Comparator, SHA, and DAC for SAR A/D

Discussion in 'The Projects Forum' started by magistrateee, May 15, 2008.

  1. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    I am making a SAR ADC and wondering if you guys can help me. I already have my SAR built and just need assistance with the others.

    For the sample and hold can I just use an NMOS as a switch and a capacitor as shown below? It will be switching with a clock at a maximum of f=200Khz but I can always cut it down if necessary to get this to work. The only other idea I have is using a t-gate. What ideal capacitor size should be used or how do you figure that out?
    [​IMG] T-Gate: [​IMG]

    For the DAC just like the comparator I do not plan to use a DAC chip. I am planning to use the R2R setup such as on this page: http://www.ikalogic.com/dac08.php. Would this exact setup be good enough? I know I can do a charged redistribution design but that would I figured the R2R is much easier.
    [​IMG]

    For the analog comparator I know I can use a comparator chip instead but I am trying to do it using CMOS, transistors (BJTs) or any other transistor based alternatives. The only idea I have with BJTs is using a differential amp then having it output to a switch shown below which is very basic. The only problem I think is the voltage swing. I think using CMOS would be easier than using BJTs.

    [​IMG]

    If you guys have any schematics I can take a look at or how I can make this I would really appreciate it.
     
  2. roddefig

    Active Member

    Apr 29, 2008
    149
    0
    Wow, neat stuff! I've used ADCs before but I've never looked at making one. However, your input stage looks very similar to the input stages of commercial ADCs. The sampling capacitor was always very small. In the ADC I am working with right now it is only several pF, however, its clock frequency is on the order of 100 MHz.

    The only limiting factor I can think of would be the RC time constant. I would make sure the capacitor has time to charge before your FET opens.

    I don't know too much about DACs, so I don't think I can be of any help there.

    I agree with you on the comparator. Maybe a Schmitt trigger on the output would help make the voltage a digital value (is that what you want?). Someone else on this forum can probably answer that question better.

    It looks like all those schematics are in SPICE, why don't you run them and see what happens?
     
  3. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
  4. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    roddefig thanks for the info. I will see if the Schmitt trigger will help. I ran some simulations without it and I believe I have to redesign it because its input voltage can only be at max around 1.5V. I will try the Schmitt trigger and try the small capacitor values as you have suggested.

    bertus thanks for the link. I will read it and see what I can get out of it.

    Thanks a lot for the help. If theres anything you can add anything else it would be greatly appreciated.
     
  5. roddefig

    Active Member

    Apr 29, 2008
    149
    0
    I'd be interested in seeing the complete ADC if you are able to post it, just for curiosity's sake.
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Go to Linear Technology's site:
    http://www.linear.com
    and download Application Note 13, High Speed Comparator Techniques.
    There are two ADC's presented; a 12-bit and a fast 10-bit using successive approximation.

    An interesting feature in the 10-bit SAR ADC is the clock speed-up at the 3rd sample.

    Something to keep in mind is your FET's gate-source and gate-drain capacitance. The smaller your sample cap is, the larger the effect your gate-source and gate-drain capacitance will have on the stored sample.
     
  7. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    roddefig, I will let you know as soon as I get it working. Currently I am working on the sample and hold. Tried running it with the NMOS and different cap values. I got some decent results but after a certain frequency it seems my scope doesn't like it. Hopefully I get this working soon. The sooner the better =p.

    SgtWookie, thanks for the info. I took a brief look at the site and can't seem to find where to download this. Do you have a direct link so I can get it or do I have to sign up?

    In the mean time I found this comparator in another forum and gonna see if I can understand it and get it to work like what I need.
    [​IMG]


    Once again thanks for the help.
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
  9. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    If you need help, I designed that comparator.
     
  10. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    SgtWookie, thanks for the link I found out where the info is. Still have not looked around the ANs and will try to get to it soon.

    Ron H, thanks for designing the comparator that I found =p. I appreciate the help. I have never used Schottky diodes before and I was just wondering if you were specifically using Schottky diodes. I am actually trying to recreate it and see it in action and I found the package of the diodes too small. I was wondering if using any other type of diodes work such as 1N4002 or if Schottky type of diodes are necessary? If the Schottky are necessary do you know of any others I can replace and use to breadboard it? Also if you could tell me what kind of voltage swing it has? Greatly appreciate it thanks.
     
    Last edited: May 17, 2008
  11. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    I ran some simulations. With little loss in response time, you can use 1N4148's instead of the high-speed Schottky diodes.
    The output voltage swing is zero to +5V volts.
     
  12. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    Ron, thanks a lot. I am gonna build and test it within the next day or so with other Schottky diodes for now. When I get hold of the 1n4148's I will try it again and see how it goes. Thanks again for the help.
     
  13. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Don't use rectifier Schottkys. They need to be small-signal, very low capacitance types.
     
  14. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    Ah I see. I just got hold of the diodes needed and gonna try it tomorrow and see how it goes. Thanks Ron.
     
  15. magistrateee

    Thread Starter New Member

    May 15, 2008
    7
    0
    Ron, thanks for the comparator. Only got to test it for a little bit. Did sin wave vs. DC voltage tests and got it working. Thanks a lot.

    I hooked up to the rest of my ADC w/o to see just how it would run got problems with my DAC output. I probably need to make more accurate with better resistors or buy a R2R ladder because at higher frequencies step (ladder?) sequence is distorted. Also probably gonna try different cap sizes with my SHA or use an alternative like a switched emitter follower or others and such.

    Still working on it. Thanks again all.
     
Loading...