BJT Can be Considered a CCCS in some situations

[John Larkin]

On Thu, 16 Dec 2004 17:01:17 +0000, John Woodgate
<jmw@jmwa.demon.contraspam.yuk> wrote:

I read in sci.electronics.design that Jim Thompson
thegreatone@example.com> wrote (in <qge3s0dt6rc3ou53rqpam8rc7h7hgjr3sn@
4ax.com&gt about 'BJT Can be Considered a CCCS in some situations', on
Thu, 16 Dec 2004:


What some seem to be clueless about is that a MODEL does not necessarily
reflect the physics, it's just a design tool.

Yes, that's the cause of most of the argy-bargy. Kevin is talking about
device physics and others are talking about circuit models.

No, he's just talking about different models.

John

 

[Kevin Aylward]

John Larkin wrote:

On Thu, 16 Dec 2004 16:47:13 GMT, "Kevin Aylward"
salesEXTRACT@anasoft.co.uk> wrote:

John Larkin wrote:
On Thu, 16 Dec 2004 09:24:13 +0000 (UTC), toor@iquest.net (John S.
Dyson) wrote:

People who persist in claiming that a transistor is physically a
CCCS are of similar ilk as those who post (publish) various kinds
of nonsense theoretical nonsense on the physics groups.

No, there's a big difference.

Nonsense. John, you usually quite together, but you seem to be
missing the point. I shall phrase it to make it really clearly what
the issue is.

What part of "...a transistor is not physically a CCCS..." did you
fail to understand?


What you may not understand is the I don't give a damn about device
physics

I don't have a problem with that. The issue is that *my* statement has
*always* been about *device physics*. For the purposes of my statement I
don't care a dam, about any higher behavioural level model, so why
people keep bringing this up is beyond me. I was never discussing
circuit models.

except as it occasionally helps me visualize device behavior
in unusual situations, like maybe c-b junction reverse recovery or
something. I don't design transistors, I design electronics, so I
simply need working models appropriate to the situation. And as Jim
notes, all device models are just varying degrees of imperfection.


The beta model works, quantitatively, in
90% of design situations,

No one is claiming that hfe is immaterial. One is discussing the
*intrinsic* device physics. What is the physical mechanism as to how
the transistor works. It is *physically* a VCCS. Period. End of
story.


I don't recall anybody discussing intrinsic device physics here.

Oh dear... Right from the bloody well start of the debate I did. It was
when I stated "the bipolar transistor is a voltage controlled device".

When anyone tried to steer it away, I restated exactly what I was
claiming. Many simply ignored this and continued debating something not
in debate.

Nobody mentioned valence structures, Fermi levels, doping profiles,
device geometry, anything remorely like that. A few people did present
simplified cookbook equations for Ic versus Vbe, to which I presented
some practical counter-cases where that model was grossly inadequate.

Oh, the "end of story" thing is un-original and tedious.

And correct.

This makes no sense at all. The full equations work very well indeed.
That is, ones that treat base current as an effect of applied
voltage.

Only at low currents, and at DC.

By this I assume you mean the the main spice models don't always have a
good fit for hfe at all current ranges.

Perhaps people who are
totally clueless will be influenced by the incorrect rhetoric, but
at least, it makes NO difference to the world that the idiots are
misled.

I suppose "idiots" includes everyone who publishes bipolar
transistor datasheets; I've never seen one that didn't specify beta.


This is completely irrelevant to the question as to how the bipolar
transistor works.

Time and time again, people just don't seem to understand the point
being made.

for the last time, I will repeat the argument:

Promise?

Yeah...

Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

 

[Paul Burke]

Rich The Philosophizer wrote:

When the Moon is in the Seventh House,
And Jupiter Aligns with Mars,
Then Peace will Guide the Planets,
And Love will Steer the Stars!

All Together Now...

Sodomy, fellatio, cunnilingus, pederasty,
Father, why do these words sound so nasty?

We inherited the LP of the show after my cousin bought it and her good
Catholic mother made her get rid of it.

Paul Burke

 

[Paul Burridge]

On Fri, 17 Dec 2004 08:36:14 +0000, Paul Burke <paul@scazon.com>
wrote:

Rich The Philosophizer wrote:


When the Moon is in the Seventh House,
And Jupiter Aligns with Mars,
Then Peace will Guide the Planets,
And Love will Steer the Stars!

All Together Now...

Sodomy, fellatio, cunnilingus, pederasty,
Father, why do these words sound so nasty?

We inherited the LP of the show after my cousin bought it and her good
Catholic mother made her get rid of it.

What show's it come from? I remember the hit single in the pop charts
around 1969 or thereabouts.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[Winfield Hill]

Keith Williams wrote...

When I did my last board I bought upwards of six figures worth
of FPGA development software. The stuff actually worked and
was supported by real people.

Exactly what did you buy and what were your experiences with it?


--
Thanks,
- Win

 

[Paul Burridge]

On Fri, 17 Dec 2004 17:10:03 GMT, Rich Grise <rich@example.net> wrote:

http://news.bbc.co.uk/onthisday/hi/dates/stories/september/27/newsid_3107000/3107815.stm

Hmmm.. Heady days indeed!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[John Larkin]

On Sat, 18 Dec 2004 11:59:34 -0500, keith <krw@att.bizzzz> wrote:

On Fri, 17 Dec 2004 13:38:16 -0800, John Larkin wrote:

On Fri, 17 Dec 2004 11:31:52 -0500, Keith Williams <krw@att.bizzzz
wrote:

In article <cpup9401sva@drn.newsguy.com>, hill_a@t_rowland-dotties-
harvard-dot.s-edu says...
Keith Williams wrote...

When I did my last board I bought upwards of six figures worth
of FPGA development software. The stuff actually worked and
was supported by real people.

Exactly what did you buy and what were your experiences with it?

Synplicity Synplify Pro and Amplify (well that's what I ended up with
[*]), ModelSim PE, and the Xilinx "Alliance" PAR tools. I had very few
problems and those I did have were taken care of quickly. Since I'd
never done any FPGA design nor used any HDL, I found the tools I chose
and VHDL very easy to learn myself. ...not that there weren't some
frustrations along the way.

[*] I made some dead-ends through a small FPGA company that went belly-
up after the boards were laid out, before parts were delivered. :-(
That was an expensive lesson, though I didn't have a lot of choice
given the I/Os (1.5V @ 200MHz) I needed.


We do some fairly hairy stuff just using the Xilinx ISE software,
which costs about $800 a year. We mostly do schematics, with VHDL
inside some of the boxes when that makes more sense. FPGAs are
great... you can change it a dozen times and the pads don't fall off
or anything. It's great to be able to fling adders and multipliers and
big blocks of ram around almost for free.

I'd likely use the same sort of mixed schematic-VHDL if I were doing it
again. Though one of the reasons I took the project was that I wanted
to learn VHDL. ;-) IMO schematics are much nicer for data-flow, where VHDL
is nicer for control logic (state machines). I've seen some tools that
translate schematic to VHDL and keep all the entities, components,
libraries, and such coordinated. They promise to reduce the verbosity
of VHDL to a managable level. ;-)

When I did this design, ISE wasn't available (perhaps it's just the
ISE WebPack that's new?). A piece of the design needed some pretty tight
timing (as I indicated, 200MHz processssor bus and SRAM with 1.5V, and
lower I/Os). Without a lot of time and a boatload of cash I opted for the
more expensive tools. During the time (~'99-'00) I was reading of all
sorts of grief spilling out from the users of the Xilinx Foundation (I
think that was it) tool users. I ran into very little with the Synplicity
tools. Synplify's schematic representation of the output logic was
critical for what I was doing. It made it easy to see what source
constructs created what piles of CLBs. For instance, it liked the
CASE/WHEN construct over the IF/THEN/ELSE for anything nested more than a
few deep. Synplicity also had some nice features like defining how a
state machine was to be implemented by tagging a process with the
"ONEHOT", "BINARY", or "GREY" attribute.

ISE is sort of the next generation of Foundation; it's been around for
a year or so. We did a bunch of designs with Foundation - admittedly
fairly simple ones - and they worked fine, without a lot of hassle. I
did one fairly complex nano-processor sort of state machine in
Foundation, for the pokey old 4000-series parts, that ran at 77 MHz.

That project taught me enough VHDL to avoid the big layoffs of 2001 and
land in processor development, where the design is a *mountain* of
structural VHDL and (custom logic) schematics and is modeled by another
*mountain* of VHDL.

I had a lot of fun with FPGAs and would go back to doing that in a
heart-beat. Got a job? ;-)

Do you mean an opening? Not just now... there's literally no room for
anybody else. However, I did just buy a Chinese fortune-cookie factory
(and sewing sweatshop) and once we move in there, we're planning on
cranking up. If you're loose in 8 months or so, keep in touch.

John