display | more...
As the name implies the Analog to Digital Converter (A/D converter, ADC) is a device for converting an analog voltage into a digital format, so it can be used by a microprocessor or microcontroller system. There exists several different A/D converter types; flashconverters, Successive approximation converters, ramping converter and sigma/delta converters.

The flashconverter consists of several resistors connected in serial. One end of this resistor network is connected to ground, the other to a voltage reference. This network works as a voltage divider with many out-voltages. These voltages are compared with the input signal with comparators. When the signal voltage is higher on one of the comparators it will give out a digital '1'. The comparator outputs are sent to a decoder circuit that converts it to a binary number, that can be read by the ยต-processor.
Simple diagram of a flashconverter.
(A real converter can have tens or houndreds
of resistor/compapator steps)
     Ref
      |
      |
Sin --+--.
     | | |    __
     | | |   |  |
      |  *-\_|  |
      *--+-/ |  |
      |  |   |D |
     | | |   |e |
     | | |   |c |
      |  *-\_|o |
      *--+-/ |d |
      |  |   |e |
     | | |   |r |
     | | |   |  |
      |  *-\_|  |
      *--+-/ |  |
      |  |   |  |
     | | |   |  |
     | | `-\_|  |
      *----/ |__|
      |
     ---
     GND

   |
  | |  Resistor
  | |
   |

  -\_  Comparator
  -/
The flasconverter has a high sampling speed(converters with up to 1 Ghz exist but up to 100Mhz is the most common), but it's accuracy is low. These are used in digital oscilloscopes and spectrum analysers where the high speed is needed.

The successive approximation converter uses a comparator to compare the input signal to the output from a D/A converter. The controller circuit steps up the D/A output bit for bit. When the two voltages match, the comparators output switches to '1'. Since the D/A output voltage is the same as the input signal voltage, and the output from the D/A is known, the voltage of the signal is also known. This type of converter is slow but very accurate, and is used in high accuracy instruments.

The ramp converter charges a capacitor for a fixed time, then discharges it through a known resistance. Since the discharge time, the capacitor's capacitance and the resistor's resistance is known, the input voltage can be computed. This A/D converter type is somewhat less accurate than the Successive approximation type, and is also quite slow. But it is cheap and is the convertor used in most digital multimeters.

The Sigma/Delta or bit stream converter is a 1 bit converter used in CD players, mobile phones and ADSL modems. The output stream is a pulse width encoded representation of the input signal. A Sigma/Delta converter output bit width is increased by oversampling, but for each oversampling the bandwidth decreases. The Sigma/delta converter is cheap to manufacture and has a good linearity.

I love weekend mornings. Each Saturday and Sunday, I wander through my neighborhood on a winding path up and down residential streets, for dumpster diving at the school and strip mall. Starting early avoids questioning shoppers and suspicious business owners.

This morning, I noticed a flickering house light. The owners had installed a pair of lights next to their front door on the inner sides of a small alcove that led to the front door. The house faced south, so the pair of lights faced East and West. Only the light on the West wall was flickering, so I imagined that a poorly made CFL was the problem. However, there seemed to be something unusual about the flickering pattern, something beyond a corroded connection or failing ballast. Its flickering seemed to be slowing down as I made the long approach on this street. The sun was rising slowly behind me as I headed West, coloring homes in a soft reddish orange light. By the time I was perpendicular with the home's front door, the light was hardly on at all. Then I realized what was happening. The lamp was probably responding exactly as designed, to turn off automatically when the sun rose. It wasn't broken, but instead was responding to the slowly increasing light intensity.

This reminded me of a laboratory lesson in undergrad engineering school about digital sampling. Our goals were to compare theoretical and measured values of an 8-bit sampler. We needed to demonstrate a digital sampler converting an analog signal into 0's and 1's, given a specific sampling range:

bits = 2^8 = 256
range = 5v
increment per bit = 5 / 256 = 0.012 V per bit


/-------------\    /-------------\    /-------------\
|             |    |             |    |             |
|   generator |----|    ADC      |----|    scope    |
|             |    |             |    |             |
\-------------/    \-------------/    \-------------/
       |                                     |
       |                                     |
       \-------------------------------------/
 
While the generator was designed to produce as smooth of a change as possible, noise is unavoidable. Electromagnetic interference from the AC power supply, other lab equipment and maybe some powerful cellphones in students' pockets added to the otherwise smoothly undulating source wave. at each junction between one bit and another, there would be a flickering or irregular transition between the sampling levels. Our group was the only one to use a low-frequency input wave, and we noticed this phenomenon displayed on our scope. we tried to investigate the phenomenon further by setting the generator output instead to a constant value at a transition point that would ride right between two sampling levels, to show the effect of noise on the sampler. it was one of those fun little tricks we discovered by trying a different approach to open-ended questions asked in our lab experiments.
/-------------\    /-------------\    /-------------\
|             |    |             |    |             |
|  rising sun |----|  houselamp  |----| wandering   |
|             |    |      ADC    |    |  neighbor   |
\-------------/    \-------------/    \-------------/
       |                                     |
       |                                     |
       \-------------------------------------/

I thank our professor for making the lab experiments open ended, and not always narrowing the objectives too much. My partner also deserves thanks for his patience with my wandering mind. We played around in lab a lot, answering questions that the activity sheets never asked. Most of the time, this playful exploration would just have us spending more time in lab, and put us behind some of the more focused groups. however, like that day, sometimes it would reveal a convenient shortcut for answering several questions at once.


Aside from playing with technology and learning important engineering lessons, college was a time when I started questioning my values and beliefs. this was where the clear edges of science melted and the I'm-not-really-sure-if-I'm-confused mush started - sexuality, faith, trust and political bent. This is where the schematics and graph paper ran out and I was left with watercolors. What am I really looking for in life and why am I here in school? How much can I trust someone who appears to be my friend? How do I express sexuality? What and who will I vote for or against?

On a walk down the running trail towards the waterfront, I was deeply conflicted about my sexuality. Looking at my history of friendships, I noticed that most of them were other males. Someone looking at a list of my friends might think that I prefer the company of men, since I hadn't taken the initiative to befriend many women. I'd never had a girlfriend. Did that mean I'm gay?

My family belongs to a traditional church with very conservative beliefs. Their views on homosexuals were well defined, a mirror of what they heard in church. While my views were disorganized, I still felt more accepting of homosexuals, and felt some element in myself. How could I address this issue? Back to the ADC idea - maybe I'm just not intensely straight. In other words, a few bits away from the 'straight' end of the sampling range, and shy. I also tolerate and accept more than the people I met there. Would you ever catch me at a gay bar or a pride rally? Will you see a rainbow button on my backpack? Probably not. But what about befriending a openly gay coworker or neighbor? Or recognizing someone's choice? Thanks for the understated and generic anniversary cards, Hallmark.

Yes, call me a nerd. I started seeing ADC's everywhere - even in places where I couldn't plug them in. One week in a psychology class, we were learning about personality disorders. people with borderline personality disorder act very friendly and trusting towards friends. They show absolute trust, loyalty and approval. However, all of this can flip in an instant. Then, they suspect former friends of plotting against them, and turn to despise the people they trusted once before. Could this be another case of a saturated amplifier or lack of sufficient bits? Looked like the same patterns.

College is a great place for making new friends. It's also a time to learn sometimes painful lessons. Who can I trust? How much can I trust someone I've just started to get close to? Here was another case where I'd overstepped my bounds, and hit the rail so to speak. Now I saw trust too in light of a graded approach.

Those classes were the last time I designed or plugged in an ADC. Months or years will pass between the rare opportunities to put circuit analysis lessons to use. All the ADC's in my life are neatly packaged in black boxes, behind 'No user serviceable parts inside' stickers. Still, playing with the general concept helped me tackle tougher issues. The idea seems powerful to me now, since it applies to a few questions past engineering. Sometimes taking the long way home has its benefits, many thanks for opening my curiosity. Thank you too, local PTA - your expired Diet Coke tastes fine to me!

Log in or register to write something here or to contact authors.