This chip I hope will be good news in time.....

  I set up a LTSpice file model to test the split in two, MC1496 for
quadrature I/Q output.   Which was demonstrated could to be possible
in a previous test file of the idea here in the group files.  Since I
have changed the circuits operation it wants to operate in a linear
region of 360mV p~p into a 50K output.  There are some losses but the
sound card will make up for that.  This seems to me to a nice peak
output level.  I hope it will please people.  Wild voltage swings and
all do not serve the new  modified way of using this chip.  Since I
internally switch off two transistors to divide the chip in halves. 
So its output impedance is now something new.  You know, higher.

  I continued to work around the two 50 ohm inputs of the circuit.  So
that might interest some folk who like to use allot of 50 ohm
broadband HF preamps.  I am thinking, from my test so far, that the
software acts like the input is in the 1K region.  So I have 50 ohms
matched up to about 1k at this point.  The best way to find out
exactly what the input impedance will be is to test the circuit on the
bench.  But we know the range now.  As I said the circuit has been
changed from its normal use to something different in view.  Its input
behavior has to be redefined.  I will do some of the software plot
math here later to see if I can calculate the input impedance on the
pins.  Get us an good idea.

  You can unmistakably see the 90 degrees between the output waveforms
and calculate it from the software plot data.  There is a big angular
rotation of the input through the circuit to the output end.  So
unless you use your math for rotation around 360 degrees, or more,
back around again,  you might be mistaken about the 90 degree phase
angle relationship in the end.  Being it has rotated allot relative to
the input reference.  I and Q will still have a difference of 90
degrees as the smaller angle between them.  Sometimes the software
shows you the larger angle between them after they have rotated around
some.  The larger angle adds up to 270 degrees.

  I also set up a mixer test.  Where I took the above circuit and
added a signal input that I can vary from 1 MHz to 10 MHz to watch the
mixing effect in the I/Q waveform outputs.  What you can not see in
the waveforms is such a thing as a small offset of 12 kHz, so you have
to examine the mixer with wider frequency offsets of several MHz to
demonstrate the effect and see the mixing levels.  Anyways it will act
as a mixer in its upper section.  You can also use the lower section
as the mixer, it makes no matter which way you go.  I think numerous
models of the ways that the IC can be used will evolve along in time.
 I see several ways to use this.  

  I will upload these files soon.  So far it seems as though the chip
will have some good separation between the two channels.  And it can
also be used as a phased modulator.

  You can also add a potentiometer and balance the outputs to the same
level.  So that is interesting.   I will show that to you in the files
soon.

  Well its working out pretty well.  LTSpice is interesting.  I make
no promise though on how it will come up after being tested on the
test bench for noise and other factors.  Yet it should work well for
some inexpensive radio ideas.  It  is perfect I think, for use in two
way software radios.  I always like to watch my sinewave forms for
linearity so that is sublime with me.  I will keep it linear all the
way to the end if possible.

Dan


Reply via email to