Lithium 2cell LED driver-ops question

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
I need someone to help me out with an operational question.

I'm going to build this LED driver ckt, and I need to know if I have made the correct assumption about exactly how it is operating. Which means I want someone to build this same exact circuit in LTspice and tell me if I'm right.

The power level to the LED increases when R3 is increased in resistance BECAUSE it is changing the DC level of the base of Q1. Even though the peak to peak amplitude of the signal at the collector is smaller the over all power to the LED is higher because it is DC biased at a higher average voltage.

Correct? Here's the ckt and waveforms I have. (Anyone who wants it may use it with my permission)

(R4 will have to be a high power 5W resistor, R3 needs to be a 1W resistor)

Edit: This circuit is a beginners Spice mistake. It only switches the transistor into conduction and the LEDs off because of that. The LED should be connected from Vcc to the collector of the transistor and not to ground.


Thanks so much for the help everyone. You guys are AWESOME :)
 
Last edited:

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
ozkey dokey. That first circuit was just a big mistake. The LED should have gone between the battery and the collector instead of battery and ground. The increase in LED power with increased base resistance was just turning the Transistor off and allowing more current to flow through the LED circuit.

Yeah, It's been a while since I did any breadboarding! :)

So I've been playing with it more and have it all decked out and ready to go. The next step is to replace the pulse generator from spice with a variable PWM circuit.

I'm sure I can do that one, but I have a question on biasing. The 5 volt square wave need to be riding on a 1.2 volt DC level for a 1.2 to 6.2 volt signal.

What would be the best(least number of fidgity components) way to get that result. My first instinct is the down and dirty battery bias above ground for the PWM. I KNOW there has to be a better way. And after getting so many responses to the first post, I wanted to ask for more help :rolleyes:

Come on. Give an old man a break. Let's hear about some new fangled methods my old school didn't know of...:D
 

Audioguru

Joined Dec 20, 2007
11,248
1) LEDs all have different voltages. LEDs should never be connected in parallel unless you measure many and sort them so that each one has exactly the same forward voltage.
2) Isn't the current in each LED much too high for them and much too high for the transistor?
 

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
I'm using what models spice provides. I have a single 3 watt LED I'll be using...blah blah blah

But I'm really interested in an answer to the question I asked.

Think maybe someone could try that and post it here?


Please?
 

Audioguru

Joined Dec 20, 2007
11,248
I don't know why you posted a schematic with 12 LEDs when you are using only a single LED.
I don't know what you have that gives a 5.0V signal to the low base resistance of the transistor without dropping its voltage.
I also don't know why you want the transistor to remain turned on (1.2V at the base) without going off between flashes.

The 2N2222 will melt if you use it to power your 3W LED. Which transistor will you be actually using? Then we can calculate the base resistor for it.
 

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
I was asking about getting a 5 volt square wave to have a 1.2 volt DC bias.

IS THERE A BETTER WAY THAN BATTERY BIAS THAT ISN'T TOO COMPLEX?


Forget the stupid circuit. I'm learning how to use spice, and that's something I 'made' with spice while working with it and trying to see what it has and what it can do. I have a signal source in spice represented by a circle and I specified 5volts amplitude with a low of 1.2 V and a high of 6.2 volts. I'm starting to become frustrated with the limitations of spice set by the limited device selections. They don't have transformers for one thing. they have no 78xx, or 79xx regulators. they only have polarized caps in the library. etc etc

A simple answer, like - YES there is a better way to bias a square wave than with a battery. You could do "x" or you could do "y", instead of using a battery to bias the circuit, would be much appreciated. If you want to know why I did something you see in the circuit I posted, how about you answer my question and then ask me yours?

Sorry I seem gruff, but I think I was clear about the question I wanted answered, and I would be happy to give you an explanation of why I did something one way or another, if you would be kind enough to answer my question. This is how its supposed to work, no? I ask a question, and someone answers it, then if they have another question for me, I answer that one.

Now I'm sure this will piss people off. "How dare that person cop and attitude with me?" etc. If you get pissed off over this then I don't want to converse with you anyway. I'm here to offer any help I'm able to others and I try to do that within the limitations of what I know and still remember from school. Would someone who is also here to help others like to answer a question from me? It still stands. I just want to know if there is a SIMPLE way to bias a square wave besides battery methods. If I need to use a 555 device to do so that would be a good thing. If I need to use logic gates to do so that would be good too.

What about it? Anything new come out of these electronics schools since Triodes and Pentodes and mercury vapor rectifiers were in the books?

Something without two dozen parts would be most welcome, and if it needs less than ten to implement I'd be tickled pink.
 

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
I'm tempted to just remove the whole circuit thingy I whipped up. LTspice is a wonderful little toy, but I doesn't have the LED I want to use; it doesn't have the transistor I want to use, It doesn't have any three terminal regulators AT ALL, No transformers, no small signal caps. It sucks for selection. The circuit is NOT what I'm going to use, but I find it useful and fun to play around with. I'm familiar with 2222 transistors. And they will take the amount of power I am sending in the circuit(if the simulation numbers can be trusted). If anyone looked AND ran the circuit they would notice it is only running at a 10% duty cycle. Lots of time to rest between pulses...

Not the point anyway. I could go research the web and perhaps settle down with a book on these new fangled IC thingy's(whatever happened to tubes anyway) :) but I like interacting with people and kinda hoped(what,with all the banter about electronics and circuit design here) that someone would offer up a simple solution without me "doing it all myself"

I know my way around a circuit board well enough to be earning a very good living wage. Military defense contractors and bases the world over have been graced with my presence. If this place is to 'elite' for me, then you guys can have it...I'm just trying to have some fun playing with stuff I have no need of in my working day life.

Some of my working life escapades this past year.







Can't whip that camera out just anywhere ya know, somethings aren't meant to be captured on film and displayed to the world...
 

Audioguru

Joined Dec 20, 2007
11,248
Spice doesn't know and doesn't care that a 2N2222 has an absolute max current of 800mA. In your simulation the current is over 1A.

Spice doesn't have voltage regulators because they are nearly perfect. In a circuit you might see the output drop a few mV with full load or add your own small output resistance to a battery in spice to do the same thing. With the correct output capacitor (as shown on the datasheet) the transient response of a regulator is fine.

To make a digital pulse go from 1.2V to 6.2V I would take a Cmos 0V to 7.4V pulse and load it until it drops to 1.2V and 6.2V. But maybe spice doesn't know the output resistance of Cmos logic.
 

Thread Starter

Kermit2

Joined Feb 5, 2010
4,162
I had attached a version of the spice file in which I was seeking the component values for achieving the 100mA peak pulse rating of the selected LED's; hence the overdrive of the 2222 transistor.

My apologies for the confusion this must have caused you!


NEW QUESTION:

I've got a flip/flop type oscillator(spice again), and found it will not oscillate with all component values matched and mirrored. Is spice really this sensitive? by selection of a different transistor for one side,or by changing an emitter resistor value on ONE transistor, everything functions as it should.

Does spice not allow oscillation without doing this? Must one model real world imbalance in a circuit for the program to tell which transistor to turn on first?
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
Spice assumes that the transistors and resistors are exactly the same so it is possible for the circuit not to oscillate (I have seen it happen).
 
Top