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.

ADS1256 with Ref5025 powered by RaspberryPi 3B

Other Parts Discussed in Thread: ADS1256, REF5025, REF5020, ADS131A04EVM, REG71055-Q1

Hi,

After realizing using a 3.3v as a reference for the ADS1256 might be a problem, I'm trying to figure out how to connect the REF5025 and the ADS1256.  I would appreciate any suggestion regarding the circuit design for connecting them.

Also, I was wondering how does the unstable 5v output of the RaspberryPi could be used to supply power to the analog ADS1256. I'm using it to convert the output of an accelerometer LIS344ALH which has a 33uv/sqrt(Hz).

 Here is a schematics of what I have so far:

ads1256Project.pdf

Cheers,

Ran

  • Hi Ran,

    The connection from the REF5020 to the ADS1256 looks okay.

    • I would recommend removing C14, R1 and R2 (replacing R1 & R2 with zero Ohm resistors). The reason for this is to keep the reference source impedance low to drive the reference input pins; even 50 Ohms series resistors could cause an additional gain error. I would remove C14 to keep the output capacitive load within the recommended range of the REF5025. C15 is not an issue, but keep in mind that it may require additional time to settle after powering up the circuit.

    • The connection between VREFN and GND is perfect as drawn, since VREFN is routed back to the ground of the REF5025 (as opposed to connecting it to ground at the VREFN pin). When creating the layout, make sure to route VREFP and VREFN as a differential pair outto the REF5025.


     

    Regarding the 5V supply...

    I know USB power can be a bit unpredictable and that some USB supply sources are better than others. To handle wide supply tolerances you could could boost the USB voltage to ~5.5V and then use an LDO to supply a clean and stable 5V supply. In some cases that might be overkill.

    If you do use the Raspberry Pi's 5V supply directly, then I would just suggest using good supply bypassing technique on the ADS1256 and REF5025 supply pins. I would also caution the use of L1 and L2 on your supply lines, as they create LC resonators with the bypass caps (possibly doing more harm than good).

     

    Best Regards,
    Chris

  • Thanks Chris,

    • I removed C14 and replaced R1&R2 to 0 Ohm (why not just remove them completely?)
    • If I understand your comment on the VREFN, I should create a separate GND net for the VREFN and the REF5025 (along with other passive components). I thought of using one big GND copper area for the whole board, since you mentioned the AGND and DGND of the ADS1256 should not be separated.
    • Should I remove the L1 and L2? replace them with something else? I was not sure if I should use them in the first place.

    I'm redesigning the board I used with a 3.3v as VREFP. Its a similar design to what I've sent.

    The 5v AVDD has sometimes 60mv drops - especially when the raspberryPi reads from the ADS1256 - I guess more power is used by the PI and less is available for the ADS1256. I think your suggestion of trying to power it up with 5.5v and using an LDO to 5v might help. What should I be looking for in a LDO? any recommended noise limit? and particular component you can suggest?

    Cheers,

    Ran

  • Hi Ran,

    Ran Nof said:
    I removed C14 and replaced R1&R2 to 0 Ohm (why not just remove them completely?)

    Either way is fine. Sorry, I starting out making recommendations as if you had already built the PCB, then I switched mindsets and was providing a layout suggestion.

     

    Ran Nof said:
    If I understand your comment on the VREFN, I should create a separate GND net for the VREFN and the REF5025 (along with other passive components). I thought of using one big GND copper area for the whole board, since you mentioned the AGND and DGND of the ADS1256 should not be separated.

    Keep the one big ground plane, as that keeps the ground impedance low. However, since the ground plane will still have some finite impedance and that impedance allows for common-mode coupling of (noise) signals, reduce the common-mode impedance between the ADS1256 and the REF5025, by routing VREFN as a trace to the GND pin of the REF5025. VREFN still connects to ground, but in this case if noise couples into VREFN through the REF5025 ground pin it will likely couple as a common-mode signal into VOUT/VREFP as well, which is rejected by the ADC's VREFP/VREFN differential input pair.

     

    Ran Nof said:
    Should I remove the L1 and L2? replace them with something else? I was not sure if I should use them in the first place.

    I know it is sometimes common-practice to provide additional filtering or try to isolate a shared analog/digital supply, but generally we recommend just placing supply decoupling caps as close as possible to the ADC supply pins. This avoids possible issues with LC resonance, large voltage spikes from transient supply currents, and sometimes IC start-up issues due to the inductor limiting current to the device.

     

    Ran Nof said:
    The 5v AVDD has sometimes 60mv drops - especially when the raspberryPi reads from the ADS1256 - I guess more power is used by the PI and less is available for the ADS1256. I think your suggestion of trying to power it up with 5.5v and using an LDO to 5v might help. What should I be looking for in a LDO? any recommended noise limit? and particular component you can suggest?

    Good supply bypassing with some larger capacitor values may help to temporarily supply current to devices during load transients. If you decide to go with the boost converter + LDO solution, pick the boost converter and LDO together. You'll want the LDO to have a high PSRR rating near the switching frequency of the boost converter to filter out the switching noise. An LDO with low output noise overall is always good, but you don't need to get the lowest noise LDO you can find, as delta-sigma ADCs are generally fairly immune to supply noise.

    I would check out the following blog post:

    Also, you can see an example of a boost converter + LDO power supply on the ADS131A04EVM. Refer to the schematic near the end of the User's guide:

     

    Best Regards,
    Chris

  • Christopher Hall said:

    Hi Ran,

    Ran Nof
    I removed C14 and replaced R1&R2 to 0 Ohm (why not just remove them completely?)

    Either way is fine. Sorry, I starting out making recommendations as if you had already built the PCB, then I switched mindsets and was providing a layout suggestion.

     

    Yes, I understand. I do have a PCB built, but I now want to revise it since you note my attention to the VREF limits. I have very limited knowledge regarding electronics, PCB etc. so I really appreciate your help here. I wish to measure the LIS344ALH signal at mainly 0-10Hz (or at most <50Hz).

    Ran Nof
    If I understand your comment on the VREFN, I should create a separate GND net for the VREFN and the REF5025 (along with other passive components). I thought of using one big GND copper area for the whole board, since you mentioned the AGND and DGND of the ADS1256 should not be separated.

    Keep the one big ground plane, as that keeps the ground impedance low. However, since the ground plane will still have some finite impedance and that impedance allows for common-mode coupling of (noise) signals, reduce the common-mode impedance between the ADS1256 and the REF5025, by routing VREFN as a trace to the GND pin of the REF5025. VREFN still connects to ground, but in this case if noise couples into VREFN through the REF5025 ground pin it will likely couple as a common-mode signal into VOUT/VREFP as well, which is rejected by the ADC's VREFP/VREFN differential input pair.

    Should I put the bypass capacitors closer to the ADS1256 or to the REF5025?

     

    Ran Nof
    Should I remove the L1 and L2? replace them with something else? I was not sure if I should use them in the first place.

    I know it is sometimes common-practice to provide additional filtering or try to isolate a shared analog/digital supply, but generally we recommend just placing supply decoupling caps as close as possible to the ADC supply pins. This avoids possible issues with LC resonance, large voltage spikes from transient supply currents, and sometimes IC start-up issues due to the inductor limiting current to the device.

     

    I do see noise at 20Hz and 40Hz (I sample the ADS1256 at 100Hz, so it might also be aliasing) and have noise spikes (could also be a result of the 3.3v VREF). L1 & L2 will be removed. Thanks.

    Ran Nof
    The 5v AVDD has sometimes 60mv drops - especially when the raspberryPi reads from the ADS1256 - I guess more power is used by the PI and less is available for the ADS1256. I think your suggestion of trying to power it up with 5.5v and using an LDO to 5v might help. What should I be looking for in a LDO? any recommended noise limit? and particular component you can suggest?

    Good supply bypassing with some larger capacitor values may help to temporarily supply current to devices during load transients. If you decide to go with the boost converter + LDO solution, pick the boost converter and LDO together. You'll want the LDO to have a high PSRR rating near the switching frequency of the boost converter to filter out the switching noise. An LDO with low output noise overall is always good, but you don't need to get the lowest noise LDO you can find, as delta-sigma ADCs are generally fairly immune to supply noise.

    I'm not sure how stable the RasperryPi power supply is and what capacitor to overcome that. I think a safer way would be to use the booster+LDO. I found REG71055-Q1 but still need to check for LDO.

    • Since the booster is a switching component, should I put it on a separate ground area?
    • What about the crystal for the ADS1256, should that be separated from the ground plan? I use 2 layer PCB.
    • Also, I have several components that needs 3.3v (the MEMS sensor, opamp and the ADS1256 DVDD). What would be the best way to supply them from a single LDO (TPS76333)?

    I would check out the following blog post:
    Blog Post

    OK so power supplies are important – what more can I do? My previous post explained the impact that power supply variation and noise can have on analog-to-digital converter (ADC) performance. Thankfully, your data acquisition systems are not doomed...

    Also, you can see an example of a boost converter + LDO power supply on the ADS131A04EVM. Refer to the schematic near the end of the User's guide: See here

     

    You might wanted to refer me to: http://www.ti.com/lit/ug/sbau259/sbau259.pdf? I couldn't find the schematics in the file.

    Cheers,

    Ran