# Question About Diode Voltage Drop

1. Jan 18, 2012

### Fischer777

Say the voltage drop of a diode is .7 volts. Once the voltage drop has been reached, and it begins conducting electricity, can the voltage be backed off to around .3 volts and it will continue to conduct electricity? Or does the amount of electricity need to be a constant .7+ volts?

2. Jan 18, 2012

### Simon Bridge

You need to be .7+volts to conduct.

3. Jan 18, 2012

### Kholdstare

Its a common misconception that you need 0.7 volt across the diode to make it conduct. In fact a diode will always conduct regardless of the bias. However the current depends on the voltage bias exponentially.

I = Is[exp(V/nVt) - 1]

In fact manufacturers define a voltage called cut-in voltage which is the required forward bias for a specific current (say 1uA). This is typically 0.7V for Silicon. The built in voltage has nothing to do with diode I-V relationship.

4. Jan 18, 2012

### Kholdstare

Its a common misconception that you need 0.7 volt across the diode to make it conduct. In fact a diode will always conduct regardless of the bias. However the current depends on the voltage bias exponentially.

I = Is[exp(V/nVt) - 1]

In fact manufacturers define a voltage called cut-in voltage which is the required forward bias for a specific current (say 1uA). This is typically 0.7V for Silicon. The built in voltage has nothing to do with diode I-V relationship.

5. Jan 18, 2012

### vk6kro

Here is a graph of current vs voltage for a silicon diode.

http://dl.dropbox.com/u/4222062/diode%20conduction.PNG [Broken]

If there was 0.65 volts across the diode (in the forward direction), the current would be about 4 mA
If there was 0.70 volts across the diode (in the forward direction), the current would be about 11 mA
If there was 0.75 volts across the diode (in the forward direction), the current would be about 27 mA

These voltages and currents apply if the voltage was previously higher or lower.

You may be thinking of neon discharge tubes which conduct at a high voltage and, after the gas has ionized, a lower voltage is sufficient to keep the current flowing.
This does not apply to silicon diodes.

Last edited by a moderator: May 5, 2017
6. Jan 18, 2012

### Kholdstare

This did not click in my mind in the first place. I was thinking about pnpn diode (Shockley diode).

7. Jan 18, 2012

### Simon Bridge

the exact graph depends on the diode.
I didn't grok that this was quoting the rule of thumb though.

8. Jan 18, 2012

### Fischer777

Okay. So is there some device out there that will not conduct electricity until it reaches a certain voltage?

9. Jan 18, 2012

### f95toli

Zener diode?

10. Jan 18, 2012

### Staff: Mentor

A relay driven by a comparator...?

11. Jan 18, 2012

### vk6kro

Most insulators conduct to some extent, but you have to be pretty pedantic to call it a current. Some of them do conduct better if they have already been forced to conduct.
Capacitors which have been forced to conduct by a high voltage will often conduct at a lower voltage because there is then an air path through the insulator.

Silicon diodes are surprisingly consistent. They have a microscopic amount of conduction up to about 500 mV and then they become quite good conductors as the voltage increases.

If the voltage is increased to about 750 mV, and if there is nothing to limit the current, then they run some risk of being destroyed.

This "turn-on" voltage range is affected by temperature. It moves downwards with increasing temperature, which is unfortunate for the diode as it produces a runaway effect where it suddenly draws a lot of current, gets hotter and draws more current.

For this reason, it is better to test diodes by limiting the current and measuring the voltage across the diode, instead of the other way around.

GROK: to intimately and completely share the same reality or line of thinking.
I did not know that.

12. Jan 18, 2012

### vk6kro

A Zener turns on quite gradually. I remember plotting the curves of some of these.
Oddly enough, the base emitter diode of some transistors works as a better Zener with a very sharp knee curve. The zener voltage is about 7 volts and it makes a good reference voltage.

Possibly a DIAC might be a good example. These conduct or not depending on a very critical voltage.
Once they are conducting, reducing the voltage does not turn them off.

13. Jan 18, 2012

### Fischer777

So I looked up a zener diode. It is my understanding that, if the zener voltage is say 5, it will not conduct electricity until the voltage across the circuit is 5v or more, and then it will conduct until the voltage drops below that level again (I am assuming that it is installed "backwards", so to speak).

14. Jan 18, 2012

### Staff: Mentor

Saying "will not conduct" is a relative statement, not an absolute statement (at least not for real devices). There is a finite reverse leakage current before Zener breakdown, and as mentioned, the knee is not all that sharp.

15. Jan 18, 2012

### Simon Bridge

Giving commercial zener diodes to students - they usually cannot detect the curve of the knee with equipment readily available in an undergraduate teaching lab.
OTOH: NZ colleges have pretty crappy equipment.

... this is more like what I see for reverse characteristics of a zener.
except I'm kinda used to seeing the sharp knee on the other side...

I mean: yeah sure you are correct. What you actually see depends on the application and the diode.

16. Jan 18, 2012

### Kholdstare

Reverse leakage current is always constant and stays that way before dielectric breakdown. However, Sandy Bridge's diagram is correct. The diode has a depletion region generation current flowing in reverse bias which gradually increases with bias. It is proportional to square root of applied reverse bias. Thus the reverse characteristics is not so sharp as it is thought to be.

EDIT: Opps It'll be Simon Bridge ! sorry

17. Jan 18, 2012

### AlephZero

Look at the data sheet before you test the component. A typical BZX... 5V zener has a dynamic resistance of a few hundred ohms at 1mA reverse current, but 100 times less at say 50mA.

If you are only plotting currents up to 8mA, you are probably looking at the "off design point" part of the characteristic through a zoom lens.

Of course raising the current too far above the "design point" will kill the diode by exceeding the max power rating.

18. Jan 18, 2012

### Fischer777

Would the reverse leakage through the diode be enough to trigger a transistor gate, even if one were to also add a resistor? Or would it depend on the size of the resistor, and type of transistor?

Forgive me if these questions sound silly. The reason I ask these questions in the first place is because am trying to build a simple circuit that will monitor the amount of electricity stored in a capacitor, and when it reaches a certain level, dump it.

19. Jan 18, 2012

### vk6kro

So, you are now talking about a Zener diode?
Unfortunately, if you charged a capacitor with a Zener across it, the voltage of the capacitor would rise to the Zener voltage and then stay there, but the capacitor would not be discharged.

Perhaps you could draw up a diagram of what you want to do?

20. Jan 18, 2012

### jim hardy

unijunction transistor 2n2646 will do what you said

it is a strange transistor with only one emitter and two bases.
it'll turn on when voltage at emitter reaches: [ voltage between its bases X it's "intrinsic standoff ratio" ] . widely used for 'relaxation oscillator'.

2n2647 is similar, slightly higher standoff ratio and tighter spec if i remember right.

altavista search is working today. check 'em out.

Last edited: Jan 18, 2012