doubts on smoothening techniques for ADC/DAC data (DSP)

this thread is to complement

http://www.8051projects.net/forum-t7667-last.html


please post your doubts about

smoothening techniques here

Hi shyam,
Is this the thread you mentioned? Good idea! :-) But please do post something soon.. am waiting eagerly. And if possible do include help in assembly and not only C. Will be a huge help for noobs like me

please visit..
http://www.8051projects.net/forum-t7667-last.html

this thread is only to discuss the doubts regarding the above link

Hey Shyam, the concept is good. but how do u implement that using a uC? and the link is closed. cant post anything in that so i posted here...

keep reading!!!..

and yes u cant post there so all doubts will be solved here..


well implementing the sample hold technique can be done thru a micro controller it is easy!!!...

but it is like fooling your client

this is the way to do it..

all you need is to make the changes in the ADC data a little slower!!..

step1 . set interval for sampling.. suppose 1 sec..
step2 set interval to display data.. assume u r making a graph so points per sec suppose 5.
step3. read data @ADC at sampling rate...
step4 .no need to read the next 4 values simply copy the values to the next 4 variables.. repeat step3 and 4.

ex..

sample 1 =0 sample 2 =1 sample 3 =2 sample 4 =3 sample 5=2 sample 6 =1 sample 7 =0

u will get the following representation




thus you will get a theoritically smooth looking graph but practically it is useless...
therefore sample and hold technique is seldom used in softwares/codes... it is mostly an hardware implementation...
simplest S&H mechanism is the operation of a capacitor..
u can add a capacitor into your input(adc) to get the S&H response.

keep reading as i will be posting more techniques..

Nice explanation Shyam........ :-) :-)

you must click on thanks button

Hey Shyam thats really nice! Could never think of that. Well now i know the difference b/w having a theoretical knowledge and its practical implementation!

But I have a doubt. By copying the data to then next 4 variables, arent we introducing incorrect data there as there is a possibility the ADC o/p may have changed at that instant.

Also, can this be used as a replacement to the barrel shift algorithm u mentioned earlier?

By using a capacitor before the i/p of ADC will surely help smoothening out the variations! Il try that for sure!!!

Thanks a million!!!

barrel shifter mechanism is way too ahead of all these techniques
i am taking u there one by one..

yes we are intro ducing incorrect data... thats why i said it shud not be implemented thru software but mostly thru hardware!!!

Hmmm... I am waiting eagerly to learn more... pls do post soon... thanks alot!!! :bye