In a (hobby) project I'm doing, I noticed that the diodes I needed were more expensive than the NPN transistors, which left me wondering if I could substitute the transistors for the diodes.
I'm aware the base-emitter junction can be used as a diode (usually tying base and collector for better properties), but this limits the reverse voltage to \$V_{EBO}\$ (typically about 6V), which is too low for my application. Therefore I'm interested in using the base-collector junction, like so:
simulate this circuit – Schematic created using CircuitLab
Researching this, I've found multiple claims that B-C can be used as a diode, but very little information on whether or not this is a good idea, and its properties. So in particular, I'm wondering:
- Is it safe to use an NPN BJT in this configuration? Is this as robust as a dedicated diode?
- Should I leave the emitter unconnected (A), or tie it to the base (B)? How does this affect the properties?
- What would the resulting reverse breakdown voltage be? \$min(V_{CBO}, V_{CEO})\$?
- What would the resulting forward voltage be? Can it come close to \$V_{CE(sat)}\$ (at similar currents)?
- What forward current could the resulting diode support? Can it come close to \$I_C(max)\$?
- How fast would the resulting diode be?
Application
I would appreciate general answers on using B-C as a diode, but for my particular application the diodes (transistors) would be used as a low-current, low-frequency full bridge rectifier. I'm looking to maximize \$V_R\$, and minimize \$V_F\$ at a current of about 10-20 mA. The transistor I had in mind is the MMBTA06WT1G.