Zener diodes and wattage

[patricks]

Hi there, I have a couple of quick questions about zener diodes.

1. Is there a maximum voltage that I should keep the supply voltage below? i.e. for a 9v zener, does it matter whether I put 20V or 200V in front of it? I checked a few datasheets and didn't see a "maximum forward voltage" rating, but I'd like to check

2. Do I calculate power dissipation using the difference between the supply voltage and the zener voltage, or just using the zener voltage. E.g., if the supply is 20V and the circuit draws 600mA, would the formula be:
Power = (20 - 9) * 0.6, or Power = 9 * 0.6

The reason I ask is that I'm planning a project that needs a 9V DC regulated supply. I know that most laptop power supplies are regulated, but the voltage they supply isn't consistent between models. I thought that an easy way to make sure I get 9V would be to stick a zener diode in there, but I thought I'd check about maximum voltage and power dissipation first.

Cheers
Patrick

[amptramp]

A zener siode is a shunt regulator which means it is connected to a power supply in parallel with a circuit and the zener and its parallel connection are driven through a series resistance which limits the current.  The zener current is set by the supply voltage minus the resistor drop which comes about because of the current through the circuit and the current in the zener.  The current that is not taken by the circuit is taken by the zener and this times its voltage is the power dissipation in the zener.  To make it simple:

Suppose you  have a 12 volt DC source and a 9 volt zener.  The drop has to be three volts.  If you have a series resistor of 120 ohms between the zener and the circuit in parallel with it and the supply.  You have a difference of 3 volts and a resistance of 120 ohms which means the current is 25 mA.  If the circuit is not connected, this current would flow through the zener for a dissipation of 9V * 25 mA = 225 mW.  If the circuit draws current, this current is subtracted from the zener current and the more current the circuit consumes, the less the zener consumes and the lower the dissipation.

The answers to your questions are:

1. The zener diode is rated for reverse voltage, not forward voltage (which is the typical silicon value of about 0.7 volts).  If the resistor used in series with the supply can keep the current below the maximum allowable for the diode, the supply voltage does not matter.

2. The diode disspation is diode voltage times diode current.

For the laptop supply you are considering, you have to use the current rating (minus the current of the circuit in parallel with the zener diode) and select a series resistor that will allow full current at the lowest voltage (and minimum zener current) and maximum current at the highest voltage.  The zener should be sized for the voltage times the maximum current after deducting the current going through the circuit.  In some cases, I design for zero circuit current so a disconnect doesn't fry the zener.

In many cases, it is preferable to use a three-terminal regulator since the total supply current becomes the current in the circuit plus a tiny regulator current.  This provides a lower total power dissipation.  In the real world, regulator IC's are not much more expensive than zener diodes, you don't use as much current and you don't have to calculate a series resistor value.

[patricks]

Thanks very much, that's extremely helpful. I had the wrong idea of how zeners are used, too, so it's set me straight on that, too

In many cases, it is preferable to use a three-terminal regulator since the total supply current becomes the current in the circuit plus a tiny regulator current.  This provides a lower total power dissipation.  In the real world, regulator IC's are not much more expensive than zener diodes, you don't use as much current and you don't have to calculate a series resistor value.

Excellent idea! I can use a 7809 regulator and that'll allow me to use a switched-mode power supply if I need to as well. One last question - when choosing a linear regulator, how much of a safety margin should I allow for the current rating? I'm guessing I should go for one rated for twice the expected current to be on the safe side.

[GibsonGM]

Well, if you told us what you expect to draw, we'd know more, patricks...start there, and work outwards...

1.5X what you expect to draw is MY 'go to' rating, for most of my projects...but that may be different.  If you're expecting to be all over the place with your power supply, 2x sure won't hurt!   Remember the voltage ratings on your caps, too...16V min, 25V better (IMO). 

For all my stompbox-type stuff, a 7809 on a heatsink will get me 1.5A if I NEED it (never do), so that's what I design around.  That would power about 40 pedals (ha ha).    They have thermal shutdown protection, so they'll tell you if they pass too much current.     Check out the data sheet for the LMXX series, and find power supplies others have put together - there is a ton of info on those chips via any search engine!  Good projects, too.   

[patricks]

Yeah, sorry about that - I had worked it out but hadn't shared the info. The build is a submini tube amp and I'm building it in a modular way so current draw will depend on which pre- and power sections I connect, but I'm expecting between 600mA and 1400mA. The larger figure would be if I connected everything at once, which I don't plan on doing, but I figure that it's better to be safe than sorry

The thermal shutdown is a handy feature, I'll check out some datasheets. Thanks!

[R.G.]

Let me add to A.T.'s comments.

Zener-regulated power supplies are very tricky once the power gets bigger than trivial. As an example, many solid state guitar amps which use 15 or 16V zeners to get the supplies from the opamps from the +/-30 to +/-50V power for the output stage. This works mostly, but these amps are plagued by the series resistors charring the PCB or even melting away their solder joints. You have to waste a lot of power in the series regulators to keep zener regulated power supplies running well, and even then there are tough cases. They fail a lot.

The problem is that a shunt regulated power supply pulls the maximum current all the time. It has to. The maximum current the load might ever need is let through all the time at minimum input voltage, and some minimal amount for the zener to keep in regulation. That means that the maximum input, which can be 30% to 50% higher than the minimum, puts even MORE current through the series resistor and zener. For shunt-regulated supplies, maximum input voltage and minimum load current is the worst case condition. On top of that, load variation is hard to deal with in shunt regulators. The more variation, the more normally-wasted power you have to put through the zener.

The load currents you're talking about, 600 to 1400ma are going to be very tough to deal with for a zener regulator.

The reason all those amps use zeners is that three terminal regulators normally can't stand inputs over 30V or so. Some of them go higher, but many are 25V. If you're running from a power supply that's always less than 20-30V, go for three0-terminals. It will simplify your life a lot.

[patricks]

That's great, R.G., thanks heaps!

I often come to this forum with an idea in my head that turns out to be the wrong one, but I love that I can get pointed in the right direction and understand why at the same time

[R.G.]

It's all about learning. Go for it~!