|
Hello, I'm trying to measure how long the A to D coversions are taking. I'm using inputs 0 and 1 in differnetial mode. I have it set up to use the call back on conversion. I drive a gpio line high when I call the ioctl with ADC_START and then drive the line low in the call back. It seems to be taking 8 microsecs per conversion and sometimes the call back routine doesn't get called. The manual says the coversion time should be 1.7 micro seconds. Also, if I'm running from my main routine in a loop waiting for the call back routine to get called, it never gets called. I had to put a nanosleep in my loop for this to happen ? Any suggestions on what I'm doing wrong. Thanks |
|
|
DSP56F803EVM ADC
Started by ●May 12, 2003
Reply by ●May 14, 20032003-05-14
|
I do not think you can measure exact ADC conversion time by using software. There are many factors you haven't taken in account. The first all, interrupt latency. The second, ioctl and callback code efficiency. To roughly measure ADC conversion rate, 1st use assembly instruction; 2nd measure the period of 8 conversions instead of single conversion; 3rd slow down ADC clock. So that you may be able to get rough cycle number of 8 conversions. Hopefully this will help. C.W. --- elowe_spsy <> wrote: > Hello, > > I'm trying to measure how long the A to D coversions > are taking. > I'm using inputs 0 and 1 in differnetial mode. > I have it set up to use the call back on conversion. > I drive a gpio line high when I call the ioctl with > ADC_START > and then drive the line low in the call back. > It seems to be taking 8 microsecs per conversion and > sometimes > the call back routine doesn't get called. > The manual says the coversion time should be 1.7 > micro seconds. > > Also, if I'm running from my main routine in a loop > waiting > for the call back routine to get called, it never > gets called. > I had to put a nanosleep in my loop for this to > happen ? > > Any suggestions on what I'm doing wrong. > > Thanks > __________________________________ |
Reply by ●May 14, 20032003-05-14
|
We have a set of ADC routines in the FAQ system that might help you. Take a
look at the following FAQ. http://e-www.motorola.com/webapp/sps/utils/SingleFaq.jsp?FAQ-11868.xml <http://e-www.motorola.com/webapp/sps/utils/SingleFaq.jsp?FAQ-11868.xml> The example 4-Sim Triggered - 2 us is quite efficient in the way it is written. If you use this type of structure of installing a static super fast ISR you can effectively measure the conversion time sicnce the interrupt latency times will be minimized. Although I guarantee it will match the times as stated in the user manual and data sheets. A common mistake for when the ADC is taking longer than expected is that unsued channels are not turned off. The above examples are written for a 805 EVM , but the general code applies to the 803. Thanks. - Bill -----Original Message----- From: Charlie W [mailto:] Sent: Tuesday, May 13, 2003 9:28 PM To: elowe_spsy; Subject: Re: [motoroladsp] DSP56F803EVM ADC I do not think you can measure exact ADC conversion time by using software. There are many factors you haven't taken in account. The first all, interrupt latency. The second, ioctl and callback code efficiency. To roughly measure ADC conversion rate, 1st use assembly instruction; 2nd measure the period of 8 conversions instead of single conversion; 3rd slow down ADC clock. So that you may be able to get rough cycle number of 8 conversions. Hopefully this will help. C.W. --- elowe_spsy <> wrote: > Hello, > > I'm trying to measure how long the A to D coversions > are taking. > I'm using inputs 0 and 1 in differnetial mode. > I have it set up to use the call back on conversion. > I drive a gpio line high when I call the ioctl with > ADC_START > and then drive the line low in the call back. > It seems to be taking 8 microsecs per conversion and > sometimes > the call back routine doesn't get called. > The manual says the coversion time should be 1.7 > micro seconds. > > Also, if I'm running from my main routine in a loop > waiting > for the call back routine to get called, it never > gets called. > I had to put a nanosleep in my loop for this to > happen ? > > Any suggestions on what I'm doing wrong. > > Thanks > __________________________________ <http://rd.yahoo.com/M%3924.3294213.4590844.1261774/D=egroupweb/S05771855:H\ M/A80983/R=2/id=noscript/*http://companion.yahoo.com/config/features5> <http://us.adserver.yahoo.com/l?M%3924.3294213.4590844.1261774/D=egroupmail/S=\ :HM/A80983/randd3112798> _____________________________________ Note: If you do a simple "reply" with your email client, only the author of this message will receive your answer. You need to do a "reply all" if you want your answer to be distributed to the entire group. _____________________________________ About this discussion group: To Join: To Post: To Leave: Archives: http://www.yahoogroups.com/group/motoroladsp <http://www.yahoogroups.com/group/motoroladsp> More Groups: http://www.dsprelated.com/groups.php3 <http://www.dsprelated.com/groups.php3 ">http://docs.yahoo.com/info/terms/> . |
