0
\$\begingroup\$

Is the cathode for schottky diode hotter than the anode, as is common for most diodes, or is it reverse polarity? A schottky is a metal-semiconductor junction, with the metal at the anode. So it should be better for heatsinking, even if the cathode is actually larger, right (i.e. reverse polarity in package nomenclature)?

I'm specifically interested in the 1N5819 and 1N5822 leaded DO-41 and DO201 package.

\$\endgroup\$
  • \$\begingroup\$ Is the cathode usually hotter than the anode? That's interesting; I'd never heard that before. (not that I doubt you; I genuinely find it interesting!) \$\endgroup\$ – Hearth Oct 16 '18 at 19:46
  • \$\begingroup\$ It seems to be that the N substrate (which is what is used for most diodes) sits on a large metal plate. This is the cathode. The anode is tiny touching the other semiconductor at the center from the other side. But in a schottky, the other semiconductor is not there. I'm not sure if "hotter" is the right term though. Maybe I should call it "heatsink ability". It is true in a thermionic diode though! \$\endgroup\$ – Indraneel Oct 16 '18 at 23:00
2
\$\begingroup\$

I have never seen a diode with a thermal pad on the anode.
Always on the cathode if one exists.

This is a Vishay SS2PH10 Schottky Barrier Rectifier used in my Buck LED driver in a DO-220AA package.

The cathode has a much larger pad for thermal dissipation.

enter image description here


And the temperature measurement point is the cathode.

enter image description here



A schottky is a metal-semiconductor junction, with the metal at the anode. So it should be better for heatsinking


It may be true that the metal is going to conduct heat better but metal characteristics are not going to be affected much by heat.
The metal anode could be smaller than the cathode contact metal.

enter image description here
Source: Electronic Notes


The cathode also has a metal layer.

enter image description here

Source: Electronic Notes


enter image description here
Source: Electronics Maker


Which Method is Best?

Even though method 3 has the lowest RΦja method 1 may be the better method.
The RΦja listed is for leads only.
Method 1 would allow an additional thermal path from the case to the board.

Method 2 could be interesting if the vector pins were copper with large vertical fins.
Kind of like this:
enter image description here

\$\endgroup\$
  • \$\begingroup\$ Well, I'm not sure what to make of it then. Is it possible the "mounting method 3" picture in the onsemi datasheet is wrong? Did they misplace the black band in that? Should it have been the cathode that is actually towards the PCB? Or does the cathode hanging in the air actually dissipate more heat? FWIW, a lot of reverse polarity stud diodes seem to exist (ones carrying 200+ amps), where the stud is the anode. Some of them are schottky. \$\endgroup\$ – Indraneel Oct 17 '18 at 18:59
  • \$\begingroup\$ It is possible to design the diode to have equal thermal characteristics on both anode and cathode. I think they used the anode because it is connected to the ground plane. They may have done that to emphasize you can use the anode as well as the cathode. Then again it could be a mistake, but not likely. The PDF was created over 12 years ago, 7/5/2006. You can't be the first to notice. I would use method 1. \$\endgroup\$ – Misunderstood Oct 17 '18 at 19:33
2
\$\begingroup\$

There might be differences, but if your concern is just heatsinking, you have to rely on the thermal mode given by the manufacturer. According to the thermal model in the datasheet, they don't specify a different thermal resistance from junction to the anode or from the junction to the cathode.

enter image description here

This is the model you have to use to make your thermal calculations. RtL and rtJ are the thermal resistances for junction and lead.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the onsemi datasheet! Had missed it (way better than vishay and diodes). Their photo on the same page (mounting method 3) shows the anode towards PCB. But on page 2 "average rectified forward current" is temperature referenced to "cathode lead 1/32″ from case". Weird. Or maybe 3 amps is worst case, and mounting method 3 will provide more headroom...? The thermal data on that page sure looks like it. I just hope they have not misplaced the black band on that picture. \$\endgroup\$ – Indraneel Oct 16 '18 at 18:26
1
\$\begingroup\$

I wouldn't assume that the metal used to form the junction's anode is actually the metal used to form the diode's anode lead. This seems very unlikely to me. There's probably a thin bond wire connecting the two, so the thermal conduction may be no different than a normal PN diode. Check the datasheet or ask the manufacturer.

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

LEDs have an ultra-thin gold wire whisker bond to the anode so the heat sink is the cathode but since the wire bond is large in these diodes, they don’t have to look small which is why they conduct heat well to both terminals. Some power diodes now use aluminium instead of copper for wire bonds or direct terminals.

\$\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.