Video Tutorial: Design and Characterization of Digital to Analog Converters

[Hugoneus]

In this episode, Shahriar presents the theory, design and characterization of Nyquist Digital to Analog Converters (DACs). After a brief overview of DAC operation and theory, the schematic of an 8-Bit R-2R DAC is presented. The R-2R DAC, which is driven by a dsPIC30F6014A Microchip microcontroller  is capable of producing ramps and arbitrary waveforms uploaded through an RS232 interface. The static integral non-linearity (INL) and differential non-linearity (DNL) is measured by using a Rigol DM3068 Multimeter through a Matlab interface program. The dynamic performance of the DAC is characterized using an Analog Devices AD6645 105MSps 14-Bit ADC evaluation board coupled with a USB FIFO interface board. By using the 'Visual Analog' software, the spectrum of the DAC output as well as the signal to quantization and distortion ratio (SQNDR) is calculated. Finally, the impact of component mismatches, operational amplifier non-linearity and timing uncertainty on the INL/DNL and SQNDR (ENOB) of the DAC is examined.

See the video here [51 Minutes]:
http://thesignalpath.com/blogs/2012/12/10/tutorial-on-the-theory-design-and-characterization-of-nyquist-digital-to-analog-converters/

More videos at:
http://www.TheSignalPath.com

[Mechatrommer]

this one video worth saving and waiting. you always kicking arse! thanks!

[Hugoneus]

this one video worth saving and waiting. you always kicking arse! thanks!

Hehe, Thanks. This one is rather technical and requires some background in DSP. There is a lot going on behind the scenes.

[BravoV]

Great learning video, definitely this kind of video that is worth downloading into my tablet as my learning material collections.

Thank you so much and just want to let you know I really appreciate the energy & time for preparation, simple thing like those tidy and nicely laid resistors at the breadboard, it shows you care for your audiences ! 

[EEVblog]

Awesome video. I can appreciate how much work goes into a video like that 
Is this setup going to be used for one of your classes?

Dave.

[firewalker]

A nice question for students. Why symmetrical +12, -12 volts to the buffer opamp? Isn't the DAC 0 to 5? Someone will ask anyway. 

Alexander.

[andyturk]

Brilliant presentation.

[Hugoneus]

Awesome video. I can appreciate how much work goes into a video like that 
Is this setup going to be used for one of your classes?

Dave.

Thank you Dave for your kind comment. This indeed was a very carefully constructed test. Even a single wrong DAC/ADC sample completely messes up the ENOB measurements.

This was a setup that I just put together for the video. I felt bad taking it apart! 

[Hugoneus]

A nice question for students. Why symmetrical +12, -12 volts to the buffer opamp? Isn't the DAC 0 to 5? Someone will ask anyway. 

Alexander.

Yes, you are right that +/-12V is highly excessive. A -1V to +6V would suffice (my OpAmp was not rail-to-rail). However, by using such large supplies the non-linearity of the OpAmp is completely eliminated from the equation.

[almoehi]

Thanks Shahriar, that was very interesting and I certainly learned something I didn't know until now.

 

Cheers,
Armin

[free_electron]

A couple of things that are important when doind those INL DNL plots :

- Make sure your meter is in FIXED range. never ever let it autorange. the problem is that , if it switches range, you get hit by the tolerance of the input scaler of the meter. this will be visible as a jump in the dnl/inl line.

for this 8 bit dac this may not be important but if you are going to do 16 bitters :

- turn of the autozero function of the meter. run one autozero at the beginning . at the end of the sweep :leave the dac at max setting , take a reading , autozero and take another reding. the delta between the two last readings gives you the zero drift of the meter. assuming constant ambient temperature this drift is linear. the number obtained can be used to adjust the slope. if you autozero every time you are throwing in 'noise' to the dnl inl.
- sync the entire setup to the powerline and measure at powercross.
- letting the meter do the oversampling is faster than doing this in software.

[Hugoneus]

Thank you free_electron for the extra information. Indeed measuring high-resolution DAC/ADCs is a tricky business. As a circuit designer, the last thing you would want to do is to design a good circuit and perform bad measurements.

[robrenz]

Excellent video, as always I learned a lot, thanks

[Hugoneus]

Excellent video, as always I learned a lot, thanks

You are welcome. I was not expecting people to be so much in favor of the technical information.

[pachuma]

Great video you've put together and a very nice test set-up. Thanks for sharing.

[Jad.z]

Hi Shahriar, you know how you mention at the end of the vid that next time is going to be less technical? well, please reconsider.
I mean, the internet is full of less technical blogs. It is ever so rare that you find such good ones like this was.

You are a very nice guy for investing so much passion and time into your v-blog. Great job 

[Hugoneus]

Hi Shahriar, you know how you mention at the end of the vid that next time is going to be less technical? well, please reconsider.
I mean, the internet is full of less technical blogs. It is ever so rare that you find such good ones like this was.

You are a very nice guy for investing so much passion and time into your v-blog. Great job 

Thank you, that is much appreciated. The videos are not rehearsed so I can't help but to drift into the more technical stuff.  O0

[oPossum]

Another excellent video! Thanks.

Doing the same characterization of 8 bit resolution PWM with first, second and fourth order filters would be very interesting. I often use PWM + filter to minimize cost and parts count.

[Hugoneus]

Another excellent video! Thanks.

Doing the same characterization of 8 bit resolution PWM with first, second and fourth order filters would be very interesting. I often use PWM + filter to minimize cost and parts count.

There is a little bit of PWM information available in two of my previous videos, including one way on how to generate them purely in analog using OpAmps.