JFET Static Charge Detector

Good experiment and demonstration.

The drain to source current flow is highest when the gate to source voltage differential is zero.

Although this is obviously nit-picking, I think you'd find, if you tried it, that it would keep increasing as you went above zero. Obviously, at some point, the gate-source diode is going to start to become forward-biased and spoil things by conducting, but you'd probably have a small amount of leeway before that happened.

In the detector circuit, the gate/antenna becomes negative relative to the source...

That's a bit vague. How does it become negative? Are you claiming that the statically-charged object adds negative charge to the gate/antenna? An alternative hypothesis might be that it causes a redistribution of the free electrons in the wire, so that there's a greater concentration down the gate end.

Good points and thanks for the comments...  I originally used the word "positive" instead of zero.  But one of my references stated the gate should not be biased positive so I changed it to zero.  Your explanation makes the reasoning for doing that and the true behavior clear.

I researched how the gate/antenna works on the internet before writing but most of what I found made no sense and seemed like hand-waving blather.  I couldn't find anything in my textbooks.  Not wanting to add to the misinformation on the internet I made it vague.  In truth, I don't know the physics but it seems more likely that electrons are redistributed in the wire than that they are somehow transferred to the wire.  I like your explanation.

I couldn't find anything in my textbooks.

I'm no academic, and my physics is a bit dodgy at times, but perhaps we might describe what you've got as an irregular capacitor. For the 'old' physics models, that us engineers use, you'd be looking at a text on electricity and magnetism, and specifically electrostatics, electric fields, and capacitance.

The eraser is one 'plate', with an excess of trapped charge, the wire the other, with charge able to move beyond the capacitor but only within the confines of the length of wire, so all to do with 'end effects' and little regular capacitor theory. It's really more about how the electric field arranges itself in space and then changes as the free charge in the wire drifts as a reaction to it. Trying to calculate anything with regular capacitor models would be horrendous - textbook examples of capacitance always consider very simple structures with lots of symmetry - so here you'd probably use finite-element methods, in a similar way as is used for the fields around an antenna, and rely on the ability of a computer to do lots of calculations very quickly, but to do that you'd need a 3-d model of the structure to work with, and it would have to be run through multiple iterations [dividing time into discrete intervals] with all the usual worries about accumulated errors and so on, until it reached an equilibrium.

In practice it's much simpler to just do what you did and try it. Or think about it in a hand-wavy way, like I did, and imagine roughly how the electrons in the wire are going to behave simply from an electrostatic point of view. I did it backwards. You told me that the potential at the gate dropped, so a surfeit of electrons, driven down by the charge on the eraser, which in turn must be an excess of electrons. Simples! - as the interwibblers say [or maybe used to say - I don't make any effort to keep up].

Keep in mind that the JFET will be prone to some leakage across the reverse-biased junction, so if you leave the eraser close to the wire, the gate potential will slowly change as charge leaves the wire. (Increasing the reverse potential will encourage more leakage.) But the JFET has a neat 'reset' mechanism - think what happens if you let the gate get back close to zero and then remove the eraser again.

But, having said all that, a further thing to keep solidly in mind with any experiment like this, done in an inside environment, is the prevalence of mains fields. Your hand will distort those fields and it's quite possible that the effect you're seeing derives from that rather than the charge on the eraser. You would need to experiment further to show that it was the eraser causing the effect.

Finally, to continue the glorious tradition of polluting the interknots, the water analogy might be something like this: the eraser as the moon, the wire as a tidal river, and the JFET as a port where the boats sit on the mud at low tide. It doesn't quite work, but no worse than the 'capacitor as a bucket of water' kind of tosh.

Hi Jon,

Unrelated (well, slightly related although it's not fully 3D for the situation here) I found this cool free software yesterday, called atlc2 that allows to determine the capacitance of any arbitrary cross-sections of conductors, and in the presence of any arbitrary shape cross-section dielectrics. I think it can't do arbitrary 3D shape though, perhaps that requires a supercomputer.. but nevertheless even 2D is useful.

Here's my first attempt with it. The input has to be a graphic file (.bmp) describing the cross-section. Here I drew the conductors as red and blue, simulating copper trace, on top of a slab of FR4. The colours have to be specific RGB values to represent particular materials.

Then, I told the program how many millimetres each pixel represented, and it thinks about it, and generates this diagram:

And it also spits out the capacitance, of (I'm pretty sure) 1m length of such a cross-section. Also the inductance of the red element of that length too. And characteristic impedance. From the limited time trying it, it's a really cool program : )

Hi Jon,

Unrelated (well, slightly related although it's not fully 3D for the situation here) I found this cool free software yesterday, called atlc2 that allows to determine the capacitance of any arbitrary cross-sections of conductors, and in the presence of any arbitrary shape cross-section dielectrics. I think it can't do arbitrary 3D shape though, perhaps that requires a supercomputer.. but nevertheless even 2D is useful.

Here's my first attempt with it. The input has to be a graphic file (.bmp) describing the cross-section. Here I drew the conductors as red and blue, simulating copper trace, on top of a slab of FR4. The colours have to be specific RGB values to represent particular materials.

Then, I told the program how many millimetres each pixel represented, and it thinks about it, and generates this diagram:

And it also spits out the capacitance, of (I'm pretty sure) 1m length of such a cross-section. Also the inductance of the red element of that length too. And characteristic impedance. From the limited time trying it, it's a really cool program : )

That is an interesting UI - using color to specify material and pixel size for dimensions. Quick and easy