6

I read that galium nitride and silicon carbide are piezo materials.Does this cause some problems? I can imagine that unwanted piezoelectricity directly in the circuit path would cause some kind of distortion or noise.

Can this be modeled as some kind of parasitic voltage source? How strong is this effect? I know some piezoelectric materials are alot more piezoelectric than others,is SiC and GaN only extremly weak piezoelectric materials ,so weak that it can be ignored?

wav scientist
  • 641
  • 1
  • 5
  • 13
  • 2
    I really don't want to be near a PCB where vibration turns on the gates... –  Apr 24 '18 at 10:40
  • I did not think it could turn on gate.I was thinking it will cause some subtle but very real distortion and noise.Also,I believe the piezo semiconductor is not in direct electric path with gate? I mean gate is isolated by oxide layer so I am not sure if it will even affect the gate. – wav scientist Apr 24 '18 at 11:03
  • Gate is isolated for DC, but there is miller capacitance, so you know. Although vibrations have low frequency, so maybe i am overreacting –  Apr 24 '18 at 11:34
  • Either way gan or sic transistors are used as switches for high power, hence a little low frequency noise there should not matter. Of course i am not sure, i only read datasheets so far. –  Apr 24 '18 at 11:35

1 Answers1

2

I imagine the actual answer may depend on the transistors you're interested in.

For example, there's no "oxide layer" in a typical GaN-on-SiC HEMT; the gate forms a Schottky contact and the device is a "planar" device with a 2DEG. Piezoelectricity has little effect here since there is little potential across the crystal (only along the surface). Incidentally, typical SiC substrates are semi-insulating, so there's no electrical contact on the back-side of the GaN, either. Without the second electrical contact, you won't get any piezo-related effects.

Shamtam
  • 3,546
  • 19
  • 27
  • And what about pressure? Either sound or vibration or maybe even the rapid heating when conducting very large current can heat the piezo fast and make it expand due to thermal coefficient of expansion.Becose the transistor package doesnt heat up as quickly as the conducting piezo semiconductor inside,I think it might be rapidly compressed.Also,correct me if I wrong,but when the transistor is off or when it is transitioning between ON and OFF state,there is potential difference in the piezo semiconductor material. – wav scientist Apr 27 '18 at 01:27