??? 12/23/04 03:52 Read: times |
#83675 - Re: Not possible comments. Responding to: ???'s previous message |
Dear Shivani,
thanks prahlad for your explanation,any comment about not possible reply is he right i am bit confused I assume you are referring to Kuldeep Singhs Reply here http://www.8052.com/forum/read.phtml?id=83635 when you said Not Possible reply. Shivani, The text in Kuldeep's reply is theoretically correct but I believe kuldeep overlooked lots of practical aspects when he say "NOT POSSIBLE". My Comments: 1. Microvolts signals can be measured with an 8 bit ADC after proper amplification. its not Impossible. 2. Selection of an ADC doesn't depend on weather signal is milivolts or microvolts but depends on least count in measurement desired along with the range desired. 3. Descrete values from and ADC = (2^bit ADC) for example for an 8 bit ADC descrete values output=2^8=256. thus n=2^b. 4. Range of measurement = descrete values from ADC * Least count of measurement. R=n*LC. 5. An 8 bit ADC gives 2^8=256 descrete values for input signal thus if you want to measure 10mV max with this ADC your least count becomes 10mV/256=39uV. 6. A 14bit ADC gives 2^14=16384 descrete values thus when measuring 10mV max signal you get least count 0.6uV. 7. The above doesnt mean 0.6uV can't be measured with 8 bit ADC It means that you can not measure 0.6uV when having a range of 10mV. If you wish to have a least count of 0.6uV with 8 bit ADC then your range becomes 0.6uV*256= 153uV. 8. Always remember that you must amplify your signal so that it gives you max. voltage to ADC for max. signal. 9. Always use the maximum reference voltage and use a precision Vref Source. 10. Working with higher voltages for signal input to ADC and reference makes the noise margin larger and gives you better measurement accuracy. 11. 1V reference 1/2V reference are theorectically correct but practically it shouldn't be done. Since doing so reduces your noise margin 5 to 10 times and reduces your measurement accuracy by same margin. 12. Amplification factors of 500 as mentioned by Kuldeep are not so easy to achieve if you really want 14bit accuracy in measurement. You need to have a very good board layout and need to consider lots of noise reduction measures for getting that accuracy. Checkout for monolithic instrumentation amplifiers by Aanlog Decives. www.analog.com if you really need that high amplification and simultaneously 12bit+ accuracy. Regards, Prahlad Purohit |
Topic | Author | Date |
adc resolution | 01/01/70 00:00 | |
Not Just Resolution but Range too. | 01/01/70 00:00 | |
do not confuse | 01/01/70 00:00 | |
not possible | 01/01/70 00:00 | |
adc resolution | 01/01/70 00:00 | |
Re: Not possible comments. | 01/01/70 00:00 | |
Not all adcs are the same | 01/01/70 00:00 | |
glass body diodes | 01/01/70 00:00 | |
Something new...![]() | 01/01/70 00:00 |