3
\$\begingroup\$

I want to use an ADS1219 to read a voltage and send the data to an Arduino board. Instead of using the IC alone, I am using the evaluation board BOOSTXL-ADS1219 so I can test the ADC easier. After reading both datasheet/manual several times, I still have some doubts:

  1. This ADC requires two independent power supplies: one for the analog part and another for the digital part. The thing is that ADS1219 datasheet specifies that voltage can go up to 7V, however from the datasheet diagram of the BOOSTXL-ADS1219 I understand that I need 3.3V for the digital part and 5V for the analog part. Is this ok? My guess is that this 3.3V is a must due to the dedicated port to connect to the PC. Anyways I will not use this port.

  2. I also don't get how to connect both power supplies to the BOOSTXL-ADS1219, because this evaluation board has printed voltage values (3.3V and 5V) and not AVDD/AGND and DVDD/DGND. Can anyone tell me how to connect the four cables?. I find weird that in one side of the board I have a 3.3V pin next o a 5V pin and one ground, and at the other side of the board there is only a ground. Which ground is the analog and wich the digital?.

enter image description here

  1. Finally, I am not sure if I need a logic level converter or not for the I2C communication. Arduino uses 5V for I2C, however reading the datasheet diagram of the BOOSTXL-ADS1219 it looks like the DVDD is 3.3V, so I would need a logic level converter to go up to the Arduino 5V I2C. Is this correct?. Anyways, I have purchased one just in case I need it.
\$\endgroup\$
7
  • 1
    \$\begingroup\$ Look at the Eval schematic in section 6.3 - AVdd goes directly to ADC pin 12; now what does the ADC datasheet say about pin 12? ADC datasheet section 6.1 Maximums says pin 12 can be <7V, while section 6.3 Recommended is <5.5V - so consider that a 5V power input. The answers are in these documents, just needs finding. Look at them again, more closely. \$\endgroup\$
    – rdtsc
    Jan 5 at 19:51
  • \$\begingroup\$ @rdtsc I really appreciate your answer! From now on I will have more confidence when reading datasheets/schematics. I am going to try 5V power input directly coming from the Arduino board which makes perfect sense for me. \$\endgroup\$
    – bardulia
    Jan 6 at 16:49
  • \$\begingroup\$ I have connected everything and my measurements sometimes have not expected changes. I think the cause is that both analog and digital grounds are connected on the evaluation board. I need to have both grounds totally separated. Can this be done easily by doing some modification on the evaluation board?. \$\endgroup\$
    – bardulia
    Jan 26 at 22:07
  • 1
    \$\begingroup\$ It doesn't look like it (image above shows header ground pins both connecting to ground plane.) So grounds are tied together here. Measurement error is usually caused by noise on the sensor or power rails. Sensor noise can be measured with a 'scope. Power noise is usually caused by "loop" currents and is a physical property of where the currents go - either conducted noise (digital pin switches --> creating a voltage perturbation, which reaches the analog section) or radiated noise (analog wires/traces too close/parallel to high-current/fast digital wires/traces.) \$\endgroup\$
    – rdtsc
    Jan 27 at 21:01
  • 1
    \$\begingroup\$ If still stuck on noise, suggest opening a new question with more details of what and how things are connected. \$\endgroup\$
    – rdtsc
    Jan 27 at 21:01

2 Answers 2

Reset to default
1
\$\begingroup\$

If you take a closer look at the schematic of the BOOSTXL-ADS1219 in the user's guide that you linked (page 20), you can see that there are only two chips that are connected to the 3.3V supply rail: The ADS1219 itself and an I2C EEPROM. I've looked up the datasheets of both parts and they can all operate on 5V.

So, as unorthodox as it might seem: You can simply apply 5V to the 3.3V supply pin of the BOOSTXL-ADS1219. It will operate just fine.

The board also only has a single ground; AGND and DGND are connected internally.

So, to sum it up:

  1. Connect the Arduino's 5V supply to the ADC board's 5V and 3.3V inputs (parallel them).
  2. Connect the Arduino's ground to the ADC board's ground (either of the pins; they're both the same).
  3. You don't need a level shifter if you power the ADC board from 5V only.
  4. Remove the jumper from JP1 to enable the board's internal 5V reference voltage supply.
  5. Use the jumper on JP2 instead to connect AVDD to 5V (position "1-2").

The only component that might get slightly unhappy about this "5V on the 3.3V rail" situation is the power indicator LED D2, but you can remove that LED from the board if you like.

\$\endgroup\$
6
  • \$\begingroup\$ I've done your steps and it works! But I have some comments. First of all, what do you mean by "parallel them"? Right now I am using a breadboard from the Arduino 5V power pin to the breaboard, and in that point I have put 2 wires connected in consecutive points that go each one to the ADC power supply inputs. Second of all, I am measuring an AAA battery and I am getting pretty good values, however the value is increasing along time: in one minute, it rises like 0.1mV (not acceptable as I need an accuracy around 0.1mV). I have checked this with a multimeter and it measures the same increase. \$\endgroup\$
    – bardulia
    Jan 6 at 18:29
  • 1
    \$\begingroup\$ What you've done - connecting the Arduino 5V pin to the ADC board's 5V and 3.3V inputs - is what I meant with "paralleling them". They're connected to the same voltage source, in parallel. You did that correctly. Second, about that voltage drift: What's drifting here is the battery voltage, not the ADC. You've already verified this with your multimeter, after all (if the ADC measures the same voltage as the multimeter, the ADC is likely accurate). A battery is not a voltage reference, its voltage will change a lot with even just tiny variations in temperature. \$\endgroup\$
    – Jonathan S.
    Jan 6 at 19:45
  • \$\begingroup\$ I have done more measurements and I can see that the drift slows down along time probably due to temperature as you have pointed out. Finally, I have a question about the ADC I2C address. I understand from the evaluation board that A0 and A1 address pins are set to DVDD (by checking eval board schematic), is this correct? There is a DNP text in there but I don't understand its meaning. I am using address 0x40 (0100 0000 in binary )in the Arduino code, so It would match if the DNP means open circuit because the A0/A1 pins would be set to GND. \$\endgroup\$
    – bardulia
    Jan 7 at 17:42
  • 1
    \$\begingroup\$ DNP means "Do Not Populate", so these resistors are absent from the board. Therefore, A0 and A1 are grounded via the other resistors (the ones without DNP). \$\endgroup\$
    – Jonathan S.
    Jan 8 at 19:57
  • 1
    \$\begingroup\$ No, this can't be done easily. If you need to separate the grounds, you have to design your own board. \$\endgroup\$
    – Jonathan S.
    Jan 26 at 22:31
1
\$\begingroup\$

  1. According to the datasheet (Section 6.3: "Recommended Operating Conditions") the manufacturer recommends a supply voltage anywhere between 2.3V and 5.5V for both supply voltages. The digital supply voltage usually depends on the interfacing device. If the interfacing device runs on 5V (and has 5V logic levels), it makes sense to supply the digital part of the ADC with 5V as well to ensure compatible logic levels. The analog supply usually is selected as high as practicable (5V in this case).

  2. The evaluation board appears to have only one common ground for digital and analog (see Figure 23 in the evaluation board's user's guide). An ADC having an AGND and a DGND pin is more about separation of the ADC's internal analog and digital sections. If the grounds would be tied together inside the chip, switching events in the digital section of the ADC would lift the ground level of the analog section especially due to the inductance of the bonding wires causing noise on the conversion result (sometimes called "ground bounce"). Having a separate analog ground inside the chip ensures that the ADC's analog ground is always close to the ground of the measured signal (i. e. the PCB's ground). This is particularly important with high resolution ADCs like this 24 Bit Delta-Sigma. On low resolution ADCs separate ground pins are often not needed. Connecting AGND and DGND on the PCB is usually no problem and in most cases even preferred over separated grounds (having separate analog and digital ground on PCB level is usually pointless).

  3. As mentioned in my first answer: Since Arduino is running at 5V, it would make sense to use 5V as digital supply for the ADC to ensure compatible logic levels. So you might want to connect Arduino's 5V to the 3.3V net on the evaluation board. However, you should double check that every component connected to the 3.3V net on the evaluation board is also compatible to 5V. I don't know about Arduino's logic level specification, but if Arduino safely interprets voltages >3V as "high" it should be ok to use 3.3V as digital supply, too. Just watch out when using logic level converters: Since I2C has a high-impedance bus state, you can't just use any arbitrary logic level conversion.

\$\endgroup\$
4
  • \$\begingroup\$ that makes sense! But I don't understand why the evaluation board uses a common ground. I say this because for my measurement I need at least an accuracy of 0.1mV. Could this common ground affect the measurement due to unexpected noise?. \$\endgroup\$
    – bardulia
    Jan 6 at 18:20
  • 1
    \$\begingroup\$ @bardulia I edited my answer to provide a little more information about separation of digital and analog ground. \$\endgroup\$
    – feynman
    Jan 6 at 19:12
  • \$\begingroup\$ Very interesting, thanks. In the following weeks I will connect this ADC to measure a different type of battery (redox) and I will compare the results with a commercial Battery Test System to see if I get a similar accuracy. I will come back here to post the results. \$\endgroup\$
    – bardulia
    Jan 7 at 17:05
  • \$\begingroup\$ I have connected everything and my measurements sometimes have not expected changes. I think the cause is that both analog and digital grounds are connected on the evaluation board. I need to have both grounds totally separated. Can this be done easily by doing some modification on the evaluation board?. \$\endgroup\$
    – bardulia
    Jan 26 at 22:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.