Amps
[Top] [All Lists]

Re: [Amps] Checking for IMD

To: <amps@contesting.com>
Subject: Re: [Amps] Checking for IMD
From: "Tom W8JI" <w8ji@w8ji.com>
Date: Fri, 11 Aug 2006 16:32:19 -0400
List-post: <mailto:amps@contesting.com>
One thing everyone needs to keep in mind is IMD products at 
one point can even cancel other IMD products produced at 
another point in the system. You don't necessarily get the 
worse case condition, you might get something that looks 
better than it really is.

You need a way to drive the amplifier with a very clean 
two-tone signal, and the detector has to be very clean.

Also there is a basic flaw with two-tone tests. Two-tone 
tests don't vary the power supply loading like a voice 
signal would. Flaws in grid voltage or high voltage dynamic 
regulation may not show at all. Almost always the error 
makes things look better than they really are.

As a general rule it would be impossible to get a good 
reading with a standard amateur SSB transceiver as the 
driver. You would have to use two radios on CW through 
attenuators into a combiner, and for the combiner to work it 
has to be terminated in 50 ohms (or whatever it is set for). 
That combiner cannot be run into the amplifier under test, 
it has to go through a VERY linear amp that has a constant 
input impedance. That amp then drives the actual device 
under test.

As for the detector, a good selective level meter will work. 
That means you could press a receiver into use if you are 
absolutely sure it is good enough at close spacing and if 
you have a calibrated attenuator that has enough steps to 
always keep the S meter at exactly the same level as you 
check various IM products against the desired carriers.

If you mix to audio and use a sound card, how will you sort 
out the opposite sideband of the mixer? How would you know 
the IM level of the sound card with all the multiple signals 
present? IM occurs a lot faster with multiple signals than 
it does with just two pure tones.

73 Tom 


_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>