EveryCircuit - 8bit adc

maxmax_66

8 years ago

Well done.

nikisalli

8 years ago

Nice i like how compact you made it!

hurz

8 years ago

Anti spike solution "analog" the digital variante is to boring ;-) http://everycircuit.com/circuit/5795114189586432

rich11292000

8 years ago

Dont we need 0-15 volts here?

roops1967

8 years ago

Cheers hurz that dampens it nicely. i was actually looking for a 'simple' (boring??) digital fix, any idea?

hurz

8 years ago

Simple, cuz you just have to sit and figure out the correct timing of latch clocks. Put enough latches to collect the results and promot the final result.

hurz

8 years ago

For me it was boring http://everycircuit.com/circuit/5042892006752256

roops1967

8 years ago

That works thanks, can get away with just buffering the most significant adc like http://everycircuit.com/circuit/4787299039838208

roops1967

8 years ago

Okay but can you or anybody think of a different (simple) solution to get 8bit adc using 2x4bit without having to fix it with additional components like above? I believe that my circuit is flawed if we have to bandage it like above, its not a simple circuit anymore

roops1967

8 years ago

Yep rich11292000 i still gotta fix the voltage range

hurz

8 years ago

@2ctiby, not really. What was the idea behind this, if any?

2ctiby

8 years ago

http://everycircuit.com/circuit/4673926633619456 brief variation...any use?

hurz

8 years ago

@2ctiby, please make at least some mini tests for your circuits.

hurz

8 years ago

Here a proposual of how you can test your curcuit, demonstrated at roops working version http://everycircuit.com/circuit/5481557619113984

2ctiby

8 years ago

@hurz... your demonstration has changed the sine wave into a pulse input, and there is a counter along with an extra DAC.... that is far removed from the original request for a simple no-extra reduction of spikes.

roops1967

8 years ago

2ctiby your example don't look right sorry. The top (least significant) adc is should be converting the modulated output from the opamp (difference between input signal and 4bit dac). On your circuit the orange output is showing the analogue input and a small amount of the 4bit dac output

roops1967

8 years ago

hurz' demonstrating the dacs are working as should be. 2ctiby examine your circuit this way you will see whats wrong. Unless I'm missing something you are trying to demonstrate then I'm sorry i apologise...

hurz

8 years ago

Thanx roops, the discussion normaly with 2ctiby starts at this point being very painful. Why, I dont know. Hope he is more listening to your words and see whats wrong on his side.

2ctiby

8 years ago

Thanks @roops1967 ...it was just a casual play from me with the new toys which I know next to nothing about (I put it forward tentatively as you can see). Sadly, Hurz does not communicate in this normal way. Hateful disparaging anger from him tends to follow, rather than the sensible, amicable discussion of yours. Bear in mind that you invited ideas from anyone with a view to not adding extra complexities from your original post....sorry that my idea was faulty.

hurz

8 years ago

@2ctiby, your fantasy must be a horror. Read this thread again. You post two times a "brief variation" of a functioning circuit from roops and do not tell us what you have changed and why you think it must be changed. Nothing. Isn't it allowed to ask for a explanation from your side? Dont be a pussy, even you requested it in your last comment.

roops1967

8 years ago

New version: http://everycircuit.com/circuit/6753980066824192

roops1967

8 years ago

I appreciate your attempt 2ctiby, thanks

hurz

8 years ago

Hey @roops1967, here a brief variation. Its a delta sigma concept (Delta Sigma Modulator), but based on a single nibble 4bit path instead of 1bit which is the std for DSM. Oversampling is 1MHz and -3dB analog corner is 20kHz audio. The precision is close or even better then a flash 8bit parallel ADC. Check this, at least its more compact. http://everycircuit.com/circuit/5284494319026176

roops1967

8 years ago

Waaay over my head sorry, trying hard to keep up with you here

roops1967

8 years ago

Okay i understand you are converting the difference (delta) between each pass, the opamp is summing (sigma) it up to get the original signal. But I'm totally lost on the rest, really wish I understood this technical stuff :P

roops1967

8 years ago

But yeah great work! Why don't you post it and share with users who will understand the technicalities :)

roops1967

8 years ago

So you say this is more accurate than 8bit standard adc. I think i understand why cos the 4bits here is only converting the difference?

hurz

8 years ago

But you are very close to understand it completely. Here on everycircuit, are not many who can follow that. Right, its just the delta/error which is given back to an integrator, which sums up the delta.

roops1967

8 years ago

Well I try to make sense of what I see. But it's hard for me to make sense of numbers and formulas now that I'm getting older :P

roops1967

8 years ago

Can you break this down more in simple terms... 'Oversampling is 1MHz and -3dB analog corner is 20kHz audio'?? That's the part I really want to understand ;)

hurz

8 years ago

Ok, lets reduce the complexity to just 1bit http://everycircuit.com/circuit/5210504380547072

roops1967

8 years ago

Thanks hurz. I see your 1bit adc is struggling to follow the input, that because it needs a higher sampling rate?

hurz

8 years ago

Higher sampling and/or higher resolution (4 instead of 1bit) everything is scalable 😁

roops1967

8 years ago

Cool!

hurz

8 years ago

Right, we do not have to increase no of bits to get a better resolution. Its also valid to increase the samplerate to get better results.