  Forum Pages: 1 sigma delta adc concept (Read 655 times)
 saralandry Junior Member  Offline Posts: 17 sigma delta adc concept Nov 13th, 2020, 12:58pm   Hi,I have a couple of question and I hope someone can help me to understand a few concepts. Let�s assume that we have a second order sigma delta modulator similar to the block diagram below. In Richard Schreier books it is mentioned that �The c2 coefficient is unimportant since the quantizer is singlebit.�.Now my question is that how to implement the coefficient C2? Shall we simply remove C2 coefficient from the calculation and remove it from the following block diagram? If so, I think it is going to impact the STF of the modulator. Is that correct?My second question is that let�s say we synthesize NTF such that the modulator is stale for 0.9 of the full scale (for example if the reference voltage is 3-V then the input can swing from 0.15V to 2.85V. Now let�s assume that we want to input goes from 0 to 3V. There is a footnote in Schreier�s book but I do not understand it. I am quoting it here�A simple transformation u' =0.9u + 0.15 would allow input voltages u in [0,3V] while ensuring that the modulator receives an input u' which is within the stable input range. Such a transformation could be implemented by changing the input-sampling/reference-feedback network.My question is how to change the reference voltage to be able to support 0to3 V.thanks Back to top  IP Logged
 bernd2700 Community Member   Offline Posts: 34 Re: sigma delta adc concept Reply #1 - Jan 15th, 2021, 3:01am   Dear "saralandry",I am new to this forum since today (15.1.2021) and I have seen your question below. Did you solve the topic already?Coefficient "c2" is for single-bit, what you have, usually also named as "quantizer" gain (more precisely the "quantizer-dac" gain since "gain" is unit-less", often referred as to "kq" or "k", check Schreier). And this also has to be modelled if you want a correct model e.g. in Simulink. How to calculate it: Check Schreier, Yellow Book, p. 144.Of course a different kq will shift the STF, too. Make a simple Matlab feedback model, so do a la:STF = Forward / ( 1 + Forward * Return )whereas the "Forward" includes the multiplication of "kq".The full-scale, input referred, is also what is the full-scale of the DAC multiplied by any gain blocks. E.g. if the gain blocks from the feedback (=DAC) are 1 and also the input gain blocks are 1, and the DAC Full-Scale (FS) is e.g. 3V, then also your input-referred FS is _in the average_ also 3V. If there is an additional gain block from the feedback, e.g. factor 2, then your input-referred FS is 6V.Got it?Nice greetings,bernd2700 Back to top IP Logged Pages: 1