Naked Science Forum
Non Life Sciences => Technology => Topic started by: Geezer on 24/02/2012 04:51:45
-
To put it another way, why can't the control field be made to "float" with respect to the channel?
-
You can get them with 4 terminals (http://www.electronics-related.com/usenet/design/show/33961-1.php) the but the body and source are usually connected internally ...
(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2F0%2F05%2FMOSFET_functioning_body.svg&hash=cb0c04d02ad3610c25499af14b9c083f)
so only three connections.
-
Maybe they are only kinda-sorta field effect devices. Is this a case of dodgy advertising? After all, a field is a field. Who cares what's really producing it?
We'll probably have to wait for some pointy-headed semiconductor guru to show up and e-splain it.
-
The nearest approach to such devices are opto coupled devices such are used as control devices in PSU,s where complete isolation between the mains supply and the output is wanted.
This also applies to Triacs via which this computer is powered.
-
For all FETs the field is created by the voltages between the terminals. For the most widely used MOSFETS (which are Insulated Gate devices or IGFETS) the most significant controlling field is between the Gate and the source. The devices are generally symmetric so the source and drain are interchangeable unless one is connected to the bulk (substrate). There are different regions of operation: 1. The Gate-source voltage is less than the necessary threshold voltage needed to cause channel formation - in this mode there is no significant source drain conduction. 2. The triode (or non-saturation) region where a channel is formed by inversion of the surface and source drain conduction can take place and 3. if the drain voltage is sufficiently high, the saturation region where the channel is terminated by the depletion region associated with the drain. The three regions have different characteristics which are usefully emplyed in design. The maths can be found on many sites on the web.
The channel formation is also influenced by the voltage on the substrate which is effectively another gate acting on the channel, though the effect here is more complicated; it is usually modelled as affecting the threshold voltage from the proper gate.
Anyway, basically the operation is affected by all the fields from all the terminals although, for many applications, we like to use only the gate-source to control the device and the other effects are more of a nuisance. The field that controls conduction is that field which inverts the surface and generates more carriers (lowers resistance). It has to act between terminals. A strong field across the channel (or any conductor) can modulate impedance but the effect is small by just modulating the ability of the carriers to flow optimally along the conductor. A magnetic field can also act in a similar way as in Hall effect devices. These effect play no part in conduction in FETs.
Ordinary FETs use the reversed biassed depletion region of a diode to control the conduction of an already conducting channel which evetually gets pinched off when the region is depleted of carriers. No surface inversion is involved. The similarity is that there is a local field controlling the conduction. In neither case is it an external field.
I don't think I explained this too well but there is plenty on the web on the subject. Need to sharpen my head more :-)
-
Thanks Graham!
I will have to read your post about ten times before it sinks in.
BTW, you'll be impressed to hear that I have actually been doing a bit of Spice modeling [8D]
-
Are you using Simetrix for Spice modelling? It's not bad for a freebie and OK for small, general purpose circuits, or even drawing nice looking diagrams.
-
I'm using something called 5Spice. There is a free version which cleverly restricts the schematic page size. If you run out of space, you have to buy a license, but it's fine for what I'm doing, and it seems to be very solid and easy to learn. Now that I have some basis for comparison, I can take a look at Simetrix too. Thanks for the hint.
I was mucking about with that Idea I had, so I decided to expose it to the afianados in the field. I was lucky to avoid the Spanish Inquisition! Someone suggested I simulate it, so I thought I better call their bluff. It's certainly a bit less painful than burning one's fingers with a soldering iron.
-
I like Simetrix. Thanks Graham!