How does a voltage regulator differ from a zener diode?

[Wrichik Basu]

I am making a circuit with Arduino. It is known that the recommended power supply of an Arduino is 7V - 12V. I am going to use a 9-0-9 (230V to 9V) transformer, convert the stepped down voltage to DC, and feed it into the Arduino through the V_in pin.

Now, there are chances that the input AC may considerably rise over 230V, thereby increasing the output voltage. I was thinking of using a voltage regulator to keep the voltage stabilised at 8V to 9V or lower. I have two options:

• LM7808-8V Positive Regulator
• A zener diode with breakdown at 9.1V

An Arduino has an internal voltage regulator. But I read that they tend to heat up very soon. So, I don't want my board to be destroyed by overvoltage. I want to control the voltage outside the Arduino, coupled with a heat sink.

Which among the two components should be better? How do they differ in their working? If you prefer the voltage regulator and not the diode, how should I connect it in my circuit?

N.B.: I calculated that the minimum voltage of my circuit would be about 6V under extreme conditions, and even if the Arduino shuts down at that voltage, I have no problem.

 

[anorlunda]

Why go to so much trouble? Use an AC adapter for and old flip phone, or PC, or some other device. They come in various output voltage settings including 9V. They include internal protections and regulation. The older ones are available almost free.

I go to the nearest car rental agency and say, "I forgot my AC adapter in the rental car, did you find it?" They always pull out a big cardboard box of forgotten AC adapters and say, "Help yourself."

 

[Wrichik Basu]

Why go to so much trouble? Use an AC adapter for and old flip phone, or PC, or some other device. They come in various output voltage settings including 9V. They include internal protections and regulation. The older ones are available almost free.

Not a bad idea. But what I am making with this Arduino is something permanent. I don't want to disassemble that ever again (until and unless something malfunctions). Since I am working with breadboard and wires, which I intend to fix in a metal box, using an adapter means that I have to arrange for another socket to plug it in. That will become a bit difficult as the box is not very large. It is supposed to contain a transformer, an Arduino (or maybe a clone to reduce cost), a small breadboard with resistors and diodes, a 16A SSR, a heat sink and a fan as per the current plans, so adding another component is not a good option here.

On the other hand, I also want to understand how these two components differ in regulating voltage, apart from using them in the circuit.

 

[CWatters]

Forget using a zenner diode for this.

Can I suggest you just buy a 9v DC mains adaptor. Much easier and safer than designing your own.

A 0-9-0 transformer has 9v RMS secondary windings. When rectified the DC voltage will be the _peak_ voltage less the loss in the diode. This will be nearer 12v if I have done my sums right. So the regulator you linked to will need to dissipate 12-8=4V times whatever the current drawn is. Depending on what else it's powering the regulator may need a heatsink.

 

[CWatters]

I haven't looked but there are probably loads of mains to 9v PCBs on ebay. Possibly even switching types that are more efficient than a linear regulator.

 

[CWatters]

On the other hand, I also want to understand how these two components differ in regulating voltage, apart from using them in the circuit.

I zenner diode has a fixed voltage drop over a range of currents. It has to be be biased to the on state with a resistor. The load is connected in parallel with the zenner. It's only really suitable when the load current is small compared to the bias current, otherwise variations in the load current can effect the voltage.

A linear regulator is like a zenner with a power amplifier attached. The zenner provides a voltage reference and the power amp drives the load at that voltage. The power amp isolates the zenner from the load so changes in current don't affect the voltage.

Switching regulators are better still because they are usually more efficient. The means they might not need a heatsink when a linear regulator would. The design of a switching regulator is too complex to go into here. You can sometimes buy assembled switching regulator modules quite cheaply on eBay.

 

[Merlin3189]

I think a regulator would have a much more stable output, which you might not need, but is generally a good idea.
It would have a current limit, which might help if things went wrong.
It's easy to design a PS with a regulator, just follow the data sheet example, but you need to understand a little to design a zener circuit.
You may find it easier and cheaper to use a regulator for larger currents. Zener circuits are easy enough for small currents, but more difficult to design for higher currents. You start adding transistors and entourage and end up designing a regulator circuit similar to, but not as good as, the one you buy for the same price as the zener diode.

If you think of the regulator chip as a zener diode (actually often a better reference than a zener) with extra circuitry built into increase the current capability, add fault protection and improve stability, which sells for a similar price as a zener diode, it's hard to see why you'd use a zener, other than for educational experience. (And you know what we learn by.)

As far as off the shelf PSU's go, I wouldn't use a cheap one (for a circuit of any value or sensitivity) without adding a regulator of known capability. There may be plenty of excelent PSUs on the market, but I've seen some terrible ones amongst cheap wall warts and unbranded laptop psus.

 

[Wrichik Basu]

Just for knowledge: from the above discussions, I believe that if I supply 4V input to a 5V voltage regulator, it will give 4V voltage as output, right? If I put in a voltage greater than 5V, it will give only 5V as output, and the rest will appear as heat (and that's why a heat sink is required). Am I right in these?

 

[Merlin3189]

A linear 5 V regulator, you need to supply over 5 V, as much as 7 V for a 7805. If you supply only 4 V, the output is not specified, but likely less than 4 V.
Once you go over the required minimum, you get the specified output and, as you say, the rest is heat.

A switchmode circuit can increase or decrease the voltage. It will have a specified range of acceptable input voltages and give the specified output for any supply within that range. It is much mor efficient for big step down in voltage, so less heat would be generated than by a linear regulator in the same place.

You can get regulators with adjustable output, both in linear and switchmode forms.

BTW I was probably exaggerated a bit about the prices of zeners and regulators being the same (I've looked and they're not from one supplier at least), but if you built an 8 V 1 A supply, the component bundle to do it with a zener might well be more than the 7808 IC regulator component bundle.

 

[davenn]

Just for knowledge: from the above discussions, I believe that if I supply 4V input to a 5V voltage regulator, it will give 4V voltage as output, right?

incorrect ... it's quite possible there will be near zero output

A linear 5 V regulator, you need to supply over 5 V, as much as 7 V for a 7805.

exactly

but if you built an 8 V 1 A supply, the component bundle to do it with a zener might well be more than the 7808 IC regulator component bundle.

yeah but at least with the linear regulator, the output voltage will be stable over the range in current that it is designed for

And that is the worst part about zeners, they DO NOT have a stable voltage output when the output load varies,
which of course causes the current to vary and that causes the voltage out of the zener to vary.Dave

 

[Wrichik Basu]

Thanks, @Merlin3189 and @davenn for your explanations.

One more question: in a datasheet of the voltage regulator, for fixed output voltage circuit, this one is given:

Why are these capacitors necessary?

 

[davenn]

Why are these capacitors necessary?

you should have kept reading to the next page

They are required to stop internal oscillations of the reg chip
The input cap, CI, for when the reg chip is a distance from the rectifier/smoothing section of the PSU ...
tho it doesn't state what that distance is
The output cap, CO, improves stability and transient response ( basically CI is doing the same)

The values are not super critical ... 0.47uF down to 0.01uF ... I commonly use 0.1 or 0.01uF (100nF or 10nF) in the input and output)
The ESSENTIAL part is to have them as close as possible to the pins of the reg chip.cheers
Dave

 

[Wrichik Basu]

you should have kept reading to the next page

They are required to stop internal oscillations of the reg chip
The input cap, CI, for when the reg chip is a distance from the rectifier/smoothing section of the PSU ...
tho it doesn't state what that distance is
The output cap, CO, improves stability and transient response ( basically CI is doing the same)

The values are not super critical ... 0.47uF down to 0.01uF ... I commonly use 0.1 or 0.01uF (100nF or 10nF) in the input and output)
The ESSENTIAL part is to have them as close as possible to the pins of the reg chip.cheers
Dave

Understood, thanks for the help.

 

[Windadct]

As to your OP - in a circuit the Zener could be considered a basic V regulator or the key component of a very basic V reg circuit, but low tech with many shortcomings...

 

[Wrichik Basu]

As to your OP - in a circuit the Zener could be considered a basic V regulator or the key component of a very basic V reg circuit, but low tech with many shortcomings...

But there is one utility of zener diode over a voltage regulator - if you want a circuit where the DC voltage will rise to a certain maximum value but not after that, then you have to use a zener. In my circuit, I have to use it at one place, because I don't want a fixed voltage always, but a cap on the voltage, so that it cannot rise over a certain value (5V in my case). So I'll use a 4.7V zener.

 

[davenn]

In my circuit, I have to use it at one place, because I don't want a fixed voltage always, but a cap on the voltage, so that it cannot rise over a certain value (5V in my case). So I'll use a 4.7V zener.

but that isn't what you stated as the requirement in your OP ... in fact it is nothing like what you initially stated

so what are you doing with this new varying voltage supply ? it cannot be for supplying DC to the ArduinoDave

 

[Wrichik Basu]

but that isn't what you stated as the requirement in your OP ... in fact it is nothing like what you initially stated

so what are you doing with this new varying voltage supply ? it cannot be for supplying DC to the ArduinoDave

No, that's an absolutely different part, not at all related to this question. I just mentioned it with respect to what @Windadct has said. The question was about powering the Arduino, and that has been solved. What I mentioned was for measuring the AC voltage, which is a different part of the circuit, and not directly influencing the Arduino power supply.

 

[anorlunda]

Are you working on that "voltage stabilizer" ?

https://www.physicsforums.com/threa...ns-stabilizer-is-removed.955711/#post-6059861

 

[CWatters]

But there is one utility of zener diode over a voltage regulator - if you want a circuit where the DC voltage will rise to a certain maximum value but not after that, then you have to use a zener. In my circuit, I have to use it at one place, because I don't want a fixed voltage always, but a cap on the voltage, so that it cannot rise over a certain value (5V in my case). So I'll use a 4.7V zener.

Just for info...many IC have diodes between inputs and Vcc to stop input going much higher than Vcc. Mainly to protect the input from static damage.

 

[davenn]

What I mentioned was for measuring the AC voltage,

you can't put a zener across an AC voltage and expect it to work

 

[Wrichik Basu]

Are you working on that "voltage stabilizer" ?

https://www.physicsforums.com/threa...ns-stabilizer-is-removed.955711/#post-6059861

Right! I'm making my own stabiliser, that shall control the appliance depending on the AC voltage (either keep it on, or switch it off using an SSR).

 

[Wrichik Basu]

you can't put a zener across an AC voltage and expect it to work

I know that. Everything is being done with DC after an initial rectification.

 

[Wrichik Basu]

Just for info...many IC have diodes between inputs and Vcc to stop input going much higher than Vcc. Mainly to protect the input from static damage.

I have zero knowledge in ICs. Can you mention some?

 

[CWatters]

Just about every CMOS device has static protection diodes on inputs. See figure 3.6...

http://www.eeeguide.com/cmos-inverter/

 

[Merlin3189]

I have zero knowledge in ICs.

Most have data sheets available online. Those from major manufacturers have a lot of detail and often application notes, which explain such things as the need for bypass capacitors etc. I think I often learned more from data books and application handbooks than from textbooks.

 

[Windadct]

Interesting the ATMega, a common Arduino basis (http://web.csulb.edu/~hill/ee346/Lectures/08%20ATmega%20GPIO.pdf does not seem to have a Diode Voltage clamp, where Raspberry Pi does. - Still in these cases I would call this protection (against an abnormal condition), and not regulation ( regular operation).

 

[anorlunda]

I know you have already decided to try Arduino, but this article may give you ideas.

https://www.eleccircuit.com/wp-content/uploads/2017/02/over-voltage-and-low-voltage-protection-circuit-600x361.jpg

 

[Rive]

if you want a circuit where the DC voltage will rise to a certain maximum value but not after that, then you have to use a zener.

Indeed, for voltage limiting on an input signal there is such use for a zener, but you have to be careful, because the actual work (dissipation) will be on a serial resistor, not on a zener. This kind of circuits should be very carefully designed, especially if you want to measure anything related to line voltage (somehow this thread has such feeling...)

 

[Lord Crc]

Interesting the ATMega, a common Arduino basis (http://web.csulb.edu/~hill/ee346/Lectures/08%20ATmega%20GPIO.pdf does not seem to have a Diode Voltage clamp

The datasheet says otherwise:

All I/O pins have protection diodes to both VCC and Ground as indicated in the following figure.

https://www.microchip.com/wwwproducts/en/ATmega328p

 

[Windadct]

I thought this was the case but could not find this it ( Ref Fig 18.1) - then assumed it was something I had seen for the Pi.

In reality I think this is actually a byproduct (semi-intentional) of using MOSFAT for the basic Digital output - as their body-diodes provide the function.

 

[Wrichik Basu]

Indeed, for voltage limiting on an input signal there is such use for a zener, but you have to be careful, because the actual work (dissipation) will be on a serial resistor, not on a zener. This kind of circuits should be very carefully designed, especially if you want to measure anything related to line voltage (somehow this thread has such feeling...)

Yes, I will measure AC voltage with Arduino, and I have taken care of what you have said. Thanks for mentioning.

 

[CWatters]

Messing with mains/line voltage can be dangerous. Can I suggest you use an isolating step down transformer and multiply up in the Arduino.

 

[sophiecentaur]

Zeners and Voltage regulator ICs. I can't find a mention of the fact that there are Series and Shunt voltage regulators. This is very relevant. A Zener acts as a shunt regulator as it sits across the load with a series resistor to the supply. It's cheap and cheerful but it requires the current supplied by the source to be significant at all times - even when the load is high resistance or off (you need a standing current in the diode in order to get the Zener volts). For anything other than a very 'light load' (such as when the Zener is used as a low power voltage reference inside a circuit) the regulator dissipates a lot of power.
Most Voltage Regulators are Series regulators because they only need to supply / take very little current in addition to the current that the load draws. No load power will mean little or no power dissipated in the regulator.

 

[Wrichik Basu]

Messing with mains/line voltage can be dangerous. Can I suggest you use an isolating step down transformer and multiply up in the Arduino.

Can you explain a bit on what you mean by "isolating" transformer? Before connecting anything to the Arduino, I shall check with my multimeter first. Is that what you mean by "isolating" transformer?

 

[CWatters]

Line/mains voltage is too high to be connected directly to an Arduino. The ADC cannot measure voltages that high. Some sort of pre-scaler is required to bring the voltage into the range the Arduino can measure. Bit like setting the range dial on your multi meter.

One way is to use some sort of diode/resistor network to reduce the voltage but in the event of a fault the whole circuit could become "live" and dangerous.

Another method is to use a transformer to step down the mains voltage to something safer first (say 5 or 10v) then use a diode/resistor network to get it in range of the ADC. If the transformer is suitably insulated then it's very unlikely any part of your Arduino circuit could accidental become live.

To be honest... if you have to ask about necessary safety precautions such as these you probably shouldn't really be attempting it.