ADC in Atmega8

I am doing a project using ATmega8. For that we need an ADC of atleast
5Msps. We first thought of using the built in ADC but I want to know
the maximum samples per second for the in built ADC? I guess it is
around 150Ksps but I am not sure. Can I increase this sample rate by
loading the prescalar value like one or 2? Could someone help.
with regards
Reply to
Loading thread data ...
I'm not aware of any of the 8-bit AVRs offering that kind of speed, but there are plenty of pipelined ADC chips that provide these speeds or faster, and which offer parallel, serial, I2C, and other interface methods. Try Analog for starters.
-- Gordon
Reply to
Gordon McComb
I think you can get up to 0.6Msps with some AVRs (13 ADC clock cycles per conversion, /2 prescaler for the ADC clock and a 16MHz system clock). You get a lot of noise running them that fast though, perhaps effective 6 or 7-bit resolution from the 10-bit ADC. You need to stay below 15ksps if you want full 10-bit resolution.
I think you'll need an external ADC.
Reply to
Tim Auton
Ignoring all the problems Tim mentions, is there any hope of even approaching the OP's need for 5Msps by parallelism?
To put it another way, is there any (relatively) simple way of using multiple AVRs (or other multiplexed ADC chips) together to increase the effective sampling rate?
I'd guess that synchronization would be a problem. Having (say) 10 AVRs at 0.6Msps all sampling at the exact same instant _does_ yield 6Msps but they're not very _useful_ samples. What else would one have to be worried about?
Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut minds pring dawt cahm (y'all) -- "Wisdom lies in taking everything with good humor and a grain of salt." -- George Santayana --
Reply to
Frnak McKenney
I'd worry about the low input impedance of multiple AVRs sampling flat-out; some decent buffering (an op-amp or 10) would likely be required.
Given 20Msps ADCs cost about the same as just one of the AVRs you'd need, have higher input impedance, lower noise, and would take about 1/20th the board space, I think one might also have to worry about ones sanity :)
Reply to
Tim Auton
And we haven't even addressed the problem of where to _put_ all those samples while we're scratching our assets and trying to figure out what exactly to _do_ with them.
"All buffers come in one size: too small.
Okay, but more than usual?
Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut minds pring dawt cahm (y'all) -- I would only say to the poets that science must dissect as one path to understanding, but never to destroy the beauty and joy of wholeness. --- Stephen Jay Gould / The Hedgehog, the Fox, and the Magister's Pox --
Reply to
Frnak McKenney

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.