Signal filters for weather balloon

E

elsey.jack

New Member
Hello everyone!

I am an environmental engineering major, and for a research project I need to build a weather balloon that will go up 100 meters in the air to get temperature measurements.

Now, to get these temperature measurements back to the ground to a receiver module, I plan to transmit a signal in binary from the balloon using a pair of walkie-talkies.

After some fiddling with circuit ideas, I decided on the following system:
  1. The weather balloon transmits an audio tone at a certain frequency to indicate to the receiver module that the temperature data is coming.
  2. Using a different frequency tone, the weather balloon transmits the temperature reading in binary.

In order to get the receiver module to be able to differentiate between the two signals, I decided to use two bandpass filters. My circuit design is below. Vs is the signal coming from the receiving walkie-talkie. One bandpass filter is on the left, and the other is on the right.



I created an excel file and predicted how each filter will handle different frequency signals. The graph of how each filter will react is below. The blue line is "Channel 1", designed to pass a signal at 100.3 Hz, and the red line is "Channel 2", designed to pass a signal at 702.7 Hz.



Everything should be gravy, right? Well, that's not quite the case. :-(

When I assembled the circuit shown above and transmitted a 702.7 Hz signal to Vs using a modified pair of walkie-talkies, this is my oscilloscope readout. CH 1 is the output from the "Channel 1" bandpass filter, and CH 2 is the output from the "Channel 2" bandpass filter.



As you can see, there are two problems:
  1. The walkie-talkies increased the frequency of the transmitted signal from 702.7 Hz to 1540 Hz. No big deal. I expected some distortion.
  2. There was no difference between the two signals! According to my spreadsheet, there should be a significant difference in amplitude between the outputs of the two filters at this frequency.

Any ideas on what I might be doing wrong?





P.S. For those who are interested, the formulas that I used to predict the behavior of the bandpass filters follow.



A bandpass filter is made by feeding the output from a lowpass filter to a highpass filter. If fb (called the break frequency) for the lowpass filter is high enough and if fb for the highpass filter is low enough, a band of frequencies passes through both the highpass filter and the lowpass filter with a negligible decrease in amplitude. The transfer function for a filter is the output voltage divided by the input voltage given a certain frequency f. To get the transfer function for the bandpass filter made with a lowpass and a highpass filter, multiply their two transfer functions together.
 
M

MFB

Senior Member
The techniques described in my recent posting to User Projects (title, Narrow-band radio telemetry) will transmit data reliably using a pair of low cost walkie-talkies. Input on the mic input of one and output on the speaker of the other. However, this may not be a legal use of this type of radio devices in all countries.
 
M

MBrej

Member
Hi, when you 'listen' for tones from a radio, normally a tone decoder would be used, such as the LM567, which is very good at picking out a set tone from alot of noise.

Also, when sending binary data over radio with tones, usually one tone is used for '1', and a different one for '0'. This would need two LM567s to pick out each tone. Another way is to use a modem IC such as the MX614, which will produce tones and also decode them on the other end. 100m line of sight isnt a very large distance however, so most of this paragraph may not be needed for your application.

Walkie talkies can be used, but they may begin to attenuate higher frequencies, which may become a problem if you are sending two tones through. You should be able to test this with the equipement you have however, and so chose two tones which are not effected, or build a treble boost filter to undo the attenuation by the radio.

Matt
 
E

elsey.jack

New Member
Polarized

Hey, thanks for the help everyone! I guess I've kind of been trying to reinvent the wheel here.

One last quick question: If I use the LM567 tone decoder, do my capacitors in the circuit have to be non-polarized?

Thanks,
Jack
 
B

BeanieBots

Moderator
Your capacitors need to be non-polarized if they see reversing polarity. Nothing to do with whatever supplies or uses the signals.
At a quick glance, all the capacitors in your circuit need to be non-polarized.
If in doubt, use non-polarized.
 
You must log in or register to reply here.
Top