In this column, Russ answers questions about
all aspects of electronics, including computer
hardware, software, circuits, electronic theory,
troubleshooting, and anything else of interest to
the hobbyist. Feel free to participate with your
questions, comments, or suggestions. Send all
questions and comments to: Q&A@nutsvolts.com
20 January 2014
■ WITH RUSSELL KINCAID Q & A
Post comments on this article and find any associated files
and/or downloads at www.nutsvolts.com/index.php?/
• LED Driver
• Remote Control
• Video and Power Over Cable
QI want to drive a 12 VDC one amp 10 watt LED. I am using a 2N3903 transistor for the output. Can you design a circuit for me using a MOSFET or bipolar transistor? I would like to try a solid-state device instead of a relay. I thought about using a
PN2222A transistor, but I don’t think it has enough power
to handle the LED.
— Jeff Miller
AYou are right about the PN2222A, it is rated at 600 mA; normally, you would want a transistor ated several times the load current of one amp. The device that I chose is rated at six amps and
is in a tiny six leaded package. Four of the leads are
connected to the drain to maximize heat dissipation. The
schematic is shown in Figure 1; all the parts are surface-mount and are hand solderable.
The MOSFET is FDC637AN (512- is a Mouser part
number). It is rated 20V, six amps, and the gate is rated 8V
max, so I put a 6.2V zener at the gate to limit the voltage.
The ON resistance is .024 ohms, so the power dissipation
at one amp will be 24 milliwatts.
Probably don’t need any heatsinking but I did add
some copper to the layout just in case; see Figure 2. The
layout is 0.95 x 0.6 inches (0.57 sq in).
QI need a frequency of 49.848230 kHz for my project. I bought a crystal from Jan Crystals but they sent one for 49.848 MHz third overtone, which oscillates at 16.616 MHz at
the fundamental. I think a VCXO will be needed because
I want to be exactly on frequency. Can you come up with
— James New
AI didn’t want to deal with Jan Crystals to replace the crystal, because I want to divide from a higher frequency — but not that high. The oscillator will have some second harmonic
distortion but the divided square wave will have zero
second harmonic. I will low-pass filter the square wave
with an elliptic filter that has a notch at the third
harmonic, so the first visible harmonic will be the fifth.
The filter response at the fifth harmonic (249 kHz) will
be far down, so there will be essentially no harmonics. I
chose a frequency of 1.5951433 MHz which is 32 times
■ FIGURE 1.
■ FIGURE 2.