This post was prompted by the recent discussion about Diode-Connected Transistors.
The Crystal Set nutters are always seeking the holy grail of a perfect diode detector (it must be passive of course).
The problem is that existing diodes are lousy rectifiers at low currents (three steps forward and two steps back).
What is needed is a diode which has a steep forward/reverse curve at very low currents (e.g. even with bias, a Silicon diode is worse than the old fashioned Germanium).
The favorite at present is a zero-bias FET, using the high Q of a tuned circuit to develop a voltage swing to drive the Gate, which switches (varies) conduction in the (low impedance) Drain/Source path.
I think the best low-level detector diode is a germanium back diode, essentially a tunnel diode with very low peak point current. People still make them.
..as visions of a straw covered alpine shack floor, with a germanium spinning wheel in the corner come to mind. Outside, the dark and foreboding clouds of the impending Silicon invasion are gathering. A silence has filled the Valley.
Ah... Germanium Valley was never as popular and successful as Silicon valley, it never really had the same potential.
Back to the original point though, JFETs with source & drain connected together are one thought. The silicon Schottky CDSH270 (replacement for the
1N270, etc) is quite good compared with old timers like the OA47..OA95,
1N34A, etc. Original germanium diodes have low Vf only at low currents (possibly if people applied some of what has been learnt since the 1950's they could make much better modern germanium devices?). I am interested in back-diodes though, to try... I thought their supply had dried up... possibly they have advantages at high speed (well, compared with "precision rectifier" circuits anyway). So where do I get back-diodes from now?
Using a JFET (or pentode?) as an infinite impedance detector (which I think was what was meant in the first post) is pretty good (tends to be a bit fussy about the resistor value, but only a bit fussy). I have used them at
68MHz (with an MPF102, extracting sound FM carrier from VHF PAL TV signals)... no RF amplification, so a lot like a crystal set in finding the right rabbit to use for its ears.
I recall "super diode" being used to describe transistors with base to collector, used as a diode (e.g. thermal compensation in output stages), but they aren't great at low signal levels... pretty good at higher currents though. Now "super diode" is used only to mean "precision rectifier circuit" in many web pages.
--------------------------------------- Posted through
A back diode is the only germanium IC-like (planar, passivated, wire bonded) process I've heard of. Too bad they don't make serious switching-type TDs this way.
The best silicon RF detectors are low-barrier schottky diodes. SMS7621 is nice... 2 volts reverse rating, 0.26 pF junction capacitance.
Jfets used as diodes are pretty horrible. Lots of capacitance, lots of series resistance. Even the leakage is mediocre compared with what you can do with the c-b junction of a good transistor.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.