EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Benta on August 26, 2024, 09:28:43 pm
-
Probably a silly question, nut...
I'm working on a system with 868 MHz sensors that are deeply embedded (read: I can't get into the hardware). The system is purely unidirectional and outputs an ASK signal (format/protocol unknown/proprietary) at 25 mW, LOS range is around 100 m.
A need has arisen for longer distances, and the supplier range extenders are f...ing expensive. They use demodulation and remodulation etc. to do the job.
So the question is:
Is it possible to make a simple Antenna -> LNA -> PA -> Antenna repeater?
I can imagine that the configuration might make either a wonderful oscillator or a wonderful attenuator, depending on phase relationship between input and output.
Using directional antennas for RX and TX pointing away from each other (to minimize interference) and controlling carrier phase between RX and TX (I imagine 90 degrees would be right), do you think it's possible to double or even quadruple the TX range? The repeater will be placed somewhere between sensor and concentrator (receiver).
Thanks.
-
Just a suggestion, place the repeater near to the original Tx to minimise the amount of gain the Rx needs, and therefore its susceptibility to picking up the new Tx signal. Place a shield between the Rx and Tx antennas and separate the electronics into two as well, separated by the same shield. Can you place a small antenna coupled to some coax right next to the original Tx and run that signal to your PA repeater?
-
Just a suggestion, place the repeater near to the original Tx to minimise the amount of gain the Rx needs, and therefore its susceptibility to picking up the new Tx signal. Place a shield between the Rx and Tx antennas and separate the electronics into two as well, separated by the same shield. Can you place a small antenna coupled to some coax right next to the original Tx and run that signal to your PA repeater?
Good inputs, Thanks..
Yes, placing the repeater RX close to the sensor is certainly an option, paired with reducing RX sensitivity to a minimum.
The RX/TX shield is alrady there, forgot to mention that, sorry.
A coax connection is not possible, the sensor is a molded unit and often placed awkwardly, but I can get as close as ~1 m.
But you think it's possible?
Cheers.
-
Just a suggestion, place the repeater near to the original Tx to minimise the amount of gain the Rx needs, and therefore its susceptibility to picking up the new Tx signal. Place a shield between the Rx and Tx antennas and separate the electronics into two as well, separated by the same shield. Can you place a small antenna coupled to some coax right next to the original Tx and run that signal to your PA repeater?
Good inputs, Thanks..
Yes, placing the repeater RX close to the sensor is certainly an option, paired with reducing RX sensitivity to a minimum.
The RX/TX shield is alrady there, forgot to mention that, sorry.
A coax connection is not possible, the sensor is a molded unit and often placed awkwardly, but I can get as close as ~1 m.
But you think it's possible?
Cheers.
Yes, I think it is possible with careful design, as you appreciate, you want to minimise coupling between the repeater Tx and Rx, to avoid a feedback situation as in a live audio situation. The coax could be used to provide separation between the Rx and Tx, distance is your friend in such a circumstance. Just something to keep in mind.
-
But you think it's possible?
Nope. I can't see it being possible to isolate the input from the output enough. I suspect that's why the manufacturer does it the way they do. Probably uses a store and forward technique.
EDIT: What about a link repeater which has different Rx and Tx frequencies at each end? Signal goes from (Device) --> (868 MHz-->xxx Mhz) --> ( xxx Mhz --> 868 Mhz) --> (receiver)
Note: Each device in ()
-
Have you exhausted passive options? High gain antennas can do some wonders even if coupled back to another antenna.
-
But you think it's possible?
Nope. I can't see it being possible to isolate the input from the output enough. I suspect that's why the manufacturer does it the way they do. Probably uses a store and forward technique.
EDIT: What about a link repeater which has different Rx and Tx frequencies at each end? Signal goes from (Device) --> (868 MHz-->xxx Mhz) --> ( xxx Mhz --> 868 Mhz) --> (receiver)
Note: Each device in ()
Yes, that is my fear, although we're not talking a lot of gain here - perhaps 5 dB. OTOH, the manuvacturer has to keep it universal. I don't.
The frequency remodulation is standard fare, but then I haven't really achieved anything.
Thanks.
-
Have you exhausted passive options? High gain antennas can do some wonders even if coupled back to another antenna.
Interesting idea. No, I haven't until now, but utilising directional gain might be a damn good option. It won't even be large at 868 MHz.
Thanks.
-
But you think it's possible?
Nope. I can't see it being possible to isolate the input from the output enough. I suspect that's why the manufacturer does it the way they do. Probably uses a store and forward technique.
EDIT: What about a link repeater which has different Rx and Tx frequencies at each end? Signal goes from (Device) --> (868 MHz-->xxx Mhz) --> ( xxx Mhz --> 868 Mhz) --> (receiver)
Note: Each device in ()
Back in the day,using Superheterodyne principles, devices like this had a standard IF frequency, so that all that was needed was to down-convert the incoming signal to the IF, then using a second LO,up-convert to the new RF frequency.
This is much simpler in concept than demodulating to baseband then remodulating onto the final frequency.
Of course, this was done by organisations with a reasonable budget, so could do it the "hard way".
Another thing that was done was to use linearly polarised gain antennas & have horizontal polarisation of the booster receive antenna, & vertical polarisation of the converted signal.
This gives you 20dB of rejection of the unwanted signal.
At high UHF frequencies, vertical separation of the receive & transmit antennas also adds to this rejection figure.
-
I'm working on a system with 868 MHz sensors that are deeply embedded (read: I can't get into the hardware). The system is purely unidirectional and outputs an ASK signal (format/protocol unknown/proprietary) at 25 mW, LOS range is around 100 m
roughly how many sensors are involved, and are you able to get access to the internals of the receiver?
i'm not an RF expert, but it seems that for every sensor you could have an associated up-converter, similar to what we see used as a 'band expander' on car radios. ie, pick up the (868MHz) signal, mix it with, say, a 12MHz signal, and feed the result through an 880MHz bandpass filter. then amplify and transmit. you want each up-converter to be relatively deaf, so it can only hear the transmission from a single sensor.
then you just need to re-tune your (single?) receiver to 880MHz.
cheers,
rob :-)
-
Yagi style antennas have a lot of gain available in a reasonable physical size at those frequencies. Coupling a pair each having 10 to 12 DBI gain may work although there will be some optimum physical distance apart. They could and maybe should be on an equal horizontal plane. Yagi's always have a back lobe which could aid in the mutual coupling. The physical spacing horizontally between the two is probably extremely important if used in a passive mode as is the length of the cable connecting them. There will be some spacing where the antennas and back lobes will all work in phase to provide some apparent path gain, and at the frequency in question changing that horizontal spacing by a mere 1/4" could cause the setup to worsen the apparent path loss. I have seen some fantastic results when 'sympathetic radiators' were used to fill in 'dark spots' directly beyond a mountain ridge. A tuned sympathetic radiator on the mountaintop provided a line of sight signal to those in the shadow of the original broadcast wave. Case in point, a 247 foot tall Rohn T.V. tower with nothing on it sitting on the mountain top!! Concrete anchor, ground radials and guy wires!! Perhaps a home built yagi looking antenna where it has maybe twenty elements all exactly the same length and resonant at your frequency of interest!! It would be a simple bi-directional sympathetic radiator!!!Wouldn't be any practical reason you couldn't go crazy with even 40 elements.
-
Thanks again to All.
I'm falling more and more in love with the passive solution. Back-to-back antennas with directional gain is a nice idea, but I'm veering towards a tuned Yagi just at the concentrator (receiver) end. As this is LOS, it should do the same job.
It would solve all problems of power supply, regulations/conformance etc.
As the sensors are mostly placed in fixed positions, pointing is easy.
Which kind of antenna would you recommend here? If I can achive 10 dB gain, it's wonderful, 6 dB could also do it.
Thanks
-
I believe the antenna would have to be hand built. Most yagi's are set up with one or two 'reflectors' a driven element with a gamma match for 50 ohm feeds and the several 'directors'. They also take on a sort of log periodic look. I suppose for one fixed frequency all of the directors could be the same length which would provide a bit more gain by sacrificing wide bandwidth. Perhaps you could use a piece of 1/2" or 3/4" E.M.T. and maybe using steel 'brake line' tubing passing through the E.M.T. as the elements. You could set up the 'reflector' elements as a 'corner reflector' like used on U.H.F. television yagi's. Yagi construction, formulas and practical examples for the ham bands are in the ARRL Antenna Handbook and tons of construction ideas and formulas lurking on the internet. Back when hams still built things it was common to see home built 'Moon Bounce' or E.M.E. (Earth, Moon, Earth) antennas. Quads of monster yagi's on a boom that looked like a giant 'H'.
-
Nah there are tons of COTS antennas for 868MHz, especially now LoRA is taking off.
Here's a couple of examples
Yagi
https://www.rfsolutions.co.uk/antennas-c8/23db-gsm-yagi-antenna-p162 (https://www.rfsolutions.co.uk/antennas-c8/23db-gsm-yagi-antenna-p162) OK a bit extreme this one
Sector panel antenna, if you have a bunch of sensors to cover. Lower gain, wider aperture
https://www.quwireless.com/product/qupanel-xr-lora-868 (https://www.quwireless.com/product/qupanel-xr-lora-868)