2
\$\begingroup\$

Basic Explanation of my System:

The proposed design is a battery-powered wearable device that integrates advanced sensing and power management features. The system operates primarily from a single-cell Li-ion battery (3.7 V nominal) and includes a USB Type-C interface for charging and external power supply.

The device uses a battery charger with power-path management ( BQ24075) to ensure seamless operation:

When USB power is connected, the system is powered directly from the USB input while simultaneously charging the battery. When USB is disconnected, the system automatically switches to battery power without interruption. A fuel gauge IC monitors battery health parameters such as voltage, state of charge (SOC), and remaining capacity, reporting these to the microcontroller via I²C.

The system includes an electrochemical front end (AD5941) with a 16-bit ADC for high-precision sensor measurements.

Sensor signals are routed through a multiplexer (MUX) to the AD5941, which interfaces with an STM32 microcontroller development board for data acquisition and processing.

Power Tree: enter image description here

Power Supply Design:

I designed the below circuit using for powering the ADC and remaining circuitry using ADP2503. My simulation results and circuit diagram are given below.The ferrite bead P/N is MPZ1608S601ATA00. The simulation results are given below.

enter image description here

enter image description here

The voltage rails are respectively 3.2551427 V (AVDD) , 3.2995143 V (DVDD & IOVDD)

  1. Can I go ahead with this power supply design.Or do I need to go with Buck-Boost + LDO

  2. Could you please check my battery current consumption calculation.

  3. If I am using internal reference, will the ripple on the AVDD line is an issue?

\$\endgroup\$
11
  • 1
    \$\begingroup\$ What do the two 80 ohm load resistors represent? Why do you think an LDO regulator would be required? \$\endgroup\$ Commented yesterday
  • 1
    \$\begingroup\$ It looks like you are saying you estimate your load to be 18mA to account for margin, but I would indicate that on your diagram. Your switcher does not have 90% efficiency at low output currents, maybe between 75-80%. A good linear regulator might provide better efficiency over the discharge curve of a lithium battery. Your electrochemistry chip can work down to 2.8V, and has internal references - it seems like the precision needed on the 3.3V rail is low. \$\endgroup\$ Commented yesterday
  • 1
    \$\begingroup\$ What is the allowable lower limit of the battery voltage? I mean you propose using a buck-boost regulator so, it could be as low as maybe 2.8 volts but, this doesn't tally with your battery current calculation. Efficiency will be more like 50% at these levels depending on what mode you select. But could be up around 75%. In short, it's unclear what you are trying to calculate. \$\endgroup\$ Commented yesterday
  • 1
    \$\begingroup\$ @Confused you can used the adjustable version of the ADP2503 and set the output to be enough for the LDO regulator. \$\endgroup\$ Commented yesterday
  • 2
    \$\begingroup\$ There's not a lot of energy left in the battery when it's below 3.3V... but you could also just run your chip at 2.8V and probably be fine, right? Then if you feel compelled to add a switcher, then you could probably find a simpler/smaller/more efficient buck converter. Just going by gut instincts, the ADP2503 feels too big and too much overkill, and the power supply voltages seem unoptimized.... Running a ADP5304 at 2.8V output would get you better runtime. \$\endgroup\$ Commented yesterday

1 Answer 1

Reset to default
2
\$\begingroup\$

Focusing on the questions at hand:

Can I go ahead with this power supply design, or do I need to go with Buck-Boost + LDO?

I think the real answer is that it would probably work, but the buck-boost you have selected and your power plan in general feels... suboptimal. This buck-boost converter is much larger than you need, and you'll pay for that in lower efficiency at (relatively) light loads. Keep in mind, a linear regulator taking 4.2V down to 3.3V is 78% efficient (matching efficiency), and with most of the discharge curve living at 3.7V... it's 89% efficient.

The 3.3V operating condition is also a choice - I bet your analog electrochemistry chip has the highest VDD requirement. I think my preferences to your supply selection would be the following:

  1. 3.3V LDO - lose regulation of the last 5% supply voltage, high efficiency. No AC PSRR concerns
  2. 3.0V LDO - about 10% less efficient power supply, gets that last 5% under regulation. No AC PSRR concerns
  3. 2.8V small Buck converter - about 90% efficient over entire voltage range

The critical decision on LDO vs. switching is probably based on the power supply rejection ratio (PSRR) for your analog ICs. It looks like the AC PSRR is about -50 to -60 dB, depending on frequency. That is how much the power supply ripple will be attenuated. This would be what dictates the filtering and power supply topology selections for me.

Could you please check my battery current consumption calculation.

Without all the datasheets, this looks plausible. However, this is all based on quiescent currents. Some electrochemistry requires significant current to execute the measurement - none of that current will be in the datasheet.

A simple example is if you had a digital buffer gate datasheet running at 3.3V. If the datasheet says the quiescent current is 0.01mA, that wouldn't include anything about your application, such as driving a LED at 20mA. Knowing nothing about your setup, I can't say if that's an issue, just giving a warning.

You may need to build and measure your system to get accurate current requirements.

If I am using internal reference, will the ripple on the AVDD line is an issue?

Restating the earlier concern, you want to look at the Power Supply Rejection Ratio (PSRR) to see how much power supply signal will end up in your digitized signal.

For example, the high power on-chip voltage reference, which looks like it feeds into the ADC, has 60dB attenuation at 1kHz. There are frequency plots for AC PSRR as well. It's not entirely clear how much of that ends up in your digitized signal, or the bias generation, but worst case would be that it gets amplified at some point. I would try to minimize this ripple, and if I needed to run a switcher I would make sure that the frequencies line up with the higher attenuation frequencies of the AD5940.

\$\endgroup\$
2
  • \$\begingroup\$ I found another buck boost with good light load efficiency(ti.com/lit/ds/symlink/tps63900.pdf).May I know your thoughts about this \$\endgroup\$ Commented 18 hours ago
  • \$\begingroup\$ First, please understand that I didn't do a detailed or exhaustive device search. I found a part from the same manufacturer that you used. From an efficiency and digital operation, it looks fine, my concern would be the variable switching frequency - that might give you inconsistent power supply rejection if your load varies significantly. \$\endgroup\$ Commented 8 hours ago

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.