About Propellers Other Products Articles Bookshop Gallery Links Contact Supercool

Propeller Dynamics

Essential reading for model aircraft contest fliers. This is the only book on the market explaining propeller theory in non-mathematical terms. A rattling good read, I know, I wrote it.


Remote Doppler Recording


The Doppler method for determining airspeed and in-air RPM has now gained wide acceptance. See previous articles on Doppler for more information. The method has application in R/C pylon race and C/L speed events, where mufflers are not required. F3D and F2A are two events that benefit particularly from in-flight evaluation of engine and propeller performance.

However, there is one disadvantage to the method, which applies during single-handed testing. This disadvantage relates to the position of the microphone. In the case of F3D, the recording microphone should be placed at least 200m past #1 pylon, and in F2A and F2C 50m from the edge of the circle. This is highly inconvenient in both cases, enough to deter even the fittest modeller, or in my case, one whose knees now creak like the doors of a haunted house!

There may be a solution to this problem, which I now propose. I have not tested this method in the field, so I cannot guarantee that it will work. So the risk is yours, a situation which suits me just fine!

There are now on the market little walkie-talkies (transceivers?) with good range, many features and a reasonable price (since you need two). The pair I have are Uniden UH-040XR units that operate on the UHF band at 476 MHz. I paid AUD99 each for them from Dick Smith some time ago. I note that recently similar units have been going for AUD89, which I would call good value for such quality items.

The idea is to place one unit at the point for sound detection, with it fixed on transmit (PTT), using perhaps a cable tie to hold the switch down. With a range of up to 3 kilometres, there should be no problem with positioning, even for F3D. This idea will not work if the unit has an internal timer which switched the unit off as a power saving measure. My units seem to be OK in that regard.

A second unit is placed in the pits, where it is set to "Monitor" and connected directly to a tape recorder input. On the Uniden unit, if the Monitor key is held down for 2 seconds, it stays on permanently, an ideal circumstance. The Uniden has a 3mm output socket labelled SPKR. An ordinary mono (or stereo) audio cable will fit that socket, and may be run directly to the MIC input on the tape recorder.

Well that's just about it. Oh, I use a Digitor tape recorder, it is very cheap, well featured and battery life seems to be much better than the old Realistic brand. The only problem that may apply to the system is whether the transmitter has a threshold level for sound input. That is one of the bad features of mobile phones when used for Doppler in F3D: they tend to cut in and out. I have not been able to find a threshold on my Uniden units, but if there is one it will soon show up in field tests!

Now a tip for F3D. There is no reason why multiple transmit units should not be used. The Spectrogram trace works as well for multiple models as it does for one, just makes identification a bit of a problem. So I suggest having 2 transmit units, one 200m past #1, and one 200m out past #2and #3. This way you get speed/RPM in the straights, and speed/RPM as an average in turns of going around 2 and 3.

Do let me know how you get on with this, and good luck trying!


Addendum to Remote Doppler Recording

In the article entitled "Remote Doppler Recording", on my website at www.supercoolprops.eftel.com, I discussed the possible use of Walkie-Talkies for facilitating the collection of sound data. I have now done some testing on this method and report the results here.

The Uniden UH-040XR is a reasonably priced (AUD100) and high quality transceiver operating in the UHF region at 477 MHz. The unit features 2 sockets, one a 3.5mm mono audio socket for connection to an external speaker, the other a 2.5mm mono audio socket for use by an external microhone.

The unit has no automatic gain control (AGC) feature. When using the inbuilt microphone, it is necessary to hold the unit within 150mm of the mouth to get good voice signal strength. 

In the case of recording model engine sounds, this means you need a really noisy situation, such as in control-line Goodyear and F2C racing. That being the case, you don't need the system at all, as a standard tape recorder handles this easily and chances are the operator will be near the flight circle anyway.
In F3D, the operator can easily be 500m away from the desired recording location, and distance from the model can be anything from 100m to 700m. To ensure good sound pickup, it is desirable to boost the sound signal going into the unit: this implies the application of an external AGC unit connected to the microphone input on the Tx.

The point of having an AGC, is that weak sounds are amplified more strongly than loud sounds, so that the AGC output is largely independent of sound level. 
Provision of this AGC unit has held up my work on the acoustic antenna for some time now. The cost of design and production of a custom AGC board is rather high. However, a solution is at hand. The electronics component and kit supply company Oatley Electronics (www.oatleyelectronics.com) offer a Laser Communicator kit, part number K073, price AUD31, which rather interestingly does not include a laser!

However, the transmitter board does feature a rather useful AGC section. Perhaps this could be modified to suit the Uniden? I spent half a day trying to make these mods work, with a rather comprehensive degree of failure. Now my electronics is rather dated (my Uni education was on valves!) but I am not that hopeless. Also, my Huang Chang oscilloscope does not tell lies, and it said I had a good signal going into the Uniden.

Just when I was ready to toss in the towel, I realised I had the AGC output connected to the SPEAKER socket on the Uniden! This degree of idiocy has plagued me all my life, so read on and perhaps you will find some more.
In modifying the Laser AGC board, it was first necessary to know just what signal level the Uniden required from an external microphone, so I purchased the Uniden hand-held Speaker/Microphone unit which plugs into the aforementioned audio sockets on the transceiver. Pulling this apart revealed an electret microphone, a 3.3k resistor, a 470pF capacitor, a switch and a speaker. No fancy electronics. This was promising, so I measured the microphone bias voltage, which was -125mV (rather less than I expected), and the output signal level. This was +-10mV, about right, and less than the expected 30mV.

This meant that the AGC had to deliver a maximum signal of +-10mV into the transceiver MICROPHONE socket. Now this could be tricky, as most audio circuitry is designed to deliver line-level signals at about 1V. And indeed, the Laser Tx was similar.

To commence the mods, I removed the unwanted components from my previously built Laser Communicator. This was no hardship as the existing laser diode I had been using was recently made illegal, and I had already bought a new toaster. 

Parts removed were VR1, Q2, R13, R14, R10 and C10. I also cut the conductor 5V supply track to R11.

VR1 was then replaced with a 500 ohm trimpot (Dick Smith part #R1763), which was a pin-for-pin replacement. 

Now the problem was to make the right new connections. The signal I wanted appeared on the output side of C4, which was now open circuit after removing R10. So I connected this point to the supply side of R11 and broke the track to supply. Bad mistake, but I didn't know that yet!

The combination of R11 at 47k and VR1 at 500 ohms now formed a voltage divider which would provide my 10mV to the Uniden mic input. Of course, this was adjustable in case I got it wrong. And I did. But not by much.
Now I didn't know the input impedance of the Uniden mic in, so I figured it would be nice to have a unity gain buffer to drive my 10mV sound signal. As it happened, in removing Q2, I freed up one half of the LM358 dual op amp, so this was a candidate for the buffer. Big mistake.

Anyway, I hung a 1.5k resistor across the buffer output and neg in, mainly because I don't know any better, and put a 1k resistor and 1uF cap in the output line to the Uniden.

The 1k was an effort to protect the Uniden input from over-current and the cap was to block any stray DC, including the -125mV from the Uniden.
So I fired it up and started looking at waveforms. The input to plus-in on the buffer op-amp from the C4 cap was fine, and easily adjusted to 10mV at high sound level. So far so good. 

But the ouput from the buffer was chopped off, only giving the top-half of the signal!! Disaster! What could be wrong? The problem was at the ouput pins of the op amp, so it had to be the op amp that was the problem. I had previously run a 741 op amp like this, with no problems, so I was both surprised and disappointed.

The Dick Smith catalogue listed the LM358, so I had a look at its specs. Would you believe its differential input voltage was plus only? How dopey. I guess it has something to do with making it a low power chip. But that didn't help me.
So I looked at how the other half of the chip was linked to the electret microphone output. There was an odd combination of resistors on the plus input, which appeared to be was some sort of level-shifter.

By reconnecting R11 to supply and running a line from C4 output via a 47k resistor to the VR1 pin at the R11 junction I was able to level shift and get the full-wave output from the buffer. Any DC was blocked by the output cap, so we were back in business.

After connecting the AGC to the Uniden, strapping the Uniden Tx switch to transmit and turning on the FM radio for some music, I went outside with my second Uniden and was rewarded with a strong clear signal. 

Success at last. AGC today, acoustic antenna tomorrow! 

Second Addendum!

I have now conducted some field test with the AGC connected to the Uniden 040 transceiver. A sound level at the transceiver of 70 dBA works fine, and this corresponds to a normal voice level. If any problems remain, they are most likely to concern the recording and playback quality of the tape recorder. 
I have been using a very cheap Digitor unit, which is well featured, has good battery life, and has served me well. However, I do now feel that a lot of the noise on the tape is due to the tape recorder itself. Having little experience of such devices, I do not know what to do next. Possibly a mini-disc player is the way to go, but they are expensive (at least by my standards!) . 

This criticism of the Digitor needs to be qualified. If, instead of playing back the tape on the recorder itself, it is played back on my Hitachi Stereo, then the sound quality is much better. It would be nice to find a device with digital signal processing that eliminated tape hiss.

Concerning the modifications to the Oatley Laser AGC board, there is one final point to be made.
It is a bit annoying trying to hold down the Press-to-talk button with a cable tie.
With my usual lack of alacrity, it eventually struck me as odd that the hand-held press-to-talk on speaker/microphone unit actually worked at all. All it does is switch the electret microphone into the circuit. So why did not the AGC board also cause the unit to transmit when it was in circuit?

The hand-held unit was again stripped down; this time a more thorough set of tests were done. It was found that the speaker part of the circuit had no influence on causing the Tx to transmit. Not really surprising, but as I say, this was a thorough check!

Measurements on the electret side were a bit more interesting. The DC voltages and DC current draw were measured with the following results:

Tx off -2.32 .03
Tx on -0.127 .11 

These measurements were repeated for the press-to-talk switch on the Tx itself.

Tx off -2.32 .03
Tx on -0.35 .11

Note that these numbers are generated from the unit itself, not the microphone. Perhaps it was possible to fool the Tx into thinking the hand-held was connected, when in fact the AGC was in place.

Accordingly, various resistors were placed across the AGC output (to ground on the AGC board) to produce a current draw similar to that above. It was found a value of 8k gave a current draw of .08mA, and this indeed did cause the Tx to transmit. Success!! No more switch tie down!

This is, I hope, close to the end of the story. There remains just one little idiosyncrasy, which concerns the Tx itself at turn-on. 

If the AGC is plugged in live, or without power, before the Tx is turned on, the Tx goes bananas and gives an audible error signal. It does not want to go straight into transmit mode. The cure is to install a switch in series with the 8k resistor. The order of events may then occur as follows:

1. With the power off on all units, plug the AGC into the Uniden Tx
2. Turn on the Tx
3. Turn on the AGC power 
4. Switch in the 8k resistor.

Note that there must be 2 switches on the AGC, one for power and one for the 8k resistor.

Well I hope that's about it. If any more surprises come along I will make and Third Addendum! 


 Back to Top