Stabilizing Power Supply for Electrochemical Sensors

Thread Starter

DJ_AA

Joined Aug 6, 2021
305
I am currently working on a design that incorporates electrochemical sensors powered by analog devices such as OPAMPs. The voltage output is then read using an ADC. This part of the design is functioning well.

The design also includes a GSM module, which, according to the datasheet, has a maximum current requirement of 2A. Additionally, we are using a high-capacity battery, specifically a 4.2V, 8000mAh one.

I am seeking advice on how to stabilize the power supply to the sensor and implement any necessary filtering as a best practice. It's crucial that the power of the electrochemical sensor remains constant and stable for optimal results. While I haven't experienced any issues with the power supply yet, I would like to include some level of protection and stabilization as a precaution.

My goal is to prevent any current spikes or voltage drops in the power supply. I'm considering whether to add large capacitors and ferrite beads for this protection. Alternatively, I could use a large supercapacitor, which would act as a local battery for the sensors, but I'm concerned this might be overkill.

I would appreciate any suggestions on how to best approach this.
 

Papabravo

Joined Feb 24, 2006
21,225
IMHO, powering a device with an operational amplifier is a dubious proposition. I cannot tell from your schematic if that is indeed what you are doing but I strongly suggest you consider alternatives.
 

loose_electron

Joined Jan 16, 2024
19
As said by others, using an op-amp to power anything is probably not the best idea. Look into using a bipolar voltage regulator instead. (Bipolar instead of CMOS to keep the flicker noise low.)

Other ideas:

Anyplace you create a DC reference voltage (resistor dividers) locally stabilize the voltage with a 0.1 uF (suggested) filter capacitor to keep high frequency noise minimized on the DC reference voltage.

Using those 1M resistors typically leads to noisy and unpredictable behavior sometimes. Such a large resistor value can be sensitive to dust-humidity-etc. If you can, lower the circuit impedance for those nodes.

c01f14.png
 

Papabravo

Joined Feb 24, 2006
21,225
I was following the reference design of the manufacturer's PCB design.
THAT does not answer the question. Manufacturer's do stupid things all the time. You have to ask yourself why they would do that. It is contrary to basic principles. You're not going to be effective as an engineer if you go through your career as a copycat.
 

sghioto

Joined Dec 31, 2017
5,390
I am seeking advice on how to stabilize the power supply to the sensor and implement any necessary filtering as a best practice.
I think you already have done so with the U6 voltage reference. You said yourself it hasn't been an issue so far.
 

Thread Starter

DJ_AA

Joined Aug 6, 2021
305
I think you already have done so with the U6 voltage reference. You said yourself it hasn't been an issue so far.
No problem so far, but we are looking at possible power filtering or stabilising circuit practices. This is to ensure if a power dip arises, the circuit is stable.
 

Thread Starter

DJ_AA

Joined Aug 6, 2021
305
Is the entire circuit and GSM module operating on the 4.2 volt battery?
Yes it is.
There is the possibility that the module could draw up 2A of current, which could cause some voltage drop, so this is the concern for this circuit.
 
One simpler suggestion - You got an ADC someplace in the system. If you can mux the ADC, why don't you just monitor the battery power? If the power supply goes out of bounds, just throw the sensor data away, and re-measure. In your controller, make sensor data contingent upon good power before and after the sensor sampling.

Electrochemical sensors are usually fairly slow things, so a re-sample strategy should suffice in most situations.
 

Papabravo

Joined Feb 24, 2006
21,225
One simpler suggestion - You got an ADC someplace in the system. If you can mux the ADC, why don't you just monitor the battery power? If the power supply goes out of bounds, just throw the sensor data away, and re-measure. In your controller, make sensor data contingent upon good power before and after the sensor sampling.

Electrochemical sensors are usually fairly slow things, so a re-sample strategy should suffice in most situations.
This is an excellent suggestion. In addition, I would visit Battery University Homepage to gather information on the discharge characteristics of various battery types so you can implement your battery verification protocol with some degree of confidence.
 
This is an excellent suggestion. In addition, I would visit Battery University Homepage to gather information on the discharge characteristics of various battery types so you can implement your battery verification protocol with some degree of confidence.
Yeah the BU page has a lot of good information on it. Be careful there's some stuff there that's a little out of date, but overall it's a good information source.

Better yet (yeah, I'm biased) read: Chapter 5 - Battery Power, of my book which has about 40 pages on how to design battery packs, charging systems, etc, and goes into the nitty gritty of different battery chemistry, discharge curves, how battery capacity changes with temperature, discharge rates and all that.

Shameless plug, I will admit.

- Jerry
 

Thread Starter

DJ_AA

Joined Aug 6, 2021
305
Thank You for the suggestion to do this via ADC, i can easily monitor the battery voltage. In addition, i can place some Capactors few 100uFm prior the sensors PCB.
 
Top