Li-ion charger circuit help needed

[Integrate]

Hello. I'm working on making a device that I want powered by a Li-ion battery, like this one:
SparkFun Electronics - Polymer Lithium Ion Battery - 1000mAh

I want some sort of charging circuit for my device, and I'm thinking of using a MCP73837/8:
MCP73838

I would like to be able to charge using an ac-dc adapter for fast charging, but also have the ability to charge through usb safely. I would also like to charge from a 5V solar panel that is always connected to my device. The microchip above does the switching between the adapter and usb safely.

Here is my proposed circuit (also in attached pic in case imageshack doesn't work):
**broken link removed**

I have a very poor understanding of electronics, but here is my reasoning for the design:
D1 and D2 protect the adapter and solar cell. I plan to use 5V, 9V and 12V adapters.

The mess with the 6V zener and the 6V regulator and the 2 mosfets are intended to protect the microchip. It can take a max input of 6V, but I want to use 9V and 12V adapters too. I'm hoping that the circuit will operate as follows: At 5V input, the P-mosfet's gate is grounded, so the mosfet lets current through to the chip. The N-mosfet does the opposite. At higher than 6V, the zener lets current through, and the P-mosfet's gate is now high. The N-mosfet now lets current through to the 6V regulator, which now passes 6V to the chip. D3 protects the regulator, D4 protects the P-mosfet.


I do have a few questions:

1) Is it ok to have the Load attached to the Li-ion while it is charging?

2) I don't know if my circuit will work or not. Before I go and buy the parts, can anyone see any problems or any changes?

3) I don't know if D1, D2, D3 and D4 are really needed. Can someone confirm that they should be there?

4) The Li-ion is 3.7V, but my Load requires 3.3V. I plan to use a 3.3V regulator, but that requires an input of like 5V. Is there a way around this? Are there 3.3V regulators that accept voltages below 3.7V?

5) If there is no sunlight, and the usb and 5-12V sources are not connected, would the MCP73837/8 chip be ok with the Li-ion still attached to it?

6) I intend to set the charge current to be 1A in the MCP73837/8 chip, however the solar cell cannot deliver that much, only the adapter can. Would the MCP73837/8 be ok with something like 100mA input instead of the expected 1A? I'm hoping it'll just charge slower.

Thanks for your time!

 

[Integrate]

Bump. I particularly am interested in the answer sto questions 1, 4 and 6. Anyone know, or can point me in the right direction? Maybe someone else has done something like this in the past?

 

[mneary]

1) The load can confuse the charge termination circuitry. There might be a way around this. See Microchip AN1149.
4) They are called Low DropOut (LDO) regulators.
6) See Microchip AN1149. At a glance it would seem OK.

 

[nickagian]

Integrate,

regarding your question 1 : I am not sure, but I think that you should not charge the battery while the load is connected to it. One solution is to use an analog switch, which can be left open (not connected) while the battery is being charged. Another solution is that while the battery is being charged, the AC adapter or the USB adapter also provides power to the load. You can read this or this application notes to understand what I mean. Have also in mind that this function can be performed by some ICs, specially designed for this purpose.

For question 4, you can use a buck/boost regulator like TPS63031. Other manufacturers also have chips like this. This regulator receives an input voltage of 1.8V-5.0V and produces a stable 3.3V

For question 6: I have never used or studied the MCP73837/8, however I believe that as you say, if the input current is lower than the expected, then the charging time will be much bigger than before. But the IC hopefully will not have any problem operating.

Regarding question 5: There is no problem with that. However, have in mind that there will be a small discharge current that flows from the battery through MCP73837/8 to the ground. The datasheet says exactly how big this current will be.

Regarding question 3: D1/D2 form an analog 'or' gate that selects either the solar cell or the AC/DC adapter. If the solar cell is constantly connected to the circuit, then yes you need to keep these diodes. If you plug in only one of the two power sources simultaneously (AC/DC adapter or the solar cell) from the same input line then these diodes are not needed. You can also keep D3/D4 there.

 

[MrAl]

Hello there,


I dont see any problem with having a load on the Li-ion cell while it is being charged with some chargers, but it would depend on the charge method.
For example, with an analog charge circuit the circuit would be trying to pump out as much current as it can up to it's limit untill the output reaches close to 4.2 volts, after which it would cut back the allowed current output, maintaining the output at 4.2 volts. This means that if we loaded a 1 amp charger with say 10 ohms and the battery voltage was say 4.0 volts the load would get 400ma and the cell would get 600ma, which means it would charge up, but just a little slower than usual. If the load was near 1 amp, the cell would not charge at all. If the load was more than 1 amp, the cell would discharge.
If the cell was near full charge, like 4.20 volts, and if the load was 400ma then the charger would cut back the current to near 400ma, so the load would get the required 400ma and the cell would not charge anymore. If the load was near 1 amp the same. If the load was more than 1 amp, the cell would start to discharge.
None of these things seem to be a problem except of course the usual problem of discharging the cell too far down and ruining it. There would have to be some cutoff mechanism built into the circuit if this is going to happen on a regular basis. The cutoff would have to disconnect the cell from the load. This however is a problem that is independent of the charger and could be a problem even when not using a charger at all.

 

[Sceadwian]

You technically can't charge a lithium battery when it's connected to a load because it can't source and sink current at the same time =) If the power supply that's connected to your charger is up to the job though and the charge circuitry allows for it the DC power supply can power the device while the battery is charging. The first stage of lithium charging is a fixed current until the cell reaches 4.1 volts which the extra load should not affect anything as long as the charge control IC and power supply can supply it, but the second phase is constant voltage (4.1 volts) until the current drops to 2-10% of the lithium's amp hour capacity. If the load is being measured along with this current the lithium battery could be overcharged because if 10% of the lithium's AH capacity is the same current the load draws it will never trip the charger off. That's easily fixed by making sure that the sense resistor to the charge control circuitry is only passing current to and from the battery and that the wires from the load bypass the sense resistor to the charger IC, then everything should be fine. The actual charge voltage limit for the first stage and the current shutdown of the second stage are manufacture dependent, MrAl said 4.2, I would lean more towards 4.1. The %of the cells AH capacity that the second stage is terminated I've seen listed between 2 and 10% I'd aim for 10%, under utilizing the capacities of a lithium cell can make them last as long as they would on the shelf unused as long as the discharge is conservative and the dropout volts is kept high as well. All depends on your use and capacity needs.

 

[MrAl]

You technically can't charge a lithium battery when it's connected to a load because it can't source and sink current at the same time =) If the power supply that's connected to your charger is up to the job though and the charge circuitry allows for it the DC power supply can power the device while the battery is charging. The first stage of lithium charging is a fixed current until the cell reaches 4.1 volts which the extra load should not affect anything as long as the charge control IC and power supply can supply it, but the second phase is constant voltage (4.1 volts) until the current drops to 2-10% of the lithium's amp hour capacity. If the load is being measured along with this current the lithium battery could be overcharged because if 10% of the lithium's AH capacity is the same current the load draws it will never trip the charger off. That's easily fixed by making sure that the sense resistor to the charge control circuitry is only passing current to and from the battery and that the wires from the load bypass the sense resistor to the charger IC, then everything should be fine. The actual charge voltage limit for the first stage and the current shutdown of the second stage are manufacture dependent, MrAl said 4.2, I would lean more towards 4.1. The %of the cells AH capacity that the second stage is terminated I've seen listed between 2 and 10% I'd aim for 10%, under utilizing the capacities of a lithium cell can make them last as long as they would on the shelf unused as long as the discharge is conservative and the dropout volts is kept high as well. All depends on your use and capacity needs.

Hi there Scead,

You're joking around right? Obviously the cell can not both charge and discharge at the same time, but the chargers power supply will supply the cell and the load at the same time. I thought i gave plenty of possible cases where the load is smaller and bigger, and talked about the results.
Im sure a simulation would prove this.

Dont forget the small be ever present series resistance of the cell.

I agree however that setting the termination voltage a bit lower than 4.200 volts is a good idea due to meter variances, but when talking in general about these cells the termination voltage is often quoted as 4.2 volts rather than having to continuously go through the whole explanation of how it should actually be a little lower, like you forced me to do here

I'll do a simulation if you would really like to see it, but basically the charger is a current generator until either the max current is exceeded or the cut off voltage of 4.2v is reached (note again simplifying the voltage level for cut off). Up to that point it can always put out 1 amp (for a 1 amp charger of course).

Of course the exact workings depends on the load too.

 

[Sceadwian]

The power supply for many lithium powered devices is 5 volts, the charge control circuit itself should supply the regulated voltage. Feeding a 3.3v device stock 5V power might fry it without a regulator, in this type of case the charge control circuitry itself is actually the regulator, it's output is regulated to the maximum voltage allowed by the lithium cell in question. 1 volt makes a BIIIIG difference to a 3.3V circuit when you're already working at a peak charged cells .9volts over the chips rated voltage, .9 vs 1.9 over stock voltage is the boundary between a dead or working chip.

 

[mneary]

Earlier, I mentioned the Microchip AN1149.

As I read it more carefully, it appears to address all of the concerns that are being raised here. I recommend it highly.

 

[MrAl]

Hello,


That MC app note does show a load switch to switch the load off probably during a battery low voltage state. I was thinking this was a good idea too just in case the user tried to use the battery too long without a charge. On the other hand, if they were using the cell alone before building the charger then they must have encountered this probably long ago anyway and are dealing with it. Maybe the point there then is that they still have to deal with it unless they include a load disconnect circuit somehow.
Also note that for some battery chargers the load cant be connected while charging, as the app note points out too.

 

[Integrate]

Hello. Thanks a bunch for all the help. The application note really helped, I never noticed those things on the bottom of the microchip pages, but I'll have to keep an eye out for them from now on.

I am also looking into the buck boost regulators, and it also seems to be exactly what I need. Thanks for the tip on that.

I think I'm going to just go by the application note and ditch the solar cell for now. I don't know what would happen if the charger chip receives a tiny amount of current from the solar cell, and I don't want to be smelling any magic smoke from the li-ion. I think in the future, I'll add in a capacitor, and have the solar cell charge the capacitor until it reaches a certain voltage before letting the charger chip access the stored energy, something like the one shown on page 6 here with a few modifications:
https://www.electro-tech-online.com/custompdfs/2010/06/kss_solarspeeder2-docrevfeb102009.pdf

Thanks again for all the help!