• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 47 subscribers
  • Views 976 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADS1258 overflow problem when AVSS different than DGND

Ciobanu Ion1
Prodigy 50 points

Other Parts Discussed in Thread: ADS1258-EP, ADS1258, OPA2365, ADS1282

Hello all,

I am new to this forum and to precision data converters, so I apologize if this issue has been addressed on this forum in the past.

I made a functional acquisition board around the ADS1258-EP ADC having a unipolar type of supply, but I have serious problems with the GND planes separation.

The current configuration on the power supplies and references for ADS1258 that works in my board is like this:

1. AVSS connected to DGND =GND of a battery AND to the GND of the computer.

2. AVDD=5V from a LT1763 voltage regulator.

3. DVDD= other supply of 5V

4. VREN,VREFP supplied as it is indicated in the datasheet using OPA2365 and ADR441 (2.50000 ) as reference voltage. 

5. EPAD (thermal PAD) is  NOT connected to the AVSS.

With this configuration everything works, I get a signal wave plotted on the GUI on computer, but because I am basically connecting the Ground planes (AVSS with DGND and with GND of the laptop) I get quite a nasty sinus like perturbation, the power supply interference. I thought that separating DGND from AVSS would reduce the noise a lot.

So, I tried to separate the DGND from AVSS, but I  get all the time the overflow error displayed, sometimes even a reset occurs. It also starts to scan from multiple channels even if I was plotting and even if I did set it to scan only 1 channel before. It is as if it enters in AUTO scan mode. If I reconnect the GND planes it would plot again the one channel I set it before.

So, I  tried when the ADS1258 is plotting correctly to disconnect the GNDs while it is working and then reconnect them again and the ADC "remembered" the settings (so the digital part is still working, I rarely get a reset on the digital part when I disconnect the GND planes, but  I have the overflow and the sudden multichannel scan errors).

This is the configuration that does NOT work, giving these errors:

1. AVSS connected to GND of a battery B1 separated from the DGND which is connected to the GND's laptop.

2. AVDD=5V from a LT1763 with it's input connected to the battery B1.

3. DVDD=5V from laptop. I also tried with another 5V from another battery B2 (with it's GND connected to the laptop). It still didn't work.

4. VREFN,VREFP supplied in the same way with the GNDs of the amp, ref. voltage,etc. connected like usual to the AVSS (B1's GND).

Can you please tell me where the things got wrong or what can I try? I would like to make a bipolar supply for the ADS1258, but if even the separation of the AVSS and DGND gives these errors I can't go further until I'm clear with this error.

Thank you very much for your time!

  • 0
    TI__Mastermind 44780 points

    Hi Ciobanu,

    The analog and digital grounds of the ADS1258 need to be at the same voltage potential. Internal to the ADC, the analog and digital circuits are NOT isolated. Isolating them externally will cause problems like you are seeing.

    So you may ask, "Why are there separate analog and digital supplies?".
    The reason is to partition the analog and digital return currents. (Notice I said "partition" instead of "isolate"). This helps prevent digital noise from coupling into the sensitive analog circuitry. It also allows you to set the digital logic level in some cases.

    It sounds like you have ground noise coupling into the analog circuitry. Typically, the PCB layout is the most common culprit when this happens. We usually recommend using a single solid ground plane for both the analog and digital circuitry, but partitioning the board such that you don't have digital return currents flowing in the analog portion.

    Regards,
    Chris

  • 0 in reply to Christopher Hall
    Prodigy 50 points

    Hi  Chris,

    Thank you very much for the answers! Now, that explains a looot of things.

    I looked on the ADS1282 evaluation board schematic and I didn't notice that DGND=AVSS so,  thanks for pointing this out. 

    But now, I'm not sure how the bipolar supply should be configured. Maybe like this:

    1. AVSS= - 2.5V . Here  I thought that a TPS72325 negative voltage regulator might be good, although I don't know if it is precise enough (it has 60uVrms and ~65dB PSSR, 2% initial accuracy).

    2. AVDD= +2.5V from a LT1763 (2.5V, 20uVrms, ~65dB PSSR). LT1763 supplied by the B1 battery.

    3. DGND=AVSS connected to the GND of other battery B2?.

    4. DVDD connected to a cheaper (and noiser) 5V regulator supplied by the battery B2.

    So, I'll keep the DGND=AVSS, but am I not making it to complicated with 2 batteries? Maybe I could supply from the same battery 2 separate voltage regulators (1 for AVDD and 1 for DVDD). But isn't it noiser this way?

    One last thing, I was thinking of using a Digital Isolator with Dielectric Insulation Voltage of 1kVrms (Isolation voltage). DO you think this is good enough to take out all the noise from the computer? I never used a Digital Isolator before so, I'm not sure it will solve the problem (maybe a digital isolator is inducing it's own noise into ADC?)

    Thanks,

    Ion

  • 0 in reply to Ciobanu Ion1
    Prodigy 50 points

    Hello all,

    In the ADS1258-EP datasheet,page 2,  is specified that the "AVSS to DGND" absolute maximum voltage can be between -2.8V and +0.3V. It is as if I can have AVSS= -2.5V and DGND= 0V. 

    In a bipolar supply is okay to just put AVSS and DGND to -2.5 as Chris suggested?

    Thank you all,

    Ion

  • 0 in reply to Ciobanu Ion1
    TI__Mastermind 44780 points

    Hi Ion,

    In my post above I was speaking generally when I said to keep the analog and digital grounds at the same potential.

    DGND is the digital ground, but AVSS is not necessarily the analog ground. For a +5V unipolar supply, AVSS would be the analog ground. However, for bipolar supplies AVSS would be at -2.5V (note we are referencing AVSS to the the analog ground at 0V). In both unipolar and bipolar supply cases, the analog ground reference voltage is at the same potential as the digital ground.

    Regarding your question above:

    Ciobanu Ion1 said:

    In the ADS1258-EP datasheet,page 2,  is specified that the "AVSS to DGND" absolute maximum voltage can be between -2.8V and +0.3V. It is as if I can have AVSS= -2.5V and DGND= 0V. 

    In a bipolar supply is okay to just put AVSS and DGND to -2.5 as Chris suggested?

    If (for example) you set DGND to -2.5V, and DVDD at 2.5V (assuming you use 5V analog and digital supplies), then your logic will be level shifted. A logic LOW will be at  -2.5V and a logic HIGH will be at 2.5V. You can do this, but you will most likely need to add additional components to shift the logic voltage back to 0 to 5V levels to be able to interface with your SPI controller.

    I picked 5V logic in my example above, but the same issue will be present for any logic level. Therefore, though it may require an additional supply, it is preferable to keep DGND at the analog/digital ground potential (0V).

  • 0 in reply to Christopher Hall
    Prodigy 50 points

    Thank yoy for the answers!

    A  Digital Isolator with Dielectric Insulation Voltage of 5kVrms like ISU7640FM, will be onough to solve both isolation and shifting of the voltage levels?

    Ion