SPCR

So i have done some research and would like to get the SPCR community feedback on the reliability, viability and risks of using zener diodes to reduce voltages for computer fans.

Someone please correct me if any of my statements are incorrect. So i bought some zener diodes from Radio Shack: 1watt (rated max) 48mA (test current) 4.7v (zv) Zeners.

The theory is that this will reduce a 12v line by 4.7v to make 7.3v when placed on the live wire in reverse (colored band on diode pointing towards the source).

There are some catches, the most important being that the zener diode does not limit the current passing through just the voltage and if the current drawn accross it exceeds what its maximum the diode dies. So that means you match the zener to the ratings of fan you're wiring up. Making it unrealistic to switch different fans in and out of your set up.

My first zener died instantly when i wired up a 0.85a fan to it. With a new zener i then changed to a 0.08a (80mA) 40x10mm fan and tested that it dropped 12v down to 6.85 - success!

You should remember i said the zener was 48mA and the fan was 80mA so why didnt it blow up? well because the fan after a voltage drop no longer draws the full current. The expected voltage drop was 7v which is 58.33% (12/7) of 12v and so multiplying 80mA by 0.583 gives 46mA.

Why bother? apparently low voltage zeners produce no heat. Theyre still pretty damn cheap (cents). Seems like a very simple solution to a common problem but im interested in feedback from some electronics experts.

Thoughts?

No problem - calculations are explained by cpemma here.

I put one in a printer fan circuit years ago - the printer is still running.

I havent used zeners for this. But I'm using two regular diodes in series, to lower the voltage with about 1,5-1,6 volt. One diode gives a drop of 0,6-0,8 volt, and you can solder as many in series as needed. The point is, that diodes like 1N4004 can take one amp, without cooling. They are very cheap and reliable.

Another way to increase the max posibble amount of current draw, is to put two zeners in parrallel. With same voltage rating ! This will simply make it possible to draw double current, without blowing anything.

Sligthly off topic - I use a Scythe S-Flex 1200 RPM with the onboard controller on my Gigabyte GA-MA78G-DS3H mobo. This works very good, but the fan is a little too fast for my noise taste/cooling needs, and 2 diodes in series with the fan, did the trick.

Zener diodes are generally designed for low current applications, like signal processing. It's no surprise that you burned it out with a >10W load.

If you want to efficiently drop the voltage to your fan, just buy a DC-DC regulator from somewhere like ti.com, but the savings are going to be near inconsequential relative to other hardware...

What Thomas said about using rectifier diodes is exactly correct. The 1N400x series is available just about anywhere. I've used them extensively over the years. Here's a fan controller I made.

alleycat wrote:Here's a fan controller I made.

Pretty smart; Cheap, effektive, can take powerfull fans, and without the problems which can occur with controllers you buy. I'm thinking about noise problems from PWM circuits, and that many cant go lower than 6-7 volt.

Actually, once the fan speed requirements have been worked out, most people probably only need fast/slow settings (eg summer/winter or high/low load). Then you only need a two-postion switch with the appropriate number of diodes on each position.

crankyhobo wrote:low voltage zeners produce no heat

Ahem, little nitpicking required

Of course zeners produce heat, in this circuit exactly like an ordinary resistor (or some diodes in serial). Heat output (power) can be calculated as P=U*I, where U is voltage drop and I is current.
For 5V and 80mA we get 5V*0.08A=0,4W.

There are zeners up to 50W (and probably more, I checked only 'local' online store) available, they're a bit expensive though.

Arvo wrote:

crankyhobo wrote:low voltage zeners produce no heat

Ahem, little nitpicking required

Of course zeners produce heat, in this circuit exactly like an ordinary resistor (or some diodes in serial). Heat output (power) can be calculated as P=U*I, where U is voltage drop and I is current.
For 5V and 80mA we get 5V*0.08A=0,4W.

There are zeners up to 50W (and probably more, I checked only 'local' online store) available, they're a bit expensive though.

Interesting, i got this idea from the wikipedia article on zener diodes, which says:

In silicon diodes up to about 5.6 volts, the Zener effect is the predominant effect and shows a marked negative temperature coefficient. Above 5.6 volts, the avalanche effect becomes predominant and exhibits a positive temperature coefficient.

In a 5.6 V diode, the two effects occur together and their temperature coefficients neatly cancel each other out, thus the 5.6 V diode is the component of choice in temperature-critical applications.

Modern manufacturing techniques have produced devices with voltages lower than 5.6 V with negligible temperature coefficients, but as higher voltage devices are encountered, the temperature coefficient rises dramatically. A 75 V diode has 10 times the coefficient of a 12 V diode.

Sounds like i misinterpretted it.

alleycat wrote:Actually, once the fan speed requirements have been worked out, most people probably only need fast/slow settings (eg summer/winter or high/low load). Then you only need a two-postion switch with the appropriate number of diodes on each position.

Agree. Though, personally I prefer a device like NoiseMagics NMT2, which do it automatically.

@crankyhobo

IMHO temperature coefficient indicates how "threshold" voltage (eg 5.6V) depends on temperature. If temperature coefficient is [near] zero then zener diode stabilizes 5.6V at wide range of operating temperatures - crucial for voltage stabilizers.

What about heat dissipation then you can't ignore simple formula for DC - P=U*I (for alternate current it is more complex).