Solar hot water controller

This circuit functions as a control unit in a solar hot water system (HWS). The temperature at the top of the panels is compared with that in the tank and when the Sun shines brightly enough, the pump is switched on. Water continues to circulate through the panels for as long as the temperature in the panels is greater than that in the tank.

Note that most solar hot water systems don’t require a circulator pump as the panels are mounted below the tank and the natural thermo-siphon effect is relied upon to circulate the water. However, in situations where it is more cost-effective to have the panels at the same level or higher than the tank, a circulator pump is required, hence the impetus for this design.

The circuit of the solar hot water controller uses a series of LM335 temperature sensors (TS1-TS6) to monitor the temperatures in the tank and the various panels. Their outputs are monitored by comparators IC1a-IC1c which in turn control a pump via a relay.

The circuit includes an anti-freeze feature that starts the pump when the water temperature in the lower panel drops below 4°C. With the addition of a digital panel meter, it can also be used to monitor water temperature in multiple locations around the system.

A series of LM335s (TS1-TS6) are used as temperature sensors. Controller operation is based around sensors TS1 and TS2, which measure the temperature in the tank and top of the second panel. One element of an LM339 quad comparator (IC1b) compares the voltages from these two sensors. A higher voltage on the inverting input (pin 6) than the non-inverting input (pin 7) signals a higher panel temperature. This causes the output of the comparator to swing low, switching on Q1 and energising the relay (RLY1). This in turn applies power to the pump. A 2.2MΩ resistor affords some positive feedback around IC1b, ensuring jitter-free relay switching.

A second comparator (IC1a) in the package is used to monitor the temperature in the bottom of the first panel for the anti-freeze function. The inverting input (pin 4) is supplied with a 2.77V reference, whereas the non-inverting input is connected to TS3. As these sensors are calibrated directly in °K, they have an output of +2.73V at 0°C. Therefore, once the water in the panel drops to below 4°C (2.77V), the voltage at the non-inverting input will be less than the reference voltage and the comparator output swings low. This forward-biases D3 and switches on Q1, again energising the relay and starting the pump.

A third comparator in the package (IC1c) is used to provide indication that the anti-freeze function has been activated (apart from the fact that the panels aren’t frozen!). If IC1a’s output goes low, the non-inverting input (pin 9) is pulled lower than the inverting input (pin 8) and its output goes low, turning on LED1. At the same time, current is drawn through the base of Q2, turning it on and providing positive feedback via the 100kΩ resistor to the inverting input. This causes the output to remain latched in the on (low) state, keeping the "anti-freeze" LED illuminated even after the pump has been switched off. To reset the circuit, switch S2 must be pressed, overriding the positive feedback from the comparator’s output.

A digital panel meter (DPM) provides a convenient means of displaying water temperature at various points in the system. As well as the three sensors mentioned above, the author added three more sensors (TS4-TS6) just for monitoring purposes. The output from any of these sensors can be displayed on the DPM with the aid of a 6-position rotary switch (S1). The series chokes (L1-L6) and 100nF shunt capacitors are included to filter out RF interference, necessary because the controller is situated close to a ham radio antenna.

In order to read degrees Celsius directly, the negative input of the DPM is offset with a 2.73V reference, corresponding to 0°C. This voltage originates from a REF50Z temperature-compensated precision reference. The 5V output from the reference (REF1) is divided down by trimpot VR1 and a string of resistors. The trimpot should be adjusted for precisely 2.73V between the negative input of the DPM and ground. If readout accuracy is non-critical, then REF1 can be replaced with a (cheaper) 5.1V zener diode.

As shown, the circuit is powered from a small 24V centre-tapped transformer, with regulator REG1 giving a stabilised +12V output. Take care to ensure that all 240VAC wiring is properly terminated and insulated. The project can be be housed in a plastic instrument case that’s protected from the elements.

Two basic motor speed controllers Here are two simple 12V DC motor speed controllers that can be built for just a few dollars. They exploit the fact that the rotational speed of a DC motor is directly proportional to the mean value of its supply voltage. Fig.1: a very simple motor speed controller based on a compound emitter follower (Q1 & Q2). The first circuit shows how variable voltage speed control can be obtained via a potentiometer (VR1) and compound emitter follower (Q1 & Q2). With this arrangement, the motor’s DC voltage can be varied from 0V to about 12V. This type of circuit gives good speed control and self-regulation at medium to high speeds but very poor low-speed control and slow starts. Fig.2: this slightly more complicated circuit gives better low speed control and higher torque. The second circuit uses a switchmode technique to vary motor speed. Here a quad NOR gate (IC1) acts as a 50Hz astable multivibrator that generates a rectangular output. The mark-space ratio of the rectangular waveform is fully variable from 20:1 to 1:20 via potentiometer VR1. The output from the multivibrator drives the base of Q1, which in turn drives Q2 and the motor. The motor’s mean supply voltage (integrated over a 50Hz period) is thus fully variable with VR1 but is applied in the form of high-energy "pulses" with peak values of about 12V. This type of circuit gives excellent full-range speed control and gives high motor torque, even at very low speeds. Its degree of speed self-regulation is proportional to the mean value of the applied voltage. Note that for most applications, the power transistor (Q2) in both circuits will need to be mounted on an appropriate heatsink. Ravi Sumithraarachchi, Colombo, Sri Lanka. ($50)

Op amp building blocks

Here’s a series of basic op amp circuits that have a multitude of uses as building blocks in larger circuits. They all use a minimum number of components and with one exception, component values are non-critical. All op amps are FET-input types such as the TL071/2/4 single/dual/quad varieties and all diodes are small-signal 1N4148s.

Fig.1

All circuits are derived from the basic function block shown in Fig.1, which we’ll refer to as a "MAX" function. Its operation is as follows; if the voltage applied to V2 is less than V1, the output of the op amp is close to the negative supply rail, reverse-biasing diode D1. The output voltage is then just V1, as seen through the 100kΩresistor.

Conversely, if V2 is greater than V1, the op amp’s output swings positive so that D1 is forward biased and the voltage at the inverting input of the op amp (and hence V_out) is equal to V2.

For best results, V1 should be driven by a low-impedance source such as an op amp connected as a voltage follower. The value of the input resistor (shown as 100kΩ) is not critical. In addition, any circuitry connected to V_out should have an impedance greater than about 1MΩ.

Reversing D1 gives a "MIN" function block (not shown), whose operation should be self-explanatory.

Fig.2

Fig.2 shows a precision clipper, made by merging a MAX and a MIN function block. The signal at V_in is transformed to the signal at V_out by clipping it when it is greater than V1 or less than V2. As before, V1 should be driven by a low-impedance source and any circuitry connected to V_out should have an impedance greater than about 1MΩ.

Fig.3

Fig.3 shows a precision full wave rectifier. Op amp A, resistor R1, and diode D1 form a half-wave rectifier (this part of the circuit is equivalent to a MAX function block with V2 equal to 0V). Op amp B is configured with resistors R2 and R3 to subtract the original input signal at V_in from twice the half-wave rectified signal, giving the full wave rectified signal at V_out.

This circuit needs fewer matched resistors than some other designs. For linear operation, R2 and R3 should be equal. The value of R1 is not critical. Once again, V1 should be driven by a low-impedance source such as an op amp connected as a voltage follower.

Fig.4

Fig.4 shows a precision 2-way signal selector. It is made from two MIN function blocks (op amps B and C), one MAX function block (op amp D), and an op amp wired as an inverter (op amp A). None of the resistor values are critical nor do they have to be matched to achieve linear operation. For best results, "select" should be driven by a source with an impedance of less than about 10kΩ and any circuitry connected to V_out should have an impedance greater than about 1MΩ.