how many nRF24L01+ RF tranceivers in a network

RE: how many nRF24L01+ RF tranceivers in a network

Jasper — Tue, 21 Jul 2015 14:11:11 GMT

nRF51 should have more processing power than I need. I was planning to use a ATMega328. But the nRF51 can do the job...

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 13:26:31 GMT

Yes, it is a system on chip. Possibly if you don't have high processing requirements it could handle all the things that you need on the MCU side as well. If it is not powerful enough, we've recently launched the nRF52 which boasts even lower energy consumption and higher processing power.

RE: how many nRF24L01+ RF tranceivers in a network

Jasper — Tue, 21 Jul 2015 13:16:55 GMT

transmit and receive at the same time is not necessary. Always in receive mode, if we need to transmit, then transmit and back to receive mode.

I'm planning to use an easy and cheap Atmel AVR like on the Arduino. So an existing library is preferable (like for the nRF24) I'll have a look at the nRF51. it is a SoC, right?

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 12:58:50 GMT

It would also be possible to use the nRF51, which has an extensive list of third party modules here, for your purpose, I'm not sure what kind of MCU you are planning on using, but it is not that hard to interface with the nRF51.

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 12:54:58 GMT

The antenna you would use on the robot can be used as both a receiver and a transmitter, so it would increase the range in both cases. However you cannot both receive and transmit at the same time with the same antenna.

RE: how many nRF24L01+ RF tranceivers in a network

Jasper — Tue, 21 Jul 2015 12:52:46 GMT

I have also found that pcb. But it has a range of 1km, And I prefer to have a max. range which is shorter. When using 1km, the AP will hear many tranceivers.... and then I've to expand the amount of addresses. and we get more and more interference...

Rotating antenna on the robot should be nice if we only have to send, but they also receive some data from the AP. So then I have to rotate that one as well. I probably forget to mention this.

In fact, there are many modules to do this, the nRF24L01+ is not the only one, but probably one of the best solutions as far as I know right now..

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 12:42:54 GMT

Yes, this the APs should be able to do this.

A quick google search gave this result, unfortunately I'm not very familiar with nRF24L01+ modules, so there might be more/better options.

You are right that rotating the antenna on the AP is probably not a great idea, but what about rotating the antenna on the robots?

RE: how many nRF24L01+ RF tranceivers in a network

Jasper — Tue, 21 Jul 2015 12:27:00 GMT

yes, you're right. It is possible to see more than 24 robots, but due to lower RSSI and also a different network ID, the AP should discard that message anyway, right? So with 4 dipswitches, I'm able to create 16 network IDs.

there are 500 APs in this configuration which make a really long solar panel field (+/- 100km)

The 100 meter is a must, so an amplifier should be a better option. I prefer to have a LOS range coverage of 150 meter. Not much further. Is there a reference design for the nRF24L01+ with PA or LNA. Or can I buy this as a module. I prefer the last option. RF PCB design is not my daily job.

A mechanical contraption is probably not an option. I can't move/rotate the antenna from the AP which is located in the middle.

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 12:01:09 GMT

I'm not sure that it would be a max of 24 robots, potentially you could suddenly see a lot longer range than expected and be able to see all 8 adjacent squares, however this is not necessarily a problem as the robots in your square will always have a higher RSSI than the ones from adjacent squares.

While the modules have a range of 100m described, this is almost always best-case scenario. You want an application that is robust to changing atmospheric conditions and other things that might cause path loss. What is going to happen to your robots if the range suddenly drops to 80m?

Another option is to look at antennas that have higher gain than a regular dipole, for instance since your robots are moving along a track you can have some mechanical contraption that moves a yagi antenna in a way so that it points towards the correct access point.

RE: how many nRF24L01+ RF tranceivers in a network

Jasper — Tue, 21 Jul 2015 11:37:38 GMT

Hello Øyvind,

Thank you for your reply.

The configuration you have described is nearly what I would like to have. But instead of multiple robots per row, I just have only 1 robot in a row. The accesspoint in in the middle of a square of 200x200 meter, so the next AP is also 200m further. See below.

= robot, ------ = robot track AP = accesspoint
- - - - | * - - - - - - - -
- - - - | - * - - - - - - -
- - - - | - - * - - - - - -
- - - - | - - - * - - - - - AP1 < 200m > AP2
- - - - | - - - - - - * - -
- - - - | - - - - - * - - -
- - - - | - - - - * - - - -
- - - - | - - - - - - - - *

So when a robot from AP1 is on the border of his grid, AP2 will also receive his signal. But in the worst case, an AP will receive max. 24 robots, right?

I also have to think about addressing of the units. In this case, I don't need a unique ID for all robots I think. And I can use different channels.

EDIT: When using dipswitches to set the network ID 0 to 3 for example, I can use address 0 for all AP's and 1 to 8 for all robots when using a different network ID?

Why using a PA or LNA. I have found some modules (sparkfun) which have a range of 100m with external antenna and no PA or LNA.

Best regards,

Jasper

RE: how many nRF24L01+ RF tranceivers in a network

Øyvind Karlsen — Tue, 21 Jul 2015 11:10:21 GMT

Hi Jasper,

Are you concerned about noise between the different centrals? If you have a configuration that is something like this:(* is a robot and C is an access point)

* * * * * *
* C * * C *
* * * * * *
* * * * * *
* C * * C *
* * * * * *

Then the distance between robots that are transmitting different network ids and the closest access point with a different network id would be 200-300m based on the way I am reading your post.

To get a range of 100-150m reliably with the nRF24L01+ you will need to use a power amplifier(PA) or a low noise amplifier(LNA). Since you are able to control the effect that you transmit you will be able to tune your system so that you can "hear" a robot/access point that is within your 100-150m range, but not within the 200-300m range, i.e. the loss is so high for ranges over 200m that the robots appear to be invisible.

I might be interpreting your question incorrectly, please elaborate if so.

Best regards,

Øyvind