A test of Orion's external CW keying
[Please read using a fixed with font such as Courier New.]
Introduction
============
TenTec's Orion is being described as having state-of-the-art CW keying.
ARRL's "Product Review" and "Expanded Test-Result Report" both contain
a scope printout indicating severe and varying pulse shortening.
This text describes a series of more detailed measurements proving
that keying pulse is being shortened by a significant and randomly
varying amount.
Readers not interested in details may want to skip
directly to "Discussion" below.
Setup
=====
1. TenTec Model 565 Orion S/N 10C10083
firmware version = 1.369 (Master Reset performed after loading)
set power level = 40 W (due to attenuator ratings)
30 minutes warm-up with intermittent transmitting
no controls were touched during measurements
SWEEP off, freq = 7010 kHz
internal keyer off (note: "CW Weighting" control does not affect
external keying)
front panel CW jack:
open circuit voltage = 4.19 V
short circuit current = 407 uA
Rinternal = 10.3 kOhm
2. Agilent 33120A Function/Arbitrary Waveform Generator
Rout = 50 Ohms
amplitude = 2.5 Vpp into 50 Ohms (5 Vpp open circuit)
offset = +1.25 V into 50 Ohms (2.5 V open circuit)
shape = square, duty cycle = 50 %
burst mode: count = 1, phase = 180 deg
this provides a perfect keying waveform
with independently adjustable period and pulse width
3. Narda 776B-30 Attenuator (50 W CW, 30 dB, 18 GHz)
4. Tektronix TDS 3054 Digital Phosphor Oscilloscope (500 MHz, 5 Gs/s)
Procedure
=========
Keying signal from 33120A was applied to Orion's front panel jack.
Both the keying signal and the attenuated RF output signal
were observed and recorded using TDS 3054. Timing points for
the RF pulse width measurements were set at 50 % envelope amplitude.
Results
=======
Test 1: Delay and pulse width
------------------------------
"CW Rise/Fall" control = 5 ms
keying pulse width = 20.833 ms ("dit" duration at 60 wpm)
For each row in the table a number of pulses were issued
until the steady state was reached (it could take a couple of minutes)
and then 25 measurements were performed and results noted.
keying RF observed
pulse "dit" CW QSK leading trailing pulse pulse
rate rate delay edge delay edge delay width shortening
[Hz] [wpm] [%] [ms] [ms] [ms] [ms]
-------------------------------------------------------------------------
2 5 0 20.6-27.1 14.2-19.6 9.6-16.6 4.2-11.2
2 5 100 22.4-28.4 14.3-19.9 7.5-15.7 5.1-13.3
-------------------------------------------------------------------------
10 25 0 22.6-25.3 14.6-18.6 12.1-15.7 5.1-8.7
10 25 100 22.6-25.2 15.0-19.3 10.6-16.1 4.7-10.2
-------------------------------------------------------------------------
24 60 0 25.9-29.9 14.6-17.6 7.4-11.9 8.9-13.4
24 60 100 26.2-29.0 13.9-17.3 7.3-11.9 8.9-13.5
-------------------------------------------------------------------------
Corresponding averages and standard deviations are:
keying RF observed
pulse "dit" CW QSK leading trailing pulse pulse
rate rate delay edge delay edge delay width shortening
[Hz] [wpm] [%] [ms] [ms] [ms] [ms]
-------------------------------------------------------------------------
2 5 0 23.3, 1.3 17.2, 1.6 14.4, 2.3 6.4, 2.3
2 5 100 24.2, 1.5 17.0, 1.4 13.3, 3.0 7.5, 3.0
-------------------------------------------------------------------------
10 25 0 24.0, 0.6 15.9, 1.1 12.8, 1.0 8.0, 1.0
10 25 100 23.8, 0.8 16.8, 1.4 13.9, 1.4 6.9, 1.4
-------------------------------------------------------------------------
24 60 0 27.9, 1.1 16.3, 0.9 9.4, 1.5 11.4, 1.5
24 60 100 27.6, 0.7 15.6, 1.0 9.0, 1.8 11.8, 1.8
-------------------------------------------------------------------------
Test 2: RF Rise and Fall Times
-------------------------------
keying pulse width = 20.833 ms ("dit" duration at 60 wpm)
For each row in the table a number of pulses were issued
until the steady state was reached (it could take a couple of minutes)
and then a single measurement was performed and results noted.
RF RF
pulse "dit" CW QSK CW Rise/Fall rise fall
rate rate delay setting time time
[Hz] [wpm] [%] [ms] [ms] [ms]
----------------------------------------------------
2 5 0 3 0.26 1.5
2 5 0 5 0.38 3.0
2 5 0 10 0.38 3.0
----------------------------------------------------
2 5 100 3 0.24 1.5
2 5 100 5 0.38 3.0
2 5 100 10 0.38 3.0
----------------------------------------------------
10 25 0 3 0.46 1.5
10 25 0 5 0.75 3.0
10 25 0 10 0.75 3.0
----------------------------------------------------
10 25 100 3 0.46 1.5
10 25 100 5 0.75 3.0
10 25 100 10 0.75 3.0
----------------------------------------------------
24 60 0 3 1.6 1.5
24 60 0 5 3.0 3.0
24 60 0 10 3.0 3.0
----------------------------------------------------
24 60 100 3 1.6 1.5
24 60 100 5 3.0 3.0
24 60 100 10 3.0 3.0
----------------------------------------------------
Discussion
==========
Test 1: Delay and pulse width
------------------------------
Orion performs a randomly varying shortening of the CW RF pulse
relative to the CW keying pulse. Shortening values from 4.2 ms
to 13.5 ms were observed. The ARRL's scope printout indicates
shortening of 8.9 ms for the "first dit" and 12.1 ms for the
"second dit", which is in agreement with my results. No substantial
difference was found between 0 % and 100 % "CW QSK delay" settings.
According to observed values, the keying waveform with a perfect
1:1 dit/pause ratio will be distorted as follows:
CW speed dit time dit/pause ratio
[wpm] [ms] [1 : value]
-------------------------------------
25 50.0 1.18 - 1.74
30 41.7 1.22 - 1.96
35 35.7 1.27 - 2.22
40 31.3 1.31 - 2.52
45 27.8 1.36 - 2.89
50 25.0 1.40 - 3.35
55 22.7 1.45 - 3.93
60 20.8 1.51 - 4.70
It is recommended to make all keying pulses longer by approximately 9 ms,
at the expense of the following pause (I think WriteLog can do it).
Doing so will "center" the RF pulse width at the desired CW pulse width,
although random variations will still remain.
Test 2: RF Rise and Fall Times
-------------------------------
Observed RF rise and fall times were found to be much shorter
than the "CW Rise/Fall" menu setting.
The leading edge of the RF pulse is "chopped off" and will produce
strong "clicks". This effect is reduced with a continuos stream of "dits".
No substantial difference was found between 0 % and 100 % "CW QSK delay"
settings.
Various other findings
======================
* Oscillating artefacts were observed in the rising edge.
Three scope screens were dumped to TIFF files and will be
emailed to interested parties.
* S meter has indicated the power as either 0, 15 or 60 W,
although the scope reading did not change at all between
those three indications.
73,
Sinisa YT1NT, VA3TTN
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|