Feasibility of the VERY low power wireless network?

Dear Ulf @ All,

I reviewed several low power wireless chipsets from different vendors. They typically drain about 20mA @ 3.3V when in the active Rx or Tx mode, and less then 1mA in sleep. Switching from sleep to active incurs the overhead of several milliseconds; unfortunately none of the transceivers retain the bit synchronization between the packets; there is no easy way to synchronize the nodes to the master to a fraction of the bit interval. As noted by David and Mark, it could be possible to fit the

10mW constraint if the system cycle time is several seconds; but it is hardly feasible with the cycle time < 1 sec.

Next question: are there any off-the-shelf wireless modules that could allow for the ~1/100 duty cycle operation with ~10ms packets with the average power consumption ~10mW ?

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky
Loading thread data ...

e

low

This requirement is very difficult to achieve. Just curious, what application requires such tight timing?

Reply to
linnix

Vladimir Vassilevsky skrev 2011-01-07 20:14:

If you are desperate for good power figures, then you should consider running at 1.8V.

The ATmega128RFA1 power consumption in sleep is 0.4 mA * 1.8V so less than 1 mW. With 30 mW during transmit, you can have a duty cycle of almost 33% to still achieve

Link budget is 103,5 dBm which is also quite good. If you can live with a link budget of 97,5 (still better than most competition), then the transmit power goes down to < 15 mW.

Did not check the Atmel ZigBit modules (which are two chips solutions) but they should be similar in power consumption.

--
Best Regards
Ulf Samuelsson
These are my own personal opinions, which may
or may not be shared by my employer Atmel Nordic AB
Reply to
Ulf Samuelsson

"A bit surly"??? You mean like Ted Bundy was a bit rough on women, or the Japanese are a bit aggressive with whales, or the way the Taliban is a bit ugly with women? Yeah, otherwise they are all really great guys and there is no reason to not accept their bad side. Heck, everyone has a bad side, no?

Rick

Reply to
rickman

While it might be possible to calculate the link budget for an interplanetary space link with an accuracy of a fraction of an dB, such accurate (and successful) predictions in ground based systems should IMHO be awarded with the Nobel prize :-).

In real world urban conditions, the path loss will typically increase by 10-15 dB each time the distance doubles (contrary to 6 dB for free space).

I first assumed that you are simply talking about the difference between the Tx/Rx power, but the figures does not make any sense.

If 103.5 dBm actually refers to -103.5 dBm receiver sensitivity and the thermal noise at room temperature id -174 dBm/Hz thus bandwidths larger than 1 MHz should be assumed, much larger than the OP needed (10-50 kHz at most).

Reply to
upsidedown

he

f
,
s
y

I read the whole datasheet for Atmega128rfa1. Other than the standard AVR stuffs, there is not much information on the RF controller. Perhaps someone can answer my questions:

  1. Is the Radio permanently attached to the UART?

  1. Is the Radio permanently attached to the SPI?

  2. Can the radio be enable/disable on demand?

  1. Can the transmitter power be adjusted/changed?

  2. Can the frequency be adjusted/changed?

Are there other docs to address these issues?

Reply to
linnix

I can envision one: A security system where an intrusion sensor is triggered and the master system then wants to zoom a surveillance camera in on that particular location. If the system has a latency of 10 seconds, you might get a high-res picture of a broken window minus the ski-masked perp.

Mark Borgerson

Reply to
Mark Borgerson

a

n the

oll

e to

0mW

d allow

e of

g
.

de,

ers

way

s

In that case, single on-off transmitters might be more cost effective. You can have all sensors transmitting the same frequency and any one can trigger all cameras in the group. Gate/garage transmitters are dirt cheap.

Reply to
linnix

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.