|Hawes Mechanical Television Archive||by James T. Hawes, AA9DT|
LED Driver FAQ
BASIC QUESTIONS & PROBLEMS
QUESTION.Instead of the TIP120 Darlington that you specified, I built my own Darlington with two transistors. I can't seem to make the circuit work right. Why?
ANSWER. If you expect to duplicate my performance, then you must stick to my schematic. My reason for specifying parts is to eliminate substitution. Certainly, you can experiment. In fact, I encourage advanced builders to experiment. But don't expect your experimental circuit to act like my circuit. You might achieve equivalent results, but you probably won't.
Before you built your Darlington circuit, did you study datasheets for the TIP120 and for your two transistors? If you expect your circuit to act like a TIP120, then your transistor specs must match the TIP120.
Darlingtons aren't commodity components. Two Darlingtons can differ in gain by over a thousand! You'll find that Darlingtons differ more than say, general purpose replacement transistors do. For example, an NTE TIP120 has a current gain of 2500 (maximum). Fairchild's 2N6426 has a maximum current gain of 300,000. The current-handling specs for these two devices also vary widely. The TIP120 can handle four to eight series strings of LEDs. The 2N6426 can only handle one string. For mechanical video, the Darlington's top frequency is very important. Different devices have very different top frequencies.
A specification tolerance of say, 10 to 20 percent is often okay. Anyway, I can't say for sure. I didn't build all the circuit variations that you might think of. This circuit doesn't allow for a lot of variations. Notice that the input transistor has no emitter resistor. It's operating with very little negative feedback. The only current feedback comes from the transistor itself. (The formula is 26 / Ie in milliamperes. The term Ie is the emitter current. We know this formula as "Shockley's Constant.")
QUESTION. Instead of the 12-volt supply that you specified, I run my circuit off 18 volts. I can't seem to make the circuit work right. Why?
ANSWER. Don't expect a 12-volt circuit to work properly at 18 volts. A wise engineer once told me, "circuits that run on different voltages are different animals." In my novice years, I tried to prove him wrong. I found out that he wasn't kidding. Beginners should stick to the voltage, current and resistance specifications in the circuit. Unless you want to be disappointed, there are no exceptions. Running this circuit at the wrong power voltage is a great way to barbecue transistors and LEDs. The School of Hard Knocks teaches persuasively.
QUESTION. Okay. I really want to operate this circuit at a different voltage than you specify. This time, I proportioned all the resistors for another voltage. Why doesn't the circuit work well?
ANSWER. Let me guess: Your circuit clips, it burns up parts, it doesn't have enough gain or contrast, etc. Proportioning all the resistors is no way to redesign a transistor circuit. This "method" ignores the bias voltages. You probably tried this approach because you normally work with op amp ICs. Integrated circuits don't behave the same way as transistors do. A silicon transistor amplifier requires an external bias voltage. Regardless of power voltage, this bias voltage must be a certain level. The different amplifier classes use different amounts of bias. Like most preamplifiers, my LED-drive circuit has a class-A preamplifier. For class A, the transistor base should have a no-signal bias of 0.7 volt. You measure this bias with respect to the emitter. In a PNP stage, this voltage is negative. In an NPN stage, this voltage is positive.
Here's are examples of why proportioning all resistors won't provide proper operation:
In either case, the proportioned circuits probably also include collector resistor errors. Substitution of 20-percent standard resistor values might increase these errors. The original circuit uses a 20-percent tolerance. Now your derived circuit applies this tolerance again. The new circuit might be quite far from the ideal value. Improper collector or emitter resistors can lead to malfunctions, component failures or bad fidelity.
PROBLEM. The circuit doesn't respond properly to 1-volt video signals. These signals are the NBTVA mechanical TV standard.
The bottom of this page lists the NBTVA standards:
ANSWER. I didn't intend this amplifier to conform to NBTVA club video standards. Instead, this circuit is a transistorization of a 1928 Daven television amplifier. Besides, the input signal comes from a CD. By the way, how many CD players conform to the club video standard?
Most CD players have volume controls. To save the builder's wallet, I allow him to use that volume control. (Or add your own control. The choice and expense are yours.) With the control, you can select a level that operates the display without distorting. The volume control approach also allows the amplifier to work with video that might have too low or high a level.
If you want an exact, 1-volt input, design an amplifier or attenuator. Place this new circuit before my circuit. The attenuator can be as simple as a two-resistor voltage divider. If you don't understand voltage dividers or Ohm's Law, stick to the original circuit. For a tutorial on Ohm's Law, surf to Ohm's Law.
PROBLEM. The amplifier provides too much current or voltage for my display.
ANSWER. This design isn't infinitely adaptable. If you want to drive a different display, you need a different amplifier. The amplifier might be slightly different than this one, or grossly different. I suggest using the design formulas on this site. By the way, when I say "a different display," I mean a different number of LEDs, or a different LED type. For example, I designed this amplifier to use red LEDs. You can probably use the same circuit for orange or yellow LEDs. Blue, white or cyan LEDs require a different input voltage and a different amplifier. Some green LEDs will work with this circuit. Others won't. Multiple-chip LEDs also won't work with this circuit. Check the LED voltage and current specs, or stick with the original circuit.
QUESTION. Where do I connect the negative power supply?
ANSWER. This is a single-ended amplifier. It requires no negative supply.
I receive such questions from students who have mostly op amp experience. Of course most op amp projects require a negative supply. Yet without the extra power supply, transistors work just fine. Eliminating the second power supply is an advantage of transistors over typical differential op amps. Power supplies can add considerable bulk to an otherwise compact circuit. Without the extra supply, your transistor project might be smaller than an IC project. The transistor version will also be easier to build and debug.
I recommend that you examine other transistor circuits. Hobby circuits of the fifties, sixties and particularly the seventies provide helpful examples. Used electronics cookbooks from this period are quite economical.They turn up at Amazon, Barnes & Noble and other used book sites. Do a Google search on "used books," "transistor projects," etc. Also check your local library.
Copyright © 2007 by James T. Hawes. All rights reserved.