AD5726 major code transition glitch impulse

Please refer to this document:

http://www.analog.com/media/en/technical-documentation/data-sheets/AD5726.pdf 

1. Table 3, bottom line (page 5) states that the glitch is typically 30 nV*s, under very different conditions:

AVDD = +5 V ± 5% or +15 V ± 5%, AVSS = −5 V ± 5% or 0 V or −15 V ± 5%, GND = 0 V, VREFP = +2.5 V or +10 V, VREFN = −2.5 V or 0 V or −10 V, RLOAD = 2 kΩ.

That is, the conditions defined for the Table 3 indicate that the glitch should be independent on power supply and reference voltages.

On the other hand, Fig. 23 (page 11) indicates that the glitch does depend on the power supply voltage. 

What is the actual DAC's behavior? 

2. On Fig. 23, the area under the positive part of the largest glitch (about 0.9 V peak voltage) exceeds 100 nV*ns - once again, in contradiction with Table 3. Isn't that a mistake? Can the glitch peak really reach almost 1V?

3. When you define the glitch magnitude in nV*s, do you mean to the formal integral definition (area under positive part of the glitch minus area under its negative part), or to the largest of these two areas?

Thanks in advance,

Michael

  • 0
    •  Analog Employees 
    on Oct 25, 2016 1:08 AM over 4 years ago

    Hi Michael,

       I am checking on this. Let me get back to you soon.

    Regards,

    Jonathan

  • 0
    •  Analog Employees 
    on Oct 26, 2016 11:36 PM over 4 years ago

    Hi Michael,

       As shown in table 3 of the AD5726 datasheet, the AC performance characteristics uses different supply voltages. The major code transition Glitch impulse is specified at a typical value, not the maximum or minimum. This means that, typically the area of the glitch would be around 30nV-sec.

       The figure 23 of the datasheet is a much detailed condition with +/-15V and +/-5V supply at transition from 0x7FF to 0x800 and 0x800 to 0x7FF. From this figure, we can see the max and min peak values of the major code transition glitch.

       The major code transition Glitch impulse is the net impulse glitch area.

       

    Regards,

    Jonathan

  • Hi Jonatan,

    thank you for your explanations.

    About the glitch definition - what do you mean by "impulse glitch area"? The sum of areas under positive and negative parts, or their difference, as in the standard integral definition? If a glitch is a one-period sine pulse, then what its area should be - twice the area under a half-sine curves, or just zero?

    Thanks in advance,

    Michael