High Voltage Generator

This high voltage generator was designed  with the aim of testing the electrical break-down protection used on the railways. These  protection measures are used to ensure that  any external metal parts will never be at a  high voltage. If that were about to happen,  a very large current would flow (in the order  of kilo-amps), which causes the protection  to operate, creating a short circuit to ground effectively earthing the metal parts. This hap-pens when, for example, a lightning strike hits  the overhead line (or their supports) on the  railways.

This generator generates a high voltage of  1,000 V, but with an output current that is limited to few milliamps. This permits the electrical breakdown protection to be tested with-out it going into a short circuit state. The circuit uses common parts throughout: a  TL494 pulse-width modulator, several FETs or  bipolar switching transistors, a simple 1.4 VA  mains transformer and a discrete voltage multiplier. P1 is used to set the maximum current  and P2 sets the output voltage.

High Voltage Generator Circuit Diagram

The use of a voltage multiplier has the advantage that the working voltage of the smoothing capacitors can be lower, which makes them easier to obtain. The TL494 was chosen  because it can still operate at a voltage of  about 7 V, which means it can keep on working even when the batteries are nearly empty.  The power is provided by six C-type batteries, which keeps the total weight at a reason-able level.

The 2x4 V secondary of AC power transformer  (Tr1) is used back to front. It does mean that  the 4 V winding has double the rated voltage  across it, but that is acceptable because the  frequency is a lot higher (several kilo-Hertz)  than the 50 Hz (60 Hz) the transformer is  designed for. The final version also includes a display of the  output voltage so that the breakdown volt-age can be read.

From a historical perspective there follows a  bit of background information. In the past a different system was worked  out. Every high-voltage support post has a  protection system, and it isn’t clear when  the protection had operated and went into  a short-circuit state due to a large current  discharge.

Since very large currents were involved, a certain Mr. Van Ark figured out a solution for this.  He used a glass tube filled with a liquid containing a red pigment and a metal ball. When  a large current discharge occurred the metal  ball shot up due to the strong magnetic field,  which caused the pigment to mix with the liquid. This could be seen for a good 24 hours after the event. After a thunder storm it was  easy to see where a discharge current took  place: one only had to walk past the tubes  and have a good look at them.

Unfortunately, things didn’t work out as  expected. Since it often took a very long  time before a discharge occurred, the pigment settled down too much. When a dis-charge finally did occur the pigment no  longer mixed with the liquid and nothing was  visible. This system was therefore sidelined,  but it found its place in the (railway) history  books as the ‘balls of Van Ark’.

Author : By Jac Hettema – Copyright : Elektor

Source:  http://www.ecircuitslab.com/2012/03/high-voltage-generator.html

Read More...

Automatic Switch For Voltage Converters

New applications for DC voltage converters, such as the ‘workhorse’ LT1070, arise every day. These converters can be adapted to nearly every imaginable ratio of input and output voltages. However, all of these circuits and devices have the same shortcoming, which is that they lack an on/off switch. Especially when they are used as a source of 6-V / 12-V power for a car radio, this is highly impractical. The circuit described here adds automatic load detection to the converter. For use in a car, the additional circuitry must be small and fit into a compact enclosure together with the converter.

Since the battery voltage and ambient temperature vary over wide ranges, a simple form of load detection must be used. Besides this, the voltage drop across the load sensing circuitry must naturally be as small as possible. This can be achieved by using ‘ultra-modern’ SiGe technology. The 6 V from the battery and the 12 V from the converter are combined in the MB R2545 dual diode. Consequently, a voltage of at least 6 V is always applied to the radio (for memory retention). If the radio is switched on, it draws a current from the 6-V battery, which may be around 100 mA.

This current produces a voltage across R1. If this voltage is 75 mV or greater, the AC128 germanium transistor starts conducting and charges electrolytic capacitor C1, which is connected to the gate of the BUZ10. The MOSFET energises RE1 and thus connects the supply voltage to the converter. As a result, 12-V power is connected to the radio. The resulting increased current causes the voltage drop across R1 to increase, which is undesirable, so a 10-A Schottky diode is connected in parallel. The total voltage drop is thus approximately 0.6 V. The RC network connected to the BUZ10 ensures that the transistor always remains switched on for at least several seconds, to prevent the circuit from ‘chattering’ with varying current consumption.


If the load is switched off, the AC128 cuts off, the electrolytic capacitor discharges and the relay again disconnects the voltage converter. The residual current consumption is so small that the circuit can also be connected ahead of the ignition switch. The Schottky diodes need only be rated for the necessary voltages and currents, and above all, they should have the lowest possible saturation voltage. The exact type is not critical. Two separate diodes can also be used. A small heat sink for the MBR diode won’t hurt, but this is normally not essential. Practically any type of PNP germanium transistor that is still available or on hand can be used (AC125, AC126 and AC128 work perfectly).

It may be necessary to modify the value of R1. In combination with the germanium transistor, R1 determines which level of current will be ignored (for memory retention) and which level of current will cause the converter to be switched on. With the component values shown in Figure 1, this level is between 10 mA and 25 mA. It is recommended to measure the quiescent current (at 6 V) and switch-on current of the load and then simulate the switching process using dummy load resistors. When selecting the 6-V relay, ensure that its contacts have an adequate current rating. The actual value can be significantly greater than the nominal output current. With a load of 5 A at 12 V and a converter efficiency of 70 percent, the current through the relay contacts rises to 14.3 A.

Read More...

a simple DC voltage booster

Joule Thief: a simple DC voltage booster

A joule thief circuit lit from a 0.5V AAA battery. Click to enlarge.

A Joule thief is a simple circuit that acts as a DC to DC booster, raising a supply voltage by several volts. In this iteration it uses the exhausted voltage of a alkaline battery and boosts it enough to light a blue LED that requires 2.8V to light. With an otherwise dead 0.5V AAA battery, it will light a blue LED and run for days, using (at the moment) just under 2mA of current. Its much much dimmer than using a fresh battery or running the LED with a proper current through it. Giving the Joule thief circuit 3V from two fresh batteries pulls 75mA through the LED, making it very, very bright and probably short-lived.

A schematic is below. The circuit works like this: When first turned on, current flows into the inductor and produces a magnetic field in the toroid. While this is happening, no voltage appears at the base of the transistor, so the transistor remains off. The LED sees at first no voltage and while the inductor fills up, it only sees a maximum voltage of the battery, which is not enough to pass the diode. Once the inductor is charged, the battery voltage appears at the base of the transistor, turning it on. This allows the right side of the inductor to want to dump the energy it has stored in its magnetic field as quickly as possible, and this gives us a high-voltage that appears across the inductor. When that voltage exceeds 2.8V the LED turns on and lights up until the voltage drops below, triggering the sequence to begin again. I measured the frequency of the on/off oscillation and it seems to run at about 34kHz; the multimeter said between 68 and 72kHz but a radio showed there was 34kHz signal as well, which I assume was the fundamental (and the 68kHz one a harmonic). It did change in frequency a bit while on.

Read More...