13
\$\begingroup\$

It is always said that forward voltage drop in the diode is around 0.7 volts. LED also being a diode, why does it have a greater forward voltage drop of around 3 Volts?

What is the model of LED that explains this higher voltage drop?

\$\endgroup\$
  • 2
    \$\begingroup\$ This is one of those questions where the answer is to read a solid state physics book. \$\endgroup\$ – Matt Young Jan 26 at 16:13
  • 2
    \$\begingroup\$ You know, I don't think I've seen this question asked on here before, but it seems like a fairly easy misunderstanding for beginners to get, which means it's a useful one to have on here. Good question! \$\endgroup\$ – Hearth Jan 26 at 16:23
  • \$\begingroup\$ Some decent reading: ledsmagazine.com/articles/2004/01/what-is-an-led.html \$\endgroup\$ – Peter Smith Jan 26 at 16:25
  • \$\begingroup\$ You might note that at room temperature the forward voltage of an LED can be 1.2V or so for an IR LED, 1.8V or so for a red LED or 3V or so for a white (really blue) LED. I have a datasheet here for a 245nm (UV) LED that has a typical Vf of 10V. \$\endgroup\$ – Spehro Pefhany Jan 26 at 16:32
  • \$\begingroup\$ Be aware that normal silicon diodes will change the forward voltage by about 0.058 volts, for every 10:1 change in the current. If Vforward is 0.6 volts at 1mA, expect 0.542 volts at 100uA, and so forth. \$\endgroup\$ – analogsystemsrf Jan 29 at 3:58
19
\$\begingroup\$

Different semiconductor junctions have different forward voltages (and reverse leakage currents, and reverse breakdown voltages, etc.) The forward drop of a typical small-signal silicon diode is around 0.7 volts. Same thing only germanium, around 0.3V. The forward drop of a PIN (p-type, intrinsic, n-type) power diode like a 1N4004 is more like a volt or more. The forward drop of a typical 1A power Schottky is something like 0.3V at low currents, higher for their design working currents.

Band gap has a lot to do with it -- germanium has a lower band gap than silicon, which has a lower band gap than GaAs or other LED materials. Silicon carbide has a higher band gap yet, and silicon carbide Schottky diodes have forward drops of something like 2V (check my number on that).

Aside from band gap, the doping profile of the junction has a lot to do with it, too -- a Schottky diode is an extreme example, but a PIN diode will generally have a higher forward drop (and reverse breakdown voltage) than a PN junction. LED forward drops range from about 1.5V for red LEDs to 3 for blue -- this makes sense because the LED mechanism is basically to generate one photon per electron, so the forward drop in volts has to be equal to or more than the energy of the emitted photons in electron-volts.

\$\endgroup\$
  • \$\begingroup\$ small signal is more like 0.6V <1mA I agree. yet you did not mention there are 2 major contributions Rs + bandgap eV to Vf. This is why Green can be higher Vf than Blue yet lower eV \$\endgroup\$ – Sunnyskyguy EE75 Jan 26 at 20:34
  • \$\begingroup\$ Checked digikey to see what I could find on SiC schottky diodes, and the lowest Vf I could find is this obsolete one (in quite the fancy package) with a Vf of 1.3V. I'm not sure if that's a single junction or multiple, though, since power diodes tend to use multiple junctions in series. \$\endgroup\$ – Hearth Jan 26 at 21:04
  • \$\begingroup\$ Also, do you have a source on the 1N4004 being a PIN diode and not a simple PN diode? I had always thought it was just PN. \$\endgroup\$ – Hearth Jan 26 at 21:07
  • \$\begingroup\$ @Hearth There are many Cree SiC power diodes. Since eV is higher, Vt=1V yet PIV =2kV with Vf=2V@10A or Rs=0.1Ω in a package rated for 50W so k=0.2 which is excellent \$\endgroup\$ – Sunnyskyguy EE75 Jan 26 at 21:10
  • \$\begingroup\$ @SunnyskyguyEE75 I'm sorry, I can't seem to follow what you're saying there. This doesn't seem like it's actually a response to what I said, but I could just be out of it today... \$\endgroup\$ – Hearth Jan 26 at 21:12
14
\$\begingroup\$

Fundamentals

All materials in the chemical table and molecules of different combinations have unique electrical properties. But there are only 3 basic electrical categories; conductor, insulator( = dielectric) and semiconductor. The orbital radius of an electron is a measure of its energy, but each of many electron orbits formed in bands can be:

  • spread far apart = insulators
  • overlap or no gap = conductors
  • small gap = Semiconductors.

This is defined as the Band Gap energy in electron volts or eV.

Laws of Physics

The eV level of different material combinations directly affects the wavelength of light and the forward voltage drop. So the wavelength of light is directly related to this gap and the black body energy defined by Planck's Law

So lower eV like conductors have low energy light with a longer wavelength (like heat = Infrared) and a low forward voltage "Threshold" or knee voltage, Vt such as; *1

Germanium           Ge  = 0.67eV,   Vt= 0.15V  @1mA  λp=tbd
Silicon             Si  = 1.14eV,   Vt= 0.63V  @1mA  λp=1200nm (SIR) 
Gallium Phosphide   GaP = 2.26 eV,  Vt= 1.8V   @1mA  λp=555nm (Grn)

Different alloys from dopants make different band gaps and wavelengths and Vf.

Old LED Technology

SiC         2.64 eV Blue
GaP         2.19 eV Green
GaP.85As.15 2.11 eV Yellow
GaP.65As.35 2.03 eV Orange
GaP.4As.6   1.91 eV Red

Here is a range from Ge to Sch to Si low-med current diodes with their VI curve, where the linear slope is due to Rs = ΔVf/ΔIf.

enter image description here

Newer alloys created may have similar colours at different radii but similar colours share the same band gap but may have a larger Vf yet still proportional to the eV energy which is inverse to wavelength. These are selected for reasons of improved power levels and lower series conductor resistance, Rs which is always inversely related \$R_s = \dfrac{k}{P_{max}}\$.

  • Thus a 65mW 5mm LED with a 0.2mm² chip and k=1 has Rs=1/65mW=16 Ω with a tolerance ~ +25%/-10% but older ones or rejects were +50% and better ones with slightly bigger chips ~ 10Ω yet still limited by the thermal insulation of 5mm epoxy case for heat rise.
  • then a 1W SMD LED with a k=0.25 to 1 may have Rs=0.25 to 1 Ω with arrays scaling the resistance by Series/Parallel factored by S/P x Ω and the voltage by number in Series.

k is my vendor quality related constant related to thermal conductivity of the chip thermal resistance and efficacy as well as the designer's board thermal resistance.

Yet k typ. only varies from 1.5 (poor) to 0.22 (best) for all diodes. Lower the better is found in newer SMD LEDs that may dissipate heat in the board and old Si case mounted power diodes and also improved in new SiC power diodes. So SiC has a higher eV thus higher Vt at low current but much higher reverse voltage breakdown than Si which is useful for high voltage high power switches.

Conclusion

Vf of any diode is a result of Band gap energy for the threshold voltage, Vt at the curve knee (X-axis intersection) and the conduction loss, Rs such that \$V_f=V_t+I_f*R_s\$ is a good approximation of the linear curve at Tjcn=25'C.

If we include the package power rating with some temp rise to Tj=85'C we can also estimate \$V_f=V_t+\dfrac{kI_f}{P_{max}}\$ However you never find k published in any datasheets, like many others, it is a designer's selection criteria ( or customer's Quality control variable) or Figure of Merit (FOM) like gm * nF * Ω=T[ns] for MOSFETs RdsOn.

Ref

*1

I changed Vf to Vt since Vf in datasheets is the recommended current rating, which includes bandgap and conduction loss but Vt does not include rated conduction loss Rs @ If.

Just as MOSFETs Vgs(th)=Vt= the threshold voltage when Id= x00uA which is still very high Rds yet starting to conduct and you usually need Vgs= 2 to 2.5 x Vt to get RdsOn.

exceptions

Power Diode MFG:Cree Silicon Carbide (SiC) 1700V PIV, @ 10A 2V @ 25'C 3.4@ 175'C @ 0.5A 1V @ 25'C Pd max = 50W @ Tc=110C and Tj=175'C

So Vt=1V, Rs ¼ Ω, Vr=1700V, k= ¼Ω * 50W = 12.5 is high due to 1.7kV PIV rating.

  • @ Tj=175'C = (3.4-1.0)V/(10-0.5)A = ¼ Ω , k= Rs*Pmax

    enter image description here

Here the Vf has a positive tempco , PTC unlike most diodes due the Rs dominating the bandgap senstive Vt which is still NTC. This makes is easy to stack in parallel without thermal runaway.

\$\endgroup\$
  • \$\begingroup\$ A link to the source materials would be helpful. \$\endgroup\$ – Jack Creasey Jan 26 at 20:40
  • \$\begingroup\$ you got it Jack. TY for asking \$\endgroup\$ – Sunnyskyguy EE75 Jan 26 at 20:47
4
\$\begingroup\$

The voltage drop across a forward biased junction depends on the choice of materials. A common PN silicon diode has a forward voltage of about 0.7V, but LEDs are made from different materials and so have different forward voltage drops.

\$\endgroup\$
  • \$\begingroup\$ Choice of materials, and doping concentration. Material is a more significant effect, though. \$\endgroup\$ – Hearth Jan 26 at 16:21

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.