Custom Search

Monday, October 11, 2010

Solar Relay

With extended periods of bright sunshine and warm weather, even relatively large storage batteries in solar-power systems can become rather warm. Consequently, a circuit is usually connected in parallel with the storage battery to either connect a high-power shunt (in order to dissipate the excess solar power in the form of heat) or switch on a ventilation fan via a power FET, whenever the voltage rises above approximately 14.4 V. However, the latter option tends to oscillate, since switching on a powerful 12-V fan motor causes the voltage to drop below 14.4 V, causing the fan to be switched off.

In the absence of an external load, the battery voltage recovers quickly, the terminal voltage rises above 14.4 V again and the switching process starts once again, despite the built-in hysteresis. A solution to this problem is provided by the circuit shown here, which switches on the fan in response to the sweltering heat produced by the solar irradiation instead of an excessively high voltage at the battery terminals. Based on experience, the risk of battery overheating is only present in the summer between 2 and 6 pm. The intensity of the sunlight falling within the viewing angle of a suitably configured ‘sun probe’ is especially high precisely during this interval.

This is the operating principle of the solar relay. The trick to this apparently rather simple circuit consists of using a suitable combination of components. Instead of a power FET, it employs a special 12-V relay that can handle a large load in spite of its small size. This relay must have a coil resistance of at least 600 Ω, rather than the usual value of 100-200 Ω. This requirement can be met by several Schrack Components relays (available from, among others, Conrad Electronics). Here we have used the least expensive model, a type RYII 8-A printed circuit board relay. The light probe is connected in series with the relay. It consists of two BPW40 phototransistors wired in parallel.The type number refers to the 40-degree acceptance angle for incident light. In bright sunlight, the combined current generated by the two phototransistors is sufficient to cause the relay to engage, in this case without twitching. Every relay has a large hysteresis, so the fan connected via the a/b contacts will run for many minutes, or even until the probe no longer receives sufficient light. The NTC thermistor connected in series performs two functions. First, it compensates for changes in the resistance of the copper wire in the coil, which increases by approximately 4 percent for every 10 ºC increase in temperature, and second, it causes the relay to drop out earlier than it otherwise would (the relay only drops out at a coil voltage of 4 V).

Depending on the intended use, the 220-Ω resistance of the thermistor can be modified by connecting a 100-Ω resistor in series or a 470-Ω resistor in parallel. If the phototransistors are fastened with the axes of their incident-angle cones in parallel, the 40-degree incident angle corresponds to 2 pm with suitable solar orientation. If they are bent at a slight angle to each other, their incident angles overlap to cover a wider angle, such as 70 degrees. With the tested prototype circuit, the axes were oriented nearly parallel, and this fully met our demands. The automatic switch-off occurs quite abruptly, just like the switch-on, with no contact jitter.

This behaviour is also promoted by the NTC thermistor, since its temperature coefficient is opposite to that of the ‘PTC’ relay coil and approximately five times as large. This yields exactly the desired effect for energising and de-energising the relay: a large relay current for engagement and a small relay current for disengagement. Building the circuit is actually straightforward, but you must pay attention to one thing. The phototransistors resemble colourless LEDs, so there is a tendency to think that their ‘pinning’ is the same as that of LEDs, with the long lead being positive and the short lead negative. However, with the BPW40 the situation is exactly the opposite; the short lead is the collector lead. Naturally, the back-emf diode for the relay must also be connected with the right polarity. The residual current on cloudy days and at night is negligibly small.

High Side Current Measurements

It’s always a bit difficult to measure the current in the positive lead of a power supply, such as a battery charger. Fortunately, special ICs have been developed for this purpose in the last few years, such as the Burr-Brown INA138 and INA168. These ICs have special internal circuitry that allows their inputs to be connected directly to either end of a shunt resistor in the lead where the current is to be measured. The shunt is simply a low-value resistor, across which a voltage drop is measured whenever a current flows. This voltage is converted into an output current Io by the IC.

This current can be used directly, or it can be converted into a voltage by means of a load resistor RL. In the latter case, the ‘floating’ measurement voltage across the shunt is converted into a voltage with respect to earth, which is easy to use. The value of RL determines the gain. A value of 5 kΩ gives 1×, 10 kΩ gives 2×, 15 kΩ gives 3× and so on. It all works as follows. Just like any opamp, this IC tries to maintain the same potential on its internal plus and minus inputs. The minus input is connected to the left-hand end of the shunt resistor via a 5-kΩ resistor.When a current flows through the shunt, this voltage is thus lower than the voltage on the plus side. However, the voltage on the plus input can be reduced by allowing a small supplementary current to flow through T1. The IC thus allows T1 to conduct just enough to achieve the necessary lower voltage on the plus input. The current that is needed for this is equal to Vshunt / 5 kΩ. This transistor current leaves the IC via the output to which RL is connected. If the value of RL is 5 kΩ, the resulting voltage is exactly the same as Vshunt. The IC is available in two versions.

The INA138 can handle voltages between 2.7 and 36 V, while the INA168 can work up to 60 V. The supply voltage on pin 5 may lie anywhere between these limits, regardless of the voltage on the inputs. This means that even with a supply voltage of only 5 V, you can make measurements with up to 60 V on the inputs! However, in most cases it is simplest to connect pin 5 directly to the voltage on pin 3. Bear in mind that the value of the supply voltage determines the maximum value of the output voltage. Also, don’t forget the internal base-emitter junction voltage of T1 (0.7 V), and the voltage drop across the shunt also has to be subtracted.

Measuring Inductors

Often you find yourself in the position of needing to wind your own coil for a project, or maybe you come across an unmarked coil in the junk-box. How can you best find out its inductance? An oscilloscope is all you need. Construct a resonant circuit using the coil and a capacitor and connect it to a square wave generator (often part of the oscilloscope itself) Adjust the generator until you find the resonant frequency f.
When C is known (1000pF) the inductance L may be calculated from: L=1/(4π2.f2.C) If you are also interested how good the coil is i.e. what is its quality factor or Q, you can use the oscilloscope again. If the level of the damped oscillation drops to 0.37 (= 1/e) of the maximum after about 30 periods, then the Q factor of the coil is about 30. The Q factor should be measured at the intended operating frequency of the coil and with its intended capacitor. The coupling capacitor should by comparison be a much smaller value.