Voltage Inverter using a 555

Category: 555, DC to AC Inverter, Power Supply
In many circuits we need to generate an internal adjustable voltage. This circuit shows how it is possible to use a trusty old NE555 timer IC and a bit of external circuitry to create a voltage inverter and doubler. The input voltage to be doubled is fed in at connector K1. To generate the stepped-up output at connector K2 the timer IC drives a two-stage inverting charge pump circuit. The NE555 is configured as an astable multivibrator and produces a rectangular wave at its output, with variable mark- space ratio and variable frequency.

This results in timing capacitor C3 (see circuit diagram) being alternately charged and discharged; the voltage at pin 2 (THR) of the NE555 swings between one-third of the supply voltage and two- thirds of the supply voltage. The output of the NE555 is connected to two voltage inverters. The first inverter comprises C1, C2, D1 and D2. These components convert the rectangular wave signal into a negative DC level at the upper pin of K2. The second inverter, comprising C4, C5, D3 and D4, is also driven from the output of IC1, but uses the negative output voltage present on diode D3 as its reference potential.

Voltage Inverter Circuit using a 555

The consequence is that at the lower pin of output connector K2 we obtain a negative volt- age double that on the upper pin. Now let us look at the voltage feedback arrangement, which lets us adjust this doubled negative output voltage down to the level we want. The NE555 has a control voltage input on pin 5 (CV). Normally the voltage level on this pin is maintained at two-thirds of the supply voltage by internal circuitry. The voltage provides a reference for one of the comparators inside the device. If the reference voltage on the CV pin is raised towards the supply voltage by an external circuit, the timing capacitor C3 in the astable multivibrator will take longer to charge and to discharge.

As a result the frequency of the rectangle wave output from IC1 will fall, and its mark- space ratio will also fall. The source for the CV reference voltage in this circuit is the base-emitter junction of PNP transistor T1. If the base volt- age of T1 is approximately 500 mV lower than its emitter voltage, T1 will start to conduct and thus pull the voltage on the CV pin towards the positive supply. In the feedback path NPN transistor T2 has the function of a voltage level shifter, being wired in common-base configuration. The threshold is set by the resistance of the feedback chain comprising resistor R3 and potentiometer P1.

Voltage Inverter Circuit using a 555

When the emitter voltage of transistor T2 is more than approximately 500 mV lower than its base voltage it will start to conduct. Its collector then acts as a current sink. Potentiometer P1 can be used to adjust the sensitivity of the negative feedback circuit and hence the final output voltage level. Using T1 as a voltage reference means that the circuit will adjust itself to compensate not only for changes in load at K2, but also for changes in the input supply voltage. If K2 is disconnected from the load the desired output voltage will be maintained, with the oscillation frequency falling to around 150 Hz. A particular feature of this circuit is the somewhat unconventional way that the NE555’s discharge pin (pin 7) is connected to its output (pin 3).

To understand how this trick works we need to inspect the innards of the IC. Both pins are outputs, driven by internal transistors with bases both connected (via separate base resistors) to the emitter of a further transistor. The collectors of the output transistors are thus isolated from one another. The external wiring connecting pins 3 and 7 together means that the two transistors are operating in parallel: this roughly doubles the current that can be switched to ground. The two oscilloscope traces show how the output voltage behaves under different circumstances. The left-hand figure shows the behaviour of the circuit with an input voltage of 9 V and a resistive load of 470 Ω connected to the lower pin of output connector K2.

The figure on the right shows the situation with an input voltage of 10 V and a load of 1 kΩ on the lower pin of output connector K2. The pulse width and frequency of the rectangle wave at the output of IC1 are automatically adjusted to compensate for the differing con- ditions by the feedback mechanism built around T1 and T2. Because of the voltage drops across the Darlington out- put stage in the IC (2.5 V maximum) and the four diodes (700 mV each) the circuit achieves an efficiency at full load (470 Ω between the output and ground) of approximately 50 %; at lower loads (1 kΩ) the efficiency is about 65 %.
Continue Reading...

Power LED Driver

Category: LED and Light, LED Driver
If you want to operate power LEDS with a truly constant current – which significantly prolongs the lifetime of the lamp – and avoid the power loss resulting from using a constant voltage supply with a series resistor, you need a suitable constant current source. However, the only way to achieve really good efficiency is to use a switching regulator. Altogether, this means that you need a switching regulator designed to generate a constant current instead of a constant voltage. With this in mind, the author started working on the development of a LED pocket torch with especially high efficiency.

Power LED Driver Circuit Diagram

Along with using high-capacity rechargeable batteries to maximise operating life, it’s worthwhile to be able to reduce the brightness, and therefore the operating current of the LEDs, when you don’t need full power. Accordingly, the author incorporated a dimming function in the design, based on operation in PWM mode in to reduce power losses to an absolute minimum. As you can see from the circuit diagram, the author chose an LT3518 switching regulator IC, which is a buck/boost converter optimised for LED operation. Here it is used as a down converter (buck mode). This IC can achieve better than 90% efficiency in this mode, depending on the input voltage.

According to the typical application circuit on the data sheet, its switching frequency can be set to approximately 170 kHz by selecting a value of 82 kΩ for R1. To maximise overall efficiency with this type of IC, the volt- age drop over the sense resistor used to measure the current flowing through the LED should be as low as possible. This particular device operates with a voltage drop of 100 mV, corresponding to a current of just under 1.5 A with the specified value of 68 mΩ for R2. This value proved to be suitable for the Cree LED used by the author. At this current level, a diode with a power rating of at least 6 W should be used for D1.

IC1 has an additional property that is ideal for this application: the connect- ed LED can be dimmed by applying a PWM signal to pin 7 of the IC, with the brightness depending on the duty cycle. Obviously, the PWM frequency must be lower than the switching fre- quency. The PWM signal is provided by IC2, a special voltage-controlled PWM generator (type LTC6992 [2]). The duty cycle is controlled by the volt- age applied to the MOD input on pin 1 (range 0–1 V). The resistor connected to pin 3 determines the internal clock frequency of the IC according to the formula f = 1 MHz × (50 kΩ/R3).

This yields a frequency of approximately 73.5 kHz with R3 set to 680 kΩ, which is much too high for controlling IC1. However, the PWM IC has an internal frequency divider with a division factor controlled by the voltage applied to pin 4, which in this circuit is taken from voltage divider R4/R5. The division fac- tor can be adjusted over the range of 1 to 16,384. The division factor with the specified component values is 64, resulting in a PWM frequency of around 1,150 Hz. If you want to be able to generate a PWM signal with an adjust- able duty cycle over the full range of 0 to 100%, you must use the LTC6992-1 option.

The -4 option, which provides a range from 5 to 100%, might be an acceptable alternative. To prevent the duty cycle (and thus the brightness of the LED) from depending on the battery voltage, which gradually drops as the battery discharges, IC3 generates a stabilised 1.24 V control voltage for potentiometer P1. Series resistor R7 reduces the voltage over P1 to 1 V, which exactly matches the input voltage range of the LTC6992. All capacitors should preferably be ceramic types, in particular due to their low effective series resistance (ESR) as well as other favourable characteristics.

However, only capacitors with X5R or X7R dielectric should be used; capacitors with type Y dielectric have very poor temperature characteristics. The supply voltage is limited to 5.5 V by the maximum rated supply voltage of IC2. The author used four NiMH re- chargeable cells connected in series, which yields a voltage that is just within spec. With an operating voltage in the range of 4.5V to 5.5V, you must use an LED that can operate at less than 4 V. This eliminates devices with several chips connected in series on a carrier, which is very often the case with power LEDS rated at over 5 W. These devices require a correspondingly higher supply voltage, which means more cells connected in series.

This is only possible if the supply voltage for IC2 is reduced by a 5 V voltage regulator or other means, and of course R4 must also be connected to this lower supply voltage. Finally, a few words about soldering. An exposed thermal pad must be provided on the PCB for the LT3518, and the rear face of the IC must be soldered to this pad. The author obtained good results by dimensioning the exposed pad large enough to extend beyond the outline of the IC. When assembling the board, first tin the pad and the rear face of the IC. Then heat the pad with a soldering iron. When the solder melts, withdraw the tip of the soldering iron to the edge of the pad and simultaneously place the IC on the pad and align it. After this the pins can be soldered.

By Michael Hölzl (Germany)
Continue Reading...

‘Green’ Solar Lamp

Category: LED and Light, LED Flasher, Solar Power
Energy saving is all the rage, and here is our small contribution: how much (or rather how little) current do we need to light an LED? Experiments with a super-bright 1 W green LED showed that even one microamp was enough to get some visible light from the device. Rootling in the junk box produced a 0.47 F memory back-up capacitor with a maximum working voltage of 5.5 V. How long could this power the green LED? In other words, if discharged at one microamp, how long would the voltage take to drop by 1V?

‘Green’ Solar Lamp Circuit Diagram

A quick calculation gave the answer as 470 000 seconds, or about five days. Not too bad: if we use the capacitor for energy storage in a solar-powered lamp we can probably allow a couple more microamps of current and still have the lamp on throughout the night and day. All we need to add is a suitable solar panel. The figure shows the circuit diagram of our (in every sense) green solar lamp.

By Burkhard Kainka (Germany, Elektor)
Continue Reading...

Wideband Wien Oscillator with Single-Gang Pot

Category: Miscellaneous Oscillator, Radio
This Wien bridge oscillator (after Max Wien, 1866–1938) produces a low-distortion sine wave of constant amplitude, from about 15 Hz to 150 kHz. It requires just four opamps and will work off a single 9-volt battery. Also, unlike most Wien bridge oscillators, it does not require a dual-gang potentiometer for tuning. Op amp IC2b provides an artificial ground so that the circuit will operate from a unipolar supply (9 V battery or power pack). IC2a is the main amplifier for the oscillator. The frequency range is divided into four decades by 2-pole, 4-way rotary switch SW1.

Only one arm of the Wien network is varied, but the change in positive feedback that would normally result is compensated for by IC1b, which works to bootstrap R2, thereby changing the negative feedback enough to maintain oscillation. A linear change in the resistance of the tuning pot results in a roughly logarithmic change in frequency. To get a more conventional linear change a log-taper pot is used wired so that rotating the knob anticlockwise causes frequency to increase.

You could use an anti-log pot the other way around if you prefer, but these things are notoriously hard to find. IC1A is an integrator that monitors the amplitude of the output signal and drives an LED (D2). This must be mounted facing the LDR (light dependent resistor) and shielded from ambient light (for example, with a piece of heat-shrink tubing). IC1a is then able to control the gain of IC2a so that oscillation is maintained with minimum distortion.

The maximum output amplitude of the generator is about 2 Vp-p when the LED and LDR are mounted as close as possible. Distortion is less than 0.5 % in the lowest range, and too low for the author to measure in the higher ranges. Any LDR should work, provided its dark resistance is greater than 100 kO. If you do not have an LDR with such high resistance, try increasing R5 until oscillation starts. Breadboarded prototypes of the circuit were built by the author using dual and quad opamp packages, and both work equally well.

Author: Merlin Blencowe (Elektor)

R1,R2,R3,R6,R10,R11 = 10kO
R7 = 100kO
R4,R9,R12 = 100O
R5 = 12kO
R8 = 1kO
P1,P2 = 10kO potentiometer, logarithmic law
R13 = LDR, R(dark) >100kO, e.g. Excelitas Tech type
VT90N1 (Newark/Farnell # 2568243)

C1,C5 = 1µF solid
C2,C6 = 100nF
C3,C7 = 10nF
C4,C8 = 1nF
C9-C12 = 47µF 16V, electrolytic, radial

D1,D2,D3 = 1N4148
D4 = LED, red, 5mm
IC1,IC2 = TL072ACP

SW1 = 2-pole 4-position rotary switch, C&K Compo-
nents type RTAP42S04WFLSS
K1,K2 = PCB terminal block, 5mm pitch
Continue Reading...

4 Amps Photovoltaic (Solar) Charge Controller

Category: DC Power Supply, Solar Power
The use of solar photovoltaic (PV) energy sources is increasing due to global warming concerns on the one hand, and cost effectiveness on the other. Many engineers involved in power electronics find solar power tempting and then addictive due to the ‘green’ energy concept. The circuit discussed here handles up to 4 amps of current from a solar panel, which equates to about 75 watts of power. A charging algorithm called ‘pulse time modulation’ is introduced in this design. The current flow from the solar panel to the battery is controlled by an N-channel MOSFET, T1. This MOSFET does not require any heat sink to get rid of its heat, as its RD-S(on) rating is just 0.024 Ω.

Schottky diode D1 prevents the battery discharging into the solar panel at night, and also provides reverse polarity protection to the battery. In the schematic, the lines with a sort-of-red highlight indicate potentially higher current paths. The charge controller never draws current from the battery—it is fully powered by the solar panel. At night, the charge controller effectively goes to sleep. In daytime use, as soon as the solar panel produces enough current and voltage, it starts charging the battery. The battery terminal potential is divided by resistor R1 and trimpot P1.

4 Amps Photovoltaic (Solar) Charge Controller Circuit DIagram
4 Amps Photovoltaic (Solar) Charge Controller Circuit Diagram
The resulting voltage sets the charge state for the controller. The heart of the charge controller is IC1, a type TL431ACZ voltage reference device with an open-collector error amplifier. Here the battery sense voltage is constantly compared to the TL431’s internal reference voltage. As long as the level set on P1 is below the internal reference voltage, IC1 causes the MOSFET to conduct. As the battery begins to take up the charge, its terminal volt- age will increase. When the battery reaches the charge-state set point, the output of IC1 drops low to less than 2 volts and effectively turns off the MOSFET, stopping all current flow into the battery.

With T1 off, LED D2 also goes dark. There is no hysteresis path provided in the regulator IC. Consequently, as soon as the current to the battery stops, the output of IC1 remains low, preventing the MOSFET to conduct further even if the battery voltage drops. Lead-acid bat- tery chemistry demands float charging, so a very simple oscillator is implemented here to take care of this. Our oscillator exploits the negative resistance in transistors—first discovered by Leo Esaki and part of his studies into electron tunneling in solids, awarded with the Nobel Prize for Physics in 1973. In this implementation, a commonplace NPN transistor type 2SC1815 is used.

When the LED goes out, R4 charges a 22-μF capacitor (C1) until the voltage is high enough to cause the emitter-base junction of T2 to avalanche. At that point, the transistor turns on quickly and discharges the capacitor through R5. The voltage drop across R5 is sufficient to actuate T3, which in turn alters the reference voltage setting. Now the MOSFET again tries to charge the battery. As soon as the battery voltage reaches the charged level once more, the process repeats. A 2SC1815 transistor proved to work reliably in this circuit. Other transistors may be more temperamental—we suggest studying Esaki’s laureate work to find out why, but be cautioned that there are Heavy Mathematics Ahead.

As the battery becomes fully charged, the oscillator’s ‘on’ time shortens while the ‘off’ time remains long as determined by the timing components, R4 and C1. In effect, a pulse of current gets sent to the battery that will shorten over time. This charging algorithm may be dubbed Pulse Time Modulation. To adjust the circuit you’ll need a good digital voltmeter and a variable power supply. Adjust the supply to 14.9 V, that’s the 14.3 volts bat- tery setting plus approximately 0.6 volts across the Schottky diode.

Turn the trimpot until at a certain point the LED goes dark, this is the switch point, and the LED will start to flicker. You may have to try this adjustment more than once, as the closer you get the comparator to switch at exactly 14.3 V, the more accurate the charger will be. Disconnect the power supply from the charge controller and you are ready for the solar panel. The 14.3 V setting mentioned here should apply to most sealed and flooded-cell lead-acid batter- ies, but please check and verify the value with the manufacturer. Select the solar panel in such a way that its amps capability is within the safe charging limit of the battery you intend to use.

Author: T. A. Babu (India - Elektor)

R1 = 15kΩ
R2,R3 = 3.3kΩ 1% R4 = 2.2MΩ
R5 = 1kΩ
P1 = 5kΩ preset

C1 = 22μF 25V, radial

D1 = MBR1645G (ON Semiconductor) D2 = LED, 5mm
IC1 = TL431ACLP (Texas instruments)
T1 = IRFZ44NPBF (International Rectifier)
T2 = 2SC1815 (Toshiba) (device is marked: C1815)
T3 = BC547

K1,K2 = 2-way PCB terminal block, lead pitch 5mm
Continue Reading...