[Amps] Bandwidth tests

[Tom W8JI]

> That's what I was able to see, Tom.  The IM products were 
> much stronger with white noise & max ALC than with speech 
> & no ALC or even speech with ALC.  I did *NOT* do very 
> comprehensive testing, in the lab, so I'm borrowing an HP 
> spec any tomorrow to do some more detailed testing.  What 
> tones should I use?  I read the section on your site about 
> injecting a "warbling" tone a while back -- I believe that 
> was at 100 Hz (??).  Given my audio bandpass is roughly 
> 20Hz to 4kHz @ -6dB, I can do some pretty interesting 
> testing.

Joe,

In order for ALC problems to be measured, you have to let 
the ALC actually do something meaningful. Any test that 
holds the ALC steady, such as a white noise or two-tone 
test, will not test ALC dynamics at all. The fact you 
overdrove the ALC simply means you saturated the ALC control 
system and effectively disabled it. It is no surprise a 
saturated radio gets wide.

The same is true for anyone testing a linear amplifier. A 
linear amplifier that is terrible for dynamic regulation of 
power supplies, be they bias, HV, or screen, will almost 
always look significantly better when tested with a steady 
average power source like a noise test or a two-tone test.

To actually test something that amplifies speech, the test 
must have the characteristics of speech. That includes the 
off and on characteristics like spaces between words and 
syllabic rate changes, as well as the tones.

The problem with wide-fi is best left for other forums, but 
anyone who understands IM products, receivers, and 
transmitters also knows the bass hurts BW as much as the 
extra highs. I, like many others, don't have a problem in 
the world with wide-fi on empty bands. Hogging up extra room 
on a crowded band is selfish.  It would also be nice if the 
cutting edge technical information sites for ESSB actually 
knew how RF systems and measurements worked.  It doesn't 
reflect well on a group that often claims to be cutting edge 
technically when people see single tone or non-dynamic tests 
to measure bandwidth.

73 Tom