5
\$\begingroup\$

I am trying to get a toggleable 2.0 to 2.6 ish voltage reference that can supply up to 45 mA. I simulated the below circuit and found this to be acceptable. However upon building the circuit, I found my Vout voltage to be much higher, around 3.2V. The voltage drops slightly to 3.0V depending on the load (used 100 ohms to test).

So I think the Zener isn't operating correctly. However, when measuring the voltage across the 50 ohm resistor (5.5V-3.2V)/50ohm = 46mA, I found that the Zener should be on! This Zener diode only needs like 5 mA to operate correctly. So what gives? Is the knee of this zener just really bad? What should I do to ensure the voltage never goes above 2.6V? The PNP is toggled with a GPIO pin from an MCU.

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ You need a load. \$\endgroup\$ – StainlessSteelRat Mar 1 at 3:49
  • \$\begingroup\$ I don't see V2 voltage if Q2 is saturated (fully open). However, the low voltage zeners are bad. Use some IC or build our own. Probably typical NPN at high-side with zener in base regulator would work better. \$\endgroup\$ – Michal Podmanický Mar 1 at 4:12
  • 2
    \$\begingroup\$ Why go to so much trouble when you can buy a chip like, for example, TPS7A2024PDBVR (ti.com/lit/ds/symlink/tps7a20.pdf) that does exactly what you need? \$\endgroup\$ – John Doty Mar 1 at 14:05
9
\$\begingroup\$

A 2.4V (@5mA) Zener is going to have a heck of a lot more than 2.4V across it at close to 50mA. Just the way the physics works out for low voltage zeners.

3.3V or so (typical) looks about right.

enter image description here

Use a TL431 instead.

Just as a rule of thumb, think twice about using any zener < 5.1V.

\$\endgroup\$
3
  • \$\begingroup\$ So is there no way to adjust the current circuit to get under 2.6v? Maybe upping the series resistor to lower the current. yet if I need to supply up an average of say 20 mA, at least, I might run into trouble? \$\endgroup\$ – Jirhska Mar 1 at 3:58
  • \$\begingroup\$ @Jirhska You maybe can increase the series resistor. The number you need to consider is the maximum load current and the minimum input voltage at the collector of Q2. Use a TL431, it will change about 10 or 15mV from 1mA to 50mA. With no resistors the "breakdown" is 2.495V nominal. And it's as cheap as a zener. \$\endgroup\$ – Spehro Pefhany Mar 1 at 4:03
  • 1
    \$\begingroup\$ @Sphero Well looks like it came around and bit me in the butt. It is a good lesson to learn! Thanks a bunch! \$\endgroup\$ – Jirhska Mar 1 at 4:19
4
\$\begingroup\$

If you want a voltage reference, I suggest you use a specialized IC. Those are incredibly stable w.r.t current and temperature compared to Zeneres. For instance, compare the V/T and V/I curves of TL431 with the ones of your Zener:

enter image description here

Of course, you could build the compensation circuit yourself if this is what you're interested in, but it will be practically impossible to beat the IC solution.

\$\endgroup\$
1
\$\begingroup\$

I mean something like this circuit where current thru zener is relatively constant when load changes, i.e. voltages on zener and output stays stable. Pick 2v7 or 3v3 zener diode to get desired output level. It is also possible to get any of output voltage just by placing a resistor devider across zener and connect common to base. You can turn the output voltage off with groundig MCU pin.

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ Even a 3.1V 5mm LED works better than 3V zener with a conduction resistance of about 15 Ohms at 20mA. @Spehro’s Zener is 100 Ohms by comparison. The emitter follower output impedance will then be 15/100 Ohms. A total power consumption is just the LED and get 2.3V output @ 50mA load \$\endgroup\$ – Tony Stewart EE75 Mar 1 at 5:45

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.