Sometimes you want to have specific modules in your lab that are not available off the shelf — a compact Simple DC Dimmer Circuit/regulator module, for example. That’s why I decided to develop my own module, which consists of just a few standard electronic components. The module can be used as a key element to control the brightness of DC lamps and/or the speed of small DC motors in portable hand tools like your PCB drill.
My Rough Idea
Demoed below is what I’ve designed to handle the pulse-width modulation control:
In this basic, op-amp–based, astable design, components R1 and C1 set the operating frequency as the astable operation relies on repeatedly charging and discharging the capacitor from the output of the op amp through the resistor. The other three resistors (R2, R3, and R4) control the minimum and maximum voltage that the capacitor charge comes across during a cycle of oscillation. The output at the capacitor (Pin 2 of IC1) is shaped like a sawtooth wave, but a square wave (SQW) is available from the output (Pin 1 of IC1) of the oscillator. For the given configuration, the frequency is fairly close to 5 kHz.
The “fixed” (frequency and duty cycle) square wave signal coming from the output of the oscillator is not used here (it can be used for some other application in demand). Instead, the second op amp compares the sawtooth waveform (from Pin 2 of IC1) with a variable-voltage level introduced by a potentiometer (P1) to create the final square wave signal (Pin 7 of IC1) with a fixed frequency but with a duty cycle (percentage of the on-time over one period) counts on the position of the potentiometer. The following oscillogram shows the impact of the potentiometer travel on the final output signal and two random square waves (one with 10% duty cycle and the other with 90% duty cycle). One noteworthy fact is that 0- to 5-V output from a microcontroller can replace the potentiometer!
A Bad Mistake
Even though the LM358 is a very popular and cheap dual op amp, it’s not a good fit for the proposed application! Let me explain why:
First off, LM358 is not a “rail-to-rail” op amp; hence, its voltage levels will be throttled, and my basic design may not work in every situation. Actually, LM358 allows its inputs to go down to the negative rail but does not allow the inputs to get closer to the positive supply rail than 1.5 V (i.e., 3.5 V @ 5 V VCC). Furthermore, there’s a drop in the output voltage (usually 1.4 V), so it’s 3.6 V (not 5 V). However, a rail-to-rail op amp works well at lower operating voltages, swings close to the supply rails, and offers wider dynamic range.
Next is the “slew rate” (may not be debatable in a low-frequency astable) of the op amp. Slew rate is the rate of change of the output of the op amp in a given time, and it limits the circuit operation if the slew rate demand is outstripped. Typical slew rate for LM358 is 0.5 V/μs, but another high-speed op amp with a higher slew rate would be fine for an astable circuit working in kilohertz/megahertz scale.
To elucidate this, I’m doing a back-of-the-napkin calculation: With 0.5-V/us slew rate, LM358 would take 24 µs for the output voltage to change by the 12 V between –6 V to 6 V on a standard dual power supply. The 5-kHz square wave has a periodic time of 200 μs, and the slopes of the rising and falling edges would infest 24% of the whole cycle of the wave. But for a 50-Hz square wave in the same situation, this will be just 0.24%. Yes, it does matter at higher frequencies!
Luckily, now we can see that a number of op amps meet the requisite specification, and some may also work from 5 V or even less. My next experimental trials will be with one rail-to-rail dual op amp — the LMV358 from Fairchild (www.fairchildsemi.com) and/or the MCP6562 from Microchip Technology Inc.
The Power Driver
You can use the PWM output signal (OUT) from the module to control the brightness of common DC lamps and the speed of small DC motors using an IRLZ44 (or similar) “logic-level gate drive” MOSFET. See the simplified schematic:
Dimming a 12-V (0.72-W) LED sign module with the dimmer/regulator circuit is pretty straightforward. In this experiment, I’m using one 5050-3LED sign module with blue LEDs.
