Help with current limit circuit desing for charger

[Doctorbass]

Bob, Fechter or any other that could help me to desing a current limit circuit for my DC-DC modules charger PLEASE HELP!!

Here is the situation:

-The DC-DC modules that i have can give between to 170A in short circuit and they are built to give stable and continuous 30A at 5V
-They have a internal current limit protection.. but if i use directly these modules on a fully discharged A123 M1 cell it will give the 30A and up to the cell and damadge the cell.
-I will use a 0.01ohm resistor in serie with each indemendent DC-DC output to monitor the current.

What i need is to limit the current to 15A wich is safe is i use Xs 2p or more pack.. that could give 7.5A per cell .
Now 15A would give me 150mV on the shunt resistor to say to the circuit that the 15A limit is acheived.

What i know is that i can control the output voltage using a resistor or potentiometer connected between the "trim" pin and the (-)out
And also i can control the output voltage with the sense(+) and sense(-) feedback pin.

These feedback sense pins are used to compensate for the voltage loss in the wire of the output so if the setpoint of the trim resistor is let say to 3.6V and that the sense pin read 3.5V at the end of the wire (on the load) the internal circuit will overboost the output to acheive the 3.6V at the load where the sense feedback are connected

What i think is that if i try to boost the voltage on the sense feedback pin with few milivolts higher than the preset voltage, the output voltage pin will decrease to compensate because the sense read more than the preset. So by using this method, i think i could control the output current by playing with the voltage control pin like "sense" or "trim" and limit it by this way.

Now what i wonder, is what could look like a circuit that will take the few 150milivolt of the shunt and stop the current to 15A max by using the trim or sense input pin to control this current?

Could it be a simple circuit without the need of using comparator or amp-op.. i would need to built 12 of those.. :lol:

Doc

I need CC-CV of 15A 3.6V...

 

[fechter]

That should be pretty easy with an op amp.
I don't have any schematic drawing software here, so I can't whip one up too easy right now.

You would use an op amp as a comparator and compare the voltage drop across the shunt with an adjustable reference. Once the reference voltage is exceeded, the output of the op amp goes high and pushes the trim input. As long as the trim has enough range to bring the current down to the desired level, it will operate as a constant current source until the current drops below the set point.

Do you know the voltage range of the trim input? Is there a datasheet for the dc-dc converter?

I'll have to scratch my head a while to see if there's an easier way to do it without an op amp. After all, there has to be one inside the converter already, so it might be possible to make use of it.

 

[monster]

light bulbs work well for limiting current i used that in making my battery charger

 

[Dr. Shock]

Monster, please explain a little more about why lightbulbs work, how to size, etc. Thanks, Grant

 

[Doctorbass]

I've been thinking about 12V lightbulbs but there is a problem with those.

What i need is to limit the current to a pre-determined value. that mean that if i want to have CC-CV (constant current and constant voltage mode) a lightbulb will not limit the keep the current to the max i want when it will reach the 3.6V.

let me explain:

Lets say that i want 10A max current to charge the cells.

That mean that i will charge the cell to 10A at state until it reach the 3.6V top voltage. And if the cell have 2.5V and is fully discharged or if the cell is at 3.55V, i need a stable charge current of 10A.

That mean the charge current should be independent to the load (the cell) until it reach 3.6V)

A lightbulb as a resistor in serie would decrease the current when approaching the 3.6V. so with that solution, the current would be maximum only when the cell will be fully discharged...


see the drawing i did to explain that problem:

 

[Doctorbass]

See what it happen with both case in the image below.

That's the reason why i want to use CC-CV charging method.The charger will begin the charge using a maximum preseted and limited current as long as the 3.6V of the A123 cell is not reached. When the cell will reach 3.6V, the constant voltage mode will keep it to 3.6V and will control the current to maintain 3.6V. Durint this last stage, the current will decrease progressively to finish the last 5% of capacity of the cell. When the current will reach 100mA the charge is completed.


Now, what i want is to sense the current using a shunt in serie and to control the output current. I know that on these DC-DC the current is not controlled but only protected against thermal or short curcuit. In short curcuit, it can reach 170A.. and i really dont want that.

To control the current, i will need an external shunt in serie and i will need to use the few millivolt that it produce to control the TRIM or the SENSE feedback input of the module.

The problem using comparator is that i will need an external power supply to supply these circuit FOR EACH DC-DC MODULES CONNECTED IN SERIE... FOR A TOTAL OF 12 !!

the TRIM and SENSE input are referenced to the output and each are connected in serie... that'S the problem...

maybe to use some opto-isolator... well..


Doc

 

[Doctorbass]

That should be pretty easy with an op amp.
I don't have any schematic drawing software here, so I can't whip one up too easy right now.

You would use an op amp as a comparator and compare the voltage drop across the shunt with an adjustable reference. Once the reference voltage is exceeded, the output of the op amp goes high and pushes the trim input. As long as the trim has enough range to bring the current down to the desired level, it will operate as a constant current source until the current drops below the set point.

Do you know the voltage range of the trim input? Is there a datasheet for the dc-dc converter?

I'll have to scratch my head a while to see if there's an easier way to do it without an op amp. After all, there has to be one inside the converter already, so it might be possible to make use of it.

Fechter, here is the datasheet:

http://www.sierraic.info/datasheets/ds-jw_a.pdf

What do you think about using the trim input and to use voltage instead of a resistor to control this input?

I mean, the trim need to be connected to the (sense -) input via a serie resistor. so in this case, the control is a current that goes to the trim input right?

The shunt i would like to use have 0.001ohm (from digikey) I want to choose very low ohm to avoid loosing power in heat.

if i charge to 20A, the power that it would dissipate would be 0.4W and the voltage would be 20mV

That would be ideal for me if i could add this 20mV to the 3.6V sense input to give 3.620V ... if comparated with the preseted 3.600V with the trim input, that would automaticly decrease the voltage at the output and decreasing the current ubtil the shunt milivolt + the 3.600V goes back to 3.600V threshold...

Doc

 

[fechter]

I didn't read the datasheet yet, but you should be able to use an op amp that runs on the 3.6v output. I'm sure that there will be a way to use the amp output to pull the trim down. I'm just not sure how far the trim will go and which way, but I'm sure it's in the datasheet.

The other challenge will be to make a reference to measure the shunt voltage against. Perhaps just a diode and resistor divider.

This way each module will have its own current regulator and would be independent of the others so you can stack as many as desired in series.

 

[curious]

The simple solution is to have a shunt value in the ground wire of the load selected such that at target current the voltage drop is about 0.6V. Then use a single npn transistor, connect the base to the shunt-load connection, emitter to the ground and collector to the trim input. I believe you do not even need collector resistor with this converter. Power dissipation with 0.6V drop will be noticeable though.
In order to get lower drop voltage you need few more components - either op-amp or at long-tail pair. If you prefer transistors I can probably hack some schematics in 3-4 parts total. Let me know.

 

[Doctorbass]

Here is a schematic that i've found on "applications notes" from another DC-DC suplier:

Is that you are talking Fechter and Curious? about the amp-op?

in the schematics below, it's a current source.. so to add the 3.6V triming adjustment how would you suggest to add this fonction? (CC 10A until CV 3.6V max)

Doc

 

[fechter]

Yes! that's it amost exactly.

We just don't need the voltage programmable part.

Ideally, you'd want to use a device with a built in voltage reference, then divide it down to 20mv to use as your reference.

A cheaper way to do it would be just a simple divider on the output voltage. If the battery is really low at the beginning of charge, the current would be slightly less. This could be an advantage if you have a shorted cell, the current would be limited to near zero. It won't take long for cells to get up to near the CV voltage, at which point you would get the desired current.

The value of R would be chosen to get the + input of the amp to a level equal to the voltage drop across the shunt at the desired current limit.

For a 0.001 ohm shunt, 20 amps, you want 20mv. Dividing 3.6v, R would be around 180k.

 

[Doctorbass]

Thanks fechter for your help!

I really appreciate!

In your modified schematic, i just wonder what will set the voltage to maximum 3.600V?

When the cell will reach 3.600V, the 20A should decrease to keep this max voltage.

I want to try that modified circuit you did. what should i add to set max V out to 3.600V

what amp-op would you recommand for 3.6V supply without problem? TL071, 741, NE5532?

Doc

 

[fechter]

Are your converters designed for 5v output?

According to the datasheet, the maximum current will be 170% of the built in limit (not 170A!).

The JW075A model has a 18 amp current limit already.

 

[Doctorbass]

Are your converters designed for 5v output?

According to the datasheet, the maximum current will be 170% of the built in limit (not 170A!).

The JW075A model has a 18 amp current limit already.

yes, it is for 5V but the model i have are adjustable from -60% to +10%.

what i want to do is to use them with both 3.3V (charge to 3.6) and 3.7V (charge to 4.2) li-ion

with a selectable resistor and a couple of 4pst relay, i will change the voltage of all 12 dc-dc modules.

the datasheet i've put previously is not the exact same as the artesyn i have.. but i posted it because of the application note that give more details about hot to use the trim part...

the BXB150 Artesyn is a 150W 5V 30A output adj.

what i want to avoid is to use the stock 30A current limit for two reasons:

first is that at 4.2V, the 30A would need more than 140W at the input to operate and if i use the 12 modules at the same time, i would need 1680Watts and i only have 880W of DC from my first stage power supply

at 880W, if /12 that's 73.3W per module at input. and takinh account of the efficiency around 80% for this conditions, i will have maximum 59Watts

59W when used with the 4.2V li-ion would allow me to charge to maximum 14A
and with the A123 (3.6V) that would become 16.4A

16.4 is a good treadoff between 10 and 20A i talked previously.

What you think about my questionning for the 3.6V CV limit and your circuit?

Doc

 

[fechter]

OK, disregard the first one.

Try this:


A diode between the op amp and the trim pin will allow the normal trim to the desired voltage and if the current exceeds the set point, the diode will pull down the trim even further.

The thing I'm not sure about is whether the voltage drop in the diode will allow the trim to be pulled down more than it is already. 3.6v is pretty close to the minimum output.

You could test by taking one that's adjusted to 3.6v, then put a diode from the trim pin to ground and see if the voltage drops further.

The op amp should be a low voltage, rail-to-rail input and output.

 

[curious]

Less accurate but fewer components. Voltage drop on schottky diodes offsets part of the 0.6V base-emitter threshold so that load resistor can be smaller. Some tweaking with values required. BTW the circuit with the comparator above is not very accurate either as the voltage output changes during charge by nearly 1V and therefore reference voltage for the comparator will change as well affecting the current limit. Not that it matters much though.