Tip:
Highlight text to annotate it
X
How to get a constant DC voltage respectively a constant direct current will be treated at this video.
Using batteries is a simple way to get a very smooth DC voltage, but there are some things to be considered.
The nominal voltage output this lead acid battery is 12V, but almost each time you connect a voltmeter to the terminals of the battery, you will detect a value being at least slightly different from those 12V.
The chemical mixture inside of the battery is altering during the charging respectively discharging procedure.
The potential is between the end of charge voltage, which is 14V and the end of discharge voltage, which is 11.2V.
There is another property of batteries which has to be considered when a constant voltage is required for your application: The current-voltage characteristic while connecting a load to the battery.
Without a load, 12.3V can be detected, while the potential drops below 12.1V while connecting a 12V 10W filament lamp to the terminals.
A current of approximately 800mA is coming out of the battery, generated by a flow of ions inside of the galvanic cells.
The speed of movement inside of the electrolyte is limited, hence the concentration of a certain type of ions around the electrodes is decreasing and so the potential at the output terminals is decreasing, too.
The higher the current drawn from the battery, the lower the output voltage.
The battery acts like an ideal voltage source with a resistor connected in series to one of it's terminals.
The resistance of those hypothetical device inside of the galvanic cells is called "inner resistance" of the battery.
Those inner resistance is not a constant value.
Besides the current drawn from the battery, the charge condition and the environmental temperature affects the value.
Usually there is: The higher the capacitance of the battery, the lower the inner resistance, because the dimensions of the electrodes - mainly the area of their surface - is increasing, too.
The concept of inner resistance applies to all kinds of electrical sources in a real world.
The resistance of the windings of a generator or those of rectifiers also results in an inner resistance of real voltage sources.
Here, a stepper motor is used as a tiny generator and we can also detect a decrease in potential as soon as additional LEDs are connected to those voltage source.
At the video about voltage dividers we have learned how to get a fraction of the output voltage of a power supply.
When using at least one potentiometer as a variable resistor, the voltage output of the divider can be readjusted whenever the battery voltage alters.
Another variable resistor is a transistor, which can be used to build a stabilizer circuit:
The base pin of the NPN transistor is connected to a voltage divider formed by R1 and a zener diode.
Like explained at the video about voltage dividers, the zener diode provides an almost constant voltage as long as the input voltage exceeds those value.
As soon as a load is connected to the output clamps of the circuit, the emitter-collector line of the transistor and the load are forming the second voltage divider of the circuit.
If the resistance of the load is constant, the resulting output voltage across the load is caused by the variable resistance of the transistor's emitter-collector line.
Like explained at the video about the properties of bipolar junction transistors, the resistance of the Darlington transistor used here is above 20 kiloohm around a emitter-base voltage of approximately 0.7V, while it is decreasing to just some Ohm and below around a base voltage of 1.4V.
Usually the resistance of the load is clearly below the maximum and clearly above the minimum resistance of the emitter-collector line, so the emitter-base voltage is always around 1.2V while the circuit is connected to the supply voltage.
The detected voltage at the output clamps is 5.6V, those at the zener diode 6.8V and between emitter and base of the Darlington transistor it is 1.17V.
All three potentials are correlated with each other: The output voltage plus the emitter-base voltage equals the zener voltage.
The zener voltage is constant; it is independent from the load connected to the circuit; The emitter-base voltage is also almost constant like explained before, by what the resulting output voltage is constant, too; It is also independent from the resistance of the load.
Whenever the resistance of the load is decreasing, the output voltage (which is the voltage across the load) is decreasing, too, resulting in an increasing voltage between emitter and base.
However with increasing emitter-base voltage, the resistance of the transistor's emitter-collector line is decreasing, resulting in an increasing output voltage.
The feedback loop of the circuit counterbalances the falling voltage caused by a decreasing resistance of the load.
On the other hand: if the resistance of the load is increasing, the emitter-base voltage is decreasing causing an increasing resistance of the emitter-collector line.
The decrease in resistance of the load is also counterbalanced by a decrease of the transistor's resistance.
An increasing input voltage would result in an increasing output voltage, but also in a decreasing emitter-base voltage and so in an increasing emitter-collector resistance.
The increasing resistance of the emitter-collector line counterbalances the increasing input voltage; While the potential across base and emitter is increasing slightly, those between base and collector is increasing noticeable.
Vice versa a decreasing input voltage results in an increasing emitter-base voltage and so in a decreasing resistance of the emitter-collector line.
Once more the circuit acts contrarious to the varying circumstances, regulating the output voltage to the set point
The circuit is called linear regulator, because there is a linear correlation between the resistance of the transistor and the input voltage.
A high quality linear regulator acts like an ideal voltage source with nearly no inner resistance.
By using an operational amplifier in comparator mode, the stability of the output voltage can be increased significantly:
The gain of the operational amplifier is clearly higher than those of a single bipolar junction transistor, hence the device is driven with a higher current whenever the voltage at the inverting input drops below those at the non-inverting input, which is connected to the reference voltage at the zener diode.
The resulting output voltage can be adjusted to arbitrary value between the zener voltage and (nearly) the input voltage.
When replacing the NPN transistor by a p-channel MOSFET, the inverting input has to be connected to the reference voltage, while the non-inverting input has to be connected to the potentiometer at the output.
It's a negative feedback loop once again:
If the output voltage is increasing, the potential at the non-inverting input of the operational amplifier and so the difference in potential between the two input clamps is increasing, too
A higher voltage between inverting and non-inverting input causes a higher output voltage of the operational amplifier.
A more positive, hence a less negative potential between source and gate of the p-channel MOSFET causes a higher resistance across the source-drain line and so finally a decrease in the output voltage of the regulator circuit.
The minimal dropout voltage between input and output voltage of this regulator type is lower than those at the circuits before, because the pass transistor is driven fully into saturation by attaching a negative voltage between source and gate.
The voltage between source and drain which is the minimum dropout voltage, is usually just some millivolts.
It's a low drop-out regulator.
The linear power supply regulates the output voltage by continually dissipating power in the pass transistor.
The Raspberry Pi consumes a power of approximately 2W at an input voltage of 5V, hence 2.2W of electric power are dissipated by the linear regulator when operating the tiny computer with a 12V battery.
A large heat sink is required to prevent the power transistor of the circuit from getting roasted.
At this circuit, the operational amplifier is operating as a Schmitt-trigger.
The output voltage is not as smooth as at the circuits shown before.
The curve is oscillating with an amplitude of approximately 1.0V, depending on the resistance values of R4 and R5.
The NPN transistor is turned full-on whenever the input voltage of the Schmitt-trigger reaches the lower threshold and it is turned full-off, when the upper threshold is reached, hence the transistor spends very little time in the high dissipation transitions, so the wasted energy is minimized.
Ideally, those circuit dissipates no power, however in a real world there are always losses.
The slew rate of the operational amplifier limits the time required to alter the switching state of the transistor (whose slew rate isn't infinite, too) and the switching current always causes some noise which also worsens the efficiency.
The capacitor is required to buffer electric energy at the output circuit.
The higher the capacitance, the lower the switching frequency of the transistor.
Switched mode power supplies are treated at another video and those devices provide a regulated output voltage by switching transistors on and off instead of dissipating the electric energy in the pass transistor.
A voltage regulator can't compensate the distortions at the input or output circuit with no lag of time.
For example an electric motor connected to the output circuit causes voltage peaks whenever one of it's inductors is turned off abruptly by the commutator.
The DC voltage is superposed by an unwanted AC fraction.
The faster the regulator circuit responds to the distortions, the better the output signal:
We can detect a clear ripple at the circuit with a single transistor.
The linear regulator with the operational amplifier starts oscillating.
It is a general problem of regulator circuits, that they tend to oscillate if the feedback signal is delayed or too strong.
Very complex circuits are required to damp those oscillations.
So it is a wise decision to use integrated circuits which are available on the market.
As you can see, the L4940V5 IC used here generates an almost smooth output signal.
Other features of the commercial product are a protection against excessive heat or a too high current.
A simple way of minimizing distortions is connecting a capacitor in parallel to the output clamps.
The higher the capacitance of the device, the better the effect.
Like explained at the video about RC-circuits, a low-pass filter can be used to block high frequency AC signals.
The higher the resistance or the capacitance of the linear circuit, the better the filter characteristic.
A disadvantage of the RC filter is, that the inner resistance of the whole circuit is boosted by the resistance of the low-pass.
When replacing the resistor with an inductor of low internal resistance, hence a large diameter of the wound wire, the circuit is also an effective decoupling filter.
Small coils of insulated wire, are commonly used for this purpose.
The lower the frequency of the AC fraction, the larger the dimensions of the choke.
For demonstration purposes, a veeeery large inductor is used here:
It's a toroidal transformer.
The voltage source is once more the stepper motor used as tiny generator.
Here, an RC-filter with a 15 Ohm resistor and an LC-filter with the toroidal transformer of unknown inductivity are compared with each other.
Each of the devices is switched in series to an electrolytic capacitor of 22 Micro farad, where at the voltage drop is detected.
The ohmic resistance of the transformer is clearly below 15 Ohm - both at the primary and the secondary winding.
The distortions at the LC-filter are similar to those at the RC-filter.
When switching both secondary windings in series, the filtering effect is increasing.
With increasing revolution speed of the generator, hence with increasing frequency, the peak-to-peak voltage at the LC filter is decreasing clearly.
When using the primary winding, with the higher inductivity, we get an almost perfect DC voltage.
Sometimes an application requires a constant current instead of a constant voltage.
The regulator circuit has to be altered to get another type of feedback:
With decreasing resistance of the load, the voltage output of the circuit must be decreasing, too, hence the resistance of the transistor must be increasing.
The voltage drop across the zener diode is constant, hence the potential between emitter and base depends on the voltage across R2 and so on the current running through the device.
There is a also negative feedback loop: If the current running through the output loop and so through R2 is increasing, caused by a decreasing resistance of the load, the emitter-base voltage is decreasing (remember that the zener voltage is constant, while the potential at the emitter is increasing), hence the resistance of the transistor's emitter-collector line is increasing, counterbalancing the decreasing resistance of the load.
Even when short-circuiting the output clamps, which equals a load with an extreme low resistance, the current is increasing just slightly.
An increasing input voltage would also cause an increasing current through the load, which is also counterbalanced by an increasing resistance of the transistor.
Likewise a falling input voltage is balanced by the regulator circuit.
With the help of the potentiometer switched in parallel to the zener diode, the output current can be adjusted continuously.
By using an operational amplifier, the accuracy of the regulator circuit can be improved.
The operational amplifier amplifies the difference in potential between the inverting and the non-inverting input with a high gain, hence the output voltage and so the potential at the inverting input is increasing until the difference in potential between both input clamps is nearly zero (remember that the voltage at the non-inverting input is constant).
There is a negative feedback between input and output of the operational amplifier: If the resistance of the load is increasing, the current through R2 and so the potential at the inverting input is decreasing, by what the difference in potential between both input clamps is increasing and so the output voltage is increasing, too.
With increasing output voltage the current running through the output loop and so the voltage across R2 is increasing.
Once again the difference in potential between the input clamps is decreasing to nearly zero and the current is readjusted to the set point.
When bypassing the load between the output clamps, the variation of the current is clearly lower than those at the circuit with the NPN transistor.
Fluctuations of the input voltage are also balanced far better.
Besides measurement issues, a constant current is required to control electrochemical reactions.
When electroplating a metal with a fixed thickness of the layer, a very constant current is required, which has to be independent from the shape of the surface area of the electrodes.
That's all about constant voltage respectively current for now.
There's more about this topic at the project page.
Thanks for watching and: "I'll be back!"