MP2636 not enough

[Imperf3kt]

Hi. I'm working on making my Raspberry Pi 3B portable.
I bought an MP2636 lipo charger / booster module by DFRobot.
This seems to power my Pi well enough. I measure 4.92v at the GPIO.

I have a problem though. According to the datasheet, The MP2636 can be used as a UPS, and it seems to work as such. The problem though, is that since it is limited to 2.5A, as soon as I plug in a cable, power switches from boost, to charge as it should, and the Pi begins drawing from the passed through 5v coming from my power supply. However, even though my power supply is rated for 5A (and my cable is rated for 2.4A) the Pi complains of low voltage, and checking the GPIO shows 4.76v
I assume this is because charging the battery is taking most of the current? The datasheet says it charges at up to 2.5A.

If I connect a second power cable to my power supply (it has 5 USB ports which are all connected to the same rail) and then to the Pi, will this damage anything? Does anybody know enough about the MP2636 to say how the power is passed through?

Before anyone mentions it; Yes, I am 100% positive my power supply and cable can reliably power the Raspberry Pi (I just ran cpuburn-a53 with no problems, just to be absolutely sure **). It is only the addition of the MP2636 between, that causes the voltage drop.

** Strangely, it reports my Pi as 1250MHz, despite the fact I have not overclocked it. I have a Pi3B, not a B+

[drgeoff]

I assume this is because charging the battery is taking most of the current? The datasheet says it charges at up to 2.5A.

You can verify this by (temporarily?) setting the charge current to a low value. See page 31, "Charge Current Setting".

[Imperf3kt]

Thankyou, so just to be clear, I only need to change the resistor at RS1?

I am having trouble locating RS1 since the board is not silkscreened with this information. Am I correct in thinking RS1 is the large resistor just before the battery + output terminal, marked R020.

[drgeoff]

The datasheet clearly says that the charging current is determined by RS1 and Riset. The datasheet also tells you what pins of the IC those are connected to. Given that, you have no excuse for expecting me to go searching for how the module you have is physically laid out.

[Imperf3kt]

Yes, I appreciate your help in that matter (I'm not an expert at reading datasheets, but I did work that bit out)
The problem on my end is that the traces are hidden on the module by DFRobot, and I was wondering if you (or anybody else) knew more about it.
I appreciate the help so far, and I don't expect anyone to do the work for me, so if you don't know then that is fine. I'll try to track the traces with a multimeter.

Regarding my original question, if I use it in this configuration, should I put some kind of diode between Vout and the power supply?

[beacon1993]

Hi
have you worked it out?

[Imperf3kt]

Hi.
I did figure out how to change the charging current, but the results were more problematic.
After many failings, I gave up.


Recently, I came back to this though and made this two nights ago.
https://i.imgur.com/fJTlaJq.jpg

Sorry the picture is kind of useless.

In it, a 5A power supply runs directly to the MP2636, bypassing the micro USB input. The lipo also directly connects to the pads on the MP2636.
Additionally, wires run directly from the underside of the USB-A port on the MP2636, to the Pi.
I no longer power the Pi via the GPIO header pins, instead I have soldered directly to the Pi. 5v to PP2, ground to the underside of the ethernet port. KI chose this location because I removed my ethernet port, so it was easy to access.


So far, with an overclocked Pi3b (1.26GHz) running stress tests, I have not experienced any undervolt at all.
I do get undervolt issues when switching between the battery, and external power. This lasts for about 2 seconds, after which it goes away.
I am currently trying to solve this issue.


I hope this helps?

[quotaDestroier]

Hello man, I've seen your post, did you figure out how to power up your raspi? I have the same problem, but i brought the PB1000C and now i'm searching any other option that could give to me more than 1A. I can't find any other reliable alternatives. MP2636 is a good option, but i think that is not enough. Thanks for any further response.

[Imperf3kt]

I have actually, yeah.

As long as you don't want to power the Pi at the same time as recharge the batteries, you simply need to remove the USB port and solder some thick copper wire (I suggest 18AWG) between it and the input side of the polyfuse on your Pi (or the adjacent PP pad) and a ground point.

You'll require two physical switches or one switch and removable batteries as well. I can give more details on a few hours if needed.

[Oxbown]

Do you replaced de resistor? I think that it can be replaced with a 8 mOhm resistor and the charge intensity will be dropped to 1 Amp.
Check out this board:
https://www.tindie.com/products/ceech/l ... v-outputs/
Its similar to the DFROBOT board but more flexible.