0
\$\begingroup\$

I ordered 5 nRF24L01+ chips for 2.4 GHz RF. I have not done tests with them, but doing tests with my current 433 MHz results in a low speed, few meters maximum range.

So I expect with my 2.4 GHz RF's I also will need something better.

What I would like is to built these into an enclosure, to communicate about max 10 meters, but with low latency/fairly fast speeds (like 250 kbps). These 10 meters will not have walls, but musical instruments, possibly many mobile phones from audience (also 2.4 GHz I believe), (music) power amplifiers and speakers, wireless microphones, mostly < 1 GHz etc.

My questions: - Below I put some examples, but there are so many types, which of the types below would be sufficient for about 10 meters (noisy environment) range? - Did I miss other types which might be more suitable?

  1. 2.4GHz (simple ones) These are the ones I will get soon enter image description here

  2. 2.4 GHz with PA+LNA These have PA+LNA and an antenna enter image description here

  3. nRF24L01+PA+LNA YJ-13039 These do not have antenna but have PA+LNA

See Link

  1. E01-ML01DP5 based on nRF24L01P, shielded, builtin antenna? These do not have a clearly visible antenna (but shielded)

See link

\$\endgroup\$
8
  • \$\begingroup\$ What do the data sheets tell you? Regards your poor distance at 433, is this relevant to the question? You have modules so why haven't you tested them? \$\endgroup\$
    – Andy aka
    Jul 21 '17 at 10:57
  • \$\begingroup\$ @Andy ... the data sheet doesn't tell about range, mostly about dB, power etc, but not directly related to distance. I didn't test the example 1 nRF24L01 yet since I haven't received them yet. But I don't want to spend too much money checking all 4 types (or even another type) without knowing beforehand what to expect. The remark about 433 is not so important, just that 433 is not suitable for me... Just to let readers know that without any external antenna that device barely reach a few meters. \$\endgroup\$ Jul 21 '17 at 11:04
  • 1
    \$\begingroup\$ What is a "noisy enviroment". It is all about SNR. If you are doing high-speed communication, "noisy" might mean a -100dBm noisefloor. If you do low-cost, low speed keyfobs, you might consider a -60dBm noisefloor practicly no noise at all. \$\endgroup\$
    – Joren Vaes
    Jul 21 '17 at 14:05
  • 1
    \$\begingroup\$ Also, if your 433MHz device could only reach a few meters, that is the device's fault, not the frequency. I've had those things have a robust communication for many hundreds of meters, without any fancy amplifiers. \$\endgroup\$
    – Joren Vaes
    Jul 21 '17 at 14:08
  • 2
    \$\begingroup\$ High speed communication, nowadays, is expressed in gigabit/second. That aside, I'm writing a answer so check back in a few minutes. \$\endgroup\$
    – Joren Vaes
    Jul 21 '17 at 14:13
1
\$\begingroup\$

The range you quote seems extremely low.

Friis' transmission equation tells us that, all things equal (we will come back to that) higher frequencies suffer more signal loss. It states:

$$ \frac{P_{R}}{P_{T}} = G_{T}G_{R}\Bigg(\frac{\lambda}{4\pi R}\Bigg)^2 $$

Where:

  • \$P_{R}\$ and \$P_{T}\$ are the received and transmitted powers in Watt
  • \$G_{R}\$ and \$G_{T}\$ are the receiver and transmitter antenna gain
  • \$\lambda\$ is the wavelength in meters
  • \$R\$ is the distance between the antennas (actually their phase centers).

Note that this only applies in the far field!

The dB version:

$$P_r = P_t + G_t + G_r + 20\log_{10}\Bigg( \frac{\lambda}{4 \pi R}\Bigg)$$

For now, let us ignore the \$G_t\$ and \$G_r\$ terms, since in general they will be pretty close to one.

The datasheet can give us a lot of information. To get an estimation of the range, what we need to look at is the received signal strength the device needs in order to, in the best case-scenario, operate.

On page 15 of the nRF24L01 datasheet There is a table "Transmitter Operation". It tells us for the maximum output power (typical):

$$P_{RF} = 0\ \text{dBm}$$

On page 16, we find the "Receiver operation" table. In this table, we can see that:

$$RX_{SENSE2}= -82\ \text{dBm}$$ $$RX_{SENSE1} = -85\ \text{dBm} $$

The received power is the transmitted power minus the power lost along the way. In symbols:

$$P_{R} = P_{T} - PL$$

Where

  • \$P_{R}\$ is the received power in dBm
  • \$P_{T}\$ is the transmitted power in dBm
  • \$PL\$ is the path loss, in dB (Not dBm!)

In other words the maximum pathloss the system tolerates at 1Mbps communication rates is:

$$PL_{MAX} = P_{T}-P_{R} = 0 \text{dBm} + 85 \text{dBm} = 85 \text{dB}$$

That means that

$$PL_{MAX} = 20\log_{10}\Bigg( \frac{\lambda}{4 \pi R}\Bigg) $$ $$R = \Big(10^{\frac{PL}{20}}\cdot \frac{4\pi}{\lambda}\Big) ^{-1} = 140\ m$$

That means that if you use half-decent antennas, no external amplifiers, it should work to about 140 meters. Ofcourse, in practice you will never reach this range. So, pretty much any board that uses the chip and hasn't been poorly designed will easily meet your requirements. (20 bytes within 5ms is about 32Kbit/s, so you have 968Kbit/s to spare).

What could be happening is the following:

  • A lot of other devices are causing interference
  • You put the antenna too close to metal (the enclosure?)
  • The boards you use are just really, really, really poorly designed.

I would suspect a board with extra PA can do more harm than good, since it might actually cause noise and/or oscillations.

The performance of an external antenna could allow you to mount the antenna far enough from any materials you don't want close.

On re-reading your question, I noticed you saying: "These 10 meters will not have walls, but musical instruments". Is it possible that you are trying to use these modules very close to human bodies? If so, they could be an issue.

\$\endgroup\$
8
  • \$\begingroup\$ The formulas in the first part I really need to read several times (and check with the datasheet). I only did tests so far with a 433 MHz which only can transmit 9600 bits per second max. The NRF24L01+ can do much more. If you say I could get 140 m without an external antenna that would be great. During my test (thus with 433 MHz) it was indeed close to another Arduino (from 5 cm to 1 meter), they were both not in enclosures, and yes they are cheap ones (but I see the same type everywhere on aliexpress). And yes, I will use them near human bodies (like on a music/band stage eventually) \$\endgroup\$ Jul 21 '17 at 15:05
  • 1
    \$\begingroup\$ What do you mean with "external antenna". The 140 meters is with a normal antenna - this can be eitehr on PCB (the squigly line) or a "big" one (usually a dipole). with just the pin connected to nothing, you are not going to get much useful out of them. When they are next to eachother on 433MHz, it could very well be that they were coupleing very much, and cliping the front end amplifiers. Sometimes moving them away will improve performance. \$\endgroup\$
    – Joren Vaes
    Jul 21 '17 at 15:09
  • \$\begingroup\$ Btw, what do you mean with half-decent antennas? Currently I used only the built-in antenna on the PCB, no external antenna (for 433 MHz). But maybe I should wait for the nRF24L01's and check at that time if they are good enough ... on the other hand, if I know beforehand I would need something with external antenna, I buy those. \$\endgroup\$ Jul 21 '17 at 15:11
  • 1
    \$\begingroup\$ With half-decent I mean something that was designed by someone who knew what they are doing. The thing with the small, on-pcb antennas is that they can be tricky to get right, and susceptible to surroundings, more so than the "black rod" types. I would not suggest getting a extra antenna (from a third party). At 2.4GHz you need to pay attention to impedance matching and so on, so selecting the right antenna is not straight forward. Unfortunately, just by looking at pictures I cannot say if an antenna is good - it depends on many variables that you can't get from a single image. \$\endgroup\$
    – Joren Vaes
    Jul 21 '17 at 15:16
  • 1
    \$\begingroup\$ The 433 MHz uses very short antenna, to severely restrict the field strength, by FCC mandate. \$\endgroup\$ Jul 21 '17 at 16:57

Not the answer you're looking for? Browse other questions tagged or ask your own question.