Hi, The samples are not simply saved into the buffer, but they are added to the samples already in the buffer. For every chirp, this additive recording starts from the beginning of the buffer and stops after 850 samples (or whatever your buffer size is). Yes, I can truly understand this part (in fact this is how it works) - but what I don't understand is why I'm not getting all the 16 samples (each sample contributes to 850 "samples", which are added on to the buffer). I've explicitly done some debugging to prove that I am getting less than 850 samples for a particular chirp + silence period. Any idea? Am I supposed to put the specialized Clock and Timer modules in another folder besides the AcousticRanging folder in the tinyos lib? Hoping to see some light on this problem. Thanks so much. Regards, Azhar
Subject: RE: [Tinyos-help] Acoustic Ranging: Not getting 16 samples within oneset of 16 chirpsDate: Thu, 13 Dec 2007 10:34:38 -0600From: [EMAIL PROTECTED]: [EMAIL PROTECTED]; [email protected] This is the expected behavior. The samples are not simply saved into the buffer, but they are added to the samples already in the buffer. For every chirp, this additive recording starts from the beginning of the buffer and stops after 850 samples (or whatever your buffer size is). After all the 16 chirps are received, what you should see in the buffer is the samplewise sum of these chirps, which should look like some silence in the beginning, then a big peak, followed by some smaller peaks which should be discarded. The position of the first big peak should carry the information on the distance. There is a technical report that explains how it works: http://www.isis.vanderbilt.edu/publications/archive/Sallai_J_2_25_2004_Acoustic_R.pdf Janos -----Original Message-----From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Muhammad AzharSent: Tuesday, December 11, 2007 8:16 PMTo: [EMAIL PROTECTED]: [Tinyos-help] Acoustic Ranging: Not getting 16 samples within oneset of 16 chirps Hi all, I'm currently working on Vanderbilt's acoustic localization module. From my understanding, this is how it works: 1) Actuator sends a radio message to sensor to inform sensor to "get ready" to receive the 16 chirps including the silence period (chirps are constant duration, silence period following each chirp is varied).2) There is a start-up delay for the mic to receive the chirps.3) The sound is emitted at the same time as the radio message (but due to the much slower speed of sound, the sensor receives the radio message first before receiving the sound).4) In AcousticMultiSamplerM.nc, upon each Timer.fired() event (i.e. for each specific chirp + silence period), it will call MicADC.getContinuousData().5) Each time the data is ready, it will then signal dataReady() in OutsideRangingSensorM, which will then be stored to the 850 samples in the buffer. Once 850 samples have been filled in the buffer, the sampler will ignore any more data and wait for the next chir! p + silence period to arrive. (Thus for each chirp + silence period, the data would then be added to the specific "sequence" in the buffer - i.e. we would expect the peak to appear somewhere below 200 (sequence number) in the array for short distances).6) A moving average function is then carried out. (I left out the peak detection process as I wanted to see the "raw" data of the 850 samples that was filled). However, in my case, I realised that in step (5), my ADC does not sample at the required rate - hence, I'm not getting 16 samples at the sensor side (if I wasn't mistaken, 3 sequences of chirp + silence period were required to fill up the 850 size of the buffer). My actuator is able to emit the 16 chirps as required. Thus, I really do not know what is wrong as I'm using the specific micaz folder done by VU which contains a different timer from the default timer. I have renamed the default micaz folder to micaz2, so that the micaz will now "point" to the micaz f! older for acoustic localization. Also, based on the diagram generated by typing "make docs micaz", it seems that my timer is wired appropriately to the "special" timer for acoustic localization. No virus found in this incoming message.Checked by AVG Free Edition.Version: 7.5.503 / Virus Database: 269.17.1/1181 - Release Date: 12/11/2007 5:05 PM
_______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
