2
\$\begingroup\$

We want to keep a Xiaomi Mix smartphone powered on for a very long time (>6 months) with power from the USB cable. This is not recommended by the manufacturer since the battery will be constantly at 95-100% and topped up by the charging which is bad for the LiPo cells. We have already tested leaving the phone plugged in "stock configuration" and seen batteries swelling after about 4 months, which is not good.

I don't need any "battery functionality" since the phones will be plugged in with the USB cable the entire time, so one idea to prevent the swelling is to remove the LiPo cells (and keep the battery protection circuit plugged into the smartphone) and replace them with some equivalent circuitry. My question is what would that equivalent circuitry look like?

The battery I want to "emulate" is a BM3B manufactured by SCUD Electronics. It is a 3.85 V pack with 3.3 mAh.

My initial tries with just a resistor (R = 470 kOhm) and Capacitor (C = 10 uF) in series has not worked. My idea with that circuit was that the phones charging circuit would charge up the capacitor to ~3.85V and then measure that the battery was full, proceed to drain energy from the capacitor, discover that it was draining quickly and proceed to charge it again. But it did not work, the phone powered on for what looks like a microsecond and then dies.

How can I simulate a battery?

\$\endgroup\$
  • \$\begingroup\$ Provide a link to the battery datasheet. \$\endgroup\$ – Jack Creasey Jul 25 at 15:17
  • 1
    \$\begingroup\$ Consider providing power via the battery slot rather than the USB. You'll need a high current regulator with serious filter capacitance if you use something high current like the mobile radio. You may also need to figure out a way to trick the system into turning on, particularly if power is ever lost it may be reluctant to start back up without a manual button push. In many devices this is actually governed by scripts in the initramfs. Providing weak power on the USB connection as well might be enough to wake it. Using consumer devices in ways other than intended can be a challenge. \$\endgroup\$ – Chris Stratton Jul 25 at 15:31
  • \$\begingroup\$ Personally, I think emulating/bypassing the battery is a weird solution. I would just stick a timer on the charger and call it day. Turn off at 95% charge, turn on at 15%. You'd avoid hacking up the phone and it would still work in a power outage (bound to happen in 6 months). The battery would still degrade but it will last longer than 6 months. That said, I don't know the application. This would work nicely for one-phone in a project. Probably not as nicely if you're making a cell-phone farm and want to avoid replacing the batteries anyway. \$\endgroup\$ – C. Lange Jul 25 at 17:14
  • \$\begingroup\$ @JackCreasey I asked the supplier for the datasheet but have not received any replies yet, will link it as soon as I have it. \$\endgroup\$ – gruffaren Jul 26 at 7:47
  • 1
    \$\begingroup\$ Are smartphones not like laptops which will run when plugged in even if no battery is present? Because if you haven't tested that, maybe you should. \$\endgroup\$ – DKNguyen Jul 26 at 22:21
1
\$\begingroup\$

I'll put a formal answer on here, maybe it helps, maybe it doesn't. This isn't a way to simulate your battery, but it is a way to keep your phone plugged in and alive for a long period of time. A decent smartphone battery with proper charge cycles should last 2-3 years.

I personally don't like the idea of opening up smartphones. I generally don't like opening up any consumer hardware, because unless it is already broken, I usually break it. Additionally, you're never getting datasheets or schematics for that phone so any tinkering is all best guess.

My solution around this is to simulate the charge cycles.

  1. Easy method: a simple timer. One that plugs into a wall socket. Once your phone is fully charged move it over to the timer then the timer turns back on in roughly 12 hours. That time would need calibrating but this is a very crude setup that would extend the life of your phone. Your comment that this would drift and we wouldn't know how long it is not charging for is correct. I don't know what kind of battery life drain we would see in six months. This may not work if the phone won't be monitored for six months.

  2. Hard method: A more sophisticated way could be by using the power of IoT. Using an app like Tasker (maybe?) or even writing your own app, you could monitor the battery life of your phone. Once the battery gets too low the app-in-question could send out an email or an SMS. The email or SMS gets picked up by a service like IFTTT which then sends a command over the network to a smart plug for your home. This solution would have your phone always start charging at the exact same percentage and stop at the same high-level. It would also require much more setup and the internet.

  3. The same as above, just not as nifty: you download a battery alert app -- one that plays a loud sound when the battery is low. Build a small relay circuit with a microphone. When the sound goes off, the relay closes. Maybe it can play another sound when it's full as well.

My point here is that there are probably a dozen more ways in which you can extend the life of your phone. Hacking the battery charge circuit is one of them but consider some alternatives. If simulating the circuit is really what you need then don't accept this answer. Someone else may have more of an idea about what your smartphone is looking for with its battery connections.

\$\endgroup\$
  • 1
    \$\begingroup\$ Thanks for the answer, I really like the IoT "Hard method" which is what I would probably have used as a second option. I did however find a small work-around, which is not as elegant as I first thought, and I would still appreciate if someone knew what this circuit would look like, hence I will not accept the answer, like you said. But I can add some information so that others can possibly gain something: What I did was remove the LiPo cells only and measured the voltage at full charge, then I got a small DC/DC buck and attached that to the LiPO protection circuit, which worked! \$\endgroup\$ – gruffaren Jul 29 at 13:30
  • 1
    \$\begingroup\$ @gruffaren glad to hear it worked -- you should post an answer to your own question and accept it so that this question doesn't sit unanswered. \$\endgroup\$ – C. Lange Jul 29 at 14:37
  • \$\begingroup\$ I will leave it unanswered until someone takes a crack at the way I really want to build the circuit (i.e. passive components getting charged up by the battery protection circuit) or someone confirms with "electrical" terms why this is impossible to achieve. \$\endgroup\$ – gruffaren Jul 30 at 15:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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