TIP127 current calculation....

[large_ghostman]

It was the switch led part that confused me! Laptop I am on dosnt show his drawing very well, so I couldnt see it. Would a constant current source, via say a variable voltage regulator be better with a pot on? excepting of course that the leds are all identical because he has carefully matched them first, so say set that for a max of 80mA and have a pot to adjust it down?
But I have a sneaky the driver for this will be a micro

 

[MrAl]

Hello,

"Control the current" can mean many things. It can mean to switch on and off only, or to regulate the current at some set level, or to regulate at some set level where we can change that level at any time with an adjustment either manually or through a uC chip or other. Also, regulation comes in two forms also: linear or PWM, so you have to figure out what you want to do there too.

To switch on and off, the simplest way would be to use a series resistor with the LED's and just turn the D. transistor on and off with a base current of maybe 10ma. That's the simplest and there is no way to do this using a feed forward mechanism because there is no way to know the exact BETA of the transistor(s).

To regulate the current at a set level, you have to be able to sense current through the LED's, then use that as feedback to a regulating circuit.

To regulate and be able to change the level of current through the LED's, you need a current sense resistor and feedback as above plus a way to change the reference signal to the control circuit so you can change the current level though the LED's.

So take your pick and then we can discuss further.

 

[koolguy]

There are series of 4 led red and many series are connected in parallel....
I want to control current without using so many series resistance, as before bc337 was used which was working fine for low current ~250mA but now i want more brightness by controlling its current, not using PWM because there will be many other led series for flashing on display .
i am using smps of 12v 4A

 

[koolguy]

one thing more
Transistors in parallel using bc337 and one driving transistor for driving other transistor base for ~35mA as pic uC can give only 20mA.
https://electronics.stackexchange.com/questions/77045/transistors-in-parallel

 

[rumpfy]

Mr. RK had a previous thread "BJT as Switch", and this thread faded out at post #137. This thread, so far, suggests a repeat performance. In BJT as switch, post 124 instructed the use of Ic/Ib as a criterion for establishing base drive requirements and for a BC337 this ratio was 10:1.
Again, for a TIP 127, the data sheet gives the Ic/Ib criterion as 250:1. So for an Ic of 5 amp, Ib is 20 milliamp.
This gives a base to ground resistor of around 500 ohm for a 12 volt supply rail. Problem from post #1 and #2 now solved.
Now we find that the LED are to be (presumably) pulse driven because "it is for 'flashing", and we also know the SMPS is a 12V at 4 Amp unit. It is unclear exactly what this SMPS can do. Post #23 says that the method of current limit used in 'BJT as switch' was (resistors ?) and was only 250 mA so that was OK. So in this new invention, in order to control the brightness, use will be made of the SMPS but it wont be PWM for some unexplained reason.
It now is clear that there is no pre-stated specification for this new request of this thread.
It is clear that before any progress is made, the requirements need to be stated;

What is the invention for which the thread has been started?
If it is a LED display, what type of LED will be used and is the dat sheet been read by the OP?
If the LED is to be current controlled, will the SMPS be capable of this mode of operation?
If the SMPS is a commercial design, then what is the type number and manufacturer?
If the SMPS is a commercial type, has the data sheet been read by the OP?
If the invention is to be a strobed type of display, then how is the strobing to be implemented?

If clarification is necessary, have a look at posts 23 and 24 above, then compare with posts 13 and 14 in 'BJT as Switch'

 

[koolguy]

I was saying that when i was using BC337 it was acting as current limiting as you said 1:10 when there was 20mA it will give 200mA...as 50ma will 500ma at more than 400mA the BC337 get start heated and i need more current to have good brightness for led for every series of 4 red led it should be 80mA.
The SMPS is not problem it just a power source, so, i want to avoid many BC337 in parallel operation to getting more current, for this i am searching a BJT at ~1.5A which will work fine for this purpose changing its base current also changes it collector current flowing through it...

 

[koolguy]

Experiment reading......using white led...
the bias of Bc337 is connected with 12V at 5.5k bias resistance...

Ib I1 one LEd I2 Two series LED I3 three series LED Rb series Resistance at collector with LED
2.1mA 83mA 51ma 24mA 5.5K 110 ohm
2.1mA 150mA 89mA 40ma 5.5k 55 ohm

but according to data sheet 1:10 so, at base current of 2.1mA it should reach maximum 20.1mA ..

 

[large_ghostman]

Can you explain exactly why PWM cant be used? I didn't understand that bit, how many strings are using?, many is a strange number

 

[koolguy]

Can you explain exactly why PWM cant be used? I didn't understand that bit, how many strings are using?, many is a strange number

In led display there are more than 20 characters each is to be controlled individually so, 4 PIc16f676 will be used taht why...!

 

[large_ghostman]

Still dont see the issue! But maybe that is just me not seeing it.

 

[MrAl]

Hi,

Here is a circuit that could work for your purpose.

The function of R1 is to limit current into the base of the Darlington transistor.

The function of R2 is to limit current through the LED's but the reason this resistor is needed is a little more complicated than that. It does limit current through the LED's, but another reason why it is there is because we can not depend on the Beta nor the transistor voltage drop as being constant over the whole operating time of the circuit. If we try to depend on the Beta being constant then we could get too much or too little current, and if we try to depend on the voltage drop being "just right" so that it drops the right voltage when turned on then we could end up supplying too much current to the LED's. If on the other hand when the transistor is turned on it happens to drop too much voltage, then we can not use that transistor for this circuit and would have to go to another type, probably a single transistor driven in another way.
Because of this a test has to be done to find out if the transistor can supply enough current with a low enough voltage drop to be able to drive the LED's properly.

Note we had to use a negative voltage regulator here because we had a PNP transistor to drive, but if we used an NPN transistor (or Darlington) we could use a positive regulator.

A problem that may show up here is because we are using a Darlington instead of a single transistor (with proper driver transistor) we dont even know if it will work right to start with. Because the LED's drop around 8v (from your other posts) if the transistor drops 4v then that is cutting it too close and so the output current level setting may not be reliable. It would be better to use a single transistor with another transistor driver set up differently and then use a proper size resistor for R2. A single transistor with driver may only drop 1v and so that means we can set the current level with that single resistor R2 and it should remain at that level.

 

[audioguru]

Ritesh wants to connect 4 red LEDs in series which makes one string then connect many of these strings in parallel. He thinks the total will be 8.8V because each red LED has a typical forward voltage of 2.2V. But he doesn't know that each LED has a DIFFERENT forward voltage from about 1.8V to about 2.4V.

If he does not test each LED and find four that total 8.8V and do that over and over then he will have some strings of (1.8V x 4=) 7.2V that will quickly burn out and he will have some strings of (2.4V x 4=) 9.6V that will not light.

He says he "wants to control the current in the transistor". He tried a little BC337 but found it overheating (of course, because it was a linear heater, not a switch and nothing limited the current).

Ritesh and members here trying to help are completely confused about calculating base current. They do not understand that hFE is a variable number that is high for some transistors and is low for other transistors (even when they have the same part number) and is used when the transistor is an amplifier with plenty of collector-emitter voltage. hFE also changes when the temperature changes. Because it is a variable number then the circuit usually has some negative feedback to reduce the amount of difference between different transistors. His circuit used no negative feedback.

When a transistor is used as a saturated switch then it is turned on as hard as it can get so the base current is increased to 1/10th the collector current. hFE does not work at such a low collector-emitter voltage (maybe 0.1V or less). Something (a series resistor) must be used to limit the current.

In post #27 Ritesh explains an experiment but did not show a schematic and did not say its supply voltage. He used a base current of 2.1mA and we hope he connected the polarity correctly for the collector and emitter. He wrongly thinks that the transistor will limit the current to 10 times the base current because he does not understand that it is a SWITCH. Luckily he used a series current-limiting resistor with the LEDs.
He used one white LED (maybe it has a forward voltage of 3.2V), a 110 ohm resistor and got 83mA which is enough to burn out an LED. Then the supply was probably (110 ohms x 83mA=) 9.13V + 3.2V for the LED + 0.1V for the transistor= 12.43V.

He used two white LEDs (maybe they have a forward voltage of 3.3V each), the 110 ohm resistor and transistor and got 51mA. Then (110 ohms x 51mA=) 5.61V + (3.3V x 2=) 6.6V + 0.1V= 12.31V.

He use three white LEDs (3.2V each?), the 110 ohm resistor and transistor and got 24mA. (110 ohms x 24mA=) 2.64V + (3.2V x 3=) 9.6V + 0.1V= 12.34V.

You can work it out with the resistor at 55 ohms if you want. Of course its low value limited the currents to be much higher.
His experiment shows that the transistor is turned on hard as a switch but it is the resistance that controls the current.

Ritesh does not explain what he is doing with "20 characters controlled individually". He does not say what the PIC does.

 

[large_ghostman]

Ritesh could you not use resistor arrays? they are good if space is tight

 

[MrAl]

Ritesh wants to connect 4 red LEDs in series which makes one string then connect many of these strings in parallel. He thinks the total will be 8.8V because each red LED has a typical forward voltage of 2.2V. But he doesn't know that each LED has a DIFFERENT forward voltage from about 1.8V to about 2.4V.

If he does not test each LED and find four that total 8.8V and do that over and over then he will have some strings of (1.8V x 4=) 7.2V that will quickly burn out and he will have some strings of (2.4V x 4=) 9.6V that will not light.

He says he "wants to control the current in the transistor". He tried a little BC337 but found it overheating (of course, because it was a linear heater, not a switch and nothing limited the current).

Ritesh and members here trying to help are completely confused about calculating base current. They do not understand that hFE is a variable number that is high for some transistors and is low for other transistors (even when they have the same part number) and is used when the transistor is an amplifier with plenty of collector-emitter voltage. hFE also changes when the temperature changes. Because it is a variable number then the circuit usually has some negative feedback to reduce the amount of difference between different transistors. His circuit used no negative feedback.

When a transistor is used as a saturated switch then it is turned on as hard as it can get so the base current is increased to 1/10th the collector current. hFE does not work at such a low collector-emitter voltage (maybe 0.1V or less). Something (a series resistor) must be used to limit the current.

In post #27 Ritesh explains an experiment but did not show a schematic and did not say its supply voltage. He used a base current of 2.1mA and we hope he connected the polarity correctly for the collector and emitter. He wrongly thinks that the transistor will limit the current to 10 times the base current because he does not understand that it is a SWITCH. Luckily he used a series current-limiting resistor with the LEDs.
He used one white LED (maybe it has a forward voltage of 3.2V), a 110 ohm resistor and got 83mA which is enough to burn out an LED. Then the supply was probably (110 ohms x 83mA=) 9.13V + 3.2V for the LED + 0.1V for the transistor= 12.43V.

He used two white LEDs (maybe they have a forward voltage of 3.3V each), the 110 ohm resistor and transistor and got 51mA. Then (110 ohms x 51mA=) 5.61V + (3.3V x 2=) 6.6V + 0.1V= 12.31V.

He use three white LEDs (3.2V each?), the 110 ohm resistor and transistor and got 24mA. (110 ohms x 24mA=) 2.64V + (3.2V x 3=) 9.6V + 0.1V= 12.34V.

You can work it out with the resistor at 55 ohms if you want. Of course its low value limited the currents to be much higher.
His experiment shows that the transistor is turned on hard as a switch but it is the resistance that controls the current.

Ritesh does not explain what he is doing with "20 characters controlled individually". He does not say what the PIC does.

Hi,

Thanks for pointing those things out. If we get the LED current he wants we can calculate this out completely.

I mentioned using a series resistor for the LED's in my post with the schematic. But if he wants to use say four strings then it would be even better to use four resistors, one for each LED string. So that would mean there would be four R2's as shown in the schematic, and four strings of LED's as shown in the schematic.

He also seemed to want to use a uC i/o pin to turn the LED's on and off so i showed a uC connected to the transistor and a voltage regulator for the uC, and that it is slightly unusual because we are trying to use a PNP transistor rather than an NPN (either Darlington's or not).

I also mentioned that the Darlington's voltage drop is a little harder to work with than a single transistor with another driver transistor set up differently. So i favor not using a Darlington and just using two transistors maybe one PNP and one NPN or something like that. That would give us much more predictable results and at the same time result in a circuit that is more stable over environment changes.

For example, a smaller NPN driving a larger PNP, with a base resistor for the PNP too and emitter tied to +12v, and a +5v regulator for the uC, and the uC i/o pin drives the NPN transistor base through a base resistor. That would be a better circuit to use rather than the Darlington. Of course we still need the LED strings current limiting resistors, but now the change in LED current with environment will be less so we can be more sure of the LED operating current over long time periods of operation like years.

I also suggested a 10ma base current for the Darlington which allows for high collector current even with lower gain. 10ma times 300 brings us up to 3 amps.

 

[koolguy]

In post #27 Ritesh explains an experiment but did not show a schematic and did not say its supply voltage. He used a base current of 2.1mA and we hope he connected the polarity correctly for the collector and emitter. He wrongly thinks that the transistor will limit the current to 10 times the base current because he does not understand that it is a SWITCH. Luckily he used a series current-limiting resistor with the LEDs.
He used one white LED (maybe it has a forward voltage of 3.2V), a 110 ohm resistor and got 83mA which is enough to burn out an LED. Then the supply was probably (110 ohms x 83mA=) 9.13V + 3.2V for the LED + 0.1V for the transistor= 12.43V.

The LED used is Oval which has high brightness than normal ...work on 3mA to ~150mA for good brightness i have noted the current by testing.

 

[audioguru]

Ritesh,
DO NOT use a trimpot in series with the base of the transistor like that because if its is turned down so that its resistance is low then the base to emitter diode of the transistor WILL BURN OUT!
You are using the transistor as a SWITCH then its base current should be 1/10th its collector current.

With three 3.2V LEDs, a 110 ohm resistor and a 12V supply the collector current is about 20.9mA then the base current should be 2.09mA and the base resistor should be (12V - 0.7V)/2.09mA= 5.4k ohms (use 5.1k), not a trimpot.

The shape of an LED has nothing to do with its brightness. Some LEDs are focussed into a narrow beam and appear bright only if they shine directly at you or shine on something that is small.
Some very bright LEDs are round but are made to operate at a high current and be fastened to a heatsink.

EDIT: If the current-limiting resistor is 55 ohms then the current in the LEDs is about 42mA then don't they burn out?

 

[4pyros]

In led display there are more than 20 characters each is to be controlled individually so, 4 PIc16f676 will be used taht why...!

1st off you should be multiplexing our LEDs so you only need one pic. Start here;

 

[MrAl]

The LED used is Oval which has high brightness than normal ...work on 3mA to ~150mA for good brightness i have noted the current by testing.

Hi,

Yeah a trimpot in the base of the transistor is not a good idea. It appears to work because when you are adjusting it the environment stays somewhat constant, but as soon as things change your LED current changes and you are not there to adjust it again.

If you want current control, the simplest way is to build a voltage regulator with voltage feedback, then feed the output voltage to a string of LED's with a current limiting resistor. With enough overhead voltage the regulated voltage will keep the LED current fairly constant over a wide operating range.

If you want to use a uC to control the brightness then we might look into PWM either for the LED's themselves or for the driver transistor base which would be filtered so that we get a linear transistor response (even though the uC puts out PWM).

So the LED's draw a max of 150 ma then i guess. Is that for each string then too?

 

[koolguy]

EDIT: If the current-limiting resistor is 55 ohms then the current in the LEDs is about 42mA then don't they burn out?

No, it work also at 150mA ..

1st off you should be multiplexing our LEDs so you only need one pic. Start here;

its not scrolling just flashing...

So the LED's draw a max of 150 ma then i guess. Is that for each string then too?

yes may be same for good brightness..

 

[MrAl]

Hi again,

Ok so your LED strings draw 150ma maximum, and how many strings do you plan to use?

Also, so you only want to flash the LED's right, not adjust their brightness once the circuit is set up?