I am looking in using some of these RFC-33a 2 watt 433 Mhz radio-modules (see URL below); but for licensed ham-radio use (as the 433 Mhz ISM-band overlaps with the 70 cm ham-band) and experiment with mesh-radio.
I am a bit worried about spurious emissions from these cheap Chinese modules so I plan to do a lot of testing. But in a comment I read somewhere (link lost, sorry) somebody mentioned adding a 433 Mhz SAW-filter at the end of the circuit.
I have never worked with SAW-filters before; so ... any advice would be nice.
Are these just "solder-in and use" components or are there other things I need look at. Frequency-responds? Maximum power?
I never heard of power SAW filters. I have only sen the ones used in receivers. In those, generally a synchronous detector is used. The phase shift they introduce thus produces the selectivity.
If you are talking using one to clean up the output of a transmitter, that is a whole different animal. What's wrong with regular tuned circuits ? How much damn bandwidth do you need ?
Well, that's also the conclussion I had when searching for more information on SAW filters.
Looking at the datasheets, most of them did 10 mW or so maximum power.
Nothing, except that I came across this post mentioning SAW filters so I said "hmm.. never used these things, let's see what we can do with them" :-) (always interested in learning about new stuff)
Look for helical filters, they will have a somewhat higher power handling capacity.
If you intend to use SAW filters, a good place would be between the low power transmitter stages and the final power output stage. The power levels would be manageable.
The problem with that approach is that the most efficient RF output is clas s C. If you want anything near linearity in an RF final output your efficie ncy goes in the toilet.
Best plan is to get off your dupa and design the output stage right. Unders tand how to figure out the Q of the tuned circuit at the output, and maybe even multiple tuned circuits, like for example the vestigal sideband filter s that used to be use in television transmission. A SAW filter doesn't work when you put two megawatts though it, and if your finals aren't class A or AB it does not good at the input. It has to go at the output.
No matter if the power isn't that high, You want caps ad coils when it come s to transmitting.
Get a reactance/resonance table, or (and this is from here) :
formatting link
I forgot who did that but whoever it was, thanks again.
That is a nice module. I might consider getting one for myself. However, the specs are a bit contradictory. 33dBm is about 2W, as you know, but that power should need about about 1A of current from 3.3V. They claim
300mA is needed at 3.3V for 33dBm output. It is not possible to even get
1W at that current*voltage.
Well, I forgot to include 50% duty of the Manchester encoding. Still, it would be more than 300mA average if less than 100% efficient while transmitting. But then I suppose the transmitting/receiving duty could be 50% as well.
The class C output stage is not much of an issue, since it mainly produces (odd)harmonics of the carrier, which can readily be filtered out by low pass or a band pass filter.
I think that the OP is interested in other spurs possibly generated by the device, e.g. during frequency change or heterodyne spurs. Also bad VCOs in PLLs can have bad phase noise performance. A low level SAW or helical filter will take care of much of these spurs, before entering the high power class-C final stage.
Putting the filter is useful also in repeaters, since in practice, you have to put a notch filter into the _transmitting_ chain to filter out transmitter phase noise on the _receiving_ frequency. This filter in the transmitting chain will also have some insertion loss (0.5-3 dB) on the _transmitting_ frequency.
Putting the filter after the high power (and hence expensive) amplifier, you may loose up to half of the transmitter power generated with high cost into the filter. Putting this notch filter in front of the power amplifier, the power handling doesn't have to be high. The losses in the filter can be compensated with some extra gain in the high power stage(s).
A typical SAW filter will not be able to handle the 2W (33dBm) output of such a module. However, it is easily handled by a small helical filter.
You can buy them ready-made from companies like MCV but of course as a radio amateur you can construct it yourself in an evening.
Make a small box out of PCB or tinplate and put two airspaced coils in it and two screws to tune them. Examples available when you google for 434 MHz helical filter.
A filter in this case does not only filter spurious emissions from the transmitter, but it also provides some guard against strong out-of-band signals for the receiver which is very easily overloaded by them.
(this is the same technology as used in the wellknown Baofeng UV-5R portable transceiver, wellknown for its bad overloading characteristics)
I do not know who owns which company today, but googling for "Toko" helical filters should give a lot of usage examples.
IMHO constructing a 433 MHz helical filter requires quite a lot of mechanical precision for amateur building, but a 1/4 or 3/4 wavelength stripline or microstrip filter should be easy to construct.
Ah I thought Toko had left the L/C filter business, but probably only for IF filters. These helicals appear to be still available from them.
The microstrip may be a bit large (requires a 10cm length PCB I think) but a helical that is not as tiny as those Toko/MCV filters is quite easy to build from some tinplate or PCB, a piece of copper wire, and two m3x20 bolts plus 4 nuts. Cut the sheet material in small pieces and solder a casing, solder two of the nuts on the casing and use the bolts as tuning with an extra nut to lock them. Make coils that are soldered to the casing at one side, e.g. by winding them on a 6mm drill.
I will do some additional research and discuss the issue with some fellow hams in our radio-club. (ok, once they get back from Friedrichshafen :-) )
I didn't know about the Helical filters, so that's also very interesting stuff.
Step one will be to really test the device and measure them. Once we have that, we can descide on the next steps).
Also, in answer to jurb. Some first initial measurements seams to indicate the bandwidh to be around 250 to 300 Khz
On 20-06-16 12:32, John S wrote: >>> I am looking in using some of these RFC-33a 2 watt 433 Mhz >>> radio-modules (see URL below); but for licensed ham-radio use (as >>> 433 Mhz ISM-band overlaps with the 70 cm ham-band) and experiment >>> with mesh-radio.
Well, first, I learned to read the specs of a lot of these Chinese goods with more then one grain of salt :-)
I guess it all depends on the dutycycle.
Concerning the module itself, I have been playing around with three different modules: Northern Semi nrf905, Silabs si4332 and TI cc1101).
If you look at features, the si4332 (and si4663 which also does VHF) is the most interesting as it provides the bigest range of supported modulation-scemes and bitrates. (e.g. I once wrote a 512 bps POCSAG pager with one of them with just a couple of lines of code).
The big disadvantage of the nRF905 is that it only supports one mode: 50 Kbps GMSK; ... but there are 1 Watt and 2 Watt modules for it available. For the other modules, I did not find anything over 100 mW.
One nice feature of the the nRF905 is the "auto-repeat" mode, which means that the module continues repeats a packet as long as a certain pin is kept high. This generates a quasi-constant signal, which you can then use to create CW-like signals. (as ham-radio stations, you are required to identify your station, by preference in a way ecodable by everybody).
Another nice feature is the option to disable CRC-checks in the shockburst mode, so you can use FEC in the frames to try to decode the frames anyway.
Concerning power-usage. I did notice that -even if you power the module from a 2A power-supply, the voltage drops from 5 volt to 4.5 volt when transmitting.
They probably "specify" the peak output power and the average supply current. As it is half-duplex, it transmits only part of the time. During receive the current is less, and it has a standby mode to draw even less current. In actual use, it may be able to transmit some data at an average of 300mA and an output power during the transmission bursts of 2W.
Anyway, we would have killed for such a module 15 years ago! Doing that kind of datarate on 70cm would have required a very complex WA4DSY 56k modem plus a 10m-70cm transverter, hundreds of $ of equipment.
And, they are so cheap that they allow you to think "outside the box".
I once wrote a POCSAG (paging) application using one of these modules (si4332 based) and an arduino.
As that particular module only does 100 mW, its range was limited to a few hunderd meters, ... unless you combined it with intelligent radio-nodes and create a mesh technology.
A mesh peer-to-peer network is a network concisting of multiple nodes, radio-modules, all operating on the same frequency. They operate like an ARPS network: a message transmitted by one node is picked up by neighboring nodes in the mesh to retransmit it ((depending on its destination or if the message is"to-all")
In short, it allows you to "inject" a message into the mesh in one place and have the intelligent nodes take care of "carrying" that message to the other side of the mesh..
Or, .... if you inject a message in the network containing a POCSAG paging frame, that pocsag message is actually distributed to every corner of the coverage area; .. and and pocsag pager inside the coverage area will -at some point- receive that paging-message. (actually probably multiple copies of it).
As you can build such a node for some 15 to 20 euro a piece (radio-module, stm32f1, power-supply, plastic bucket + some rope to mount the device in the top an tree), you can easyly build a network to coverage quite some area with for really little money.
:-)
We really should learn to think outside the box and start to use and build intelligent networks instead of just "one transmitter and one receiver".
I know about that technology but I would consider that a problem as you don't want your pager to get multiple copies of the same message. I would separate the data transfer network from the paging network, although it can be done by the same radios. First distribute the data through the mesh to all nodes, then at a fixed moment transmit it on all the nodes in POCSAG.
I wrote a POCSAG decoder/encoder in 56k assembly for the Alef Null DSP card 1 long ago. It has been decoding POCSAG on one of the services for many years, and I have tested the encoder using an RF test generator and a pager. It worked, but I never put it into use as a transmitter.
Today it would be possible to integrate this into the well located co-channel diversity repeater network that we operate. I have been looking at data-under-voice such as used on some satellites to send low-bitrate data like paging messages, but of course existing tcvrs would not be able to do anything with it without some external decoder (even the ones which do have APRS support).
The pager we used for these tests neither had an issue with a message being received multiple times (within a certain timeframe) or it being inside a larger packet. So the idea was -if you use a radio and a mesh network to distribute your message to the nodes- why not use that to broadcast to the pagers too.
Nice.
Concerning co-channel diversity, I have been thinking about that. the timing element seams to be doable, but -at least using these radio-modules- getting them all to operate on exactly the same frequency is probably a much bigger challenge.
Maybe the pager firmware rejects direct duplicates? Not impossible as some networks also re-sent the same message at some interval to increase the chance of reception.
I understand that, but maybe I would have opted to have a separate (and higher-speed) protocol to do the mesh and not receive and relay the POCSAG messages themselves.
This even is a problem with our repeaters. It works quite well on
2m where we use Tait TB-8100 base stations, but on 70cm where we use Spectra Engineering MX800 base stations the results are not as good.
Both operate off an external 10 MHz reference provided by a GPSDO, but the MX800 has FM modulation inside the synthesizer PLL and it introduces artifacts like jitter on the modulation. The users report heavy audio distortion in overlap areas, which is confirmed by measurements of the different transmitter signals. When doing the same measurement on the TB-8100 network the results are much better and this is confirmed by the users.
We are now considering using SDR to generate the FM signal.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.