Driver to drive?

[Jim Thompson]

On Sun, 05 Dec 2004 19:23:15 GMT, Fred Bloggs <nospam@nospam.com>
wrote:

PaulCsouls wrote:
I'll second that. I have the third edition. It provides some good
general guidelines and explanations.

Paul C

Yep- easy to read pablum for schmucks...no wonder you like it..

Goddammit Fred! Take you meds, or be declared extinct ;-)

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

 

[Fred Bloggs]

Jim Thompson wrote:

On Sun, 05 Dec 2004 19:23:15 GMT, Fred Bloggs <nospam@nospam.com
wrote:



PaulCsouls wrote:

I'll second that. I have the third edition. It provides some good
general guidelines and explanations.

Paul C

Yep- easy to read pablum for schmucks...no wonder you like it..



Goddammit Fred! Take you meds, or be declared extinct ;-)

...Jim Thompson

I am going to amuse myself with a simple jig for measuring VAF...later.

 

[John Larkin]

On Sun, 05 Dec 2004 12:43:18 -0700, Jim Thompson
<thegreatone@example.com> wrote:

On Sun, 05 Dec 2004 19:23:15 GMT, Fred Bloggs <nospam@nospam.com
wrote:



PaulCsouls wrote:
I'll second that. I have the third edition. It provides some good
general guidelines and explanations.

Paul C

Yep- easy to read pablum for schmucks...no wonder you like it..



Goddammit Fred! Take you meds, or be declared extinct ;-)

...Jim Thompson

Yeah, what a sweetheart.

John

 

[John Larkin]

On Sun, 05 Dec 2004 19:46:44 GMT, Fred Bloggs <nospam@nospam.com>
wrote:

Jim Thompson wrote:
On Sun, 05 Dec 2004 19:23:15 GMT, Fred Bloggs <nospam@nospam.com
wrote:



PaulCsouls wrote:

I'll second that. I have the third edition. It provides some good
general guidelines and explanations.

Paul C

Yep- easy to read pablum for schmucks...no wonder you like it..



Goddammit Fred! Take you meds, or be declared extinct ;-)

...Jim Thompson

I am going to amuse myself with a simple jig for measuring VAF...later.

What's a VAF?

(Coupon for one good free insult attached.)

John

 

[Terry Given]

Joerg wrote:

Hi Nicholas,

It's not the ears they are after ... can't make a dollar bill from
a pig's ear.



Actually you can. Maybe not a Dollar, but at least about half that minus
processing and other costs. Just ask our dogs, they love them. Pig ears
sell for around $12 per 25-pack at the wholesale store ;-)

Regards, Joerg

Hi Joerg,

Crikey, what do they feed the pigs - caviar? I buy dried pigs ears for
NZ$0.50 one-off, and Holly just loves them. Kinda like potato chips for
dogs

Cheers
Terry

 

[Jonathan Kirwan]

On Sun, 5 Dec 2004 06:13:21 -0500, Active8 <reply2group@ndbbm.net> wrote:

On 4 Dec 2004 20:07:20 -0800, Winfield Hill wrote:

Active8 wrote...

He also calls beta a "figure of merit", like I suggested.

Beta can be increased reducing the transistor's base thickness,
which in turn reduces the Early voltage. Remember Bob Widlar's
super-beta transistors? The figure of merit is the product of
beta and Early voltage. The product is a constant for a given
Fab transistor process - it's a tradeoff. Is that what he says?

No. What's this FOM called?

You were going to post a link to this, "Modeling the Bipolar
Transistor," by Ian Getreu (when he was working at Tektronix and
probably with their STS (semiconductor test systems) group when
it still existed.)" Right?

That was Jonathan. I referenced the GP_whatever doc where the "fudge
factor" Jonathan mentioned as

Vbc
1 + ---
Vaf

was rather

Vce
1 + ---
Vaf

"Appendix 4

"The Accuracy of the EM3 and GP Basewidth Modulation Models"

"In the derivation of both the EM3 and GP basewidth modulation models,
assumptions were made that appear to be rather inaccurate. The merits of these
assumptions are described here.

"a) EM3 Model

"The derivation of the EM3 basewidth modulation model is given in Appendix 2.
The major assumptions in the derivation are: (i) the neglect of second-order and
higher terms in the Taylor series expansion of the basewidth as a function of
V_bc and (ii) the extension of the constant base-doping results for I_s, beta_F,
and tau_B to the general, non-constant doping case.

"The validity of these assumptions is difficult to assess theoretically. Their
main justification lies in realizing that basewidth modulation is itself a
second-order effect. Therefore, only a first-order model of it is required,
since any second-order effects in the modeling of basewidth modulation are
third-order effects in overall performance. A second, perhaps more convincing,
argument is the fact that the simple EM3 model for basewidth modulation has been
used successfully in many applications. For those interested in a more accurate
model for basewidth modulation, the following analysis of the GP model gives the
details on how it is obtained.

"b) GP Model

"The EM3 model made the linear assumption at the beginning of the analysis. The
GP model, however, makes its major assumption (constant junction capacitance) at
the end of the analysis. Therefore, unlike the EM3 model, the GP model shows
what needs to be done if a more accurate model is required. This is illustrated
here.

"In the GP model, basewidth modulation is modeled by q_c (the Early effect) and
q_e (the Late effect). The expressions for q_c and q_e are

V_B'C'
/
| C_jC(V) dV
/
0
q_c = --------------- (A4.1)
Q_B0


V_B'E'
/
| C_jE(V) dV
/
0
q_e = --------------- (A4.2)
Q_B0

"In both cases, the integral is replaced by an average capacitance which is
assumed to be constant.

"Therefore
_
C_jC * V_B'C' ~ V_B'C'
q_c = ------------- = ------ (A4.3)
Q_B0 V_A

_
C_jE * V_B'C' ~ V_B'E'
q_e = ------------- = ------ (A4.4)
Q_B0 V_B

"where V_A (the Early voltage) and V_B (the inverse Early voltage), given by:

~ Q_B0
V_A = ---------------------- (A4.5)
V_B'C'
1 /
------ * | C_jC(V) dV
V_B'C' /
0


~ Q_B0
V_B = ---------------------- (A4.6)
V_B'E'
1 /
------ * | C_jE(V) dV
V_B'E' /
0

"are assumed to be constant and independent of junction voltage. Normally, the
assumption of constant V_A and V_B is acceptable because basewidth modulation is
a second-order effect (i.e., V_A and V_B are large). The accuracy of the
assumption of constant V_A and V_B is actually a function of the ranges of
voltages over which they are applied and of how they are measured. If they are
measured for one particular applied junction voltage they will, of course, be
most accurate at that voltage. However, if they are averaged over a range of
junction voltages, the accuracy over that range may be improved. This point is
illustrated now by assuming that V_A and V_B have been measured at zero applied
junction voltage (a convenient standard applied voltage) and by looking at the
accuracy of the constant-V_A and constant-V_B assumption under reverse and
forward bias. Therefore, it will be assumed that

Q_B0
V_A_meas = ------ (A4.5)
C_jC0


Q_B0
V_B_meas = ------ (A4.6)
C_jE0

"and that q_c and q_e are modeled as

V_B'C' C_jC0 * V_B'C'
q_c = -------- = -------------- (A4.7)
V_A_meas Q_B0


V_B'E' C_jE0 * V_B'E'
q_e = -------- = -------------- (A4.8)
V_B_meas Q_B0

"The effects of this assumption are shown in Fig. A4.1, which is drawn for both
the reverse-bias case, (a), and the forward-bias case, (b). As illustrated in
the figure, the integrals in Eqs. (A4.1) and (A4.2) represent the area under the
actual C-V curve while Eqs. (A4.7) and (A4.8) approximate this area by the
rectangles shown. Since Fig. A4.1 refers to _both_ Q_C and Q_E, the subscripts
C and E have been dropped.

------- Fig. A4.1

C_j
^
|
|
|
C_j0 * V_bias | .
(rectangle) | .
/ | .
/ | .
/ | .
/ |.
_ _ _ _ _ _ _ _ _ _ _ _ .....|
|+++++++++++++..........xxxxx|
|++++.........xxxxxxxxxxxxxxx|
...|....xxxxxxxxxxxxxxxxxxxxxxxx|
|xxxxxxxxxxxxxxxxxxxxxxxxxxxx|
<---------/-----------------------+-----------------> V
/ |
/ V_bias v
/ /
(x'd area) = | C_j dV
under curve /
0



C_j
^
| V_bias .
| / |
| | C_j dV .|
| / .+|
| 0 \ (under curve).+++|
| \ .+++++|
| \ .++++++++|
| \ .++++++++++++|
| .++++++++++++++++|
| .+++++++++++++++++++++|
|._+_+_+_+_+_+_+_+_+_+_+_+_+|
. |xxxxxxxxxxxxxxxxxxxxxxxxxxx|
. |xxxxxxxxxxxxxxxxxxxxxxxxxxx|
. |xxxxxxxxxxxxxxxxxxxxxxxxxxx|
|xxxxxxxxxxxxxxxxxxxxxxxxxxx|
<------------------+--------\-----------------------> V
| \
v \
\ C_j0 * V_bias
(rectangle)

Fig. A4.1. The difference between the accurate evaluation of the integral and
the C_j0, V_bias approximation for the reverse bias case (top) and the forward
bias case (bottom).

------- Fig. A4.1

"(i) Reverse Bias Case

"As can be seen from Fig. A4.1a, the constant-capacitance approximation is
relatively valid when the junction is reverse biased. A better agreement would
be obtained in this case if the measurement of the appropriate Early voltage was
made at some voltage between 0 and V_bias. However, this would require a prior
knowledge of V-bias. Without this information, the use of Eqs. (A4.7) and
(A4.8) appears to be relatively reasonable.

"(ii) Forward Bias Case

"Obviously, as seen from Fig. A4.1b, the constant-capacitance approximation can
cause gross errors when the junction is forward biased. As with the
reverse-bias case, a better value of V_A or V_B would be obtained if the
measurement was made at some voltage between 0 and V_bias. The justification
for still retaining Eqs. (A4.7) and (A4.8) lies in the assumption that under
forward bias the value of q_e or q_c is very much less than unity. If this is
not so, the integration should be performed, as shown below.

"In the EM2 model, the variation of C_j with V was assumed to have the form (in
the SLIC and SINC programs)

C_j0
C_j(V) = ------------- for V <= phi/2
(1 - V/phi)^m
(A4.9)

= 2^m * C_j0 * [2*m*(V/phi) + (1-m)] for V >= phi/2

"Integration yields for q_e and q_c

/ \
phi 1 | / V \ (1-m) |
q = ------ * ----- * | 1 - | 1 - --- | | for V <= phi/2
V_meas (1-m) | \ phi / |
\ /


/ \ (A4.10)
phi 1 | / 1 \ (1-m) |
= ------ * ----- * | 1 - | --- | |
V_meas (1-m) | \ 2 / |
\ /


2^m / phi \ / V m \
= ------ * | V - --- | * | m * --- + 1 - --- | for V >= phi/2
V_meas \ 2 / \ phi 2 /


"The more accurate Poon-Gummel equation for C_j(V) is not used since the effect
of excess mobile carriers in the space-charge layers are not included in q_e or
q_c (but in q_f and q_r)."

That's all there is in that appendix.

Jon

 

[Paul Hovnanian P.E.]

Winfield Hill wrote:

Paul Hovnanian P.E. wrote...

Chris Carlen wrote:

Frank Bemelman wrote:
What a beautiful country where people *fear* being sent to prison
for shipping a handful of tubes of Atmega8515.

Imported, of course.

That's why companies like Boeing are pushing all their engineering and
s/w development offshore. If you want to export something to a
particular country, you've got to jump through all the hoops. If one of
your partners (wholly owned, of course) in China, India or Russia
develops it, the restrictions are less onerous. Then they are lobbying
for permission to purchase equipment for defense department contracts
from foreign firms where such equipment can be found at a better price
overseas. Although current restricted technology can't be exported, by
funneling the money out of the US and into these firms, eventually their
R&D will catch up. And what they develop will be beyond the reach of
ITAR. If it is outside of ITAR authority, the market for the products is
worldwide.

Read it and weep. I struggled with this stupidity 25 years ago with my
company's exports, but then it was merely a painful inconvenience, with
a financial hit against me. Now it's a financial suicide for our country.

If the US military establishment is smart and they want to keep
sensitive technology here at home, they need their own engineering and
manufacturing capability. That would solve a lot of their contracting
problems with missles as well.

--
Paul Hovnanian mailtoaul@Hovnanian.com
------------------------------------------------------------------
Telemark: If it was easy, they'd call it snowboarding.

 

[Richard]

"Dmitri(Cabling-Design.com)" wrote:

How do you guys keep track of your parts?

Physical sorting for SMT:

Keep them in the clear 1"x4" plastic baggies the suppliers ship them in,
that have a label attached with the part number & specs (at least
Digi-Key does this, and it's handy)

Clip similar groups together with a jumbo "binder clip", sorted by value

Throw the lot in a gallon-sized freezer bag (clear, heavy gauge), or
whatever suits you for bulk storage.


For tracking parts in general:

Keep a complete parts list in an Excel spreadsheet and update inventory
either as you consume a significant number of a part, or
across-the-board before re-stocking. (This isn't really for tracking
quantity on-hand - it's mostly about making cost estimates and
re-ordering quick & easy.)

* Store mfr name & part#, supplier name & part #, price each, on-hand
qty, to-order qty (with cell notes for price breakpoints and MOQ). Can
also store hyperlinks to PDFs, order pages, etc.

* Easy to sort the sheet by supplier + part number to place orders

* Add a column for qty required per project, so you can easily calc the
exact parts cost for a project


I have to say, I like the idea of using the 35mm clear storage pages to
store / sort SMT parts.

If anyone's interested, I'll post an example XLS on a.b.s.e

Regards,
Richard

 

[ChrisGibboGibson]

kensmith wrote:

In article <5kj7r09t5sdl6kik5vfgs3gtgc0lo6fbek@4ax.com>,
Jim Thompson <thegreatone@example.com> wrote:
On 06 Dec 2004 03:11:54 GMT, chrisgibbogibson@aol.com
(ChrisGibboGibson) wrote:

Tell you what......

This is something that has cropped up on this NG with monotonous regularity
over the last few years (do a quick search).

We'll settle it once and for all.

I'll build a circuit using a tranny of my choice. I'll give you *any* data
you
want from the datasheet but I won't tell you the tranny. I'll give you the
measured base current and the measured base emitter voltage and I'll also
give
you the case temperature after 15 minutes steady state. I won't exceed any
manufacturers maximum parameters.You give me the collector current.

I know *exactly* where my money is on who gets closest. It certainly won't
be
the hfe brigade.

Who's up for the challenge?

Gibbo

What's the question?

Finding the collector current given any other numbers except the part
number is, I believe, his question.

I think for added points we should put in the following additional
question:

What package is the transistor in?

Yep, I think that's fair enough. I'll go for that. You can have the package
too.

I don't think anyone will get very close. But I think the VC brigade will get a
lot close than the CC brigade.

Gibbo

 

[John Larkin]

On 30 Nov 2004 23:59:12 -0800, eliben@gmail.com wrote:

Hello,

We have a circuit that generates signals with an extremely slow
rise/fall time (milliseconds). This circuit should feed a digital
IO sampler that has a CMOS input (AHCT245/645...), and it's known
that slow rise/fall times are bad for CMOSes (forcing them to spend
long time in a conducting state, thus generating power and heat
that may cause damage).

The common solution for this problem is placing a shcmidt trigger
(74VHC14) on the circuit's output, to feed a fast signal to the CMOS.

However, one engineer presented the following appnote: "Implications of
Slow or Floating CMOS Inputs"
(http://focus.ti.com/lit/an/scba004c/scba004c.pdf) and claims that this
article doesn't mention that Schmidts themselves don't suffer from the
slow input problem, and that a Schmidt may be damaged from inputs as
slow as 1 ms. He proposes to use a comparator (that has hystheresis) as
the most robust design. On the contrary, many people say that the
Schmidt is enough, and 74VHC14'th data sheet poses no limit on the rise
time.

What is your opinion ? It's an important subject that surely raises
in many designs.

My opinion: don't worry about it. The input fets are small and the
brief crossover power dissipation won't damage the chips. The TI
appnote showed only 4 mA shoot-through (not sure for which part,
exactly) and had to conjecture 36 continuous overlaps in one part to
demonstrate damage.

And the idea that a Schmitt gate will be damaged by a slow edge
presumes that the designers are total idiots. Or something.

John

 

[K8JLF]

In the USA, you can receive various signals from NIST transmitters (10 MHz
"exactly" is one, I think). However, you must be aware of ionospheric
effects that can cause an apparent frequency shift near sunrise and sunset.

The stations in question are WWV (10Mhz) and WWVH (15Mhz) as far as I know. WWV
broadcasts from Fort Collins, Colorado, and WWVH is based in Honolulu, Hawaii
if I recall correctly. They at least used to be on other frequencies as well.
Mostly good for setting a watch by. They broadcast a tone every second, and a
different one on the minute. I have a piece of equipment somewhere in my garage
that uses their signal to generate a timing output to control something else
(designed for traffic control I think)

 

[CBarn24050]

Subject: Re: BJT transistor beta, vs Ebers-Moll, Gummel-Poon, and the other
tools God has given us
From: radio913@aol.com (Dr. Slick)


So it's easier and more measureable
to use Beta, or current gain, in
describing a bjt.

You can forget datasheets for small signal gain if you use the transistors
transconductance which for silicon is 40ic/v.

 

[Ken Smith]

In article <toMsd.18$li7.1870@news.uswest.net>,
Larry Brasfield <donotspam_larry_brasfield@hotmail.com> wrote:
[...]

Methinks that common ceramic capacitors are not
actually piezoelectric.

I think, you think wrongly here. Ceramic capacitors with high K materials
are often enough piezoelectric to make them a bad idea for small signal
work. They make a voltage if you thump on them. This makes your circuit
into a bad microphone.



--
--
kensmith@rahul.net forging knowledge

 

[K8JLF]

DHS is very interested in analyzing statistical data patterns to
identify bio-terrorist attacks. We (Pacific Northwest National
Laboratory) have a sizeable project with DHS to analyze purchase
patterns from - guess who? - WalMart. The idea is that if there is a
bio attack one of the first indicators will be people purchasing
over-the-counter meds. The type of meds will point back to symptoms,
which point back to the nature of the attack.

Or just let you know that somebody's about to cook up a big batch of meth from
pseudoephedrine.

 

[K8JLF]

Am I correct to assume that using a tweezer iron would help mitigate this,
since heat would be applied at both ends?

Hand soldering can be very hard on SMD ceramic capacitors.
The high temperature gradiant created by applying heat suddenly
at one end can fracture the ceramic. This can lead to excess noise
or a tendency to break down at a lower than rated voltage as
moisure gets into the crack(s). The insidious aspect of this kind
of damage is that it can show up in the field, quite some time
after the parts perform alright in initial testing.

At Siemens Ultrasound, we learned this the hard way, then had it
confirmed by at least one vendor's examination of abused parts.

 

[Ken Smith]

In article <UMNsd.22549$9A.388340@news.xtra.co.nz>,
Terry Given <my_name@ieee.org> wrote:
[... using 2 33pF caps on a XTAL speced for 20pF ...]

From the point of view of the crystal (IOW ignore the ground
connection) those two capacitors are in series, so 16pF total. That
capacitance is in parallel with the crystal shunt capacitance, so they
add up to (hopefully) 20pF.

Don't forget the input capacitance of the amplifying device(s). That
capacitance appears in parallel with one of the two 33pF capacitors. When
all the capacitances add up to the makers rating and the amplifier is
fast, the crystal will oscillate at the makers specified frequency.

--
--
kensmith@rahul.net forging knowledge

 

[ChrisGibboGibson]

Tell you what......

This is something that has cropped up on this NG with monotonous regularity
over the last few years (do a quick search).

We'll settle it once and for all.

I'll build a circuit using a tranny of my choice. I'll give you *any* data you
want from the datasheet but I won't tell you the tranny. I'll give you the
measured base current and the measured base emitter voltage and I'll also give
you the case temperature after 15 minutes steady state. I won't exceed any
manufacturers maximum parameters.You give me the collector current.

I know *exactly* where my money is on who gets closest. It certainly won't be
the hfe brigade.

Who's up for the challenge?

Gibbo

 

[John Larkin]

On 5 Dec 2004 17:49:52 -0800, Winfield Hill
<hill_a@t_rowland-dotties-harvard-dot.s-edu> wrote:

Second, one will never find Ebers-Moll transconductance parameters
in a datasheet, because being absolutely predicted by kT/qIc,

A 2N2222 has a transconductance roughly 1/15 this value at 500 mA. Vbe
is around 1.3 volts, according to the old green Motorola databook.

Beta, on the other hand, is only down to about 0.4 of its peak value
at the same current.

John

 

[Ken Smith]

In article <20041205230202.21770.00001557@mb-m14.aol.com>,
ChrisGibboGibson <chrisgibbogibson@aol.com> wrote:
[...]

What package is the transistor in?


Yep, I think that's fair enough. I'll go for that. You can have the package
too.

No, I was suggesting that we come up with the package type based on the
other numbers.

--
--
kensmith@rahul.net forging knowledge

 

[Richard Henry]

"Norm Dresner" <ndrez@att.net> wrote in message
news:vGSsd.1039376$Gx4.339941@bgtnsc04-news.ops.worldnet.att.net...

There are many inexpensive, so-called "atomic clocks" in the mass market
which AIUI use something like the NIST radio sites to do the time setting.
Is it possible to hack into one of these and extract a signal that could
be
used for calibration of instruments. I'm not looking for "NIST Traceable"
but just something to calibrate a bunch of frequency meters to a
reasonable
standard.

In the USA, you can receive various signals from NIST transmitters (10 MHz
"exactly" is one, I think). However, you must be aware of ionospheric
effects that can cause an apparent frequency shift near sunrise and sunset.