I just had a new idea! Can I use the circuit from those old power banks I took the batteries from? It has a USB A output 5V 1000mAh, so I could use that to power my 7 segment displays and also feed it into an AMS1117-3.3V voltage regulator to supply all the 3.3V modules.
The power bank circuit should also handle over discharge and all other safety features for the Li-ion batteries. Also it has a USB micro input 5V 500mAh, so I can even recharge the battery if I want. But I still want to be able to swap the battery quickly.
I don’t know the exact model of my old power banks, but I used them for a while and never had any issues. Have like 5 of them.
There's a potential problem with power banks, which is that they shut down when load current falls below a certain level. But I suspect with all you have going on, that won't be a problem.
Just put the batteries back into the power banks, and then switch out power banks when one taps out. Then you could be recharging at the same time.
It would depend on the current it has to supply. But in general, the 1117 is an old design, with poor efficiency and high dropout voltage. I don't know what package you would want, but something like a Torex XC6222 will do 700mA with low dropout.
I did some research but couldn't find any Torex XC6222 modules, correct me if I'm wrong. It seems like I'd have to design my own circuit for it, which I'm not really a fan of and I don't think I have enough knowledge to do so. Looking at the AMS1117 again, I found some decent modules that provide around 800mA which is more than enough for my needs. Realistically how inefficient is the AMS1117 and how bad is its voltage dropout?
Based on my rough calculations in the worst case scenario I would need about max of 200-300mA for all the 3.3V circuits
I did some more research and discovered the MINI560 3.3V DC-DC step-down regulator. It should be more efficient than the AMS1117, but I also found that it produces some noise on its output.
“As a switching regulator, it produces some ripple and switching noise.” - chat gpt
Is this true, and how can I filter this noise? Also is added complexity worth the efficiency gain compared to ASM1117?
Opinions will differ on this. I usually just use a linear regulator. But it would depend on how much difference it makes. If the voltage drop across the regulator isn't very great, then a linear regulator may work well enough. But if it's a large differential, then you'd be wasting a lot of battery power generating heat. Another consideration is that switching regulators may lose their advantage when load current is low, so if your circuit sleeps a lot, a linear regulator could actually be more efficient.
If it were easy to filter out the noise, they would have done it on that module. But see the EEVblog video on capacitance multiplier.
Anyway, you'll have to do some number crunching. How much current will the regulator have to supply on average? What is the voltage drop across it, and how do those losses compare to the roughly 85% efficiency of a buck converter?
