Member
Joined 2009
Paid Member
SQLGuy,
It would be exceedingly distressing to use a period denoting full stop after that confused statement. More relevantly, it states that two variables cannot be calculated simultaneously (anywhere in the universe). What to expect from an analysis where linearity and time are mathematical constants and highly influential parameters have ideal values?
None. Both lack proper propulsion means so they will be having a hard time getting any speed whatsovever.
Seriously, "faster" isn't well defined enough to make much sense. The beta=10 sure has a higher turnover frequency, but the beta=100 has higer or equal current gain at all frequencies. It depends on the application.
Sorry. It's called emphasis. It's something we do in English. I didn't realize it would be confusing.
You misunderstand what Heisenberg's principle says. It says that, due to the effects of measurement, e.g. the bombardment of electrons that it would take to illuminate and measure the position of a particle, the very act of measurement disturbs the item being measured. More specifically, it states that you cannot know with pefect precision BOTH the position and velocity of a very small particle. You can know one or the other, but not both. As you increase the precision of one measurement, you reduce the precision of the other.
However, this principle doesn't just define the limits of what can be measured; as far a current modern physics can tell, it defines the limits of what actually exists. In other words, unobserved quantum particles, like electrons, are not point objects, but smears of probability. When observed, the electron will collapse to some point within the space defined by the Schroedinger probability equation. It will probably show up at the point of highest probability, but it could show up anywhere within the "smear." This is how tunneling works. In a tunnel diode, for instance, there is an insulating layer through which electrons cannot pass; but, since the probability fields of the electrons extend beyond this barrier (probability field shapes are not at all influenced by external objects or pressures), some electrons will appear on the other side of the barrier.
The biggest thing to keep in mind, and that I think you are missing, is that this all applies only to VERY SMALL things or VERY SHORT periods of time. Planck discovered and calculated the applications of Heisenberg's principle to define the minimums of things that could be measured. A Planck length, for example, the minimum length that can be measured, is 1.616252×10−35 meters.
Please explain, now, how any of this is going to affect any of LTSpice's very macro analyses.
Thanks,
Paul
Last edited:
If that helps, alpha=1-Tr/Tn where Tr is the transit time through the base and Tn is the minority carrier lifetime in the quasineutral base. Recall beta = alpha/(1-alpha) and, of course, low Tr means high Ft.
Now you draw the conclusion. Whatever that is, I'll tell you in advance it's wrong for at least two good reasons:
- The base concentration grading is essential for defining Tr. As such, this is way more important than the simple Tr to alpha relationship, and can easily reverse the conclusion.
- While Ft is a small signal related number, where the device models are already linearized, "speed" is usually related to the large signal model where the nonlinear charge injection and transport mechanisms are dominant.
Comparing Ft with the "speed" and correlating to beta is at best as correct as comparing small signal and large signal slew rates in an amp. A relationship may exist, but it doesn't always hold for all topologies and is ultimately largely irrelevant.
Okay....
So far we have Claude who is sure the answer is b) and doesn't wish to ask a friend. Megajocke who thinks it depends on application. Lumba Ogir who thinks it is entirely to do with fT and syn08 who thinks, erm, well...that something is irrelevant.
So what am I to go on? The respondent with the clearest choice sounds likely to have a good reason, so perhaps Claude would elaborate.
BTW, I deliberately didn't define "faster" so as to allow this to be defined as required. Of course this thread is meant to be talking about BJT output transistors and we have been touching on NFB and so on, so the definition of "faster" should take that into account.
So far we have Claude who is sure the answer is b) and doesn't wish to ask a friend. Megajocke who thinks it depends on application. Lumba Ogir who thinks it is entirely to do with fT and syn08 who thinks, erm, well...that something is irrelevant.
So what am I to go on? The respondent with the clearest choice sounds likely to have a good reason, so perhaps Claude would elaborate.
BTW, I deliberately didn't define "faster" so as to allow this to be defined as required. Of course this thread is meant to be talking about BJT output transistors and we have been touching on NFB and so on, so the definition of "faster" should take that into account.
Member
Joined 2009
Paid Member
I like choice a). My reason is that the higher hfe implies less base current required and this reduces the constraints on the drivers and improves the chances that they won't contribute to distortion.
Of course Heisenberg does impact spice, since it runs on a computer and that computer relies on solid state which relies on the properties of electronics and materials...and the output of the program requires an observer to be meaningful and that's us and we have brains and our brains are thought to employ quantum phenomena (ever read The Emperors New Mind ?). I award 1 point to Lumba !
Of course Heisenberg does impact spice, since it runs on a computer and that computer relies on solid state which relies on the properties of electronics and materials...and the output of the program requires an observer to be meaningful and that's us and we have brains and our brains are thought to employ quantum phenomena (ever read The Emperors New Mind ?). I award 1 point to Lumba !
Last edited:
So, by extension, Heisenberg uncertainty makes everything you could ever use, see, or otherwise experience useless? Your SET amp built by first principles of meditation, good karma, and feng shui, although it may sound great today, may sound terrible tomorrow because Heisenberg messed with it (or your brain)?
I'd definetely go for choice a) if it's about output transistors in an audio amplifier if I don't need the higher high-voltage SOA of the low-gain types. Less load for the previous stage is good in this case. On second thought, most any position in a typical audio amplifier topology would benefeit from higher beta with fT unchanged.
Less load for the previous stage is good in this case. On second thought, most any position in a typical audio amplifier topology would benefeit from higher beta with fT unchanged.
Okay....
So far we have Claude who is sure the answer is b) and doesn't wish to ask a friend. Megajocke who thinks it depends on application. Lumba Ogir who thinks it is entirely to do with fT and syn08 who thinks, erm, well...that something is irrelevant.
So what am I to go on? The respondent with the clearest choice sounds likely to have a good reason, so perhaps Claude would elaborate.
BTW, I deliberately didn't define "faster" so as to allow this to be defined as required. Of course this thread is meant to be talking about BJT output transistors and we have been touching on NFB and so on, so the definition of "faster" should take that into account.
My response was an off the cuff answer based on small signal criteria. That may not be a valid assumption. Anyway, here goes.
With both devices having the same ft value, obviously the one with beta = 10 has a higher break frequency than the device with beta of 100 (beta is low frequency value). This, in the small signal domain, is modeled as a single pole with break frequency "fb", which denotes the frequency where beta starts to decline (-3 dB, -45 deg phase).
So, when analyzing the small signal model, the lower beta device has its pole at higher freq, hence its bandwidth is higher. By strict definition, the BW is the freq where the transfer function is down 3 dB for single pole response. That is the BW definition, and what I interpreted as "faster". Of course, the unit with beta of 100 has more raw gain at low freq. But BW is lower as the pole is lower.
Two amplifying networks have differing gain. Although one has more lf gain, its break freq is low. The one with less gain at lf breaks at higher freq. By definition, the low beta high break freq unit has higher BW, hence higher "speed".
Did I miss anything? Off the cuff, that was my thought process. Here is an analogy.
Two op amp networks have differing lf gains. One has a gain of 20, the other a gain of 2 (closed loop). The high gain amp has a single pole at 1.0 kHz, the low gain amp has its pole at 10 kHz. The first has a BW of 1.0 kHz, and the second has BW = 10 kHz. By definition, the 2nd unit is faster.
Syn08's post is totally correct and he is explaining well the confusion people have in mixing large ( physical) and small signal variables.
Of course astronomy and astrology both are dealing with stars.
JPV