Sherwood on Roofing Filters,Transmitted IMD and Receiver Performance

[Bill W4ZV]

Rob Sherwood's recent Boulder ARC presentation, slides in .ppt or .pdf and audio in .mp3, are now available on his website.  This is a very good primer for anyone interested in the above issues.  Download both and you can view the slides while listening to the audio.

http://www.sherweng.com/

73,  Bill  W4ZV

[Bill W4ZV]

Bill W4ZV wrote

Rob Sherwood's recent Boulder ARC presentation

http://www.sherweng.com/

Some of Rob's best comments are in the discussion following the formal talk.  It will take a long time to listen to it all of but it's well worthwhile!

73,  Bill

[alsopb]

One thing stuck out.  "Inserting filters in one direction in his testing for some rigs resulted in an 8db difference relative to the opposite direction in a performance parameter."

Are K3 filters properly polarized so this won't happen?  Has the filter technology improved so this isn't an issue?  Or is it a crap shoot with a 50/50 chance of getting it right even with polarized filters?

One other thing was really interesting.  Manufacturers have forgotten what's important and has been learned in the past regarding transmitter IMD, keying waveform, ALC, AGC.  The new generation of designers are all "digital" and don't have any training on the nits/importance of getting the analog part right.  Since most of them are not hams.  They don't listen to their product and have no clue on how it is actually working.  Lab tests only take one so far.

The fact that anything we hear is analog seems to have been lost.

It was good to see reported in one place that both the transmit and RX performance of rigs has gotten progressively worse, until recently.   Of course, the cost in constant dollars has dropped a huge amount.

The TS850 still stacks up quite favorably.  Not a surprise since it was the rig of choice for contesters for many years.

It would be interesting to have a digital modes evaluated-- especially the rapid on/off chirp type modes.  My perception is these chirp type modes are producing really bad "clicks" and wide IMD products.  Also the broader digital modes which already are 1KHz or more wide need to be looked at.  The sound card may be putting out tones at all the right amplitudes, but what is the xmtr doing to them?

K3KO

Bill W4ZV wrote

Rob Sherwood's recent Boulder ARC presentation, slides in .ppt or .pdf and audio in .mp3, are now available on his website.  This is a very good primer for anyone interested in the above issues.  Download both and you can view the slides while listening to the audio.

http://www.sherweng.com/

73,  Bill  W4ZV

[Barry N1EU]

K3KO wrote

It was good to see reported in one place that both the transmit and RX performance of rigs has gotten progressively worse, until recently.  

I agree.  However, it's hard to generalize on a downward trend of RX performance - if you look at his numbers, RX dynamic range performance doesn't generally correlate with design age if you take the Orion/SDR5K/K3 out of the picture. That being said, the close-in DR data of recent Yaesu radios (FT1K/2K) HAS gotten worse.  

Rob does mention two other issues on RX performance that has generally gotten worse: 1)newer radios with i.f. DSP generally have a problem with AGC over-responding to short duration pulses (Rob excepts the K3) 2)ssb rx audio quality ("listenability") seems to suffer in the newer digital designs

73,
Barry N1EU

[Ed K1EP]

I think that that comment was taken out of context.  In his talk, I believe
that he remarked that some xtals lacked quality control.  I think it was
said or implied that both the manufacturing process and testing process
wasn't always perfect and that some xtals in the filters had some
contamination or process defect that would cause them to have higher IMD
products than they should.  If that particular xtal was at the high signal
end (input) of a ladder filter, it would produce IMD products.  However, if
it was at the other end of the filter, most of the strong signals would have
been filtered out by that point and you would not see the problem.  Since
the filters were supposed to be symmetrical, theoretically it shouldn't
matter which direction they were used.  But given the lower quality of some
of the individual xtals in the filter, he noticed the difference.  That is
what I got from his talk.  I may not have interpreted it correctly.  The K3
filters only plug in in one way, so if they were tested properly and the
xtals manufactured properly, there shouldn't be any problems.  Yes, the
problem can still exist if the xtals are not manufactured properly.


On Wed, Mar 12, 2008 at 8:33 AM, K3KO <[hidden email]> wrote:

>
> One thing stuck out.  "Inserting filters in one direction in his testing
> for
> some rigs resulted in an 8db difference relative to the opposite direction
> in a performance parameter."
>
> Are K3 filters properly polarized so this won't happen?  Has the filter
> technology improved so this isn't an issue?  Or is it a crap shoot with a
> 50/50 chance of getting it right even with polarized filters?
>

_______________________________________________
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

[Vic K2VCO]

Ed K1EP wrote:
> I think that that comment was taken out of context.  In his talk, I believe
> that he remarked that some xtals lacked quality control.  I think it was
> said or implied that both the manufacturing process and testing process
> wasn't always perfect and that some xtals in the filters had some
> contamination or process defect that would cause them to have higher IMD
> products than they should.

Elecraft's filters are either built by them (the 5-pole filters) or
purchased from INRAD. The INRAD filters go through an additional testing
process at Elecraft to make sure they meet specs for the K3.

So yes, there can be problems and Elecraft is taking steps to avoid them.
--
73,
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco
_______________________________________________
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