectrogram and it confirms what I
thought I heard with my ears and was similar to the results you got. It
was basically flat on USB but was sloped about the same amount on LSB.

73,

John, KD8K



>
_______________________________________________
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

[John A. McCabe]

Wow, that is dramatically different then what I am getting. Here is the
response I am getting using a bandwidth setting of 0 to 6000:

http://mysite.verizon.net/respk8uz/k3/id1.html

I am open to any suggestions, I have tried changing FL FRQ and
re-calibrating the synthesizer.

73,

John, KD8K



Barry N1EU wrote:

> My SpectraPlus plots are viewable at
> http://www.n1eu.com/images/K3_2700usb_lsb.gif 
>
> LSB is the acqua trace and USB yellow
>
> As you can see, they're very close and both fairly flat in the passband.  RX
> EQ was zeroed for the measurements and config menu FL FRQ setting was
> adjusted for best match between the modes.
>
> 73,
> Barry N1EU
>
>
> John A. McCabe wrote:
>  
>>  >On USB the trace was substantially flat with the -3dB points at 430Hz
>> and 2.77KHz. Not too bad >and could probably be improved by moving the
>> carrier offset frequency.
>>  >On LSB, however the trace was far from flat having a slope from audio
>> LF to HF of about 6dB. >Not very good at all...
>>  >I wonder if others have plotted their 5 pole filters.
>>
>> Stewart,
>>
>> I just ran a similar test using spectrogram and it confirms what I
>> thought I heard with my ears and was similar to the results you got. It
>> was basically flat on USB but was sloped about the same amount on LSB.
>>
>>
>>    
>
>  

_______________________________________________
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