difference between bipolar and mosfet

Nope. You quoted this half-baked definition so it's quite proper for others to infer you're happy with it.

[snip juvenile insults]

That's rich, coming from someone who doesn't even understand transistor action. Your concept of it would have Shockley turning in his grave.

Translation: "Holy shit! I've made the kind of dumbass error in that LED thread typical of a bonehead like me. Let's hope I can get away with this half-baked excuse before Miles spots it and I get exposed as an even *bigger* bonehead!" :P

It would, as you no doubt know, take me a great deal longer than one minute to write out everything I know about Beta. But were I so inclined, at least the OP would have an extensive, complete and accurate grasp of the subject, rather than the false sense of competence you attempted to instill in him by means of your own. HTH.

Nicely put. I doubt it'll be the end of the story, though. I suspect John Fields may be a troll. No rational person could conceivably be so obtuse.

On Wed, 12 Jan 2005 08:06:17 GMT, "Kevin Aylward"

--
Well, Kevin, it's only analogy...

--
I _am_ happy with it, regardless of whether you choose to take issue
with it or not.  After all, your objections to it are largely
unfounded in that from the context of the definition everything being
described was quite clear.

--
Beta is simply collector current divided by base current.  At the OP's
current  level of understanding, that's all he needs to know, so
why would it take you longer than a minute to come up with that?

--
And you're the exception that proves the rule?

Sigh. If I might interject a couple of thoughts, just to hopefully redirect this little p***ing contest into something more productive.

Could everyone please note the sign on the door, here? This is sci.electronics.BASICS. While a number of us here are, in fact, Professional Electrical Engineers of Long and Revered Standing (at least, we have a diploma and someone actually pays us to do this stuff), the questions here are more than likely going to come from people who do NOT have such a background, are never going to get into these phenomena down to the quantum-mechanics level, and whose questions will be more than adequately satisfied by the simple, "classical" explanations.

Are those explanations, in many cases, "wrong" in the sense that they give what is to some degree a false or misleading understanding of the fundamental physics underlying the operation of these devices? Of course they are. But so what? If you're trying to answer a question posed by someone who does not now and likely never will care about his or her ability to derive the ideal diode equation from basic principles, these simple explanations are very likely still going to be the right choice. I strongly suspect that just about all of us who currently have some sort of "Engineer" title on our business cards started with just the same level of understanding, and it didn't stop any of us from getting where we are today. As long as it is made clear that the explanation being given IS a simplified look at things, and that later on - IF they choose to go further - they will learn more about what is really going on, I for one do not see the harm in starting out at this level, as opposed to effectively dumping a graduate-level solid-state physics texts in their laps and saying "there you go." That sort of approach, IMHO, does NOT serve the purpose of this newsgroup.

All too often, we seem to have requests for such basic information posted by someone who is quite clearly an amateur/hobbyist, followed by a response by someone apparently interested primarily in demonstrating their own command of the intricacies of the topic in question, rather than actually saying something that the original poster would find helpful. I would like to respectfully suggest that such an approach is, to say the very least, idiotic.

We now return you to your irregularly-scheduled chest-beating...

Bob M.

The fact that current through a bipolar transistor and a diode are described by basically the same equation points to the fact that a bipolar transistor is really just a diode, in which many or most of the electrons that would normally escape through the base are diverted to the collector by the geometry and chemistry of the device and voltage on the collector. Since a diode is clearly a voltage controlled device, it is clear that the voltage across the PN junction of the base to emitter that causes electron flow.

However, saying that current through the base has nothing to do with this is just wrong. Because of the fact that a diode is involved, the only way to keep the voltage of the base up is to pull electrons out of it through the base lead. The ratio of electrons out the base lead to electrons out the collector is 'fairly' stable, enough to be printed in datasheets as the famous hfe or beta parameter.

Because of this, the base current is yet another feature of the bipolar transistor that can be used to roughly predict the collector current, and thus to design circuits with. It is also easier to use, in my opinion, owing to the fact that the relationship between base and collector current is generally linear, within certain ranges.

On the other hand, in my experience, the voltage to current equation is far more less use, since Is is not generally published in datasheets, and is severely temperature dependent to boot (just like beta). Predicting the current through the collector of a a 2N3904, given a base voltage, is practically impossible. Predicting the current through a

2N3904 using beta is simple, if somewhat imprecise. By guessing a beta of 100, one can easily see that 10uA through the base will give about 1mA through the collector.

Both beta and EM have their place in one's toolkit. Why toss out tools that can be useful?

--
Regards,
   Robert Monsen

"Your Highness, I have no need of this hypothesis."
     - Pierre Laplace (1749-1827), to Napoleon,
        on why his works on celestial mechanics make no mention of God.

--- Speaking of obtuseness, you seem to have (after all the ass-kissing) missed the part that was missing. That is, the role that Vce plays in generating the collector current.

Kevin states that: "It is the electric field at Vbe that causes both base current and collector/emitter current"...

While it's true that Vbe causes Ib to flow and is, of course, a contributor to the flow of Ic, the implication is that all of the collector current flowing through the base region is due solely to Vbe, and that with a given Vbe and no Vce, the emitter current will be what it would have been had the collector been connected to a source odf voltage. In other words, the collector current which would normally have been making its way to the collector because of the attraction of Vce drawing it away from the base will now make its way to the base because of the lack of Vce. Such is not the case.

Consider this case with Ic flowing:

183mA--> 2N4401 [HP6285A]-----[FLUKE 8060A]----------C E-----GND B 1.8mA--> | [HP6216A]-----[WAVETEK 27XT]-----------+

and this, with no Ic flowing:

2N4401 C E-----GND B 15.1mA--> | [HP6216A]-----[WAVETEK 27XT]-----------+

Both power supplies were set up as voltage sources with no current limiting and adjusted to give the readings obtained in the upper example. The collector supply was then disconnected from the collector and the reading shown in the lower example was obtained. About a tenfold increase in base current because the collector current was no longer causing a voltage drop across the base-emitter diode, but certainly not the hundred-fold increase one would expect if Vbe were the sole contributor to the cause of Ic.

BTW, since I find you mildly annoying I decided to take a look at your posting history to see what you're about, and I found that other than about the 20 posts archived at Google (where you also seem to have an attitude) you seem only to have posted here, so welcome to the swamp.

While looking, I found this rather interesting post on this NG:

Um, yeah, but they drop voltage _according_ to current! If they're dropping 0,7V., they're not passing much current! Diodes are a crap way to drop voltage unless the load is light and predictable!!!

miles

which seems to belie your claim that you're an electrical engineer with 35 years of experience.

Continuing on, we find, from Danny T:

says

"There is a small voltage across a conducting diode, it is called the forward voltage drop and is about 0.7V for all normal diodes which are made from silicon. The forward voltage drop of a diode is almost constant whatever the current passing through the diode so they have a very steep characteristic (current-voltage graph)."

-- Danny

To which you replied:

The forward voltage drop is entirely dependent on temperature (the junction temp. of the p/n junction; which is in turn dependent upon the current passed.) Higher currents equals higher temp. equals lower voltage drop. It's a well known effect which can eventually destroy the diode altogether. The physics of diodes is actually more complex than a lot of texts would have you believe.

--- Indeed, but if you think the forward voltage drop is _entirely_ dependent on temperature, you seem to have missed reading some of the more fundamental ones.

For example, while it's certainly true that the voltage across the junction can be described by:

kT / If \ Vf = ---- ln ( 1 + ---- ) q \ Ir /

and that when T is equal to zero at 0°K, Vf will be 0, you've neglected to mention that current passing through the bulk resistance of the diode, at any temperature, will cause a drop across the junction which is dependent on the resistance and the charge flowing through the diode.

More importantly, perhaps, you pooh-pooh'd Danny T's idea to use a diode as a bad one merely because of your opinion, which was erroneous. Diodes are _often_ used as voltage dropping elements in the real world because of the small change in Vf caused by If. Moreover, your example of the negative TC of a diode destroying it would more closely describe a diode with a voltage source connected across it allowing the diode to get into thermal runaway. Such a condition would not happen with the load limiting the current through the diode and the diode sized to carry the required current under the required environmental conditions. Furthermore, depending on the diode, above a certain current the tempco becomes positive, something else you "neglected" to mention.

What's most disturbing, however, is that with Danny T admittedly being a newbie and asking for information, you deliberately sidestepped the issue when he presented you with the [valid] information he found which supported Andrew Holmes' suggestion to use a diode in order to keep from having to admit that you were wrong in stating that: "Diodes are a crap way to drop voltage unless the load is light and predictable!!!

For shame, sir! :-(

-- John Fields

--
Decent!

Indeed you did.

No I didn't. I specifically denied that. I said the *ELECTRIC FIELD* between the base and emitter causes charge to flow from emitter to the base region.

*THE FLOW OF CHARGE BETWEEN EMITTER AND BASE DOES NOT CAUSE THE FLOW OF CHARGE BETWEEN EMITTER AND COLLECTOR*

I can't say it any plainer.

All of the above. I repeat what I said here.

"Applying an *electric field* to the base emitter injects carriers from the emitter into the base region. Once the carriers are in the base region, they are attracted by the *electric field* of the collector and are swept up (collected) by the collector due to this *electric field*. Some of the emitter carriers just don't make it, and are picked up via the base terminal. This base current is an *effect* not a cause, and is incidental to the base emitter *electric field* injecting carriers."

*Nowhere* does this claim that the flow of charge causes another flow of charge.

Not at all. A correlation does not have to be causal. This is basic 101 statistics.

For example, suppose those that have cats have less stress, should we recommend that people get cats? Or is it that less stressed people simply have a side effect of liking cats.

No. This is why the tobacco industry had a legal claim against smoking causing cancer. It might be that those more likely to get cancer also had a side effect of just wanting to smoke.

One needs to show that a correlation is *actually* causal. In this case, it is Vbe that causes both collector current and base current. Base current does not cause the collector current, but it is correlated.

Your argument above says that you do disagree.

Your argument above says that you do disagree.

Because this analogy must account for the fact that there is a direct connection from the base to the emitter.

Its not like the tube. The tube has no connection from grid to cathode or anode.

Fair enough.

Kevin Aylward snipped-for-privacy@anasoft.co.uk

SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design.

just

and

Hurray Bob! :-)) I started to write something along these lines, but couldn't make it sound as nice as you just have. I think it's important to also note that everyone's model is wrong in the strictest sense of definition of correctness. They are all just models, and like all models, they serve to be a "reasonable" facsimile of the real and absolute truth; whatever that may be. They are not perfect or they wouldn't be called models. I'm reasonably sure that there are people that could come in and make a mockery of every model that has been presented here by demonstrating their shortcomings and failures. I suspect some of those people are actually present here reading this, but not wishing to add fuel to the fire.

--
LOL! Just as I thought; another lame-ass dodge!  

BTW, there was no error.  I just wanted to see if you could come to
that conclusion and, obviously, you couldn't.  No surprise, really,
after reading some of your other trash.

--
Well, Sherlock, you _do_ have a remarkable grasp of the obvious!

Not in this context it isnt.

No its not stable. Hfe varies all over the place. In contrast, the gm equation of a transistor is the same for all transistors.

Any decent design has to be done such that the huge variations in hfe are overcome.

No it most certainly cant be used to predict the collector current. For switching circuits it is used to calculate the *minimum* base current required, with the collector current being set by the collector loop not the base loop.

For ac circuits, hfe is designed out.

Ho hum...you don't do much transistor design do you?

This is a non starter.

Oh dear...in your experience...well, yet again, at the risk of sounding arrogant, my experience is rather more extensive then yours in the design of complex transistor level circuits. I have been professionally designing very large transistor count i.c. and board level circuits for well, some time now.

The gm equation is absolutely indispensable. It forms the heart of serious transistor level design. For example, see

e.g.

re = 1/40Ic

Av = Rc/re

max gain Av=Va/Vt

or the design of current mirrors, multipliers, bandgap voltage references. The list is truly endless.

Sure, if you just apply a raw voltage for DC conditions, but one doesn't usually do this. For ac conditions, ic= vb.40.Ibias

Oh dear, this approach is useless for anything but switching circuits, and as noted above, it doesn't determine the collector current. This is shown here

No one is claiming that beta has no use, one is pointing out that the bipolar transistor is a voltage controlled device.

Kevin Aylward snipped-for-privacy@anasoft.co.uk

SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design.

--
Actually, it's "What's the point of whipping a dead horse?" (Not you,
Kevin, the subject material. ;)

--
Then why did you snip it?  Bandwidth is cheap.

--
I should go on a wild goose chase at your insistence?  Forget it.  If
you have something to say, include it in the current post.