2
\$\begingroup\$

I have about 100 cheap cell phones that need to be on 24 hours a day. Unfortunately (likely due to the cheap nature of these phones) the batteries end up bloating after a couple months of being connected to a charger continuously.

I’m exploring the idea of ditching the batteries and powering the phones off a large DC power supply connected to the phones battery terminals.

The batteries say they output 3.8 volts. But should I be concerned about the over amperage? If I connect a 3.8v 30a power supply to a single phone, is the phone going to fry?

\$\endgroup\$
5
  • 1
    \$\begingroup\$ if they're cheap, try it! \$\endgroup\$ Commented Aug 23, 2017 at 19:46
  • \$\begingroup\$ This isn't going to work as you don't have a battery but a battery pack. These packs have a charge monitor chip built-in so the phone knows the charge state (it also prevents overloading). You had to fake that chip to make the phone detect a connected battery. \$\endgroup\$
    – Janka
    Commented Aug 23, 2017 at 19:55
  • \$\begingroup\$ @Jasen The phones are cheap, the power supply, and my time are not. I'd rather ask questions and learn something.... As per the point of this forum. \$\endgroup\$
    – box maker
    Commented Aug 23, 2017 at 20:07
  • \$\begingroup\$ @Janka Powering a phone from a power supply will work. It's a common practice. And, a simple resistor can fake the chip, but it's no necessary for my use case. But my question was about amperage. Do you have any knowledge about that? \$\endgroup\$
    – box maker
    Commented Aug 23, 2017 at 20:14
  • 1
    \$\begingroup\$ The phone will draw the current it needs. The current rating of a power supply is the maximum it can supply. You want to be careful about accidental shorts, however. 30A is enough for welding. \$\endgroup\$
    – Janka
    Commented Aug 23, 2017 at 20:34

2 Answers 2

3
\$\begingroup\$

Do you need to transmit?

Used to do this at work, and I made several fake batteries and battery charge/discharge emulators at my old work.

Powering all phones can work as you say can work fine as long as you have a good 3.8V regulated supply. Here's some issues you will run into that needs to be addressed by your design.

1) If you're using GSM phones, GSM transmit can pull 2 Amps but for a very short time, about 1 millisecond. We used 1,200 uF electrolytic capacitors at the cables very close to the phone to prevent browning out during transmit. Another company I worked with made a fake battery blank that had 27 x 100 uF low ESR tantalum capacitors because they wanted it flat like the battery.

Even if you are not transmitting, it's a good idea to have some bulk capacitance at the phone, maybe at least 100 uF of low ESR tantalum or electrolytic caps (or ceramic if cost isn't an issue).

Validate that your power is OK by hooking up an oscilloscope and trigger for dips below say 3.4 Volts. Look at least at 1 ms/division or faster.

2) Use large enough wires. Calculate the loss based on the longest run at the maximum load. If you don't know the load, measure a phone doing various things (booting up, making call, etc.). Boot current can be high so make sure you test current doing boot.

3) The phone batteries typically have a pad that connects to a thermistor to GND to measure battery temperature. You need to emulate room temperature resistance or else the phone may not power on (or typically it powers on, sees invalid temperature and shuts down). This is typically the 3rd pad on the battery.

4) Some phone batteries have a phone pad which is internally connected to a resistor to GND. You would have this if your battery has 4 pads or connections. The resistor identifies the type of battery. You will also need to add this same resistance in your fake battery circuit.

A few phones have smart batteries that do digital signaling over this 4th ID line. Your cheap phones will not do this. If it did, all bets are off because the phone may need to authenticate the battery and shutdown if it cannot.

Another word of advice, even though you could possibly run 100 cheap phones off a 30 Amp supply, I would not do it for several reasons:

1) If all the phones happened to draw significant current at the same time, they could brown out the supply. That basically puts it down to 300 mA per phone, which is not a lot. Granted they're not smart phones, but there's peaks of CPU and radio activity that may take it well over this.

2) If someone accidentally changes the voltage, or the supply suffers a failure, you could overvoltage 100 phones and blow the whole batch. Happened at my old work once. Only like 10 phones were damaged but these were rare, brand new prototypes so management was not happy.

They real way we did it was to distribute 12V and we had individual 12V to 5V on the fake batteries or debug power system for each phone.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Bounty worthy answer. \$\endgroup\$
    – Passerby
    Commented Aug 23, 2017 at 21:18
1
\$\begingroup\$

I'm not sure how physically apart these devices are, but I would try to distribute a higher-voltage rail, and then down-convert where needed. If you can make logical "groups" of phones (like clusters of 4-8), you could get a simple DC/DC module for each that takes in 12V and kicks out 3.7V at a decent current level. With 12V rails (you can buy a cheap PC ATX supply and get 12V@30A very, very easily), the current flow in wires is at relatively low, "safe" levels.

You could also try a big 5V supply, and do a linear regulator down at each phone, but that's going to dissipate a good amount of heat at each regulator (maybe 1.2W or so at 1A).

Connecting a 3.8V, 30A supply to a single phone won't kill it, but you're going to have a really bad time if there's any short-circuit whatsoever; as indicated above, that's enough to spot-weld terminals on the spot.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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