Transistor question -theory
in [Basics]
|
Prev: Integration w/o Smoothing; Demodulate, Integrate and then Take the Quotient
Next: Integration w/o Smoothing; Demodulate, Integrate and then Take the Quotient
From: lerameur on 19 Feb 2010 15:11 Hi, I have been analyzing the 'how to' of a transistor recently. Electron flow, holes, dope material etc... I still do not understand why a transistor can be reversed, anyway reverse based from all those web pages. They all show a transistor either N-P-N or P-N-P, lets take NPN or emitter-base-collector why cant we change it to collector-base- emitter, isnt the emitter and collector both N, dopes with free electrons?, shouldn't they act the same way?? this will lead me to another question, is there such devices that will work they way I mentioned, meaning when you activate the base, the current could flow in either directions ?? just like a mechanical relay. thank you ken
From: Jon Kirwan on 19 Feb 2010 15:33 On Fri, 19 Feb 2010 12:11:19 -0800 (PST), lerameur <lerameur(a)yahoo.com> wrote: >I have been analyzing the 'how to' of a transistor recently. Electron >flow, holes, dope material etc... I still do not understand why a >transistor can be reversed, anyway reverse based from all those web >pages. They all show a transistor either N-P-N or P-N-P, lets take NPN >or emitter-base-collector why cant we change it to collector-base- >emitter, isnt the emitter and collector both N, dopes with free >electrons?, shouldn't they act the same way?? Yes.... but the geometry of the emitter and the amount of dopant for the N material in it may differ from the collector's. Very old transistors were, I seem to recall, made from quite similar N materials for both the collector and emitter and were manually fused together. They were expensive, but they worked about like you suggest -- roughly equally either way -- if I got that much right. Perhaps someone with better knowledge about this can comment, though. But modern BJTs aren't built that way. They can diffuse different concentrations of dopants for the emitter vs the collector and the physical shape/size is also different. Which emphasizes one orientation/use over the other. You can still wire them up the other way, but the beta will probably be a lot lower. Some other features may be better, though. I suppose some designers take advantage of that. Not so often that I've seen it done much, though. Jon >this will lead me to another question, is there such devices that will >work they way I mentioned, meaning when you activate the base, the >current could flow in either directions ?? just like a mechanical >relay. >thank you >ken
From: Tim Wescott on 19 Feb 2010 15:38 On Fri, 19 Feb 2010 12:11:19 -0800, lerameur wrote: > Hi, > > I have been analyzing the 'how to' of a transistor recently. Electron > flow, holes, dope material etc... I still do not understand why a > transistor can be reversed, anyway reverse based from all those web > pages. They all show a transistor either N-P-N or P-N-P, lets take NPN > or emitter-base-collector why cant we change it to collector-base- > emitter, isnt the emitter and collector both N, dopes with free > electrons?, shouldn't they act the same way?? The doping levels are different between the emitter and collector, and in many transistor processes the emitter is smaller than the collector, making it harder for the carriers to 'find' the emitter when it's operated in reverse. Most bipolar transistors _can_ be operated in 'reverse', but with a severely degraded current gain -- I think a current gain less than unity is not uncommon. > this will lead me to another question, is there such devices that will > work they way I mentioned, meaning when you activate the base, the > current could flow in either directions ?? just like a mechanical relay. Most FETs operate this way, at least to a first order. However, many power FETs have diodes that are 'built in' as a side effect of the manufacturing process, so you can't just freely turn any old FET around. "Electronic switches" are commonly implemented as complementary small signal FETs in parallel, with appropriate gate control to turn them both on at the same time. Now, if I'm lucky Jim Thompson hasn't answered by the time I press "send", so I can get credit for an incomplete explanation before someone who actually knows his stuff comes on line with the real deal. -- www.wescottdesign.com
From: George Herold on 19 Feb 2010 15:47 On Feb 19, 3:33 pm, Jon Kirwan <j...(a)infinitefactors.org> wrote: > On Fri, 19 Feb 2010 12:11:19 -0800 (PST), lerameur > > <leram...(a)yahoo.com> wrote: > >I have been analyzing the 'how to' of a transistor recently. Electron > >flow, holes, dope material etc... I still do not understand why a > >transistor can be reversed, anyway reverse based from all those web > >pages. They all show a transistor either N-P-N or P-N-P, lets take NPN > >or emitter-base-collector why cant we change it to collector-base- > >emitter, isnt the emitter and collector both N, dopes with free > >electrons?, shouldn't they act the same way?? > > Yes.... but the geometry of the emitter and the amount of > dopant for the N material in it may differ from the > collector's. Very old transistors were, I seem to recall, > made from quite similar N materials for both the collector > and emitter and were manually fused together. They were > expensive, but they worked about like you suggest -- roughly > equally either way -- if I got that much right. Perhaps > someone with better knowledge about this can comment, though. > > But modern BJTs aren't built that way. They can diffuse > different concentrations of dopants for the emitter vs the > collector and the physical shape/size is also different. > Which emphasizes one orientation/use over the other. You can > still wire them up the other way, but the beta will probably > be a lot lower. Some other features may be better, though. I > suppose some designers take advantage of that. Not so often > that I've seen it done much, though. > > Jon > > > > >this will lead me to another question, is there such devices that will > >work they way I mentioned, meaning when you activate the base, the > >current could flow in either directions ?? just like a mechanical > >relay. > >thank you > >ken- Hide quoted text - > > - Show quoted text - Caveat (I probably know less than Jon about transistors) But Ive always wanted to try that experiment. I think the biggest thing to look out for (besides the much reduced beta) will be the reverse bias Vbe voltage. Is this a zener breakdown? So keep the supply voltage low and it might work..... Sounds like a project for a late Friday afternoon.... George H.
From: Tim Wescott on 19 Feb 2010 16:16
On Fri, 19 Feb 2010 12:33:14 -0800, Jon Kirwan wrote: > On Fri, 19 Feb 2010 12:11:19 -0800 (PST), lerameur <lerameur(a)yahoo.com> > wrote: > >>I have been analyzing the 'how to' of a transistor recently. Electron >>flow, holes, dope material etc... I still do not understand why a >>transistor can be reversed, anyway reverse based from all those web >>pages. They all show a transistor either N-P-N or P-N-P, lets take NPN >>or emitter-base-collector why cant we change it to collector-base- >>emitter, isnt the emitter and collector both N, dopes with free >>electrons?, shouldn't they act the same way?? > > Yes.... but the geometry of the emitter and the amount of dopant for the > N material in it may differ from the collector's. Very old transistors > were, I seem to recall, made from quite similar N materials for both the > collector and emitter and were manually fused together. They were > expensive, but they worked about like you suggest -- roughly equally > either way -- if I got that much right. Perhaps someone with better > knowledge about this can comment, though. > > But modern BJTs aren't built that way. They can diffuse different > concentrations of dopants for the emitter vs the collector and the > physical shape/size is also different. Which emphasizes one > orientation/use over the other. You can still wire them up the other > way, but the beta will probably be a lot lower. Some other features may > be better, though. I suppose some designers take advantage of that. Not > so often that I've seen it done much, though. IIRC from my college days, TTL logic uses the reverse gain of the input stage to suck charge out of an intermediate stage's base, speeding up one of the transitions. -- www.wescottdesign.com |