433MHz RF Range

Hi!

I'm having problems with the range of my RF transmitter/reciever.

I get about 30 meters range with this unit which has a transmit power of 10mW.

I have lowered the transmit speed to 32 bps, since I'm only sending a float value (4 byte = 32 bits).
I am sending a packet of 4 byte every 2 seconds.

I have tried with a basic radio antenna of about 30cm from the transmitter which increased range by about 10 meters.

I have read that the optimal antenna length is 17cm for this wavelength, but some recommend getting a WiFi antenna (what length?!).
Given that I solder an antenna wire of 17cm to the board, how would I solder it? There is no given spot marked "Ant" or "antenna". There is an onboard antenna, should I solder it to that one?

Lastly, I can increase from 5V to 12V on the transmitter, but that will be my last boost since I believe the antenna problem is the biggest one.

Also, how does the packet size / interval of messages affect the range? If I send packets of say 80 bytes instead of 4 increase probability of a packet being recieved? Does sending a packet again after 2 seconds at so low speeds interrupt the last package?

Thanks in advance, I'm really in over my head here. But I always have fun learning more.

Yes there is a spot for soldering antenna. It's the one spot right next to those copper coil antennas on your boards.

You can increase voltage and get better transmitting range. I use 9V and I got better range than 5V.

One big problem that affects this unit is interference. AC wiring nearby will give massive interference and reduce the range.

How much range do you actually want?? Do you have walls, or metal grills or is it an open area?

Thanks for answer.

Okay, I'll try an solder a copper wire of about 17cm to that spot, do you mean these?

I'll definetly try connecting the VCC to a 12v instead, and keep it out of interference.

It's going to be outside, 65 meters straight. But down a slight hill with some trees but not many. They do not have line of sight due to the curvature of the small hill. Should not be too much of radio shadow.

If you want long range look at LoRa devices.

Allan

Those modules are typically good for over 300 meters line of sight, if you use a balanced dipole as shown in the photo below (on both transmitter and receiver). Each wire is 17 cm long, one connected to ANT, the other the GND.

linusneuman:
Okay, I'll try an solder a copper wire of about 17cm to that spot, do you mean these?

http://i.imgur.com/QxSyuvS.jpg

Yeah, it's correct for the Receiver (long module) but for the Transmitter it's the one near the corner edge of the pcb and not the one in the middle written ANT.

Here's a picture for easy reference.

For me I get around 18 meters through a concrete reinforced wall with an simple 15cm long copper telephone wire a antenna. I didn't test for more range because my project requirement needed only that much distance.

You can get the distance you wanted if you are patient enough to tweak your antenna. Also try and identify any interference if they're present.

You have bought the worst possible variant of those cheap 433 Mhz receivers, which is the super regen type.
If you have to use these receivers, then at least get the superhet type , which can easily be identified as they have a crystal or saw resonator on the board.

mauried:
You have bought the worst possible variant of those cheap 433 Mhz receivers, which is the super regen type.
If you have to use these receivers, then at least get the superhet type , which can easily be identified as they have a crystal or saw resonator on the board.
https://sc01.alicdn.com/kf/HTB1DZ8xLpXXXXaEXXXXq6xXFXXXG/4ch-Superheterodyne-Receiver-board-433-92MHz-wireless.jpg_350x350.jpg

Acknowledging that, but a really sensitive receiver is futile if there is a lot of RF noise from the processor and peripheral devices attached to it (especially if it's with loopy jumper cables) nearby.

linusneuman:
I have read that the optimal antenna length is 17cm for this wavelength

You will hear that often, and whilst 17cm is the length of a 1/4 wave antenna at 433Mhz that does not mean that length of antenna will always be the optimum length for a particular module.

You will almost always find you can get more range by trimming the antennas, shorter or longer, for maximum signal.

However the practicality is that tuning the antennas for optimum length is not easy, so very few people bother.

With UHF comms as soon as you loose line of sight the range drops dramatically, so 65M through some trees and over a hill might be a very big problem, for simple modules.

allanhurst:
If you want long range look at LoRa devices.

Allan

Long range for me is around 100 meters.
Do you have an example of LoRa device? Thank you.

jremington:
Those modules are typically good for over 300 meters line of sight, if you use a balanced dipole as shown in the photo below (on both transmitter and receiver). Each wire is 17 cm long, one connected to ANT, the other the GND.

Thanks, I'll try it. You just soldered the other antenna wire straight to the GND?

Noobian:
Yeah, it's correct for the Receiver (long module) but for the Transmitter it's the one near the corner edge of the pcb and not the one in the middle written ANT.

Here's a picture for easy reference.

For me I get around 18 meters through a concrete reinforced wall with an simple 15cm long copper telephone wire a antenna. I didn't test for more range because my project requirement needed only that much distance.

You can get the distance you wanted if you are patient enough to tweak your antenna. Also try and identify any interference if they're present.

Thanks, I'll try that and see how much difference it makes.

mauried:
You have bought the worst possible variant of those cheap 433 Mhz receivers, which is the super regen type.
If you have to use these receivers, then at least get the superhet type , which can easily be identified as they have a crystal or saw resonator on the board.
https://sc01.alicdn.com/kf/HTB1DZ8xLpXXXXaEXXXXq6xXFXXXG/4ch-Superheterodyne-Receiver-board-433-92MHz-wireless.jpg_350x350.jpg

There's not much info on this so I bought one I found. But that's not true either, there are boards with a strength of just 1mW for instance.
When you say "if you have to use these", what do you really want to say? Is there another type that is better?

srnet:
You will hear that often, and whilst 17cm is the length of a 1/4 wave antenna at 433Mhz that does not mean that length of antenna will always be the optimum length for a particular module.

You will almost always find you can get more range by trimming the antennas, shorter or longer, for maximum signal.

However the practicality is that tuning the antennas for optimum length is not easy, so very few people bother.

With UHF comms as soon as you loose line of sight the range drops dramatically, so 65M through some trees and over a hill might be a very big problem, for simple modules.

Alright, thanks. I'll try and adjust the antenna length. Pointing the antenna towards the receiver and vice versa should be the best, right?

linusneuman:
Alright, thanks. I'll try and adjust the antenna length. Pointing the antenna towards the receiver and vice versa should be the best, right?

The RX and TX antennas should have the same orientation, normally vertical.

Alright, thanks. I'll try and adjust the antenna length. Pointing the antenna towards the receiver and vice versa should be the best, right?

No, that's actually the worst possible orientation. e-m waves are transverse, so the dipole needs to be at right angles
to the line of site. The receiver and transmitter dipoles/antennas should be either both vertical, or both horizontal,
so that they have the same polarization.

srnet:
The RX and TX antennas should have the same orientation, normally vertical.

MarkT:
No, that's actually the worst possible orientation. e-m waves are transverse, so the dipole needs to be at right angles
to the line of site. The receiver and transmitter dipoles/antennas should be either both vertical, or both horizontal,
so that they have the same polarization.

That's interesting, good to know. Thanks.

When you say "if you have to use these", what do you really want to say? Is there another type that is better?

Yes, a +transceiver+ with programmable modulation and packetization support - generally these have a
full superhet receiver and are designed to get rather closer to the noise floor (which can be thermal for
UHF - unlike the HF bands). The main advantages are having stable channel frequencies (quartz crystal
reference) and being able to select TX and RX bandwidths/modulation types. Sensitivity + selectivity.

And of course with a transceiver you can talk both ways (to acknowledge reception for instance)

For instance the RF12's and RF70's are tiny cheap modules with all this in.

You just soldered the other antenna wire straight to the GND?

Yes.

Here is an example of a Arduino 328P sensor board I have been working on, it has two Mikrobus sockets, shown here with a LoRa module and BME280.

Sleep current is around 12uA and with the 3 x AA battery pack shown you should be able to send a sensor reading once a minute for 5 years or more.

LoRa is very long range and highly configurable. You can opt for a very slow data rate of 100bps or less and get line of sight ranges into the many hundreds of kilometres. Or you can go to a fast data rate up to 38kbps and still get kilometres of range but with the benefit of transmissions only taking 10mS or so and thus highly power efficient.

MarkT:
Yes, a +transceiver+ with programmable modulation and packetization support - generally these have a
full superhet receiver and are designed to get rather closer to the noise floor (which can be thermal for
UHF - unlike the HF bands). The main advantages are having stable channel frequencies (quartz crystal
reference) and being able to select TX and RX bandwidths/modulation types. Sensitivity + selectivity.

And of course with a transceiver you can talk both ways (to acknowledge reception for instance)

For instance the RF12's and RF70's are tiny cheap modules with all this in.

jremington:
Yes.

Great, I'll check some of those out if I can't get what I need from what I have right now.

What do you guys think about the packet size? Is there a greater probability of a packet being caught if its larger than 4 bytes?

Longer packets have more risk of failure. Its the packet pre-amble that affects packet detection
reliability

linusneuman:
What do you guys think about the packet size? Is there a greater probability of a packet being caught if its larger than 4 bytes?

Depends what you mean by a ‘packet’.

In my World a ‘packet’ has a preamble\sync, header, data and CRC.

Now if you had 16 bytes to send, and errors occur at random, it makes sense to send one 16byte packet rather than 16 one byte packets, less chance of errors.

But why dont you try it ?

A bit of RF fundamentals.

You’ve got 10mW ie +10dBm.

Path loss for far field in free space is about 43+ 20log( number of wavelengths) between antennas .

Antenna gain for a dipole is about 3dB if correctly oriented. So tx and rx amount to 6dB.

The required signal/noise ratio for simple AM for a decent bit error rate would be about 20dB

The universe gives you about -174dBm/Hz at room temperature.

The bandwidth depends on the data rate - lets say 3kHz

A typical super-regen receiver ( ghastly things) has a sensitivity of around > -100dBm with a 3kHz bandwidth. A good superhet ought to have a noise factor of < 5dB which implies a 20dB s/n sensitivity of about -174 +5 + 20 + 10log(3000) ie about -114dBm

OK … tot all this up for a range of 300M.

Path loss : including antenna gain : ~90 dB. So with a +10 tx power the receive power is -80dBm. That gives you a gain margin of 20dB with a super-regen, or 34dB with a superhet.

But you’re not in free space!

the environment you describe could well be more than another 20-30dB loss.

You might just get away with a good superhet receiver…

Allan

PS you can make a simple balun with 2-3 turns each side of thin wire through a ferrite bead as a transformer- better than merely earthing one half of your dipole.