• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 47 subscribers
  • Views 605 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Driving ADC from high impedance source?

Ralph Loader
Prodigy 70 points
Other Parts Discussed in Thread: ADS41B49, ADS62P49, ADS5483

Hi,

Is there any reason I shouldn't use a source impedance of several hundred ohms with a TI ADC?

The TI datasheets typically show ADCs being driven from a 50ohm or 100ohm source, but the actual input impedance (without external resistors) of the ADC input is a few k ohm (E.g. ADS41B49 input is 10k at DC, 3k at 200MHz).

The few pf of input capacitance will become more significant, so I'll have to be careful about that. Using a buffered device should eliminate input switching noise, so that should not be an issue.

My motivation is that my input signal is fairly weak. I could boost the signal voltage by step-up in an input transformer; this will increase the impedance seen on the ADC side of the transformer, hence my question about impedance.

Or would it be more sensible to just use a separate amplifier?

Cheers,

R.

  • 0
    TI__Mastermind 29220 points

    Hi,

    Our EVMs are usually set up with 1:1 ratio transformer coupled input, and from a 50 ohm single ended signal generator for the analog input the termination is often two 25 ohm resistors to the common mode voltage. 

    it is common to use a 1:4 ratio transformer for a nearly 'free' step up in voltage, and then the termination in this same setup would be 200 ohms differential.  The only caution i would point out for this is that some of our *unbuffered* data converters do not isolate the sampling caps from the outside world, and so they kick back sampling glitches back on to the analog input signal.  The typical input circuit shown in the data sheet for these devices often include an R-C-R filter external to the ADC to filter out some of this noise, and sometimes a caution to not use too large a termination to the common mode voltage.  For example, on page 58 of the ADS62P49 data sheet there is the caution:

     

    "It is also necessary to present low impedance (< 50

     

    Ω) for the common mode switching currents."

    The device you mentioned, the ADS41B49, has a buffered input and so this one caution would not apply.  A larger termination would be fine, and often the devices with buffered inputs are internally biased to the common mode voltage so the external termination does not need to be two resistors to VCM, it could be a single larger termination resistor.

    There is also an example circuit in the User's Guide for the ADS5483 EVM where we show an option of using an amplifier instead of the transformer coupling, and in that circuit i believe the load resistors are on the order of 200 ohms.

    Regards,

    Richard P.

     

     

  • 0
    TI__Intellectual 1960 points

    A little shameless self promotion.  Please see the following article on this topic.

    http://www.ecnmag.com/Articles/2010/11/Hardware-Corner/Buffered-High-Speed-ADC/

     

  • 0 in reply to Philip Pratt
    Prodigy 70 points

    Nice article, worthy of a little shameless self promotion :-)

  • 0 in reply to Richard Prentice
    Prodigy 70 points

    Excellent, that's exactly the answer I wanted :-) I'll try 800ohm and drop back if I get problems - with the buffered devices, it sounds like it should be OK as long as I pay attention to parasitic capacitance and the transformer performance.