He, he. Then they aren't Schottky diodes. They are insects that are from the Amazon. :lol: :lol:

__________________
Uncle $crooge

 

not saying that at all. I am just pointing out the bad points abt Schottky's
WhenI first learnt abt them at Uni they sounded great (low volt drop, fast switching) and I wander why they were not used that much.

The fact is they are great and have their place, when you need to have as much volts as possible in a situation where the diode-effect is needed - the 0.3V is too attractive.

So you have a switching cct one with a diode and one with a Schottky

The diode one will have a high on-state losses since it has a higher on-state voltage, so the Schottky ends up winnning since its conduction losses are 50% better

So the diode is blocking it has a leakage current, the Schottky also has one that is worse then a standard diode, but I am talking abt 100s of uA so might not be significant

So you come to turn-OFF of the diode and turn-ON of your switch. As stated the diode will "dump" its stored charge to provide the reverse-recovery current. This extra blat of current will increase the switching losses of both devices (the diode has to turn-off with that extra current and the FET/IGBT must switch into an extra current of that amount)

diodes tent to have switching losses in the order of single mJ so an extra few hunder As wont do much. The problem exists for the switching device who (as seen in my piccy) at its turn-on has to turn-on into extra current, be that extra current for a finite time, the time it take to fully turn-on

All that means is the controlled devices switching losses will go up. ie cool it better

Attached Images

waveform.png (20.6 KB, 163 views)

__________________
Nothing is impossible.
Once a problem is realised, the rest is just details

 

Styx, I see your points, and I appreciate being enlightened. I guess that with the relatively recent availability of really fast recovery PN diodes, the Schottky advantages tend to pale. I was only dimly aware of these advances in diode technology until I started researching this issue.
Nevertheless, I still saw a lot of Schottkys being used in low-voltage freewheeling and flyback applications when I was researching this. Your company apparently believes that PN diodes are superior for freewheeling applications. Do you think that they are a clearly superior choice, or is it still pretty close?

I have a more detailed question. If the forward current in the diode has gone to zero (inductor has fully discharged) before the next turn-on of the controlled device, then it seems that the PN diode would be a clear choice, because stored charge would be essentially zero, and the junction capacitance is much less than that of a comparable Schottky. Is that a fair statement?

__________________
Ron

 

no problem, we are all here to learn.
Schottky's are still great and as I have said they have their place. I was just putting forward a couple of points abt Schottky's

As to hte nevative points of a Schottky (leakage, REverse-recovery,blocking voltage)
The only one of major concern to where I work is the blocking voltage
leakage might be higher then a PN, but we are still talking abt 100's of uA so not that significant (if used in power ccts)
The reverse-recovery as has been shown exist for a Schottky as well as a PN and if the benefits of a Schottky are needed then cope with the slight higher (then to a PN) with better heatsinking.
The only major problem is the blocking voltage - 200V is far too low

Now for SMPS from 12V, 28V Schottky's are by far the superior option, but for higher voltage stuff then they cannot even be touched. this is changing with Silicon Carbide - We have just been offered a 1200V 3ph H-bridge IGBT module with SiC for the diodes !!!! looking very good

that is a fair statement, leaving the only deciding decision between PN and Schottky being down to the volt-drop (be it for the conduction losses or for trying to recover as many volts as possible)

__________________
Nothing is impossible.
Once a problem is realised, the rest is just details