Long range radio communications

M

moxhamj

New Member
Sending radio messages 10 metres or so is quite easy but when the distances go over about 200 metres then the power requirements rise. This means transmitters require licences and the whole exercise gets complicated. This project was designed to send a signal 500 metres through trees and over a hill using multiple repeater modules spaced about 100 metres apart. The signal is passed between modules until it reaches the destination.

Each module is solar powered and self contained. In order to keep costs down, each module runs on a single 1.2V rechargeable battery which is stepped up to 5V for the picaxe and 9V for the radio transmitter.

Two picaxes are used - one to manage the solar charging and one to manage radio transmissions.

The complete description is at http://drvernacula.topcities.com/315_mhz_solar_powered_radio_rptr.htm
 

Attachments

  • SmallRadioRxTx.jpg
    SmallRadioRxTx.jpg
    25.1 KB · Views: 347
Very nice project indeed! I also liked your PCB.
I really liked your uC implementation of a charging circuit. I'll study it in more detail and probably use it in a 'green-energy' project.
 
Picaxester - the stepup is in the schematic on the website, but essentially it is a potcore with 20 turns and 35 turns of wire. I have a whole lot of surplus potcores lying around ranging from 1cm to 4cm in diameter. They all seem to work. Some didn't even have formers, so I just wound the wire round a piece of 13mm pipe in a rough sort of bundle then slipped it off into the potcore. Put the potcore together then dunked it in a tin of varnish then let it dry. You can buy a commercial one www.jaycar.com.au and search pot core. Interestingly small toroids didn't work - the inductance probably wasn't right.

There is another circuit at http://www.talkingelectronics.com/projects/PowerSupply5vSolar/PowerSupply5vSolar-2.html and this may be slightly better as it uses a coil rather than a potcore, and coils of 0.082mH at <=1.4 ohms are commercially available already pre-wound. I'm going to look at this circuit more and combine the overvoltage circuit from my working circuit so the regulation is more efficient (this second circuit wastes energy when it gets to full volts).

Max constant current output is 10mA so it is a bit flea-power. But with 0.047F supercaps then bigger currents up to amps can be drawn for brief periods.
 
Tx - Rx receiver interface with PICAXE question

I notice on the schematic for the PICAXE/RF units that they connect directly to the picaxe pins.

Is any type of interface required?

What type output/input is required for the RF units?

Thanks, Stan
 
Pretty much all the small radio modules can interface directly with picaxes. The input to the modules is a 0-5V waveform and the output is also a 0-5V waveform.
 
DR_A--in your text you say "The 5V regulator is an ultra low quiescent current module, as standard 78L05 devices consume 3mA...". That part is labelled 7805L on the schematic, but I am unable to locate such a part (at Jameco)--is that the full part number? Where were you able to obtain it?
 
My mistake. I used a LM2936 in the end.
I should add that this is still a work in progress as I think I can improve the range using something like Manchester coding and by siting the transmitter module some distance from the board as there is some RF coupling back into the circuit. Also getting it 3 metres above the ground makes a big difference to the range. Will post more as results come in.
 
Last edited:
Have updated the website after doing a number of experiments http://drvernacula.topcities.com/315_mhz_solar_powered_radio_rptr.htm Have found a number of things that make a huge difference to the range - i) need 2 metres clear air round the Tx in all directions, ii) needs to be at least 2 metres off the ground, iii) the number of 1s and 0s must be equal and iv) The Tx unit needs at least 1 metre of wires joining it to the board (3 wires, Gnd, 9V and data). I have no idea why (iv) is true. Maybe the wires act as a ground plane. Maybe they are an aerial. But mounting the 1000m Tx unit on the board gives a 10 metre range and mounting it on a metre of wires gives a 200 metre range. I can't get the 4000m units to work without huge storage caps - 47,000uF supercaps are still not enough. But the 1000m unit is going almost as far.

Re Haxby - am sending tank level data, temperature, a sensor to detect a car going through the front gate, turning on pumps and turning on lights with PIR sensors.
 
Dr_Acula

I suspect if you trimmed the antenna to 468/315 feet you would get better results. Your program appears to use enough UUUUUs to sync the receiver. So not sure about the equal 1 and 0 comment. My experience has been with ZeeBee operating at a higher frequency. In order to remove clashes you do a random time delay before transmit which is good, however, carrier detect would be even better.

I ordered some of the RF modules so do not have direct experience with them yet. However, ASK is not the best way of keying, FSK would be much better. If there is a lot of RFI or external noise sources, which could occur when several wireless sensors all trying to get the channel I can see it would be very difficult to read the desired ASK modulated wireless sensor.

It sounds like your electronics may be current straved, ?

How well does your system work with just a single or dual remote sensor?

One thing that has worked for me in the past is when the receivers are listening we would build a goto route table of who they can hear and the one with the highest number would then be targeted as the one we would use as the repeater. If it did not relay the data then the next one in line would be used. Not sure if the PICAXE has enough memory to handle this more complex program or not. If a small number of remote sensors this works very well. Sometimes you get funny routing but usually 100% communications between nodes and control.

Just some suggestions on where to possible look for answers.

Cheers, Stan ak0b
 
It is current starved for the 4000m module (100mA 9V) but not the 1000m (40mA 9V) module. These are about 900mW and 360mW respectively - a lot more than car locking units which are only 10mW. When I get only 200m from a module that is listed as 1000m I think maybe I am not doing something right, but I think the real reason is the manufacturers are claiming a range based on how far you can hear a signal rather than how far data can be sent. Attaching a small amp and a speaker to the Rx, one can certainly hear something a lot further than 200m. But data requires a clean signal with full 0-5V swing. With ASK as the signal fades it doesn't swing full rail to rail. I have built other circuits to try to pull out a noisy signal and amplify it but they are complex and don't add more than another 50m of range.

I think what this is really showing is how marvellous the ear is at detecting signals in noisy environments.

The antenna at 315Mhz is 23.8cm. I tried an 95.2cm antenna and it increased the range another 50m or so. I think the wires going to the module are acting as the ground of a dipole, so I might try tuning the length of these power/ground wires to a multiple of the 1/4 wavelength.

I might also try using a real ground with an earth stake and also a quad as this does not need a ground plane.

There is no interference at all - I've listened for hours with a speaker and it is just white noise out there. Not like poor Stelios who has a ham radio enthusiast next door.

Certainly it is nice to see a signal bounced 200m to one unit and then another 200m to the next unit knowing that second unit is out of range of the control signal.

Am using about 5/6th of the memory of an 08M but might be able to squeeze in a round robin. The problem is that each node can only hear nodes near it so distant nodes can't hear the control directly.

I think this might end up a hybrid solution with lower power nodes round the house and the higher power nodes handling the backbone of longer distance comms. One interesting thing about a wireless mesh is that adding more nodes increases reliability.
 
Last edited:
receiver type

Sure sounds like you are doing everything correctly. Are you using the superhet or superegen receivers?

A couple of tricks we have used for the pulse amplitude problem is to make multi-sampling and not just the 1 sample done by a built-in UART function. I doubt if we can do that with Basic programming. But might be worth while to consider a PIC using the analog input for a total assembly lang software UART that has voting.

Also might be able to apply the receiver data to the D input of a FF and run the output through a simple RC low pass filter in front of the PICAXE creating a regenerated pulse.

It is always tough to achieve 100% communications with a low dollar budget, but the RM1SH and TM1000-1 coupled with a PICAXE sure look interesting if one can pull it off.

Might also consider using a J-pole antenna in place of the vertical that would give a little gain. I doubt if we can FSK the SAW elements, that would give us 3 db if we could pull it off.

Interesting project

Cheers, stan ak0b
 
Do you have antenna restrictions for those radios, either legal or practical? If not, you might try something I just remembered.

Back in the 70s, I lived in an apartment where I had antenna restrictions. I built a 3 half waves in phase vertical collinear and hid it in a tree outside my 2nd floor apartment. It worked very well on 146 MHz.

If you don't have restrictions, you might try one of these. Easy to build from a bamboo fishing pole, hookup and 12 ga wire, electrical tape, and coax cable. Non-critical tuning.

Get it up as high as you can, in the clear.

I found the design in an old ARRL Antenna Handbook. If you can't find one at your local library, let me know. I'll dig into my trusty ARRL Handbook and see if I can re-create the design.

Tom (AB9B)
 
433 MHz &amp; Swan's Law

For those who've just come in,the simple 433 telemetry "Swan's Law" repeater => www.picaxe.orcon.net.nz/all3.jpg,rustled up a while back, may be of interest. As proof of concept,rather than acid testing, we managed reliable comms at SLOW DATA RATES ( 300 bps) to ~½km OK. Jaycar rx/tx units (~US$5 each) were found the most satisfactory performers (especially for receiving) & assorted trials with a 6dB "cotanga" Yagi => www.picaxe.orcon.net.nz/yagi433.jpg doubled range.

Keep in mind the data being sent was just °C temperature (via a DS18B20), so occasional wonky readings( not that I recall them arising) could be lived with. In Dr-Acula's case however this could lead to false alerts & all manner of tele-command woes. Visions arise of kangaroos triggering sensors at will & bounding onto the property to browse at his lettuces, or worse to shower under his stored water! What happens if a koala nibbles a sensor?

The Asian 315 MHz units, sent by Dr-A to us here in NZ & tested by both myself & Andrew "Brightspark", sadly were not convincing performers I'm afraid. Stan
 

Attachments

  • all364.jpg
    all364.jpg
    91.3 KB · Views: 85
Last edited:
Thanks manuka. I have found putting in a checksum halves the reliable distance. For reference, a 1000m module will go about 150 metres reliably all the time with a checksum, will go about 300m with no checksum about half the time, and there are certainly places at 700m where one can occasionally pick up a signal. There are some trees here and there, and clearly comms across water with a clear line of sight and parabolic antennas can go a lot further as manuka has shown.

The problem with a wireless mesh is that there is no preferred direction, one wants the signal to go in all directions (except up), so one has to live with a lot less antenna gain.

One good thing about this circuit is that there are lots of modules out there that one can use.

I might go out to the "brains trust" picaxe community and see if it is possible to find other modules that are cheap and are in the 100mW plus range. Suggestions anyone?
 
I'm still looking for the best way of interfacing UHF radios with data communications.

DTMF, while extremely reliable, all be it slow, is starting to disappear. Its starting to get difficult to find DTMF chips, as most microprocessors are able to send and receive DTMF internally these days.

But UHF is a great way to go if it can be audio interfaced to. The local bargain shop around the corner has 1/2 Watt UHF sets for $25 each!
 
manuka - can you please post some more about the tests on the 315Mhz units?

I have done some more tests - 10mW Jaycar 433Mhz units = 50 metres with dipoles. Manuka is getting many kilometres with the similar units with high gain antennas.

I don't think any of the units I'm testing are too far out of spec. 50m for 10mW. 200m for 300mW and 400m for 900mW. But these results are all through trees and are most definitely not line of sight. Tomorrow I am testing the mesh with 7 units.

Re Haxby - I started with UHF radios and then got inspired by manuka's setup above. I've hacked into these units and certainly got slow tones going through but they have very strange filtering and square waves come out the other end as biphasic sine waves. The 3 second rule means the data has to be 1200baud or above and then the distortion means no data goes through. Maybe it is possible to work out the phase change from the filters and introduce the opposite phase change to cancel it out. These $25 radios certainly are the cheapest high power RF solution around if they could be persuaded to work.
 
DTMF rules OK!

A 6dB (& thus range doubling) "cotanga" Yagi is not really high gain in my books! Most 315MHz testing was done by Andrew "Brightspark"- I'll get him to email Dr.A directly maybe. The limiting factor is often the receiver module- some are VERY deaf. A general purpose UHF scanner helps verify this. A Kiwi firm is selling professional & powerful (4 W) UHF wide band coverage (~430- 500 MHz) Chinese ZARES 2 way tx/rx sets at bargain prices. See => http://homepages.slingshot.co.nz/~c_day/index.html

Yes- that 3 seconds is not a typo ! Laid back NZ/Aus indeed have a near crippling 3 secs an hour data reg on 477 MHz UHF CB! I keep meaning to rustle up a "PRS" ( UHF CB) approach for SiChip in fact, as these powerful & sensitive 2 way sets retail at the small change level. They're so cheap ( 4 for US$70 here in NZ) that folks dismiss them as toys, but I've easily managed 10km cross harbour links with them, & even in heavy NZ bush find ~½km feasible. Initial articles using the likes of Hellschreiber & SSTV (Slow Scan TV) met VERY keen reader response, but stern words from both Oz. regulatory types & spoil sport radio hams => http://www.siliconchip.com.au/cms/A_104977/article.html served to "dampen curiosity". Not that us Kiwis are fainted hearted types of course- especially when dealing with rule book waving Australians- as at one stage we were working on direct PICAXE "Hell" generation...

DTMF is easy to interface (even acoustically) & is so astoundingly robust that usually it can be electronically decoded in the noisiest links. Although slow (~10 cps) it's something I found most convenient inbuilt when working with Basic Stamps ~10 years back,& really miss on PICAXEs. Dedicated DTMF chips are still around- I was sent a grab bag via BrendanP- but they're now not available widely enough to support an article. However PICAXE enthusiast Brian "Wulfden" in VT,USA advises his popular Zarlink MT8870D based DTMF decoder kit is back in stock again => http://www.wulfden.org/TheShoppe.shtml.

There are numerous other toned based codes around, & I even ponder that sub-audible CTCSS (Continuous Tone-Coded Squelch System) tones (~60-257Hz) may have some PICAXE wireless mileage as well, since these would NOT be heard by casual rascals keen to tinker with your data link. Most UHF CB sets have decoding for these inbuilt now of course too. Stan.
 
Last edited:
Maybe we should lobby Rev-Ed to include internal DTMF encoding/decoding in their next line if chips... Or some sort of "audio coupled" data transmission method, which would open up a whole range of possibilities.


I have been looking at the TinyTrak APRS module. Maybe inspiration can be taken from the way that this unit interfaces through the audio input of Ham radio transmitters (or any transmitter for that matter). If only a similar solution could be had to RECEIVE these signals, it would make data communication a lot easier.
 
Never mind 300 bps! I'd a background in ham radio RTTY (radio teletype) where even 50 baud was "fast". Naturally this used modem style audio tones (AFSK = audio freq. shift keying) & it still has merit - dedicated ICs are available to handle 1200 bps "Caller ID" etc. Hams are still doing great work with V E R Y slow data - just a few ch..sec in fact

However there may even be mileage in using greatly underclocked PICAXE's at each end to handle serial data, which then sounds almost akin to "slow Morse". Initial wireless trials made the ASCII data characters quite distinctive - you could tell the differences between the chs. - although beyond direct reading by ear/eye ! Stan
 
baud rates comment

Slowing down the baud rate allows one to utilize a software UART for recovery of the data pulse. These China modules use ASK which means only one pulse is transmitted. I have not received mine yet but I expect they are transmitting the 0 state (inverted data). Anyhow back to the reason for slower baud rate. A hardware UART be it uP or chip only samples once per pulse, with a software UART you can sample 3 or 5 times depending upon the speed of the uP you are using and the program language and then vote 2 or more samples of YES it is a pulse out of 3 samples or 3 out of 5, etc.

I think other things can be done with the PICAXE - for example add ACK to the software, when the data string is received if incorrect checksum then echo a Nak to the sender or a Ack to the sender if checksum is ok. If the sender does not receive a response within X seconds or receives a Nck with its ID it should automatically resend the entire data string. This feature should fit in the 08M without any major problems.

Also adding some hardware to the output line of the rcvr to regenerate the pulse could help. This might be as simple as a RC low pass filter or a D triggered FF with LPF. Get away from the depence upon pulse level. I am an old RTTY ham operator and have done both of the above and worked stations around the world with negative level signals. However, RTTY does not require you receive all the letters to be able to read it where you need all of the data on remote sensors.

Just ideas - in a couple of weeks when my hardware shows up I will try them out. I want to run a mesh network with at least five remote sensors.

Stan
 
Hi Stan,

I'm interested in the idea of using an underclocked picaxe as per your example above.

Did you go far into attempting to receive the signal through a picaxe?

I wonder if two underclocked picaxes would be able to communicate via slooooow serial strings if they had a DTMF encoder on one side, and a decoder on the other. (using only one tone)

That would certainly make the code very easy to write.
 
Dual uPs

I think the two micros is the solution. With the wireless interface being a PIC or similiar micro and the PICAXE acting as the controller.

My current thinking is to use an analog input to set a threshold level and then a software UART with a FIFO buffer. Program the PIC in assembly and still be able to do the sensor and control functions in PICAXE basic. Solution would remain very low cost and the PIC assembly should not be very many lines of code. A simple ack/nak could be included in the PIC code. Really need that or the system could fail. Do not care for the tones, done that with security equipment, the hardware becomes complex quickly and as someone noted the chips are becoming obsolute.

Using the PIC UART opens up a lot of registers in the PICAXE for additional and more complex sensors. I ran out of registers when I tried GPS to PICAXE interface with sensors. Also I do not like shuting down the controller when doing serial in/out functions. A good wireless system will require a lot of I/O reading, accepting and rejecting of data streams.

Once those chores were corrected then one could look at the RF side of the equation and get rid of the ASK modulation upgrading it to FSK.

ZeeBee works very well but has considerable overhead to do the software at the start of the project.

We need a 20 MHz PICAXE, :-)

Stan
 
An example of slow serial transmission is at http://drvernacula.topcities.com/slow_serial_data_transmission.htm - takes a couple of seconds to send a byte. One huge advantage of doing this with bit banged code is that a timeout can be added and thus there is no serin hang because serin is not being used.

I'm not sure this will work with RF modules though. They work on audio frequencies - ie 20Hz to ?10Khz. One byte every 2 seconds is 0.5Hz which is too slow. The only way around this is to send tones at audio frequencies and then use filters and op amp rectifiers to filter the signal. It can be done with one LM324 chip but does add a bit of complexity.

If signals are not getting through reliably it means the unit is right on the distance limit. Rather than send and resend data and have those signals going back and forth, with a mesh system it is better to add another transciever unit half way between the units that are not communicating properly.
 
Help me out

Hi all,
I am trying to communicate between the outpin5 of PICAXE40X1 and UART of MicaZ mote.I connected them up using normal wires like Tx to Rx and Rx to Tx between them. But Once I use this command serout 5,9600_8,(b8) it doesn't send any data to the mote.Otherthing is, baude rate of MicaZ mote is 5700 baud. Does anyone think that do I need any extra coding for communicating between Serial and UART. Or it is just the problem of baude rate.I am really stuck here.if anyone has any solution,please tell me.Thanks
 
More about the MicaZ

Hi,
Thanks for quick reply.Basically, I am trying to communicate between MicaZ(Yes!!! Zigbee type) wireless mote and PICAXE40X1. If i can get any data from the mote then i controll other motes wirelessly using one mote as a base station connected with PICAXE40X1. Now the problem is, IS IT POSSIBLE TO CONNECT UART AND SERIAL DIRECTLY.(i have done it already but doesn't work)? I have used higher baud rate 9600 than MicaZ(5700 default).Another issue
may be the cause is voltage of MicaZ is 3.3v where picaxe has 5v. I used voltage divider and made micaZ 2.5 even though it is not working.Only test is left to change the baud rate of MicaZ.If you think any other possibilites then tell me.
 
mmr90 said:
IS IT POSSIBLE TO CONNECT UART AND SERIAL DIRECTLY
Click to expand...

Yes. Not sure exactly what you mean by "UART" ( on the PICAXE or MicaZ ? ). The issue is that the baud rates must match and so too must the polarity of the signals. You may need inverters to change signal polarities.

mmr90 said:
Another issue may be the cause is voltage of MicaZ is 3.3v where picaxe has 5v.
Click to expand...

You can run the PICAXE at 3V3 and that usually makes things much simpler.
 
Some more thoughts.

1) I have ordered some TPS61200 chips from Texas Instruments. These do a step up from a single solar cell (0.3V - 0.45V) to 5V. This is quite useful, as single solar cells are usually much cheaper per watt than panels and also broken cells can be wired in parallel. Broken cells are available at very cheap prices. http://focus.ti.com/docs/prod/folders/print/tps61200.html

2) One of the problems with sending data longer distances is the way the signal degrades. A serout/serin link assumes that a square wave into a Tx comes out as a square wave 0-5V on the Rx. As the signal fades, the volts out decrease, and then the waveform degrades. Furthermore, sending a square wave through el-cheapo handheld UHF radios doesn't work as there is too much phase change. A sine wave works much better, and then the problem is how to generate sine wave tones and detect them at a reasonable price. Second order filters are needed and really 4th order is necessary and then one has to move into switched capacitor filters and the whole exercise gets quite complex. However, I have discovered a very interesting chip MT8888 which is a complete DTMF transceiver on a chip. http://www.rockby.com.au/DSheets/28264.pdf This sends data as touch-tone codes and has all the filtering on board. One chip does both transmitting and receiving and it has 7 pins for microprocessor control. The tones are 50ms long and because DTMF is sending hex nibbles, it should be possible to send 10 bytes per second, or 30 bytes in the 3 second allowable limit per hour. I've ordered some chips and hopefully can extend the range further using DTMF rather than just increasing RF power. I've certainly had handheld UHF radios sending morse code speed tones over 1Km, but the data rate was only about 0.2 baud so I could only get one byte through in 3 seconds. DTMF ought to get the baud rate up to something useful.
 
Last edited:
Increasingly THE DTMF problem relates now to just sourcing the specialised encoder/decoder chip sets for PICAXE type applications. We're really 10 years too late for this, as they looked to have peaked pre cell phone mid 90s, & modern devices seem to have COB (chip on board) DTMF instead. I've even pondered ripping into landline phones for such DTMF chips as the UM91214B ( ref pix below )

Zillions of other DTMF IC offerings abound(ed) of course, & some even have inbuilt oscillators & reduced numbers of pins. (Yah!). Springing to mind areHolteks HT-9170/90, Oki's ML7005, Philips SSI-202 , CD22204 & even PIC based systems. Sound card based decoding, such as Scot's => http://www.qsl.net/kb5ryo/Repeater.htm can ease setup & proof of concept pains.(He's also written a nifty SI MPLEX repeater that may have some mileage on UHF CB => http://www.qsl.net/kb5ryo/simplex.htm if regs. allowed! ).

There's a nice recent ( but "duel" !) DTMF project list at => www.discovercircuits.com/D/dtmf.htm Stan
 

Attachments

  • dtmfencoder.jpg
    dtmfencoder.jpg
    25.2 KB · Views: 69
Last edited:
2400 baud is 2400 bits per second = 300 bytes per second. Preamble + message packet = 30 bytes = 1/10 of a second.
 
I bought a CRO. Suddenly it is possible to debug all sorts of problems!

1) The "1000m" units work but I'm having trouble getting more than 80 metres range. I suspect it is the high tension powerlines only 300m away.

2) The "4000m" units work better. But they draw 200mA at 9V and the capacitors needed to store a charge for a packet transmission are getting too big. I have moved over to 8 NiMH batteries in series giving 9-10V. I found some neat 4.5V 50mA solar modules on ebay out of Hong Kong - they cost about $4 each and 3 in series work well charging the batteries. 50mA charging a 600mAH NiMH comes in at trickle charge rates so no electrolyte is lost. Thus complex charging circuits are not needed. The volts are divided by 3 (200k/100k) and the picaxe monitors the volts. I also tapped off the midpoint of the battery and divided that by 3 as well so as to detect any reverse batteries and put the picaxe into low power mode until morning if the batteries are going flat.

3) Layout is absolutely critical. The Rx module is happiest if its antenna is surrounded by air and no wires of any kind. So the antenna sticks out the bottom of a 1.2m length of 90mm downpipe. The Tx unit seems to need to be at least 1m away from the rest of the circuit. I've used 1m of 3 wires - 9V, Gnd and data. I suspect the 1m means this is forming the earth of a dipole. But officially a dipole should be 1/4 a wavelength but that put the Tx too close to the picaxe causing it to reset. Next thing - the wire must be absolutely straight. I was stuffing it into a 50cm piece of downpipe and the coils in the wire made the Tx signal completely disappear. RF truly is a black art and a CRO has helped so much. Best $120 I have spent in a long while. So I've got about 1.2m of downpipe with everthing in it to keep it waterproof.

4) Some interesting experiments with DTMF, sine waves and pulse coding. First, sending a square wave into a Tx module gives a square wave out as expected. But sending a sine wave biased at 2.5V also gives a square wave out. The Tx units seem to have threshold detectors at about 1.5 and 3.5V. Second thing - putting 0V into a Tx gives 0V out. Putting 5V gives 5V out but it goes to 0V after 0.1 second. So the minimum frequency these units can transmit is about 10Hz. Third experiment - how do the signals fail. They seem to fail with pulse widths getting narrower and a square wave ends up being just narrow pulses. Hence sending data will get corrupted. So I tried sending a 1Khz tone and detecting it with an op amp 2nd order bandpass filter and an op amp rectifier. This doubled the range but the baud rate went down to about 8 baud or 1 bit per second. I tried two circuits - a 100Hz/2Khz dual tone system but the narrowing pulse width meant the 100Hz tone failed too early. Then I tried a simple 1Khz tone with simple on/off. A low pass filter with a cutoff of about 0.1hz detected the average value and with a series of 01010101s at the beginning to bias the low pass filter it was possible to extend the range to double. Detecting frequency at this range is not possible - there are too many spurious spikes so all DTMF/phase lock loop solutions no longer work. I am concluding that the very slow speed does not justify the doubling of range.

Have posted a new schematic at http://drvernacula.topcities.com/315_mhz_solar_powered_radio_rptr.htm at the bottom of the page. Design is now getting quite simple - 1 metre of 90mm downpipe with the Tx sitting in the top on 1m of wire. The batteries and board sit at the bottom with the antenna for the Rx sticking out the endcap through a small hole. Two wires going to the 3 4.5V 50mA solar panels coming out the bottom. All mounted on a 2m wood stake which is tied to either a fence post or a star dropper. This gets the Tx about 3m off the ground which makes a big difference to the range.
 
Last edited:
LOL that is a completely picaxeable project easily buildable by many of the forum readers. However, to build a rocket...
 
Hmm ... Tomy Baby Monitor and a Wi-Fi Router for comms ... That'll work well with the lid on then I guess.

"The only thing you may have trouble laying your hands on is a large metal sphere with whiplash aerials poking out" ... have youngsters not heard of beach balls, papier mache, cooking foil and wire coathangers ?

No mention of strapping a pet into it I note :-)
 
Antenna Mismatch

Regarding your varying results with differing cable lengths to your antenna. It could be down to an impedance mismatch. Not sure if the devices you use have any automatic matching circuits but it may be worth checking the SWR. I'm sure a local HAM would be happy to oblige if you don't have the necessary eqpt. At the low o/p levels, any %age loss of tx power due to reflection would be fairly critical and could also explain the effect on the pic. Just a thought.
 
I ordered an SWR meter off ebay and it arrived yesterday. There are some more experiments to do. More selective shielding. Different frequencies (1Mhz, 6Mhz, and 2.4Ghz). Anything over 200m is a real challenge at low quiescent current consumption. Commercial equipment (eg a CB radio) assumes there is lots of power even when the unit is just listening.
 
You must log in or register to reply here.
Back
Top