Si4330 and Hope RFM31 RF Receiver.

[srnet]

Does anyone have any experience using the Hope RFM31 RF data reciever, I believe this module uses the Silicon Labs Si4330.

I just cant get a data link to work at all, my transmitter (the matching RFM42 RF transmitter) is sending out something, I can hear it on an audio receiver, but the RFM31\Si4330 stays mute. I know how to drive the RFM42 transmitter in direct modulation mode via one of the GPIO pins, that works.

These are not simple devices, lots of options for transmitting preambles, headers and packets.

There is some sample code in C, which I have (I think) extracted the register setup sequence to be used in Basic, but no joy.

Just wondering if someone else has tried to get these devices working ?

 

[eclectic]

Perhaps Stan can help, when he comes back online

http://www.picaxe.orconhosting.net.nz/txrxconf.jpg
http://www.siliconchip.com.au/cms/search/index.html?scope=&keywords=HOPERF

e

 

[srnet]

To put this into context, I have the RFM42 working already as a Morse tramsmiiter.

I have also built a prototype portable receiver for that which decodes the morse from the received audio and prints the text onto a modified (code only) AX133Y OLED. This works. The prototype receiver also has a socket for a RFM31 so I was trying to get the data telemetry working that the RFM42\RFM31 pair should be capable of. The objective here to transmit using direct data for GPS Lat & Long location that my lost model locator (based on the RFM42) is capable of and display it on the OLED.

 

[manuka]

Sorry - my Hope experiences from some years back were mainly based around their HM-TR transceivers. I've not used either the RFM42 or RFM31. This looks an interesting application! What sort of ranges are you expecting ? Stan.

 

[srnet]

The original intention was to implement the lost model locator with morse, and that works well enough. With the lost model locator at ground level (my garden) in the middle of a city (Cardiff) I could easily read the morse transmission of location from a nearby hill about 5kM away, so from that aspect job done.

But as the RFM (Silicon Labs) RF devices are also capable of telemetry, that was an obvious thing to investigate. Not expecting any specific range really, but I doubt it would be as good as the morse.

The RFM data telemetry capability, whilst apparently very flexible, also seems very complex to setup. The manuals are not easy reading at all.

The RFM31 receiver is working, I do get a valid RSSI reading from it, so the device would be capable of being used as a programmable wide band (240-960Mhz) scanning signal strength device.

 

[srnet]

Well, I have to report some success.

Despite the data sheets for these devices having a pile of errors and some critical registers not even mentioned, I have it working !

My 20X2 & RFM42 Lost Model locator fills up the RFM42 FIFO with a message, 'The quick brown fox' etc and waits for a button press, transmits the packet and LO ! its picked up error free by a 28X2 & RFM31 receiver a few feet away and printed on a serial terminal.

Just using OOK at the moment at 10Kbps, 1mW.

Once the FIFO and packet handler are setup correctly, the RFM42 transmiter and RFM31 reciever deal with all the generation and detection of preamble, sync lock, headers, packet length and CRC automatically. From a PICAXE software point of view its a compact transmit and receive routine.

To transmit a packet you fill the FIFO, in this case with a message stored in EEPROM like so;

'the FIFO is filled with a burst mode write to register $7F
low NSEL 'enable RFM42
HSPIOUT($7F)
'now we will send the message stored in EEPROM from location $00, its 44 characters
for loopv1 = 0 to 43
read loopv1,var1
HSPIOUT(var1)
next loopv1
high NSEL 'disable RFM42, this stops the burst write also

And turn on the transmitter with a single register write, packet gone.

Receiving is simply a matter of polling the RFM31s Interrupt output, no need for interrupts as such, the NIRQ output stays low till you clear the status registers. Then a similar burst read of the FIFO and you have your packet.

Now to try it on narrow band FM, to see what sort of useful range I get.

 

[srnet]

Range:

434Mhz, 25khz deviation, 10kbps, 1mW TX output, just loose bits of wire for antennas, gives a range of 60M LOS.

Dunno if that good or bad, not tried any tweaking at all, TX is capable of 100mW ..............

 

[manuka]

Is that 60 metres LOS ( line of sight)? As each 6dB (thus x4) power boost doubles range I'd say several hundred metres may be possible with legal limits (usually 10 or 25 mW). The limiting factor however is usually the companion receiver- many are pretty deaf. For enhanced 433 MHz performance go for GFSK (Gaussian Freq. Shift Keying). Stan

 

[srnet]

Yes, Line of sight.

I have the PCB for the receiver and am building it now. The plan is to put it in a box, fit an SMA socket and then mount a geniune 430Mhz type antenna, got to be an improvement over a random bit of wire (which is however the correct length). And then do tha same with the LML transmitter.

Ought to be easy then to take the lot over the local park and do some more sophisticated range resting.

I think the RFM31 receiver has reasonable ears, it claims sensitivity of -118dB, how does that compare with other units ?

 

[manuka]

it claims sensitivity of -118dB

That figure is about as good as it gets. However such RX performance is usually at slow data rates (~1200bps etc) rather than your 10kbps.

My simple Morse approach, based around a Jaycar Keymark/Spiriton tx ,was good for a few 100m if tracked with a decent receiver, but it's TX power of just a few mW was pretty puny. Perhaps it's time for an update using your approach? Keep us informed!

Stan.

 

[srnet]

As you can adjust the power of the RFM modules down to 1mW, all you need is a large enough open space to be able to carry out reasonably accurate comparisons of the various modes, assuming you keep the transmitter fixed in position. Easy with PICAXE to have several different programs ready set up with various modes or data rates.

You can convert the distances reception fails into dB differances with a few sums.

My bet it still with Morse by a fair margin.

One other factor to consider is that transmission by Morse of Lat & Long with morse takes a long time, sent as a 10Kbps packet it takes a fraction of a second. That is hundreads of times more efficient in terms of battery power.

 

[manuka]

Agreed-that's however if it's actually received at the faster data rate! With a lost model (in a static location) data speed may not be as crucial as signal detection,and battery life could be extended with solar PVs anyway. Slower transmissions generally have superior performance of course, & they may even be detectable when buried in noise. Such QRSS techniques can make for astounding ranges, with sub mW lower freq. transmissions spanning the Atlantic.

 

[srnet]

Its a difficult balancing act. I looked at the QRSS stuff, and whilst it is very attractive from the potential range it offers, its not readily a field portable solution.

The main compromise is available battery power, the backup battery has to be small, I use a 420mahr Lipo, and a QRSS approach even at the lowest RF power probably uses more battery power than a much more powerful but a lot shorter transmission. I will definetly give it a go though.

With the telemetry working and a portable field receiver the LML will be able to constanly transmit distance, direction and hieght (from the GPS) and have it displayed on the OLED. Its always handy to know how far away and how high the model you are flying is.

I realise that GPS altitude is not that accurate, but when a plane is several hundread metres up in the air, GPS reception tends to be rather good.

 

[srnet]

Did not take long to get the QRSS working, the LML direct mod to the RFM42 trasmitter was already being fed from 20X2 pin b.4 (by design!!) so I just chnaged the sound command to hpwm commands.

With a morse dit period in the 8 second region, Spectran shows up the morse clearly when its only just audible in the audio, so weak you would defiently not be able to detect standard speed morse. But it takes so long to transmit that it uses a lot of battery power.

Solar cells have been considered, but most planes get lost because they crash, and who knows if a solar cell would end up pointing the right way to collect useful energy.

 

[manuka]

But it takes so long to transmit that it uses a lot of battery power.

OK on the tradeoffs,but maybe consider the DFCW (Dual Freq. CW) dual tone morse variation, or just go QRSS "on the hour" & stay higher speed normally?

Add extra PVs at diverse orientations too-perhaps the paper thin flexible & light weight types from Pager Motors and Flexsolarcells. These amorphous types are only ~5% efficient BUT are great performers in subdued light. A sample one I've here (their SP3-25) gives 2-3 V output at up to ~22mA, yet weighs just ~¾ gram! (Compare that with a sheet of A4 photocopy paper which weights ~5g). Suitably matching such a PV ( perhaps the 4.2V @25mA SP4.2-37) to your 420mAh Lipo would keep it charged nicely I'd say.

Yet another technique may be to use a LEGAL lower freq beacon (~27MHz?) which'd have both more punch thru' obstructions and also ionospheric reflection prospects.
Stan. (ZL2APS)

 

[srnet]

As for the choice of frequency, that was pretty much self selecting.

At 434Mhz UHF trancievers are cheap and easy to find, I already had a Yaesu VR120 scanner and a Hora C408 tranciever but I have also bought an el-cheapo UHF tranciever off eBay, its small and light and a good deal more reception sensitive than the scanner. I have no intention of using it as a transmitter.

High gain directive antennas are readily available too, I have a 5 element ZL special for long distance searching. Small enough for wandering about searching.

The Hope modules for thier many sins are rather small and feature rich, the completed 18F26k22 beacon (shown below) is rather compact, an essential feature for RC model use.

My scanner covers 27Mhz, AM\FM\WBFM so for a comparison is there any 27Mhz modules you could recomend I try ?

 

[manuka]

I'm unsure about 27MHz legalities in your country,but the old CB regs & (now) quiet 11m band may be tolerant enough for lost model beacons. It's possible the TX & crystal from dirt cheap 4 transistor kids "spy ring" walkie talkies could form the engine of such a transmitter. 6 channels are usually internationally used => 26.995, 27.045, 27.095, 27.145, 27.195 & 27.255 MHz. Forget their receiver however, as they're quite hopeless!

Such units abound in toy shops at ~US$10-$20 a pair, but they chew thru' 9V batteries & are very uneconomic to run. Simple remote control cars usually use 27 MHz as well. Talking Electronics has lots of circuit ideas

Of course 10m 28MHz ham band beacons are extremely well established, but these need a ham call. Stan (ZL2APS)

 

[srnet]

Of course 10m 28MHz ham band beacons are extremely well established, but these need a ham call. Stan (ZL2APS)

Indeed, but I was rather keen to avoid a TX that required a Ham license.

 

[srnet]

I'm unsure about 27MHz legalities in your country,

27Mhz RC control is still used for el cheapo toys, and some land based stuff. Most of the air based RC is now on 35Mhz FM or 2.4Ghz.

I have some 27Mhz RC transmitters, and a range of suitable crytals so knocking up a single transistor transmitter would not be hard.

 

[manuka]

Go for it-there's no shortage of circuits around,although most pre-date PICAXE modulation! Check the likes of www.circuitdiagram.org/27_mhz_transmitter.html ,which is so crusty that it even mentions a carbon mike!

Note- a significant 27 MHz issue may relate to the need for a suitably sensitive receiver. This part of the RF spectrum is now thankfully usually very quiet, BUT scanners (which are essentially VHF-UHF devices) can often be pretty deaf at 27 MHz. Far superior comms receivers are both costly & pretty bulky of course! It may be hence worth designing a matching single channel tracking receiver as well.

All manner of factors & tradeoffs often need pondering for tracking setups,and the brilliant offerings now available at 433 MHz may justify sticking with this UHF band after all. Although transmitters are legally only allowed low power (~25mW),receivers can be extremely sensitive.Some of the best offer -120 dBm receiver performance, which often ~20dB better than cheapies. As every 3dB gain is equiv. to doubling power this is equiv. to a 2x2x2x2 or 16 times power boost at the transmitter. 16 x 25mW = 400 mW equivalent but with no battery drain concerns!

Naturally a significant 433 MHz benefit may also relate to UHF antenna dimensions being much more compact (and with good gain & direction) than bulky ~27 MHz types. In tough terrain this alone may be a real clincher. Stan.