Hi, With a packet every 10ms (and assuming that you're using the default packet size) the Tx block will remain active for about 1.5ms. This means that the Tx block has a 15% duty cycle while the Rx block has a 100% duty cycle.
What stumps me is why they all run out at approximately the same time, since node 1 would have 45% Tx duty cycle, node 2 would have 30% and node 3 would have 15%. Even if we consider that Tx and Rx power are approximately the same, the numbers are sufficiently different that the observed lifetimes should be different too. Is it possible to check what topology your network converges to in CTP. I have a hunch that even though you're using MHOscope, that CTP is converging to one-hop for each node to the sink. A quick way to check this would be to observe the THL field of the CTP header. Could you check this for me please. Regards, Zainul. On Wed, Jun 17, 2009 at 1:01 AM, ma. pratama<[email protected]> wrote: > Thx for your reply.. but im still confuse about the active TX block you > mention before, and what it has to do with the Rx power? For your > information, i set all node to transmit the packet every 10 ms. How can we > know which node tansmit more data than the others?and how could possibly the > node transmit more data than the others if i already set all node to > transmit every 10 ms?? in my case its node 3 ran out the power first. > > --- On Wed, 6/17/09, Zainul M Charbiwala <[email protected]> wrote: > > From: Zainul M Charbiwala <[email protected]> > Subject: Re: [Tinyos-help] ask about energy consumption > To: "ma. pratama" <[email protected]> > Cc: [email protected] > Date: Wednesday, June 17, 2009, 3:48 AM > > It is hard to tell whether the batteries for each node had the same > capacity to begin with. Assuming they were fresh batteries with equal > capacity, they would all die at pretty much the same time. This is > what you see and its because: > > 1. Transmit cost and receive cost on the CC2420 are about the same so > it doesn't really matter whether you are closer to the sink. The node > that transmits more often will tend to die a little faster (which > contradicts your observations) because even though the Tx power is > lesser than Rx, while the Tx block is active, Rx is also active (so > that is is ready for the ACK), resulting in almost double the > consumption during transmissions. > > 2. If you're not using LPL, then all the nodes are burning pretty much > the same amount and the high constant Rx cost overshadows any > differences you would see in Tx costs. > > 3. Battery drain is highly non-linear, batteries have manufacturing > variations and they are quite environmentally sensitive. Along with > this, tiny variations in node components will cause an unequal power > consumption. For example, the LED bias resistors may be unequal (they > are 10% tolerance on MicaZ) which may cause different current draws. > > > Hope that helps, at least partially. > > Zainul. > > > On Wed, Jun 17, 2009 at 12:19 AM, ma. pratama<[email protected]> > wrote: >> Hi all...i recently try to run multihop program from crossbow with 3 node >> and 1 sink (im using Micaz mote and MIB520 board programmer). and after 16 >> hour nonstop, node 3 ( the longest from the sink) started to run out of >> energy to transmitted the packet. and then after a several minutes, node 2 >> started to ran out energy too. And finally, node 1(closest to the sink) >> lost >> its power too.. i wonder why node 3 lost its power first?? from >> information >> i read the node that close to sink will lost the power first because it >> receive and transmitted the data to the sink?? please correct me if im >> wrong. >> >> Anyone know why it happen like that?? >> >> >> _______________________________________________ >> Tinyos-help mailing list >> [email protected] >> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help >> > > _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
