Hi! I'm working to a simple didactic project about sampling.

this ancient series is still available today. they're designed for 8080 processors, so you'll have to figure out how to drive the RD and WR lines properly. but they're so old that many example builds can be found on the web.
http://www.ti.com/lit/ds/symlink/adc0804-n.pdf

Your explanation does not make much sense. To begin with, do you want an ADC, a DAC or both? Do you want to show how the ADC works internally and do you want a sample and hold before it? If not, then you usually just feed the (filtered) analog signal to the ADC and out comes the digital results. Not much to see here.

One way to achieve the stepped input followed by 8b conversion would be LF398 + ADC0800. ADC0800 even requires sample and hold to operate properly. If the ADC itself is too much a black box, you can build one from a SAR register, a DAC and a comparator.
If you want to skip the S&H, AD7820 does not need it if the input frequency is less than 7kHz. There are other options too, of course.

If you then want to convert the signal back to analog, something like DAC08 would do the job. It does not have any filters, so you'll get the stepped waveform out and can then filter it to produce the final result.

>>1052126
>http://www.ti.com/lit/ds/symlink/adc0804-n.pdf

Thanks for your answer, but I forgot to say that i'd like to keep the project in audio range. It's easier to show and to understand in my opinion.

File: aliasing.png (146KB, 610x379px) Image search: [Google]

146KB, 610x379px

>>1052133

I just want an ADC, the DAC will be made with a really intuitive resistor ladder and an op amp as a buffer.

My intent is to show the Nyquist Shannon theorem so I don't really want to explain deeply how an adc works, but i want to change the sample rate freely. A simple s&h is pretty easy to build on a breadbord, and i'm already planning to make an oscillator to to clock the ADC, the goal is to change easelly the Nyquist frequency, so this is not really an importa point to chose the ADC.

>If not, then you usually just feed the (filtered) analog signal to the ADC and out comes the digital results. Not much to see here.

Thats what i want. And I also want to skyp the filter, caus i will feed maily sine waves or more complex wave via a digital function generator, so i will add the filter here. The aim is to show aliasing in practice, the reflection of the spectrum ( pic related ) and the importance of the reconstruction filter.


I'm going to read the datasheet of the ADC0800. Ty.

>>1052149
If you want to demonstrate sampling and aliasing only, then there's no need for ADCs and DACs. The sample and hold chip alone (suitably clocked) is enough.

>>1052160

Yeah, that's right. But i prefer to add a dac to avoid any confusion and to show also exactly how quantization works. It might be worth?

Anyway the rough s&h solution is so cheap that could be a plausible. Thanks.