CW Rise Time Settings and Key Click Effects
with Icom IC-756ProII HF/6M Transceivers
by John Seney, WD1V
Many Icom IC-756ProII radio users discover that there is a "CW Rise Time
Setting" menu that enables them to adjust and set the rise time value for
each keyed CW envelope. Too fast a rise (or fall) time can cause "key clicks".
Key clicks are essentially analog distortions in a keyed CW waveform that
contain out of band products that produce noise (clicks). They most often
occur at the very beginning of the transmitter being keyed. If the keying
circuit lets the transmitter go from key down zero power to full power too
rapidly, the transmitter's output becomes distorted. Distortion is easier
to understand and deal with when it can be seen vs. heard.
Actually Seeing Analog Phenomena with Digital Recording
Seeing and analyzing "clicks" with an analog scope is difficult. To capture
single shot transient events requires a digital scope. But getting good
envelope recordings with a digital storage scope can be a challenge due to
typical memory depth limitations (many scopes have 50,000 points or less). Short
memory reduces sample rates over long recording times. For instance, at a 2 ns
per sample (500 MS/sec.) rate:
500 samples records for 1 microsecond
1000 samples records for 2 microseconds
50,000 samples records for 100 microseconds
500,000 samples records for 1 millisecond
2,500,000 samples records for 5 milliseconds
If a long enough record is obtained (in this case at least 3 milliseconds
or more is needed), another challenge is to display all of the data.
Many digital scopes limit their displays to just 500 data points. So how do you
observe 2.5 million or more of anything at once?
A Digital Viewing Solution
The following screen shots and measurements give some indication of what effects
various rise time settings actually create. The data was taken with a 1 GHz BW
Teledyne LeCroy "WaveRunner" digital storage scope Model LT-584 set to .5 ms (time per
division - 5 ms total recording time on each screen).
The default Teledyne LeCroy display shows a compaction of the entire contents of memory
using a patented max/min binning algorithm. Thus the envelope of 2.5 million samples
can be seen as 1 continuous sweep.
Each trace was recorded as a single shot capture using 2.5 million samples at
500 MS/sec. of a 3.558 MHz CW carrier. Several shots were taken to verify stable
triggering and consistent results for each subsequent rise time setting change on the
IC-756ProII. RF power was set to minimum (viva QRP!) output and fed directly to the
scope's 50 ohm input.
This first screen displays 4 traces annotated as follows:
1 @ 2 ms rise time setting
B M2 @ 4 ms rise time setting
C M3 @ 6 ms rise time setting
D M4 @ 8 ms rise time setting
Note the overshoot and undershoot effects in the first 2 traces.
The "too fast" rise and overshoot create the unwanted "key clicks" effect.
The next screen shot are the same 4 traces of data but now displayed "overlayed"
- stacked on top of each other.
Note that since the trigger point was in the middle of each 2.5 million point array,
the shapes of the RF building up on the left hand side of the display makes comparative
times from key down to full carrier output easy to see.
Digital Measurement Forensics
An additional scope challenge is to process the data contained in long records and make
it useable. A Teledyne LeCroy can extract the positive trend amplitude of each 2.5 million data
point envelope, filter the trend, and then automatically measure its rise time in less
than 1/10th of a second using built-in DSP math functions. An alternative is downloading
the waveform data via an Ethernet connection and using MatLAB functions to extract
desired measurement values.
The next screen shot captures the envelope keyed at the 2 ms setting, filters the
envelope, and measures the filtered response rise time of trace B.
The next screen shot captures the envelope keyed at the 4 ms setting, filters the
envelope, and measures the filtered response rise time of trace B.
The next screen shot captures the envelope keyed at the 6 ms setting, filters the
envelope, and measures the filtered response rise time of trace B.
The final screen shot captures the envelope keyed at the 8 ms setting, filters the
envelope, and measures the filtered response rise time of trace B.
Bi-Modal Effect Based on Keying Speed
Note these rise time response measurements are proportional to IC-756ProII settings but
about 4 times faster than radio settings indicate. A difference was noted between single
shot vs. multiple random CW characters. When any 2 characters were sent in rapid succession,
it approximately doubled the rise time - reducing the effect. Thus shorter rise time
settings may produce "speed of keying dependent" clicks.
Conclusions, Comments, & Feedback
While taking this data, I also listened to the zero beat on a separate receiver (Elecraft K2)
and could easily detect clicks at the 2 and 4 ms. settings. but not at 6 or 8 ms.
I've had my IC-756ProII since November, 2002 (sold in 2007) and find it to be an incredible radio. One of the
fortunate aspects is the phenomenal range of its many settings - such as noise
reduction. Listening to a IC-756ProII is an unusual "radio" experience. Tuning stations in clear
spectrum is trivial. But it becomes possible to activate DSP (filters and noise reduction) that
enables hearing specific QSOs in spectrum that is full of traffic, QRM, and or QRN. Radios without
IF based DSP don't get you there. The sensation is like listening to something that you know is highly
processed. Since you hear what it is like without the DSP, its very, very cool. It reminds me of the
spy movies where "gadgetry" resolves a single conversation in a crowded restaurant. Bond. James Bond.
I'm not certain what value the 2 and 4 ms. rise time settings afford but that one can legitimately
tune out of a possible problem is the way it ought to be. Perhaps the default rise time setting should
be 6 or 8 ms.
I welcome your comments on this data at wd1v@arrl.net
73
John Seney WD1V [FN42]
FIST #5014 CC #1151
K3/P3 Flex 5000A
http://www.wd1v.net
Return to Main Menu
