Calling somebody who knows about semiconductor physics

[Peter Snowberg]

I picked up a bunch of lumps of silicon from a local tourist shop that sells minerals. I'm sure it's junk from semiconductor manufacturing but I have no idea if its doped or not. When I put my meter on it looking for diode Vf, I can get values from about 0.7V to 2.2V where conduction seems very stable. I'm wondering what process I've observing since there seems to be a Vf consistent with a silicon PN junction, but I'm going metal->nugget->metal.

Anyway.... After a couple of other projects are out of the way I'm going to make a dual mode clipper out of two chunks.... "Silzium Overload"

[puretube]

sounds "un-doped" to me... kinda "analogue/bipolar"... tried inverted polarization?

[Peter Snowberg]

The test probes are made of the same metal and it doesn't seem to matter where I put them. I tried swapping polarity and using (roughly) the same contact points and no change in behavior.

I saw one site where a chunk of silicon was being used as a crude photodetector. The notes said it acted like there was a cap in series. DC was generated when light exposure started but then went away. Opposite polarity DC was generated in a similar pulse when removing the light.

Strange days indeed. Most peculariar mama, whow.

[puretube]

tried to apply light/mechanical strain/temperature and measure the outcome?

[you`re probably one of a few here to have heard the words Rochelle & Peltier before  :wink: ]

ps: been tinkering for 30 years now, how to seriously win electrical energy out of geo-tectonic forces with the help of a grain of salt...

[brett]

I can get values from about 0.7V to 2.2V

An educated guess, but I suspect those are the values you'd get through highly doped and low doped Si, respectively.  Think of low doping as moving towards non-conduction.  Of course, to get forward conduction, you'd need to be getting a piece of P doped on one probe and N doped on the other.  I assume that having any given combinations of NPNPNPNP regions in-between would have no net effect.

How interesting.  Let us know what you do with it.

[puretube]

reminds me of those first detector radios, where you had to find the right point to put the needle... ("pin the nail tonthe funky")

[Peter Snowberg]

A year and a half ago we were talking about esoteric diodes and right away the crystal "cat's hair" detector came to mind. I think Paul Perry added it to the list.

The thought of having to poke around for two good diodes before a show is not suitable for "mission critical" stompboxes. Studio & experimental use only.

I was wondering if maybe these nuggets were a slurry of P and N type silicon, but if I understand the manufacturing process, the main crystal starts off pure and undoped.

I can't seem to identify regions... anywhere the probes land it looks like I have a diode.

[Ge_Whiz]

If the silicon came in lumps, yes, it's probably native to silicon valley. The big stuff is usually grown undoped and only doped in later processing (as wafers). However, scrap silicon like this is probably the end of a boule, cut off where the impurities accumulate and thrown away.

The diode Vf values you are measuring are as a result of the silicon-metal junction, like a cat's-whisker detector. Varying Vf values will be due to a variety of factors - local doping, nature of the metal, crystal facet at the contact point, pressure etc.

This kind of silicon usually makes a reliable, if insensitive, detector for crystal radio. Don't rely on it for a gig, though.

You could also try galena and, better still, iron pyrites ('Fool's gold').

[brett]

I was wondering if maybe these nuggets were a slurry of P and N type silicon, but if I understand the manufacturing process, the main crystal starts off pure and undoped.

I doubt that manufacturers could be bothered starting with highly pure silicon.  The aim of P and N doping is to have an *excess" of holes or free electrons in the material, not to have them to the exclusion of the other.  I'm guessing, but base material might be fine if you start out with 1X of N dope and 1Y of P dope if you are going to add 10X of N dope to make N material and 10Y of P dope to make P material.

Very low levels of doping may also explain the large voltage drops you described above.

Again, I'm making guesses here that may be way off.

[Vsat]

Peter,
There is also a not-very pure type of silicon called "ferrosilicon" which is used in metal refining. The broken surface has a coarse-granular appearance (sort of  like broken cast iron), and a bright silvery luster. It is usually supplied in lump form.
Cheers. Mike

[Peter Snowberg]

Thanks for the replies everybody.

Mike, you may have it there.... I googled for images and these could be ferrosilicon nuggets.

Now to finish a few other projects first..... :lol:

[Narcosynthesis]

i think it will be acting as a photodiode in photovoltaic mode, ie creating a voltage from incident light

at least one of my lecturers (i am an engineering student) mentioned something very similar a couple of days ago

David