Transistor based Step Up Circuits

[ViReN]

Hello Friends,

I am new to this forum. I am looking for Transistor based Step-Up Circuits for a small Project involving 1 - 2 AA Cells and having 3 - 20 White LED's.

I am planning to make a LED Flash Light from a 2 AA Cell Torch. I have made PCB and connected about 15 LED's in parallel. The Torch works fine, but the problem is that it has less output. LED's require 3.4 - 3.8 V and at my end i have only 3 V that too is going to diminish in value.

I cannot use SMD components as i dont have that kind of experiance.

Can any one suggest a simple way to drive these LED's ?

Many thanks in advance.

Regards,
Viren

 

[jtr1962]

I find this circuit very handy:

A description of the circuit can be found here.

By substituting low saturation voltage, high gain transistors such as those from Zetex for the generic 2N3904s you can increase the value of R2 by a factor of maybe ten. When used with a low resistance, high Q inductor this circuit can be over 90% efficient.

While this circuit is non-regulating, I have devised a way to make it regulate current by adding another transistor and a resistor between the (-) terminal of the LED and ground. The resistor's value is computed by using the equation R = 0.7/I(LED). For instance, for 20mA you would use R = 35 ohms. The base of the extra transistor is connected to the (-) terminal of the LED, the emitter is connected to ground, and the collector goes to the base of Q2. A generic 2N3904 or PN2222A works fine for the extra transistor.

 

[ViReN]

Wow Great Help. I will surely do some R & D on this.

Can you please tell me if it is possible to use same circuit or are there other circuits for multiple LED's (5- 20 of em) connected in parallel.... Also can this circuit handle more than 1 cell at the input.

thanks in advance

Regards,
ViReN

 

[jeff1500]

Here's my set of notes. Look for the BRINKMANN circuit. I've had good luck with that one.

 

[jtr1962]

You're quite welcome. /ubbthreads/images/graemlins/smile.gif And to answer your questions:

The only limit to the number of LEDs in series is the maximum voltage rating of Q2 (this is the Vce parameter). For this type of circuit putting LEDs in parallel usually doesn't make much sense unless doing so will exceed the Vce of Q2. For example, I used this circuit for my bike headlight (20 LEDs). Putting them all in series (~60V) would have exceeded Q2's Vce rating so I paralleled two series strings of 10. In general driving multiple LEDs in series is much better since you don't have to worry about matching forward voltages.

Yes, the circuit can be used with a wide range of input voltages. My bike headlight uses 4 AA NiMH cells (~5V). Resistor R2 controls the current going to the LEDs. Increasing R2 decreases the current and vice versa at any given input voltage. However, with any given value of R2 LED current will increase as input voltage is increased so a value of R2 that works fine with one cell will heavily overdrive the LEDs with four cells unless you add the third transistor for regulation. The key is to play around with R2 until you get the current you want at the input voltage you're using. Since R2 is heavily dependent upon the transistors used and the inductor there is really no hard, fast rule of thumb, just trial and error.

Three other caveats. When the input voltage exceeds the output voltage the circuit no longer functions and the LED current is limited only by the series resistance of L1 so don't use this circuit to drive one LED when the input voltage may exceed 3V or so. The second thing only applies when using the modified (i.e. three transistor) version. With this circuit the input voltage range isn't as wide. For example, I've built a regulating circuit that starts regulating at around 1.5V but above about 4 volts it just shuts off. I haven't figured out why yet. The third thing relates to C1. You can start with 220 pF as recommended and decrease it in small steps until the circuit ceases to operate, then go back up one or two steps. The smaller C1 is the more efficient the circuit but under a certain value the circuit just won't work.

One thing I did also was add a Schottkey diode and capacitor to the output to turn the choppy output waveform into smooth DC to drive the LEDs. This should prolong LED life, and it also allowed me to measure the circuit's efficiency. BTW, most of the time the circuit will start at 0.8V, and if it is running the LED will stay lit down to about 0.6V. Also, don't forget to use a decent inductor. An inductor with high series resistance (more than maybe 0.5 to 1.0 ohms) can hurt efficiency by 10% or more, and in extreme cases the circuit may not work at all.

 

[ViReN]

You Guys are truely Inspirational. I will be working around with the things that you have mentioned. Will report to you about the things soon.

Many Many Thanks Guys.

Its Amazing, how you come forward for help for a new person like me.

And also Many Thanks for CPF Founder. Without him / her, I could have never got to this site.

Best Regards,
ViReN

 

[ViReN]

I have tried my hands on the circuit but i m afraid /ubbthreads/images/graemlins/frown.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/eek.gif /ubbthreads/images/graemlins/jpshakehead.gif what went wrong !!!...

I am using Eagle 4.11r2 for designing and layouting the PCB

Please view the layout and pcb here...

Can any one guide me ? What could posibly go wrong ?

Thanks & Regards,
ViReN

 

[MrAl]

Hello there,

I've simulated this circuit and it does look like it
works. I got about 15ma through the LED.

What can go wrong:
1. If your inductor doesnt work right you could
try two of the same in series, that might help.
You could also check at 3v input with larger R2 (1k).
2. Check all polarities of transistors, LED, etc.
3. Check each individual transistor to make sure
it will turn on and off. You can short the base
to ground and look at the voltage of the collector
to see if it turns on and off ok. If not, replace it.
4. Make sure the LED still works...test with resistor
and battery.
5. Sometimes if the circuit wont start, repeatedly
connecting and disconnecting the battery real fast
will show some light in the LED. Try this too.

Take care,
Al

 

[jtr1962]

Everything looks fine to me as well. Yes, sometimes this circuit is finicky and won't start unless you connect and disconnect it from the power source very quickly. This can be overcome by using a better transistor such as those from Zetex or by increasing the value of C1 (at the expense of some efficiency) until the circuit works reliably. The inductor is also crucial-if it has too much series resistance or saturates the circuit won't work at all, or will work poorly.

 

[Doug Owen]

Artwork looks good to me as well, as it should, isn't this an autorouting program?

Otherwise, what's the details? No light? Are you sure the LED is right way around? What sort of DC current are you drawing?

Doug Owen

 

[slipstream4ph]

@jtr1962
Can I use an RF Choke (resistor type package) instead of an inductor on this circuit?

/ubbthreads/images/graemlins/confused.gif

 

[ViReN]

[ QUOTE ]
Doug Owen said:
Artwork looks good to me as well, as it should, isn't this an autorouting program?

Otherwise, what's the details? No light? Are you sure the LED is right way around? What sort of DC current are you drawing?

Doug Owen

[/ QUOTE ]

Since I am a newbie in this section, using a breadboard was kinda difficult, as it is difficult for me to understand the connections physically and schematically. So i searched for a program that can do this job for me, there are many programs like Or-Cad and Protel (that we used to use in college days).. they are costly !... . And, for experimenting with simple circuits i thought of using this FREEWARE you can download it at http://www.cadsoft.de/download.htm I can directly print on laser printer.. and then use the Toner Transfer method to get the PCB... its some what time consuming, but it's a neat way to do.... makes things more clear to me atleast.

<hr>
Circuit Behaviour Details

When I connect to a 1.5 V Alkaline, there nothing happens (perhaps.. the switching is not occuring)
When I connect it to a 3 V Lithium Cell... the LED Glows (this confirms atleast.. that the LED is connected with a right polarity), DIM though ... dimmer than if directly connected to Lithium.

I have used the best possible components (available locally) including the inductance. Since i was not able to get a 220 pF capacator... I managed to connect 3 capacators in parallel to get about 210pF .... (is that a problem? /ubbthreads/images/graemlins/thinking.gif)

I will try to play around with the circuit... by connecting 2 inductors in series (suposed to give more boost) ... as well as changing values of R1 & R2...

but i still need your expert views... and suggestions... so as to make this circuit work at 1.5 V....

Thanks in Advance
Regards,
ViReN

 

[jtr1962]

[ QUOTE ]
slipstream4ph said:
@jtr1962
Can I use an RF Choke (resistor type package) instead of an inductor on this circuit?

/ubbthreads/images/graemlins/confused.gif

[/ QUOTE ]

You can but you might have problems with the circuit starting reliably and/or poor efficiency. RF chokes are designed to operate at higher frequencies (in the MHz range) than this circuit runs at (several tens to a hundred kHz). Also, those small resistor type chokes have a higher DC resistance than power supply inductors so all other things being equal, you'll lose some efficiency there. In addition, many of these small RF chokes will saturate at even a few mA of current (this will again decrease efficiency or possible even cause the circuit to not work).

All that being said, I have used small 47uH and 100uH RF chokes with this circuit to drive one or two LEDs. Efficiency isn't what it could be but sometimes an RF choke scavenged from something else is all you have available. The beauty (and also the headaches) with this circuit stem from the fact that every part affects circuit operation to one degree or another. If you have a large collection of spare parts it's fun to switch things around and see what the circuit does.

 

[jtr1962]

[ QUOTE ]
ViReN said:
When I connect to a 1.5 V Alkaline, there nothing happens (perhaps.. the switching is not occuring)
When I connect it to a 3 V Lithium Cell... the LED Glows (this confirms atleast.. that the LED is connected with a right polarity), DIM though ... dimmer than if directly connected to Lithium.


[/ QUOTE ]
If the voltage of the lithium cell is higher than the forward voltage of the LED then what is occuring is that the circuit is not operating even on the lithium cell. Rather, the LED is receiving DC current from the battery but with the resistance of the inductor in series (this is why the LED is dimmer than being directly connected). Try putting two LEDs in series when connected to the lithium cell. If they both light then the circuit works and it looks like you'll need to decrease the value of R2 to make it work at 1.5V. If it doesn't work then I suspect a problem with one of the transistors. Are you using power transistors? If so that might be the problem since many have a gain of only around 15 to 25. Make sure you're using small signal transistors such as the 2N3904, PN2222A, or anything from Zetex. Also, measure the resistance of your inductor with a multimeter. It shouldn't be more than a few tenths of an ohm.

 

[Doug Owen]

[ QUOTE ]
slipstream4ph said:
@jtr1962
Can I use an RF Choke (resistor type package) instead of an inductor on this circuit?

/ubbthreads/images/graemlins/confused.gif

[/ QUOTE ]

Yes, absolutely. That's what I use for such circuits.

Be careful of the maximum DC current rating, it doesn't do to 'run out of iron' and saturate. In the ranges we're talking about the inductors look like resistors typically. Quarter or half Watt size, but with too many rings and low DC resistance reading.

Doug Owen

 

[Doug Owen]

[ QUOTE ]
ViReN said:
When I connect to a 1.5 V Alkaline, there nothing happens (perhaps.. the switching is not occuring)
When I connect it to a 3 V Lithium Cell... the LED Glows (this confirms atleast.. that the LED is connected with a right polarity), DIM though ... dimmer than if directly connected to Lithium.



[/ QUOTE ]

Do you have a meter? What DC current does the circuit draw with 1.5 in (and no light)?

With the black (minus) lead of the volt meter on battery minus, what are the voltages on the Base and collector (pins 2 and 1) of Q2? Just looking at it, I'm not happy with the values shown, nor the lack of conventional bias (which may make it hard to start?). I'd expect a higher R1 (maybe closer to the one K Al suggests for 3 Volts in) and a lower R2 as I'm not sure there's enough base drive for Q1.

The light from the LED also tells us the inductor is still there. Much more useful would be some current or voltage readings.

This is the sort of circuit it's nice to prototype before you try to build it up. If I get a chance this weekend, I'll try to build it. I'm behind of this sort of stuff, so it may not happen. I've still got to dig out the SMT transistors and try these spiffy LDO regulators Harry found and Al hot rodded.....

Doug Owen

 

[ViReN]

Re: Transistor based Step Up Circuits IT WORKED !!

THANK YOU EVERY ONE.....

this worked too..... again those FAULTY TRANSISTORS......

I would like to present here... my FIRST EVER LED MOD.... yes... its ugly... but I hope to improve upon this in later stages....

LED's are bright, I think, they can be more brighter... will be playing around with things... untill i get the best out of it....

Guys... U ARE AWESOME... in helping such a newbie like me....

My Next step is ... to look for a better circuits... like... the regulated ones.... and then later on will switch over to the DC-DC Switching IC's .... and ultimately... the Lx's mod's....

Can you please tell me ... what to go in for next ?.....

Best Regards to Every one....
Thanks once again...

ViReN

 

[MrAl]

Re: Transistor based Step Up Circuits IT WORKED !!

Congrats! And welcome to the world of switching
supplies :-)

Take care,
Al

 

[ViReN]

Re: Transistor based Step Up Circuits IT WORKED !!

Mr Al, ..... I am thinking of making a better & regulated circuit... (of course).... it should run with 1.5 V source.... are there any other circuits.... that i can try my hands on... (after making first circuits working, i feel confident)... is there a way where in i can modify this circuit to make it current / voltage regulating ... the onboard components should be at a minimum.... and of course no SMD's at persent.... (i am not well prepared for em...)...

THanks & Regards,
ViReN

 

[MrAl]

Re: Transistor based Step Up Circuits IT WORKED !!

Hello again,

Yes, you can modify it to regulate very well, but it
will cost adding an op amp. Of course doing this
means you'll get very good regulation.
The idea is to monitor output current with a very
small resistor and amplify the signal to use as feedback
to an error amplifier (another op amp, second half of
ic chip).
The ic im talking about here is available in 8 pin
dip so you dont have to worry about SMD parts.

Take care,
Al