# Power Management Unit: How to consider limit to connect electronic component to power rail

I am a bit confused about the way to dimension Power Management Unit, and power rail for electronic system.

I understand quite well LDO and DC-DC in terms of voltage, but i don't understand well about current (although I know kirchoff law). Particularly I am confused about the maximum total current that a source can deliver/supply to load in the case of battery powered one. Normally, I suppose that at equal voltage we should add up current to get total current needed, but how to put this in face of battery. Does Lipo battery got maximum current it can deliver to supply many loads ?

Let's say we power a system with a Lipo battery 3.7V-4.2V, 1000mAH.

The system got: - 1 MCU such as a MSP430 (http://www.ti.com/product/MSP430FR5969) - one auxiliary chip such as RF chip (let's say CC1310, datasheet here : http://www.ti.com/lit/ds/symlink/cc1310.pdf) - one module such as GPS (quectel L70, http://dl.roboticngo.com/uploads/Quectel-L70-Hardware-Design-V2-0.pdf ) - one another module such as GSM (quectel M95 http://www.soselectronic.com/a_info/resource/c/quectel/M95_HD_V1.0.pdf) - one EEPROM such as ST one - one TF Card such as SD CARD.

How do i verify this is achievable to supply all those chip and module with our Lipo battery and how do I build our power management architecture?

How do I verify this system can work properly and how would you organize the power rail in this situation and with which PMU ?

Thanks

If you add lots of bypass capacitors to ensure a smooth clean supply rail, you could use a single big buck-boost DC-DC regulator. Everything on your list will run on 3.3V, as the voltage of a lithium battery can be both above or below 3.3V depending on how charged it is, a buck-boost of at least 2A would be able to give you 3.3V under all reasonable conditions. Otherwise, use a buck or LDO to give you 2.8V for everything except the GSM module (2.8V is a flat battery so the battery voltage should always be higher than that under normal conditions) and use a 2A boost regulator to feed the GSM module, a fully charged li-Ion/LiPo has around 4.1-4.2V across it, if your boost regulator is set to that voltage, the GSM module will always get 4.1V or so.

• 3.3V is the min range of the GSM module. i am worry that going to 3.3V we are border line to minimum acceptance, if we have a very small drop, the GSM module will stop working... – chris May 15 '16 at 3:13
• Can you explain how you consider the limit of what the battery can supply in term of current ? – chris May 15 '16 at 3:25
• Unless it's a real dodgy GSM module, 3.3V is more likely the minimum recommended voltage as opposed to the absolute bare minimum, a few tenths of a volt are unlikly to be a major issue one way or another. For example, most modern CPUs are running around the 1V mark but even they can tollerate several tens of mV drop, besides, if you're using a decent buck-boost regulator and have lots of bypass and bulk capacitors, the GSM module's go9ng to be hard pressed to move the voltage away from 3.3V by any meaningful amount. – Sam May 16 '16 at 5:32
• Now with the battery, I was taking a punt as most half decent LiPo batteries can supply an awful lot of current (e.g. turnigy 6600mAh 90C cells can do 750A for a half a second). Even with most batteries if you're drawing around 1C (1xmAh capacity) worth of current or less you're usually fine, even rubbish lipos can manage at least 5C (i.e. 5A from a 1000mAh battery). Just be aware that if you pull ~2A from your battery, it'll go flat in ~ half an hour (1Ah/2A = 0.5h). – Sam May 16 '16 at 5:37
• Thank you very much for your feedback. The GSM module will be M35 or M95 from Quectel – chris May 18 '16 at 15:59