0
\$\begingroup\$

The design involves a single base station that communicates with multiple peripheral units. Peripheral units should be able to communicate back to the base station but are not required to communicate between them.

It would be best if some mechanism exists so that when a peripheral unit tries to send data to the base station while another is sending data it will get into a queue.

The amount of data is not great and not continues, more like bursts of 50 bytes or so once in a few minutes / hours.

The system should be low cost as possible with ranges up to 100m in open space under ideal conditions.

If I could pick a close example from wired bus world I guess it would be I2C where each peripheral has an address.

Thanks!

\$\endgroup\$
  • \$\begingroup\$ So, what have you found so far? Also, the search has one limiting factor - you need a device with a queueing algorithm to avoid clashes. \$\endgroup\$ – Solar Mike Oct 12 '17 at 5:58
  • \$\begingroup\$ I'd look at 315/433 MHz modules. This is what is used for home weather stations with a remote temperature and/or wind sensor. The modules for 315/433 MHz are very cheap like $ 1.20 (from China). Queuing and anti-collision is complex, if the data is send in short bursts at random times and repeated a few times then the chance of a collision becomes very small. Under ideal conditions 100m should be easy provided you get a sensitive receiver. For that, I suggest a SYN480R based module, a superheterodyne receiver or similar. \$\endgroup\$ – Bimpelrekkie Oct 12 '17 at 6:12
  • \$\begingroup\$ No, avoid the cheap 433 MHz modules - they're quite bad. If you are going to use that frequency in an application where you need range, at least use proper radio chips on both ends, see the SiLabs or TI transceiver offerings - Si446x, CC11xx, etc. A step up would be LoRa on that or 868/915 MHz (depending on what is allowed in your region). \$\endgroup\$ – Chris Stratton Oct 12 '17 at 6:24
  • \$\begingroup\$ The host device has to assign proper timings for each device, in order to avoid colision. There is no way to avoid implement some intelligence on the system itself, instead of trusting this task to the communication arrangements, otherwise you will have to pay more. \$\endgroup\$ – aluis.rcastro Oct 12 '17 at 6:26
  • \$\begingroup\$ @SolarMike I actually only looked on ESP8266 because it has an added benefit of having an MCU inside which I can use for an almost single chip solution. \$\endgroup\$ – user34920 Oct 12 '17 at 6:26
0
\$\begingroup\$

I run a system using 2.4GHz WiFi where the remotes (clients, not APs) are ESP8266 so no limit as per @dandavis, the base station(s) are OPi 1 (wired Ethernet) and the wireless base station is a router which I already had in place for internet connectivity. Central unit(s) schedule/polls remotes (using http protocol) so allows for bases:peripherals N:M. Its not MCU to MCU, but the central unit is $12 delivered which is actually cheaper than most other wireless solutions I examined. 100m might be a challenge depending on your router and antenna arrangement, but 50m is easily achievable.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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