[TenTec] meaningful test results

[Tom Rauch]

> The ARRL labs reports are certainly good as far as they go. But the
> bare numbers don't go far enough. 

The important thing to remember is ARRL testing is getting better, 
and they are willing to improve it even more IF we can give them 
reasons for changes.

Even a year ago, they measured and published IM3 and blocking at a 
nearly-useless 20kHz.  That test inflated the real-world performance 
of radios because only the RF amplifier and first mixer of most 
receivers was being tested.  
 
> Steve's suggestion of trying to copy CW through an artificial QRM
> generator would help but I would rather see a slice of 20 Meters on WW
> weekend recorded and then played back to provide the background - and
> a canned message to be played for copying at various signal levels and
> freqencies. Grade the test rig on how weak a signal can be copied
> "under" a 250 microvolt signal, and how close a 1.0 microvolt signal
> can be and still be copied.  

The value of two-tone tests is they clearly show the amount of 
garbage generated by two signals and you can compare pecking orders 
with good assurance the real-world results would be the same so far 
as order of performance. 

Unfortunately two-tone tests produce limited results if the tone 
spacings are not correct (like for example 20kHz when we are 
interested in CW). A second problem is we sometimes forget that 
multiple weaker signals can cause the same problems as a few strong 
signals, since it is the overall power being amplified that causes 
overload problems.

While a tape would be closer to the real world, it would be hard to 
quantify results or to apply the "tape-test" uniformly to all devices 
under test. A better test would be to simply add a third tone and 
pulse it on and off.

In voice transmitters, two tone tests almost always produce better 
results than real-world voice results because the tones are "steady". 
CATV systems and other systems required to handle more than two 
information channels are tested with multi-tone testing.

Adding the third tone works on transmitters quite well, even though 
very few testers use that method. It shows defects missed by steady 
level two-tone tests and should work just as well on receivers, 
especially if it is pulsed so AGC is included. (In transmitters, 
pulsing the tones off and on will show ALC defects hidden in two-tone 
tests. That's the equivalent of AGC defects on receivers.)

It certainly seems to me a controlled third tone and pulsing it on 
and off, would show virtually all hidden defects. The tone could be 
increased in amplitude and have varying on and off rates, so the 
tester could find the worse case condition.
 
> Any serious distortion in a reciever degrades copy ability, I would
> like to see a standardized test of copy ability with diminishing
> signal strengths. All the way to what ever the recievers noise floor
> is supposed to be. If you can copy a quarter microvolt signal the
> reciever is pretty clean.  

A controlled multi-tone test with CW-rate pulsed keying of tones 
would do just that, and allow the results to be quantified uniformly. 
It would test AGC and other slow rate dynamic problems, as well as 
cover the things two steady tones already test.73, Tom W8JI
W8JI@contesting.com