[Elecraft] XTAL AND DSP simply is the best design

[WA2JJH at aol.com]

You look at the TS-870, 746pro, and many other DSP rigs all make the same  
stupid design flaw.
 
  Dual conversion DSP!  No multiple pole xtal filter.
No computer algorythm can emulate a multipe pole IF filter. These are your  
$2000-$3000 rigs
 
  The TS-850 had dual xtal IFs. However there was much junk from all  the 
CPU's and yes up conversion, then downconversion and demod after the 2nd IF  
filter adds to the noise floor.
 
  How the R-7A did it was interesting. A 4 pole Fixed Xtal filter  followed 
the upconverter.
It has a pass band of 12kcs. That's right a 12Khz wide IF. This gives this  
collectors radio
its used price tag of $2000. The 2nd IF used those 8 pole can  filters.
 
  I found the K-2 has crystal clear SSB with the 1.8khz filter  position. The 
R-7A the SSB gets muffled at 1.8khz. I use 1.8 khz only on the K2.  The TR-7 
used a 2 pole filter after up conversion. 
 
  If you look at the schematic of the TS-850, you see the dual IFs are  in 
the back of the receiver. So after upconversion the passband is 50kc wide.  That 
is a large window for synth junk to get in. HOWEVER IF YOU USE CW filters  in 
both IFs, your CW reception will be as good as a K2
 
  The K2 wisely put much non xtal filtering right after the antenna  jack. 
The band pass filters are wide enough for the Ham bands and very little  general 
coverage. You lose general coverage ability of an R-7A. However do  you 
really need a short wave rcvr in a ham rig?
 
  So yes one can compare the K2 with up to $3000 dollar rigs.
However on SSB TX the K2 comes up average. It is average because the SSB  
adapter is a simple design. The all important audio chain is mostly in 2 chips.  
However I always get good audio reports. A RF proc of the TS-850's design 
would  make the K2 rock.
 
  There is a cheap way of doing this. There are a few speech procs  that go 
between the mic and rig. They are true RF procs. Your audio is up  converted to 
500kc. The 500Kc RF is processed by using an AGC like amp. At the  RFstage 
more compression can be used with out AUDIO DISTORTION. Any RF clipping  by 
products are filtered out. The DSB signal is then down converted back to  audio 
with the exact same phase it entered the device.
 
  So one is feeding in audio that was RF processed. One can use up to  12db 
of RF processing. Audio processing gets distorted after 6db. One company  makes 
this device for $60. It can be put in the mic or in the rig.
   I am ordering one. I will give on air checks for those that  want too.