| The problems of collecting
engine/flight data for Doppler
analysis have been analysed previously on this website.
The placement of the microphone is important in interpreting
the data obtained; however its positioning is determined more
often by convenience in operation than ideal placement. For
this reason, the use of Walkie-Talkie 2-way radios has been
proposed by Supercool, with the addition of automatic gain
control units to improve sensitivity. These permit ideal location
of the microphone, combined with convenient positioning of
the tape recorder. But let us take one thing at a time, as
we detail the field tests of these units.
In the case of F3D, by far the most convenient,
simple and low cost approach is to place a tape recorder in
the pits. The RPM and airspeed, thus determined, represent
some sort of average coming out of #3 pylon and entering #1.
These are useful parameters, although they are unlikely, for
example, to indicate the true speed and RPM in the straights.
Fig 1: Click image to see full size
A sample recording for this position is shown
in Figure 1 (172 KB). The tape recorder was a cheap Digitor
unit, which suffers from internally generated noise. Despite
this, it has very good sensitivity, incorporating an automatic
gain control circuit (AGC), with good battery life.
The Spectrogram figure shows a multiplicity
of harmonics, mixed in with considerable wind noise. No attempt
was made to shield from the 60kph wind, nor the 35degree heat.
The data are readily reduced to RPM and airspeed, so things
look good for this approach.
For the uninitiated, Supercool operates out
of Perth, Western Australia: a location best described as
a US Naval Warfare Base. In January, the weather pattern comprises
a strong morning Easterly, warming to about 35 degrees by
noon and then cooling as a strong South Westerly (the Fremantle
Doctor) arrives about 1:00 PM. Perth is, I believe, second
only to Auckland as the worlds windiest city. Thus take-offs
are rapid, landings short. Wind corrections are required for
Doppler data, as the wind alters the speed of sound as seen
by the tape recorder.
We then switched to Uniden transceivers fitted
with Supercool external AGC's. These latter were placed in
woolly socks to cut down the wind noise. This appeared to
work quite well, with the units placed out past #1 and past
2 and 3 pylons in an attempt to capture 2 traces for additional
information from the one flight. This failed to work, as the
wind blew one of the mounting tripods over; my age-imperfect
sight could not detect this anomaly from 200 metres away.
In addition, an amplifier was placed between
the Uniden units used as receivers and the tape recorder,
to give a boost to the signal. Note that 4 Uniden units were
in place here, 2 pairs, for receiving and transmitting on
2 different frequencies. The receiver outputs were tied together
with a mono-audio 3.5mm T-piece (for want of a better name).
Fig 2:Click image to see full size
The result is shown in Figure 2 (162 KB). Rather
then the set of harmonics seen above, the fundamental frequency
was strongly represented, with the harmonics nearly absent.
Why this should be, I have no idea, and I'm not running back
to that hellish flying field to find out. Analysis of this
fundamental harmonic is again straightforward. Interpretation
is improved, as being out past #1, these data give a good
indication of the speed and RPM in the straights (results
which rather appeal to me, as the speed everywhere else will
be lower)
Finally, at an earlier testing
session, we obtained data from the Unidens without using the
rather expensive Supercool AGC's. Again, they were placed
out past #1 and pylons 2 and 3, but this time the transmit-keys
were held down with rubberbands. The Uniden units appear to
have a battery life approaching 10 hours in the transmit-mode;
when using 1700 mA/hr LiMH batteries, continuous transmission
during a testing session is no problem. This saves all the
walking to turn them on and off.
The receiver outputs were
again tied together, but no amplifier was added ahead of the
tape recorder. Consequently the signal recorded was a little
weak, as can be seen if Figure 3 (137 KB)
Fig 3: Click image to
see full size
The traces can be seen to
be interlaced. Normally, this figure would represent 2 models
up together with one microphone. However, in this case, there
is just one model and 2 microphones.
Accordingly, one trace from
#1 position gives the speed in the straights, while the second
trace gives the speed and RPM rounding 2 and 3. Lots of useful
data from just one flight.
I think it fair to say that
the field tests show all the methods to be useful, it's just
a matter of judging the best approach for your own testing.
And you are quantifying
your F3D performances, are you not?
Note that in all these cases
the microphone is placed outside the course: this means that
the caller cannot operate the tape recorder. The recorder
must not be placed inside the course. For Doppler to work,
we require both the frequency with the model headed directly
at the microphone, and directly away from the microphone:
this is achieved automatically when outside the course.
They correspond to the maximum
and minimum frequencies in the Spectrogram trace. These frequencies
are obtained using the very valuable Spectrogram program from
Richard Horne's website.
Until next we meet, earn
lots of money so you can afford the very nice Supercool props
and AGC's, so that Supercool can have some fun too.
|
