Guys by the way, I have couple question related to serial port communication which is due today, how do we handle the difference between real time side thread and user process side data rate
Sent from my iPhone On May 15, 2013, at 8:19 AM, Martín René <[email protected]> wrote: > Thanks in advance > > ---------- Forwarded message ---------- > From: Rodrigo Carbajales <[email protected]> > Date: Thu, May 9, 2013 at 2:33 PM > Subject: Re: [Tinyos-help] Z1 High speed i2c accelerometer sampling > To: [email protected] > Cc: Martín René Vilugrón <[email protected]> > > > Hi Antonio, My name is Rodrigo and I work with Martin Rene Vilugron at CNEA, > Argentina. > Thanks for the solution proposed. Now we can reach a sample frequency of 1KHz! > I want to make another question, is it possible to use a faster timer? I use > TimerMilli but it is not possible to use Timer32khz or TimerMicro? > Regards, > Rodrigo > > > On 17 April 2013 10:06, Martín René <[email protected]> wrote: >> Aceleración >> --- >> Martín René Vilugrón >> San Carlos de Bariloche >> Patagonia Argentina >> >> >> ---------- Forwarded message ---------- >> From: Antonio Linan <[email protected]> >> Date: Wed, Apr 17, 2013 at 5:53 AM >> Subject: Re: [Tinyos-help] Z1 High speed i2c accelerometer sampling >> To: Martín René <[email protected]> >> Cc: TinyOS Mailing Lists <[email protected]> >> >> >> Hello again, >> >> Using the TestADXL345 example in /apps/tests/z1/Accelerometer I got some >> numbers for you, basically you can: >> >> - Comment out the delay at /tos/chips/adxl345/adxl345P file in line 787. >> - Edit the example and start the Accelerometer only once (in boot only), >> then periodically sample 1 axis >> >> As it is the ubr (in /tos/platforms/z1/chips/msp430/usci/Z1UsciP.nc is set >> to 800, as SMCLK is sourced from the DCO with /1 divider, this gives a rough >> 10KHz clock (I measured this to be actually 9.8724Khz), giving you a total >> reading time of 4.84ms (measured just before calling .read() and in >> readDone() event). Changing the ubr to 80 you can get a 87.912Khz clock >> (measured), with a reading time of 0.711ms. >> >> Hope this helps. >> >> Regards, >> >> --Antonio >> >> >> On Tue, Apr 16, 2013 at 10:29 PM, Antonio Linan <[email protected]> wrote: >>> Hola! >>> >>> Yes, there might be room for improvement, if you look at the code it has an >>> active waits (around line 787) that slow the readings but were necessary >>> at the time, also one could try and configure the I2C module with different >>> settings (UBR for example) to see if more speed is allowed. I have planned >>> a revision of the usci driver, so things may change, in the meantime if you >>> wanna dig a little you are welcome, che :) >>> >>> Saludos! >>> >>> --Antonio >>> >>> >>> >>> >>> >>> >>> On Tue, Apr 16, 2013 at 7:16 PM, Martín René <[email protected]> wrote: >>>> Sorry that i didn't make the interface problem clear... >>>> >>>> The accelerometer is digital using i2c interface, and doesn't provide a >>>> readstream interface. >>>> >>>> With another sensor, we reached high sampling speeds using the processor's >>>> ADC, but the bus does'nt seem to reach the necessary speeds, maybe delays >>>> or something else slows down the sampling... >>>> >>>> >>>> Saludos y muchas gracias! >>>> >>>> --- >>>> Martín René Vilugrón >>>> San Carlos de Bariloche >>>> Patagonia Argentina >>>> >>>> >>>> On Tue, Apr 16, 2013 at 1:58 PM, Michael Schippling <[email protected]> >>>> wrote: >>>>> There are some intriguingly named files containing "Stream" and "DMA" >>>>> in the ...tinyos-2.x_main/tos/chips/msp430/adc12 directory. Some of >>>>> them even have comments which reference TEP documents that might >>>>> explain what they are trying to do. Maybe one of them will allow >>>>> you to run the ADC at higher rates... >>>>> >>>>> In T1 with the mica's (what I know....) one could free-run a single >>>>> ADC at low audio rates by bypassing most of TOS. >>>>> >>>>> MS >>>>> >>>>> >>>>> >>>>> On 4/15/2013 1:27 PM, Martín René wrote: >>>>>> We are currently using the Z1 Motes and sampling at 200Hz without >>>>>> problems, but we can't raise the sampling speed above said speed without >>>>>> starting to have problems. We need to sample at 1000Hz and we can't get >>>>>> there. >>>>>> >>>>>> I know that the resources of the platform are scarse, but i know that >>>>>> the ADC can manage those speeds, and as we don't have a analog >>>>>> accelerometer we neee to speed up all we can the ADXL. >>>>>> >>>>>> Is it possible to sample the accelometer at those speeds? >>>>>> If it is, ¿how? >>>>>> >>>>>> Thanks! and saludos >>>>>> >>>>>> --- >>>>>> Martín René Vilugrón >>>>>> San Carlos de Bariloche >>>>>> Patagonia Argentina >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> 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 >>> >>> >>> >>> -- >>> -- >>> Antonio Liñan Colina >>> R+D+I Engineer >>> @: [email protected] >>> @: [email protected] >>> ------------------------------ >>> Advancare >>> T: +34 93 582 02 70 >>> http://www.advancare.com >>> http://www.zolertia.com >>> http://zolertia.sourceforge.net >>> http://webshop.zolertia.com >> >> >> >> -- >> -- >> Antonio Liñan Colina >> R+D+I Engineer >> @: [email protected] >> @: [email protected] >> ------------------------------ >> Advancare >> T: +34 93 582 02 70 >> http://www.advancare.com >> http://www.zolertia.com >> http://zolertia.sourceforge.net >> http://webshop.zolertia.com > > > _______________________________________________ > 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
