From: Paul Harden, NA5N (
Date: Mon Dec 06 1999 - 14:45:30 EST

by Paul Harden, NA5N


NOTE: This is a text version of an article appearing in the Summer
1997 issue of "QRPp." The article contains numerous illustrations and
photos of oscilloscopes displays, which unfortunately can not be
included in a text file.

Phase relationships between two signals at the same frequency can be
measured with 2-5 degree accuracy with a scope, although more suited
for a dual-trace scope. The REFERENCE signal is applied to CH. 1 and
the signal to be phase measured to CH. 2. For proper phase measure-
ments, ensure your dual trace display is in the CHOPPED mode, not
ALTERNATE mode. (Alternate mode can effect the triggering position
for the second, or CH.2 sweep).

There are many methods to do this. One is to stretch out the signal
so it takes 4 full divisions, so each division is 90 degrees of
phase. By measuring from a common point on one signal to the other
(zero crossings or from the peaks), the phase can be measured.

For example, say you are making a phased array antenna system, in
which one feedline must cause a 90 deg. delay. You calculate the
electrical length for a quarter wavelength [L=(246/f) x velocity
factor] and cut the coax to that length. You are now working on
blind faith that you have exactly 90 degrees. With a scope, you can
measure it fairly accurately by injecting a signal into one end with
a signal generator (at the frequency of interest) and a 50-ohm load
on the other. Connect the scope CH.1 to the coax input (signal
generator end) and CH.2 to the load end and measure the phase.
Trigger the scope and move the horizontal position and/or the time
base vernier so the positive peak of the CH.1 sinewave is on the first
vertical graticle line and the
second positive peak is on the : : +117 degree
fourth vertical graticle, as --->: :<--- Phase Shift
shown in the illustration to : : based on PEAKS
the right. Now measure the **-----|------|------|-----**
phase by noting where the | * | : | | * |
first positive peak on CH.2 | * | : | | * |
occurs. Say it occurs about CH.1-|------*------|------*------|
1.3 divisions to the RIGHT of | | *: | * | |
the CH.1 positive peak. Since | | :* | * | |
one division is 90 degrees, |------|-@@@-***-----|------|
using this method, then | @| : @ | | |
1.3 div. x 90 deg. = 117 deg. | @ | @ | |
YOUR DELAY LINE IS TOO LONG! CH.2-|-@----|------|-@----|------|
Cut off an inch or two at a @ | | @ | |
time until the CH.2 peak is | | | @| @|
one division from the CH.1 |------|------|------|-@@@--|
peak (or on the 2nd vertical DUAL-TRACE PHASE MEASUREMENT
graticle as shown in the
illustration) for precise tuning of the delay line.

Another method is to make the CH.1 signal to be two divisions high,
and center it between the two divisions, such that the zero-crossing
points are on the middle graticle line. Where the CH.1 sinewave
signal crosses zero going positive is the 0 deg. REFERENCE; the
positive peak is 90 deg.; the
negative going zero crossing is
180 degrees, etc. For CH.2 to be 0 90 degrees
90 degrees delayed from CH.1, the : :
CH.2 sinewave should cross zero, : :
going positive, right under the |---**|-----|-----**----|
90 degree peak of the CH.1 signal. | * * | *| * |
If the CH.2 zero crossing is farther | * |* | * | * |
to the right from the CH.1 positive |Z----|-Z---|--Z--|---Z-|
peak, the phase shift is MORE than * | * | * | *|
90 degrees. Back to the example of | | * |* | *
the coaxial delay line, you would |-----|----**-----|-----|
cut an inch or two at a time until Z=Zero-crossing points
the CH.2 zero crossing is directly
underneat the CH.1 positive peak
(the 90 degree point).

And still yet another method of comparing the phase between two
signals on a dual-trace scope is to accurately measure the period
it takes for one complete sine wave on the CH.1 reference channel.
Say it is 140nS (that would be 7.14 MHz, by the way). Now say the
CH.2 signal is 50nS delayed from the CH.1 signal. The phase shift
would be:
    Phase = 50ns/140ns x 360 degrees = 129 degrees

One thing you must remember is how to "read" phase shifts on an
oscope. When comparing two signals as described above, remember
that if the CH.2 signal peak is to the RIGHT of the CH.1 peak, then
the CH.2 signal is OCCURING LATER IN TIME than the CH.1 signal,
because time is traveling from left to right. If the CH.2 peak is
say 90 degrees to the LEFT of the CH.1 peak, then the CH.2 signal
occured in time BEFORE the CH.1 signal. This would then be a -90
degree phase shift, or 270 degrees. Think about this carefully
before you start cutting the coax on that delay line!

Phase measurements can be made on a single trace scope as well.
First, connect the REFERENCE signal, using a BNC "T", to both the
VERTICAL INPUT to the scope and the EXTERNAL TRIGGER and select
EXTERNAL as the trigger SOURCE. Adjust the TRIGGER LEVEL so the zero-
crossing occurs at the beginning of the trace on the first vertical
graticle. Now remove the reference signal from the scope's vertical
input (but NOT the external trigger input) and connect the signal to
be phase measured to the vertical input ... WITHOUT altering the time
base or trigger level. The sinewave of the signal to be tested should
be on the CRT, with the trace being triggered from the external
trigger input, or the reference signal.
The sinewave now on the CRT
likely will not have it's zero- |----**-----|-----|**---|
crossing starting at the first | * |* | * * |
vertical graticle as the reference | :* | * | :*| * |
signal did, but some place else. |-*---|--*--|---*-|----*|
On the illustration to the right, |*: | * | *: | *
the zero crossing occurs about * : | *| * : | |
0.3 divisions to the RIGHT. This |-----|-----**----|-----|
can now be converted to the phase : : :
angle in degrees. In the -->: :<--.3 div. :
illustration, one complete cycle :<----------->: 2.5 div.
takes 2.5 divisions, and the phase
delay from the reference is 0.3 div. SINGLE TRACE SCOPE
The phase shift is therefore: PHASE MEASUREMENT
  Phase shift = 0.3 div/2.5 div. x 360
              = 0.12 x 360 = 43 degrees

The SINGLE-TRACE scope method is a little easier to do if you make
the sine wave of the reference to be 4 divisions for one cycle, thus
making 90 degrees per division. The phase angle can be guestimated
a little quicker with the signal to be phase measured.

It is noteworth to mention that the above examples, measuring the
phase through a delay line at 7 MHz, would require an oscope with a
20MHz or higher bandwidth, if for no other reason, then just assure
that the time base is fast enough to display 1 or 2 sinewave cycles
on the CRT. If at your fastest sweep speed, the 7MHz signal is
displayed as many cycles, then obviously the accuracy that you can
determine the phase angle will be highly degraded.

If your scope has a limited bandwidth of only a few MHz or less,
there are still useful phase measurements that can be performed.

One interesting experiment is to measure the phase shift of the
audio signal at different frequencies as it travels through the
stages in a CW audio filter. This is done by putting the input to
the CW audio filter on CH.1 and the output on CH.2. What is the
phase shift of the wanted vs. unwanted frequencies? Recall that an
audio filter works by cancelling out (180 degree phase shift) the
unwanted signals, while re-enforcing (0 degrees) the frequencies you
wish to pass. There will only be one audio frequency for which
there is a 0 degree phase shift. This will be the "pole frequency"
of the active filter, or the frequency you wish to have the maximum
gain. For CW QRP rigs, this should be around 700 Hz.

And finally, on a limited bandwidth scope, the phase angles of higher
frequencies can be determined by applying the reference signal to the
vertical input and the signal to be phase measured on the external
horizontal input. This will form a lissajous pattern, the angle or
tilt will signify the phase angle.


This is the last part of this series of articles on OSCOPES posted
to QRP-L. There will be another series on scope measurements posted
in the future (many months) that will include some advanced
techniques, such as measuring sideband rejection, tuned circuits,
filter responses, group delay, VCO phase noise, etc. This will be
the contents of part 2 of the oscope article for the Winter QRPp.
I haven't written it yet or made hardcopies of the scope displays.
But following publication in QRPp, I will convert it to a text file
and post it to QRP-L as I did this one.

PS - making those waveform illustrations really sucked swampwater!

72, Paul Harden, NA5N

Search QRP-L Archives

[ QRP-L Archive | ]
[ 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 ]


This archive was generated by hypermail 2b29 on Fri Jun 02 2000 - 11:42:46 EDT