K3 roofing filter choices

[Dave, G4AON]

Having spent a few hours conducting measurements of the 3rd order mixing
products with my K3, I'm not convinced of the benefit of several roofing
filters for the K3. If other owners with access to two signal generators
and a hybrid coupler would like to try their K3 with whatever filters
they have I would be most interested in the results.

Try as I might with very strong dual signals into the K3 from two signal
generators and a hybrid coupler, changing the bandwidth by altering the
width control between 400 Hz and 450 Hz there is mostly no difference in
the 3rd order products except for an anomaly at 2 KHz signal spacing
where my 400 Hz 8 pole filter seems to be about 10 dB worse!

I've tried the shift control, very careful measurements at higher signal
levels with spacings from 1 KHz to 20 KHz, etc. There is a difference of
perhaps 3 or 4 dB where the narrower filter has the advantage with 3rd
order products in the spacing range 5 to 20 KHz but it's so small as to
be insignificant and gives no measurable difference near the noise
floor, only being apparent with signals some 10 dB higher than the
standard ARRL test levels to measure the 3rd order products. The levels
required to produce the small difference are hitting the end stop of the
K3 and are - 20 dBm, levels not likely to be encountered in practice for
most of us. The only filters I have at the moment are 2.8 KHz 8 pole and
400 Hz 8 pole.

I also noticed the dynamic range appears better by a small margin with
the pre-amp turned on... Although it should be pointed out that using a
pre-amp improves the MDS (minimum discernible signal) and raises all
signals by around 10 dB, so it's not increasing "signal handling".

My measurements give a 3rd order dynamic range of around 100 dB at 2 KHz
signal spacing with a 400 Hz 8 pole filter and slightly better with a
2.8 KHz 8 pole filter (around 105 dB).

73 Dave, G4AON
_______________________________________________
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

[Bill W4ZV]

Dave G4AON might be reluctant to "toot his own horn" but I simply must call attention to the excellent review he has written on his K3.  Rather than just subjective opinions, he has included a lot of valuable measurements which I've not seen before (like CW waveform measurements).  Well done Dave!

http://www.astromag.co.uk/k3/

73 & Happy New Year!

Bill  W4ZV

[dj7mgq]

Hallo Dave,

 > http://www.astromag.co.uk/k3/

Very nice indeed!!!

Close to what we have seen on my K3, but we have not been as extensive.


 > not convinced of the benefit of several roofing filters for the K3.

This is something I think we will be looking at, when we do another
round of measurements.


 > an anomaly at 2 KHz signal spacing where my 400 Hz

We did not see this on my K3 using the 400Hz roofing filter.


 > My measurements give a 3rd order dynamic range of around 100 dB at
 > 2 KHz signal spacing with a 400 Hz 8 pole filter and slightly better
 > with a 2.8 KHz 8 pole filter (around 105 dB).

We measured a dynamic range of c. 106db (+/- a few dB) from 100kHz down
to 2kHz with the 8 pole 400Hz filter on 20m. So both measurements (by
DK4YJ with K3 #00067 and G4AON with K3 #00080) are in the same ballpark.


vy 73 de toby


PS: http://www.astromag.co.uk/k3/ -> distributors in Germany and Japan

I seem to remember Peter Zenker writing several months back that his
company would not be selling the K3 in Germany. They do offer other
stuff by Elecraft.



Dave G4AON wrote:
 > Having spent a few hours conducting measurements of the 3rd order
 > mixing products with my K3, I'm not convinced of the benefit of
 > several roofing filters for the K3. If other owners with access to two
 > signal generators and a hybrid coupler would like to try their K3 with
 > whatever filters they have I would be most interested in the results.
 >
 > Try as I might with very strong dual signals into the K3 from two
 > signal generators and a hybrid coupler, changing the bandwidth by
 > altering the width control between 400 Hz and 450 Hz there is mostly
 > no difference in the 3rd order products except for an anomaly at 2 KHz
 > signal spacing where my 400 Hz 8 pole filter seems to be about 10 dB
 > worse!
 >
 > I've tried the shift control, very careful measurements at higher
 > signal levels with spacings from 1 KHz to 20 KHz, etc. There is a
 > difference of perhaps 3 or 4 dB where the narrower filter has the
 > advantage with 3rd order products in the spacing range 5 to 20 KHz
but > it's so small as to  be insignificant and gives no measurable
 > difference near the noise  floor, only being apparent with signals
 > some 10 dB higher than the  standard ARRL test levels to measure the
 > 3rd order products. The levels required to produce the small
 > difference are hitting the end stop of the K3 and are - 20 dBm, levels
 > not likely to be encountered in practice for most of us. The only
 > filters I have at the moment are 2.8 KHz 8 pole and  400 Hz 8 pole.
 >
 > I also noticed the dynamic range appears better by a small margin with
 > the pre-amp turned on... Although it should be pointed out that using
 > a pre-amp improves the MDS (minimum discernible signal) and raises all
 > signals by around 10 dB, so it's not increasing "signal handling".
 >
 > My measurements give a 3rd order dynamic range of around 100 dB at
 > 2 KHz signal spacing with a 400 Hz 8 pole filter and slightly better
 > with a 2.8 KHz 8 pole filter (around 105 dB).



--
DD5FZ (ex 4n6fz, dj7mgq, dg5mgq, dd5fz)
K2/10  #885
K2/100 #3248
K3/100 #67
DOK C12, BCC, DL-QRP-AG, JN58td
_______________________________________________
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

[Klaus Koppendorfer]

just tried to do also some dynamic range measurements

two HP 8640 Generators (out +20dBm) attenuated to 0dBm into
an 6dB coupler attenuated to -10 dBm into an stepp-attenuator

done on 20meter with 4pole 500Hz filter
IMD with pre  -140 dBm
IMD no pre   -134 dBm

dynamic range  no pre
spacing   IMD-Level   Dynamic
20 kHz      -31 dBm     103 dB
5 kHz       -31 dBm     103 dB
2 kHz       -34 dBm     100 dB
+/- inaccuration

very happy
73
klaus
OE6KYG


> My measurements give a 3rd order dynamic range of around 100 dB at
 > 2 KHz signal spacing with a 400 Hz 8 pole filter and slightly better
 > with a 2.8 KHz 8 pole filter (around 105 dB).

>We measured a dynamic range of c. 106db (+/- a few dB) from 100kHz down
>to 2kHz with the 8 pole 400Hz filter on 20m. So both measurements (by
>DK4YJ with K3 #00067 and G4AON with K3 #00080) are in the same ballpark.

_______________________________________________
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

[Klaus Koppendorfer]

sorry.. IMD has to be "MDS" !!!!!!!!

done on 20meter with 4pole 500Hz filter
> MDS with pre  -140 dBm
> MDS no pre   -134 dBm

OE6KYG

>
> just tried to do also some dynamic range measurements
>
> two HP 8640 Generators (out +20dBm) attenuated to 0dBm into
> an 6dB coupler attenuated to -10 dBm into an stepp-attenuator
>
> done on 20meter with 4pole 500Hz filter
> IMD with pre  -140 dBm
> IMD no pre   -134 dBm
>
> dynamic range  no pre
> spacing   IMD-Level   Dynamic
> 20 kHz      -31 dBm     103 dB
> 5 kHz       -31 dBm     103 dB
> 2 kHz       -34 dBm     100 dB
> +/- inaccuration
>
> very happy
> 73
> klaus
> OE6KYG
>
>
> > My measurements give a 3rd order dynamic range of around 100 dB at
>  > 2 KHz signal spacing with a 400 Hz 8 pole filter and slightly better
>  > with a 2.8 KHz 8 pole filter (around 105 dB).
>
> >We measured a dynamic range of c. 106db (+/- a few dB) from 100kHz down
> >to 2kHz with the 8 pole 400Hz filter on 20m. So both measurements (by
> >DK4YJ with K3 #00067 and G4AON with K3 #00080) are in the same ballpark.

_______________________________________________
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