My Battery charger

[neptune]

Hello,

I am building myself a battery charger, with level indicator. I have built a block diagram for you guys to see. i hope this works, but haven't tried this.

what happens here is a Flip flop opens two transistors in a complimentary fashion. it opens transistor T1 for 9 sec. and T2 for 1 sec , so that we can sense how much battery is charged through same cable as used for charging it.
5 levels are for Indicating Voltage level from 1 to 5V.

my question is will it affect battery to charge in ON-OFF-ON-OFF... fashion , although the durations are large for ON (9 sec)

I'm looking for some insight

 

[MrAl]

Hi,

That will probably work but T1 has to have a higher base voltage than the expected emitter voltage.

Also, it would help to know the battery chemistry to make a better judgment.

 

[neptune]

Standard battery, Li-Ion 1500 mAh USB charged

 

[MrAl]

Hi,

Oh so the transistor T1 goes to a USB charger then?
I dont think you can measure the voltage of the battery back from the charger can you?
Try it...charge the battery a little then disconnect and see if you can measure back.
If not, you'll have to monitor the output of the battery itself. You might be able to multiplex that (that's what this technique is called).

 

[alec_t]

Charging lithium batteries safely is important if you want to avoid a fire! Read up about that before proceeding. Does your USB charger include proper charge control circuitry (it should if it's a commercial one)?

 

[audioguru]

A Lithium battery is charged to an absolute maximum voltage of 4.20V. But when its voltage reaches 4.20V it is only about 70% charged. More charging time is needed with the voltage at 4.20V until the charging current drops to a low current, then a current sensor shuts off the charger.

Read all about it at www.batteryuniversity.com .

 

[MrAl]

Hello,


The way i understood it he was feeding a charger that takes a 5v input and outputs to a Li-ion cell, so all the charge function design is already done and self contained and wont be messed with. Only the input 5v signal will be sometimes shut off. Neptune if this isnt correct let me know.
That's why i was wondering how he would get a reading back through the charger in reverse. One of the sampling transistors would have to be connected to the Li-ion cell too though so the input line could be multiplexed with the output (cell positive) line.

The other trick that is sometimes used is to charge a capacitor and use that capacitor to supply the Vcc power while the multiplexer is in the read state. In the charge state, the capacitor is recharged along with the normal (in this case charging) function.

 

[neptune]

Hi,

Oh so the transistor T1 goes to a USB charger then?
I dont think you can measure the voltage of the battery back from the charger can you?
Try it...charge the battery a little then disconnect and see if you can measure back.
If not, you'll have to monitor the output of the battery itself. You might be able to multiplex that (that's what this technique is called).

i connected an old micro usb to phone, as i feared there is no back current in it,
while there is current in Big USB

 

[twister]

edited

 

[bountyhunter]

A Lithium battery is charged to an absolute maximum voltage of 4.20V. But when its voltage reaches 4.20V it is only about 70% charged. More charging time is needed with the voltage at 4.20V until the charging current drops to a low current, then a current sensor shuts off the charger.

+1 That is what is commonly called CC-CV charge design. It is CC (constant current) when the cell voltage is below 4.2V, then when it reaches 4.2V the charger switches to CV (constant voltage) mode and tapers off the current. When the cell voltage is 4.2V and the charge current is very low (like less than 3% of C rate) the cell is charged.

 

[neptune]

Hello,
The way i understood it he was feeding a charger that takes a 5v input and outputs to a Li-ion cell, so all the charge function design is already done and self contained and wont be messed with. Only the input 5v signal will be sometimes shut off. Neptune if this isnt correct let me know.
That's why i was wondering how he would get a reading back through the charger in reverse. One of the sampling transistors would have to be connected to the Li-ion cell too though so the input line could be multiplexed with the output (cell positive) line.

The other trick that is sometimes used is to charge a capacitor and use that capacitor to supply the Vcc power while the multiplexer is in the read state. In the charge state, the capacitor is recharged along with the normal (in this case charging) function.

The phones which use mini USB architecture give back current and have phone charger that turns light red from green indicating full charge
while phones using micro USB A architecture dont give back currents.
now what should i do ? i wanted level indicator on my charger

 

[MrAl]

Hi,

Well then you'd have to monitor the cell voltage itself. However, monitoring the voltage probably wont give you what you want anyway because the voltage doesnt change much near the end of charge yet the cell is not charged yet. What chargers do is monitor the current.

When the cell voltage reaches the limit for that charger the charge current decreases or else shuts off. So if you measure the voltage of the cell itself you'd see the voltage drop a little.

To multiplex the voltages you'd have to do the input to the charger and the output of the cell, hopefully they share a common ground. That means you'd have to have a connection to the cell itself as well as the input to the charger.

 

[neptune]

To multiplex the voltages you'd have to do the input to the charger and the output of the cell, hopefully they share a common ground. That means you'd have to have a connection to the cell itself as well as the input to the charger.

hello,
According to the graph posted by AuDIO GuRu the Li-ion battery charge can be measured by both voltage and current.
at stage 1 we can use voltage and at stage 2 we can use current to see how much battery is charged !
what do you say ?
**broken link removed**

 

[audioguru]

hello,
According to the graph posted by AuDIO GuRu the Li-ion battery charge can be measured by both voltage and current.
at stage 1 we can use voltage and at stage 2 we can use current to see how much battery is charged !
what do you say ?

You do not need to measure the charging voltage because the charger takes care of it.
Instead you simply use a current-sensing circuit that detects when the charging current drops low or is disconnected by the charger. The current-sensing circuit can light a "fully charged" LED. It will work even if you try to charge a battery that is already charged.

 

[MrAl]

hello,
According to the graph posted by AuDIO GuRu the Li-ion battery charge can be measured by both voltage and current.
at stage 1 we can use voltage and at stage 2 we can use current to see how much battery is charged !
what do you say ?
**broken link removed**

Hi again,


If you are happy with measuring the current then by all means do that instead. You can measure the current into the charger and that will give you some indication of how much the cell is charged.
If you dont really care about voltage then measure current instead.

 

[neptune]

You do not need to measure the charging voltage because the charger takes care of it.
Instead you simply use a current-sensing circuit that detects when the charging current drops low or is disconnected by the charger. The current-sensing circuit can light a "fully charged" LED. It will work even if you try to charge a battery that is already charged.

I want 5 level indicators - indicating 20% , 40% , 60%, 80% , 100% of battery is charged , can it be done through measusrement of current

 

[audioguru]

In the charging graph I posted the voltage slowly rises and the current stays at maximum for one hour. How much is the cell charged? 50%? 60%?
Then the voltage rises only a small amount to maximum while the current stays at maximium for 1/3 hour. How much is the cell charged? 65%? 75%?
Then when the current drops to a low amout the charging is 100%.

Maybe you can use the voltage of the charging cell to guess at 20% and 40%. Note that the graph wrongly shows the voltage less than 3.2V.

 

[MrAl]

Hi,

Do you need those exact points?

The current is not linear with the state of charge. In fact, for a part of the charge process it is constant.
However, charge itself is current I times the time t, so you could measure current and how long it remains at that current and use that as a guide.
For example, if it charges at 1 amp for 1 minute, that's 1 ampere minute. Charging at 1 amp for 2 minutes is 2 ampere minutes. Charging for 0.5 amp for 2 minutes is also 1 ampere minute. So what you could do is measure the current every minute and sum all the measurements.
For example, say it charges at 1 amp for the first minute, that's 1 ampere minute, then 1 amp for the next minute, that's another ampere minute for a total of 2 ampere minutes. Now say the current drops to 0.9 amps for the next minute, that's 0.9 ampere minutes so the total now is 2.9 ampere minutes. So the charge is accumulating over time here.
If the cell rating is 2 ampere hours that's equivalent to 120 ampere minutes, so if you've only accumulated 2.9 ampere minutes the cell is only charged to 2.9/120 (2.4 percent) of its capacity. When the ampere minutes accumulate to 12 ampere minutes thats 12/120 or 10 percent, so when they accumulate to 24 ampere minutes that's 20 percent and that's your first target point (you said 20, 40, 60, 80, and 100). You have to wait for another 24 ampere minutes to accumulate before reaching 40 percent.

So all you do is measure the current once every minute and then sum the results and then divide by the cell rating in ampere minutes.

There is a catch here however, and that is that the cell has to be fully depleted in order for it to take a full 120 ampere minutes of charge. If it is not fully depleted, you'd have to compensate for that by starting the algorithm with a non zero charge accumulation. For example say it has 10 percent charge left, you would then start with 12 ampere minutes instead of zero ampere minutes and accumulate the same as before.

This brings up the next issue, estimating the state of charge. To do this with an Li-ion cell you measure the voltage and do a little formula and that gets you somewhat close to the initial state of charge assuming the cell had been sitting unloaded for at least 12 hours or so. You can look for the formula on the web, if you cant find it i'll check my personal notes.

 

[bountyhunter]

Maybe you can use the voltage of the charging cell to guess at 20% and 40%.

I did that for some NI-MH and NI-CD cells to make a "charge monitor" and it is possible but not very accurate. I did it by trickle charging the celss up to 100% charge, then discharging a known amount then measuring cell voltage and charge current. The problem I saw was that temperature affects it and as the cells got older the voltage points shifted. Cell voltage is not reliable for measuring state of charge.

 

[neptune]

Thankyou MrAI for that brief study, i will have to study more from net

So all you do is measure the current once every minute and then sum the results and then divide by the cell rating in ampere minutes.

what happens when cell ratings are different, eg - 1200mAh ,1500mAh

estimating the state of charge. To do this with an Li-ion cell you measure the voltage and do a little formula and that gets you somewhat close to the initial state of charge assuming the cell had been sitting unloaded for at least 12 hours or so. You can look for the formula on the web, if you cant find it i'll check my personal notes.

I am working on phone batteries that charge via USB , USB has set standard of 5V, so measuring voltage shall not be a problem.