0
\$\begingroup\$

Does anyone know if its possible to wire up a MOSFET transistor as a zener diode? I have been happily making normal diodes from my MOSFETs by shorting the gate and the source

schematic

simulate this circuit – Schematic created using CircuitLab

But now I need a zener diode for a voltage regulator an am struggling to come up with anything! Any ideas would be great...

\$\endgroup\$
  • \$\begingroup\$ What's the advantage on using a mosfet AS a zener, instead of using a zener? \$\endgroup\$ – Sergio Jul 24 '14 at 11:04
  • 1
    \$\begingroup\$ Chances are you aren't "making a diode" by shorting the gate and source, you are disabling the (n-channel) mosfet and using the internal body diode. \$\endgroup\$ – akohlsmith Jul 24 '14 at 12:15
  • 2
    \$\begingroup\$ Why the bleep don't you just use a zener? A diode is going to be much cheaper than an IRF530... unless you get to sweep the floors at International Rectifier. \$\endgroup\$ – George Herold Jul 24 '14 at 12:24
  • \$\begingroup\$ Maybe my first comment sounded condemnatory, but it was really out of curiosity. Do you mind explaining why you intend to do this? I just think it would enrich the question, since without some reason, it's complete nonsense. \$\endgroup\$ – Sergio Jul 24 '14 at 13:11
  • \$\begingroup\$ As much as anything it was out of curiosity, just 'is it possible' and because also I have transistors on hand and not any zeners. If its not a possibility its no worries, just thought I'd find out if anyone who knows a lot more than me had any ideas! \$\endgroup\$ – troughton Jul 24 '14 at 13:34
4
\$\begingroup\$

This works: -

schematic

simulate this circuit – Schematic created using CircuitLab

As the drain voltage rises from 0V D1 is not conducting until it reaches its zener voltage (simplistic words). Then a gate-source voltage starts to develop as the drain voltage rises further. At some point (maybe 1V - 4V higher than the zener voltage) the drain starts conducting current and therefore keeping the drain voltage about constant.

\$\endgroup\$
  • \$\begingroup\$ This works as the voltage regulator but I'd really like to not use an actually zener diode, but rather approximate it some how using more MOSFET transistors and resistors \$\endgroup\$ – troughton Jul 24 '14 at 11:15
  • 1
    \$\begingroup\$ You can use a resistor instead of the zener - it won't be as sharp but it'll be usable - consider that to start to turn a MOSFET on you might need 2 volts to get maybe 1 uA. At 3V this will be more like an amp so the resistor instead of the zener will work - try it or simulate it. \$\endgroup\$ – Andy aka Jul 24 '14 at 11:44
  • \$\begingroup\$ Thanks, that is a pretty good approximation, as you say not as sharp but could suffice in a pinch! \$\endgroup\$ – troughton Jul 24 '14 at 13:38
  • \$\begingroup\$ @Andy aka .You presented what could be called an amplified zener circuit +1 .I found that high power zeners were expensive and hard to find when I was at school.I would use 2N3055 for amplified zener because they were in my 1976 junkbox .Your mosfet is valid .I have done this when power handling was more important than accuracy and sharpness.These days people are more likely to have Powermos on hand than power BJT .I am a little squeamish about using the latest gen fets in linear mode .The neg tempco of Vgs threshold does not get balanced by the pos temp co of RDs on at low currents \$\endgroup\$ – Autistic Jun 18 '17 at 2:21
1
\$\begingroup\$

You can't make a Zener diode from a commercial MOS transistor. Zener breakdown (effect) relies on quantum tunneling of electrons trough the thin space-charge region formed in heavily dopped pn junctions [1]. In common MOSFETs, the only heavily dopped regions are the source and drain, while the bulk is usually a lightly dopped p region. Thus, making pn common junctions is easy, but heavily dopped junctions are not available.

Such quantum effects show up in short channel MOS transistors (bellow 65nm), but such devices are not commercially available. Nonetheless, you may try to use the base and emitter regions of a npn bipolar transistor as a Zener diode, due that the emitter is usually a heavily dopped region, but you may expect a higher Zener voltage than the common 5.1 V.

[1] Principles of electronic materials and devices, Kasap S.O., McGraw Hill, pag. 444.

\$\endgroup\$
0
\$\begingroup\$

Mosfets with heatsinks can handle hundreds of watts. 10 watt zener diode cost about 9£ while 100 watt mosfet is about £0.7. Also if you produce many different power supplies and other products, you only need couple different mosfets in large quantities. If you used zener diodes worst case is that you need a different zener for each product.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.