I have a 555 timer producing a signal to pulse an LED. It works great with one or two, but I need it to power multiple (around 20 LEDs). The 555 timer can't produce the current for this, any ways to use an op amp or transistor to amplify the current without changing the signal voltage? I have an LM324 op amp available to me.

 

Forget the Op-Amp, use a transistor.

 

Opamps generally have wimpy output stages when it comes to driving LEDs. What you want is an N-channel MOSFET.

 

How much current do you want to drive the LEDs with, what is the power supply voltage for the LEDs, and what is the power supply for the NE555?

 

Often a Switching Transistor is necessary to provide power when the control circuit is unable to. Instead of the control circuit driving the powered control element it instead drives a switching transistor to switch on power from another source of sufficient power. Either from within the original circuit (if available) or from an outside source. Don't expect control elements to always be able to source the entire power required for the controls as most have limited sourcing capability. A logic level MOSFET preferably or BJT is a good solution in most cases.

 

There are three common ways a 555 can drive an LED. Some are easier to boost than others. Please post your schematic so we know which one you are using. Be sure each component has a unique reference designator so we can discuss them.

ak

 

This is the circuit I have right now. I have it built on a breadboard probed with an oscilloscope, and the output looks like a triangle wave with peaks from around 5v to 3.5v, pulsing the LED. An electronics store near me has an ECG152 BJT transistor in stock, would that be a viable option?

 

What is the purpose of your circuit and that of the 10k resistor, 220u capacitor and the 2N3904 transistor? Where is your current LED connected?

 

The circuit produces a square wave, the 10k resistor, 220u capacitor, and transistor convert the square wave into the aforementioned triangle wave. The LEDs are connected to the emitter of the transistor. Ideally, any amplified current should come from that output.

 

The circuit produces a square wave, the 10k resistor, 220u capacitor, and transistor convert the square wave into the aforementioned triangle wave. The LEDs are connected to the emitter of the transistor. Ideally, any amplified current should come from that output.

Why do you need a triangle to drive an LED? A simple on/off is all thats needed.

 

Why do you need a triangle to drive an LED? A simple on/off is all thats needed.

Not if you want the LEDs to fade up and down. That is what the original circuit does, and he said it works fine. That indicates to me that an up-down fade is a goal of the project.

ak

 

An electronics store near me has an ECG152 BJT transistor in stock, would that be a viable option?

Maybe. Typical gain of 60, rated for 7 A collector current (with heatsink!) -looks good on paper, but there is a problem.

I couldn't find a full datasheet online, but we can take a swing at things.

Assuming your 20 LEDs have individual current limiting resistors and the LEDs are running with 20 mA peak current, that is 400 mA of load current. Divided by the typical gain of 60 yields 6.7 mA of base current. A 555 can supply that easily, but not through a 10K resistor - the voltage drop across the resistor will be too large for the circuit to function.

Decreasing the 10 K base resistor (reverence designators - !) to 100 ohms solves that problem, but a complimentary 100x increase in the ramp capacitor is impractical.

One solution is to change the transistor to an NPN Darlington type. This will reduce the required base current, so your other components should be fine. Where are you located, and what components do you have available? If you can't get a Darlington, stay with the ECG152 and create your own Darlington by adding an NPN small signal transistor such as a 2N2222, 2N3904, or 2M4401.



By the way, there is a variation of the standard 555 astable circuit that produces a nearly-perfect 50/50 duty cycle using only one resistor instead of two.

ak

 

I'm located near Chicago, the electronics stores nearby are a Micro Center and a store called American Science and Surplus, which has limited parts available, like a ECG153, ECG157, and a few other types of transistors. There's always Amazon, which I'd like to avoid because parts can only be bought in bulk or for huge markups on single parts.


If this was the case, what transistor would go where? what kind of resistor would I use, is there a formula or something to be had?

stay with the ECG152 and create your own Darlington by adding an NPN small signal transistor

 

If you want a fade up and down using a 555 I would suggest placing an Op-Amp in follower mode with the input connected to the timing cap, then amplify that signal with a transistor also in follower mode on the output of the Op-Amp.

Then feedback the control voltage pin with a 1k resistor from the output pin, that will change the charge/discharge range of the cap. (making it wider)

Use a Darlington or the suggested pairing to reduce the load on the Op-Amp. (this may not be needed as a transistor in follower mode uses little base current)

 

The #12 schematic shows how to connect them. The power transistor is to the right, and the small signal transistor is to the left.

To get the maximum voltage change across the LEDs, there is another transistor arrangement called a Sziklai pair. This is similar to a Darlington, but has a PNP power transistor connected to a NPN small-signal transistor. A bit messy to explain how it works, but it does. More later.

ak

 

The #12 schematic shows how to connect them. The power transistor is to the right, and the small signal transistor is to the left.

To get the maximum voltage change across the LEDs, there is another transistor arrangement called a Sziklai pair. This is similar to a Darlington, but has a PNP power transistor connected to a NPN small-signal transistor. A bit messy to explain how it works, but it does. More later.

I found a few 2N6498 and MJE1123 PNP transistors for the Sziklai pair, as well as some TIP125 and 2N6668 Darlington transistors laying around, how do you think those would work? The datasheets say they're PNP Darlington types, so I doubt they'd work.

I do appreciate all your help thus far, you've been really helpful.

 

How much current do you want to drive the LEDs with, what is the power supply voltage for the LEDs, and what is the power supply for the NE555?

We really need this kind of information

 

For a Sziklai pair your circuit, you want 1 small-signal NPN and one PNP power transistor. Neither of these can be Darlington devices. A good PNP part would be a TIP30, 32, or 34. The MJE1123 might work, but the data sheet says it is an "application" specific part designed to work with a Linear Tech regulator.

To prove that the circuit does what you want, jumping in with a Sziklai pair is not necessary. Just about any NPN power transistor with a driver, or NPN Darlington, will work.

ak

 

/

Not if you want the LEDs to fade up and down. That is what the original circuit does, and he said it works fine. That indicates to me that an uo-down fade is a goal of the project.

ak

OK, I missed that. Hmmm, a quick re-read: " I have a 555 timer producing a signal to pulse an LED. ", " the output looks like a triangle wave with peaks from around 5v to 3.5v, pulsing the LED ", says nothing about fading....

However, varying the voltage to 20 LED might not have the same result as to one - varying voltage to an LED is not a good way to vary brightness and the chances of them all fading identically is small. Would be better if they were all in series so they all operate at the same current; that would be more likely to get the desired result since light output is proportional to current...

 

I replaced the single NPN with a Sziklai pair using a 2N6489 (NOT a 2N6498, I made an error) and it seemed to have the desired effect of pulsing all 20 LEDs. I don't have any power NPN's at hand, hence the use of the Sziklai pair.

Would be better if they were all in series so they all operate at the same current; that would be more likely to get the desired result since light output is proportional to current...

A teacher of mine (masters in electrical) said I should have them in parallel, what changes would having them in series make?