Hello,

I'm trying to design a power system for a desk robot. The idea is that it's kind of like a small roomba that organizes your desk before returning to a charging station. My battery (placed in the robot) is the Ovonic 1550mAh 100C 14.8V Lipo.

I have 3 loads to power:

1. 5-7V VIN to ESP32. Right now the lipo connects to a LM2754-ADJ switching regulator which limits the ~16V input to 5.1V. The ESP32 is provided with this voltage and draws around 890mA of current.
2. Two 15-24V 0.4A max stepper motors.

Total Current: 0.4A+0.4A+.89=1.69A, which should be fine considering the 100C battery.

I'd like to charge this system with a 12V wireless qi charger (acting as the charging station) and boost the output voltage to 14.8V with a boost converter at the receiver. The lipo will constantly be attached to a corresponding BMS/charger PCB to make sure it charges correctly. But I'm pretty lost on how to set up the system considering the battery is not pass-through. Is there some way to set it up so that the battery only charges when the transmitter is below the receiver (aka the robot drives over the charging station), but otherwise it just powers the loads?

I've never done much with robotics or power supply so any advice is greatly appreciated.

 

It's technically possible, but it would be huge and quite expensive.
Wireless-Charging is terribly inefficient.

I would start trying to figure-out using Magnetic-Contacts rather than Wireless.
.
.
.

 

It's technically possible, but it would be huge and quite expensive.
Wireless-Charging is terribly inefficient.

I would start trying to figure-out using Magnetic-Contacts rather than Wireless.
.
.
.

What would make it huge and expensive? If I boost the output enough and add a relay to determine when it charges, would that not be enough for it to work?

 

If You don't want to believe me,
create your own experimental projects that will demonstrate what the probable outcome may be.

Start-out by making a list of your performance requirements ..........
How much Current will be acceptable, and at what Voltage ?

Current times Voltage, of course, equals Power,
and You must have at least ~40% more power than required by your initial calculations
because there will be inevitable losses in the Charging and Regulation Circuitry.

So how much total Power do You need to be transferred "wirelessly" ?, ( including all possible losses ).

Is terrible efficiency acceptable for this project ?

How much time is acceptable for completely charging the Battery from a completely discharged condition ?

How much space will be required to accommodate the Power-Transfer-Coils and their associated hardware ?

Magnetic-Electrical-Contacts are looking more practical, and much easier, every step of the way.

BTW, how is a Robot going to clean and organize the surface of a Desk ?????
.
.
.

 

Welcome to AAC.

Your biggest limitation will be heat. The losses in inductive charging are pretty high and that comes as heat. So you would need some active cooling, at least.

I thnk it is worthwhile to stop and reconsider the design based on your goals. The problem you have how to charge your robot, not how to do it wirelessly. A reassessment of the situation without the assumption that the charging will be wireless will be very productive.

Consider the Roomba, which you suggest is a model for your robot (at least close enough that you find it useful to compare with. What does it use for charging?

Consider the costs involved in both time and money. What percentage of your budget are you committing to this function? What problems will you have to solve that will not result in a working robot, only in overcoming obstacles in creating a wireless charging device?

Imagine that desk robots were already a thing, and you decided to create a product—wireless charging. How would you pitch it to the buyer? How could you justify the cost? Assume the robots already use the Roomba-like scheme of DC contacts, what benefits could justify the complicated and inefficient wireless charging add on?

Consider the cost of pogo pins, a 3D printed "funnel" to make alignment easier, and some software versus what you are proposing—how can you justify it? Don't let your knee jerk solution to a problem replace the problem itself—this is a common and corrosive pitfall in design.

Good luck.