[1000mp] MP vs MkV vs K2

Tue, 2 Jul 2002 10:34:42 -0400

> GOOD or proper tests would normally be made with both signals INSIDE 
> the roofing filter. The ARRL has to be careful to do this, or there 
> is little reason to do the test.

ARRL's testing of DR is from 2 to 100 kHz spacing, in reasonable steps.  We
also do an intermod test that is entirely within the 2500 Hz passband of the
final selectvity. When the 20-kHz spacing was initially selected oh, so many
years ago, receiver design was such that both tones were almost always well
within the roofing filter, or at least it was deemed at the time. 

This is no longer true.  ARRL wants to keep that 20-kHz testing to give some
standard basis of comparison between rigs of old and rigs of today, but has
also added testing at a number of spacings, preamp on, one band, and 5-, 20-
and 100-kHz spacing for the more visible presentation in the Product Review
test-result columns.  

What is interesting to see is how receivers play when both signals are NOT
at the same level where the intermod is occurring.  In that case, the
third-order response does NOT follow a 3:1 amplitude relationship.  Few
receivers actually know that they are supposed to follow theory that
closely. :-)

73, 
Ed Hare, W1RFI
ARRL Lab
225 Main St
Newington, CT 06111
Tel: 860-594-0318
Internet: [email protected]
Web: http://www.arrl.org/tis

> -----Original Message-----
> From: Tom Rauch [mailto:[email protected]]
> Sent: Sunday, June 30, 2002 8:27 AM
> To: [email protected]
> Subject: Re: [1000mp] MP vs MkV vs K2
> 
> 
> > Thanks for the reply.  So when these tests are run (5 or 20 
> spacing),
> > do they then have (for example) one of them tuned into the center of
> > the IF passband, and the other one 5 or 20 KHz. away (outside the IF
> > filter, but inside the roofing filter)? 
> 
> GOOD or proper tests would normally be made with both signals INSIDE 
> the roofing filter. The ARRL has to be careful to do this, or there 
> is little reason to do the test.
> 
> Say the IF roofing filter is 70.000 to 70.010 kHz and the local 
> oscillator is at 78.005 kHz with an 8 MHz second IF.
> 
> If we tested with 5kHz spacing and a signal centered in the 8MHz IF, 
> even with a 10kHz roofing filter, the results would be flawed. The 
> problem is the test signals would be well down on the slope of the 
> roofing filter when the testing operator tuned in the IM product 
> (which would be 5kHz above or below the desired signal), and this 
> would inflate the results.
> 
> 5kHz is almost certainly too wide for many receivers, and still would 
> inflate the results.
> 
> What we really want to do is test as close to the real-world problem 
> as we can, and always do a test that keeps BOTH signals well inside 
> the roofing filter. That would mean we would have to test at a much 
> closer spacing than the roofing filter bandwidth.
> 
> The ideal test spacing in a multi-conversion receiver with a roofing 
> filter system (like most modern transceivers) would be just outside 
> the limits of the narrow filters, perhaps at 1-2 kHz using a narrow 
> CW filter.      
> 
>  If you are right, then the
> > offending element is pre-IF, such as the first mixer (70M), 
> or some RF
> > amplifier before the IF filters.  I guess it's possible.
> 
> It is not only possible, it is that way virtually 100% of the time.
> 
> 73, Tom W8JI
> [email protected] 
> 
> _______________________________________________
> List Moderator: Richard Lubash N1VXW
> 1000mp mailing list
> [email protected]
> To Change Options or Unsubscribe:
> http://mailman.qth.net/mailman/listinfo/1000mp
> 