2S battery charger chip

[anda3243]

Hello Forum

I hope everyone is doing great.

I am looking for a chip similar to BQ25883. It is a 2S battery charging chip with built in balancing, with 5V input voltage.
https://www.ti.com/lit/ds/symlink/bq25883.pdf?ts=1654409288777&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FBQ25883
It would be perfect to use, but I have some issues with it.
It is quite pricey, and the availability is not great where I am, in China
I would prefer a bigger package and footprint. It would make it a tiny bit easier for the manufacturer. 5x5 mm footprint is OK.
I dont need that high current, I will max have 1A 5V input voltage.

I have spent some hours searching Mouser and the usual places, but with little luck. Most of the components I find are even smaller or more specialized, and they have BGA solder balls instead of pins.

Help is much apreciated

[mariush]

You'll have a hard time finding one because such charger IC would also need to have a built in step-up voltage regulator to boost your 5v to the voltage required to charge the batteries.
The charger needs at least 4.2v x 2 = 8.4v to charge the two batteries in series.

Probably the easiest would be to have a step-up regulator in front of your battery charger IC.

For example MCP73844 of MCP73842 can do 2 cell charging but needs input voltage between 8.7v and 12v
Here's the xxx4 version : https://www.digikey.com/en/products/detail/microchip-technology/MCP73844T-840I-MS/593753
Here's the xxx2 version (they make them in 2 versions, one that does 8.2v max battery voltage, one with 8.4v set, this one is 8.2v version) : https://www.digikey.com/en/products/detail/microchip-technology/MCP73842-820I-UN/593742
MCP73842 just has an extra 2 pins for a temperature sensor otherwise is pin compatible with the x4 version so the same 10pin footprint could have both.

So an option would be to boost 5v to around 9.2..9.5v (8.7v plus some margin to account for ripple and voltage drops on traces and temperature variations) and send that to the charger IC.

Another example would be MP2615 : https://www.monolithicpower.com/en/documentview/productdocument/index/version/2/document_type/datasheet/lang/en/sku/MP2615GQ-Z/document_id/1423
It has a built in step-down regulator to produce the voltage sent to batteries unlike the above Microchip IC which uses a linear regulator, so it can handle a much wider input voltage, it works from 8.75v all the way to 18v

Another option you may have as a work around to adding your own step-up regulator would be to implement charging only with QuickCharge compatible chargers and add an IC that signals the charger to switch to 9v or 12v when you want to charge.
You can also do it without an IC by setting some voltages on the data lines to some specific voltages (ex for 9v  you  need to have d+ in 0.325v...2v range and d- disconnected or tied to ground (below 0.325v) for at least 1.25v then you need to connect d- to that voltage between 0.325v ... 2v, that will signal the charger that you support QuickCharge 2.0 and more precisely the 9v level. For 12v, D+ must go above 2v when D- goes up.
 

[anda3243]

Thank you for a good answer!
I think I will go with MCP73844 and boost the voltage to 9.5 V from the USB 5V. 

What would be a good way to do cell balancing? It is only 2 cells. How critical is it to balance these?

I am a little surprised that there are not more chips for balancing and charging 2S battery cells.

[tooki]

I think that for a great many applications, it’s just as easy to use a 1S battery and then buck/boost the output as necessary. That has the added advantage that changes in the battery voltage don’t pass through to the load. And you eliminate the need for balancing altogether.

[mariush]

Yeah, I agree.
You could have the 2 batteries in parallel instead of series, and have the voltage regulator AFTER the batteries, configuring the voltage to whatever your circuit needs.

This way you can power the battery charger directly with 5v and charger IC can easily charge up the batteries to 4.2v.

There are even charger ICs with load sharing features or you can add your own load sharing feature easily, here's an example : https://blog.zakkemble.net/a-lithium-battery-charger-with-load-sharing/

Microchip also has an application note about load sharing which is quite detailed : http://ww1.microchip.com/downloads/en/AppNotes/01149c.pdf

[Peabody]

I agree with putting the two batteries in parallel, using a cheap-ass TP4056 charger module, and an MT3608 boost converter, to get your 5V.  Then you don't have to balance anything.  And you might even find that you don't really need 5V, and can use a buck regulator to get 3.3V, or if you really get lucky, you might be able to power your circuit directly from the batteries.

I think the reason there aren't more 2S modules available is because balancing adds a level of complexity that can be avoided most of the time by just using a 1S configuration with a boost converter.

Also ditto on load sharing if your circuit is 1S.  It just adds a mosfet, a schottky diode and a resistor, and can head off a safety issue depending on the project's current draw.

[anda3243]

Thank you for all the replies.
I am aware off all the advantages of a single battery design. The main reason going for 2 batteries is that we will need to drive a motor, we need 6V and peak current at 2.5A. The peak is very short, but we need to be able to provide that current.

Our single cell battery was barely able to deliver this current, and if we step up the voltage to 6 V, the current would be even higher.
My thinking is that 2 batteries in series will be able to deliver the current at the needed voltage in a better way.

The downside is that the charging circuit is more complex. The cell balancing being the main problem at the moment.

[mariush]

Shouldn't be a problem with the right boost/step-up dc-dc converter.

A standard 18650 lithium battery has around 2200-2500mAh capacity, and usually the recommended continuous discharge current is around 2A, but they can do more for short bursts.

In parallel the current pulled from batteries will be even so each battery will supply half the current.

Let's say peak 6v 2.5A - that's 15w. 
When fully charged at 4.2v , you'd have 15w / 4.2v = 3.5A  or 1.75A per cell.
When discharged down to 3v, you'd have 15w / 3v = 5A or 2.5A per cell.
The step-up regulator will not be 100% efficient, but can be above 90% so not a problem.

[anda3243]

You are making a very good argument.

However we also have some size constraints. so there is no way to fit a 18650 or more than max 1000 mAh. We are kinda pushing the limits on this.   

Shouldn't be a problem with the right boost/step-up dc-dc converter.

A standard 18650 lithium battery has around 2200-2500mAh capacity, and usually the recommended continuous discharge current is around 2A, but they can do more for short bursts.

In parallel the current pulled from batteries will be even so each battery will supply half the current.

Let's say peak 6v 2.5A - that's 15w. 
When fully charged at 4.2v , you'd have 15w / 4.2v = 3.5A  or 1.75A per cell.
When discharged down to 3v, you'd have 15w / 3v = 5A or 2.5A per cell.
The step-up regulator will not be 100% efficient, but can be above 90% so not a problem.

[wizard69]

If you are using two batteries in series of the same type as the single cell, how would you get more current?   You would get more power but that is not the same thing.

Thank you for all the replies.
I am aware off all the advantages of a single battery design. The main reason going for 2 batteries is that we will need to drive a motor, we need 6V and peak current at 2.5A. The peak is very short, but we need to be able to provide that current.

Our single cell battery was barely able to deliver this current, and if we step up the voltage to 6 V, the current would be even higher.
My thinking is that 2 batteries in series will be able to deliver the current at the needed voltage in a better way.

The downside is that the charging circuit is more complex. The cell balancing being the main problem at the moment.

[tooki]

They’re saying that the cell used can deliver the necessary current, but is at its limit. Boosting the voltage would require even more current. By not needing to boost the voltage, the current available is sufficient. At least, that’s their reasoning. I think there should be headroom for current, since the current capability of the cell when less than full will also be lower.

[anda3243]

Yes, this is the issue. We believe that 2 batteries in series will keep us on the right side of the right side of the current limitations. We spend the volume that a boost circuit would take up on batteries instead. And waste some of it on battery charging and balancing circuit.

They’re saying that the cell used can deliver the necessary current, but is at its limit. Boosting the voltage would require even more current. By not needing to boost the voltage, the current available is sufficient. At least, that’s their reasoning. I think there should be headroom for current, since the current capability of the cell when less than full will also be lower.

[anda3243]

Hello Everyone

Thank you for the previous answers!

I have finished what I think can be a decent charging and balancing circuit. Battery capacity is 400 + 400 mAh. Fast charging is desired. Any kind of suggestions for improvement would be great. I am restricted to chips that are available in China.

I use a HT7991 chip to boost the voltage from 5V to 9V
https://item.szlcsc.com/252182.html

I use MCP73844-840I/MS chip for charging. As suggested by mariush.
It has been difficult to find a good alternative to this chip. Its a good chip, but the availability is not great where I am. I have looked at TP5100 but I am not convinced. Has anyone used this chip? https://voltiq.ru/datasheets/TP5100-datashhet.pdf

I use 2 pcs HY2213 chips for balancing. These are commonly used in many of the shelf balancing circuits.

As mentioned, please let me know if I have made some mistakes or if there is anything that can be improved.

[anda3243]

Anyone has experience of the MP2639A chip? Seems this chip can make the design much easier.
https://www.monolithicpower.com/en/mp2639a.html
The bidirectional charging can be disabled.