Rechargeable 30 LED burns up on AC

[5arid]

Can you please suggest the necessary corrections so that I can test it.
Also I can provide the Proteus ISIS or Multisim files.
I tried to make it a little more clear :

 

[andre_luis]

Can you please suggest the necessary corrections so that I can test it.
Also I can provide the Proteus ISIS or Multisim files.

Try converting to LTspice, I just have this simulator tool, due to be free.

 

[FvM]

I lately noticed that the schematic has errors. G3 is actually NPN and G2 should be flipped.

The basic problem of missing charging control still needs to be fixed, e.g. by a 4.8 V shunt regulator in parallel to the battery.

 

[c_mitra]

The basic problem of missing charging control still needs to be fixed, e.g. by a 4.8 V shunt regulator in parallel to the battery.

Most of these simple units do not have any regulator. The rectified output of the transformer is used for charging. The diodes are 1A rated and 3 diode drop is sufficient regulation for a battery charger. They happily last one year.

 

[FvM]

An unregulated transformer charger has at least a kind of "R" characteristic which almost stops charging for a full battery. Overcharging a 1.2Ah lead acid battery with 150 mA constant current will hardly give one year life time.

 

[c_mitra]

These small transformers are more R than L; I have never measured but they just warm up under short circuit.

 

[Audioguru]

I fixed parts of the schematic:

 

[c_mitra]

Sorry! This is direct line driven; three diode drops and 2.2uF reactance in series. They are more like constant current source, as you say. Sorry for the goof up.

 

[5arid]

How much V and I are needed to charge a 4v 1.2A acid lead battery?
There is no C2328A, nor S9014 in the ISIS (not even in the Multisim), what can I use instead of them to simulate the circuit?

 

[Audioguru]

Most of us simulate with LTspiceIV not Multisim. We use modern American and European transistors that are in the software, not obsolete Oriental transistors.

 

[d123]

Hi,

It's fairly traumatic to do the first few times, but maybe your simulator allows you to set transistor parameters based on the datasheet characteristics - but I find for a beginner such as myself it is a fairly inaccurate method as often some parameters are hard to find and/or understand what they mean in the simulator list as not all terminology coincides either.

Try search engining the term: "C2328A substitute parts", and the same for the other transistor. Or look on manufacturers product pages for substitute parts also.

 

[Audioguru]

I think the 2SC2328A transistor is too small to power 30 LEDs and the same for another transistor with the same spec's.

 

[5arid]

Maybe this device is designed for 110v and it is mislabeled by 220v?!
It doesn't make seance that they produce a defective circuit.
Please suggest me two transistors, or tell me the specs of them so that I can search them in the digikey.

 

[Audioguru]

The transistors can be found at www.datasheetarchive.com . The 2SC2328A is rated with a maximum allowed current of 2A. If the battery is fully charged at 4.6V and the LEDs are 3V white ones then each LED will have a current of (4.6V - 3.0V)/10 ohms= 160mA which is stupid (the maximum allowed current for an ordinary 5mm diameter LED is 30mA). Then all 30 LEDs will use 160mA x 30= 4.8A and the poor little 2SC2328A will be destroyed. But of course the transistor does not have enough base current for it to conduct 4.8A but maybe enough for 2.4A and the heat (1.92W) will destroy it before the current does.

 

[5arid]

Here is the main spec of the two transistors according to their datasheet :

• S9014

Transistor Type : NPN
Ic : Current - Collector (Ic) (Max) : 150 mA
Vceo : Voltage - Collector Emitter Breakdown (Max) : 50 V
Vce(sat) : Vce Saturation (Max) @ Ib, Ic : 250 mV
Icbo : Current - Collector Cutoff (Max) : 50 nA
hFE : DC Current Gain (hFE) (Min) @ Ic, Vce : 100 ~ 1000
Pc : Power - Max : 625 mW
fT : Frequency - Transition : 60 MHz​

• C2328A

Transistor Type : NPN
Ic : Current - Collector (Ic) (Max) : 2 A
Vceo : Voltage - Collector Emitter Breakdown (Max) : 30 V
Vce(sat) : Vce Saturation (Max) @ Ib, Ic : 500 mV
Icbo : Current - Collector Cutoff (Max) : 100 nA
hFE : DC Current Gain (hFE) (Min) @ Ic, Vce : 160 ~ 320
Pc/Pd : Power - Max : 1 W
fT : Frequency - Transition : 270 MHz​

- - - Updated - - -

I think they calculated the 10 ohm res for LEDs like this :

This is the type of battery they used in this device :


It is 4v 1.2Ah
This is written on the device : 20 Hours Light
So the battery is 1.2Ah / 20h = 0.06A
White LEDs are 3.2v ~ 3.6v
4 - 3.4 = 0.6 / 0.06 = 10 ohm

I don't know if it is correct!

- - - Updated - - -

And maybe that 2 A transistor like this :
0.06 * 30 = 1.8 A

 

[betwixt]

I think before delving too deeply in to LED currents, you should establish whether the schematic is correct. Everyone is telling you something is wrong with it.
Please post a photograph of the top (component) side side of the board again, as close up as you can manage before you lose focus and then flip it sideways and take a picture from the track side. The circuit is obviously very simple so it shouldn't take long to reverse engineer it. If you can take a picture or type the exact numbering on the large capacitor it might help us to establish the voltage it was designed to run on.

Brian.

 

[FvM]

The circuit should be basically operating with the modification by Audioguru. It has been previously clarified that the circuit was burned because you disconnected the battery which isn't supported.

"20 hours light" is another impossible specification. Would be achieved with one instead of 30 LEDs running at 60 mA. Realistically, the transistor saturation voltage has to be put into calculation, all in all the per LED current will lower (e.g. 40 or 50 mA) and the transistor can hold it.

There's no essential difference between 110 and 220V operation, charging current is halved but still not suitable for permanent charging. Battery would dry out after a few weeks, maximum months.

 

[Audioguru]

I betcha the cheap poorly designed flashlight does not have a guarantee.

 

[5arid]

I hope this can shed some light :

 

[betwixt]

Thanks, that answers a lot. Your schematic was wrong, I've drawn it correctly and it makes more sense. It looks like there was originally a regulator and protection but the manufacturer removed it and linked across it to cut costs.

The labels look wrong too. The LEDs should be connected between E- and D- and the battery between D+ and E+. Polarity markings look suspicious too!

Brian.