Inferno #09 31 июля 2006

Железо

Likbez - Characteristics of pn junctions at low current.

       Characteristics of pn junctions

             with weak current


   In practice, sometimes you want to get
bias voltage for the job is defined
parameters and modes of operation nodes. When
This is necessary, its relative
stability, and a significant current in the circuit
bias is not trebuetsya.Takzhe there are cases
when the excess current is not (in the design
cost-effective, low-power nodes), and then
require more mikromoschny source
bias. For the displacement and stabilization Oba
chno use zener diodes, and stabistorov
simple diodes, and pn junctions transistors
tors.

   Characteristics of these components
(I-V - volt-ampere characteristic) has
em obstruction and the exponential dependence
VCC. This is due to comp
NNAC electric field forming
potential barrier at the pn transition
stroke (a zener anyway), despite the fact that
value of the field is very constant. However, in
early characteristics at low current / at
voltage of the junction, the dependence nestabi
flax and poorly suited to permanent displacement
scheniya.Vse conventional zener and stabisto
algebra work well with currents from milliamperes
and unstable part of their characteristics
located to a fraction of milliamper.Poetomu for
low-current bias is easier to use
low-power LEDs with currents of tens of mA,
respectively, the beginning of their characteristics
will be in microamps. In addition (as
follows from the data), the voltage on
low-power transition at the same current is higher
What a powerful and more stable (abutment
elk).

   But to find a low-power diodes (and transitive
blinds) is not easy. Improved
technology - and produced more powerful
compact and durable "instruments, with good
Shimi harakteristikami.Vprochem, endurance
import samples gives our (stock
strength of our larger), but the parameters
our often very inferior to imports. In
recent years have become available chip instru
of the (small components that have no conclusions
transitions), and they are more suited to the role of small
schnyh, although their marriage is more common than
in the traditional (in modern tiny
mikruhah - too), and should once again about
believe the device before use.

   I am sometimes surprised by a large "no
literacy is "many mAsterovlomAsterov who put their details to 
the right and left, confident in their "as applied

universality. "They can not normally
present advantages and disadvantages of steam
meter details and often underestimate the properties
sites and work as a whole, although in most
cases, the activities of these "masters" op
ravdyvaetsya. These are often "open" their
ignorance, without knowing it, getting
trapped on the simple, classic things.
But enough of these "experts".

   As people are reasonable, we understand that to
what, and the search for low-current bias leads
leads us to the classical string vyvodu.Lyubaya
ktura system and the node optimally serves
some golden mean job conditions, and
at the beginning and end of the range of these conditions
manifest all sorts of negative factors
algebra, and therefore use midpoint.

   In our case, the top features
pn junction contains fluctuations of the electric
electric field barrier, multiplied by the defects
the transition itself, and does not provide the required 
stabilization lnosti, and at the end of characteristics 
manifested are lost (with a substantial current), and

exponential curve stabilization nak
ladyvaetsya direct active component -
resistance of the semiconductor. However, to
the end of us away and we will search the site
closer to the beginning ...

   Not to go crazy with an absurd number
honors the measurements, I decided to limit the range
Search within the 30 dB (10 mA - 10 mA).
To control / current setting used
old tester TS20-05 (its minimum
limit is comparable to my digital test
ramie, and precise - digital - setting does not
required). To reduce the error has decided to
take on two samples of the same type of diodes.
The voltage is measured at Sinometr'om VC9808
limit of 2 volts. All junctions were soldered
a series circuit and through podstroech
nickname connected to the battery (run down
D 0.125). Podstroechniki solder at about
waters so as not to heat the transitions. In fact,
a small stream, the possible error of those
cluster, despite the huge input resistance
resistance. On the big - the error of time
number of battery, though I for compensation
podstraival.V Total current accuracy and
spread (between the measurements of diodes) way
is called the 5% and I have not tried it stabilizes
synthesized. In the beginning I was wrong - by Meryl
slot voltage, and the first series was on
current of 5 mA, and in the end did not have the EMF for 10
mA, and I decided to restrict the new limit:
5 mA - 5 mA. Then, for accuracy, to ten
measurements range 5 - 10 - 20 - 50 ... add 9
intermediate (15 - 35 - 75 ...), but in the end
decided to measure the leakage current, reverse
the inclusion of ...


          Now about the contenders



  1. Small kvadratnenky, presumably
Indeed, the RF switching diodik flat
legs on one side and the numbers of carbon
Lamas housing, which together can be read
as 5646. I hate to obscure markings and
 lack of data references ...

  2. Gray balloon with incomprehensible to
lustertym symbol and a thick nozhkami.Kor
empty instantly warmed by adherent
of a soldering iron, and keep your hands it is impossible
but. Presumably, the high-frequency switching diode
average power. Steel-gray band at
 cathode.

  3. As above but with a blue stripe at the
 cathode and in the middle.

  4. KD104A. High-voltage rectifying
diode as a small drop of super soft
 ribbon pin 300 V, 0.1 A.

  5. Classic collector junction of
pulyarnogo transistor KT315 (Table
the letter of the sample) 0.1 A, 25 (A) or 35
 (F) B.

  6. Emitter junction KT315, it can be
assume sverhslabotochnym among all Prete
ndentov, with conditional data of 6 V, 10 mA.

   Specifies maximum reverse voltage
of, according to the reference data. Incidentally, in the
references often meets with incorrect
sequence in the designation of the maximum permissible
of the emitter-base voltage. Do not indicate that
a reverse voltage, rather than direct. WPRO
What, all so it should be clear. For about
sample 5 the maximum current is essentially
return (mode of operation - Transistor), but
Judging by what we're missing is the
Coy same magnitude current through the same transition,
it is a normal assumption ... Table
indicated millivolts.



   Writing the data into the computer and traced
for increasing the voltage drop, I found
three errors and correct the results.

   To measure the reverse current, I used
call of 20 V by connecting through the restrictive
1 kOhm resistor to the anode (soldered all devices
a common point). But since the leakage current too
is very small and is not fixed at the limit of 2 mA,
I used mode DCV 200 mV (M832) -
as nanoampermetr by dividing the reading on
input impedance (1 Mohm). But these data
nye include only a very approximate
mi, due to very low voltage and noise
to a negligible current. Readings plate
Wali, and I took the average value is very conditions
doubtedly his count.

   Based on these data can not be done
conclusion of belonging to a unit Vysokovo
ltnym or low voltage, although some
connection is traced. Not counting the transistors
ditch, the KD104A - the smallest leak, but
300, it should be 3 uA - there is a nonlinear
linear dependence of the inverse curve. And
it should be. And here is the smallest leak -
collector junction - did not correlate
tsya high voltage. Large leak
characterized samples, calculated on the value
siderable current, but again manifests itself
not since the beginning of characteristics, while mean
flax stress - when the breakdown starts
transition. Given this, we can say that
the first 3 sample is unlikely to withstand hundreds of
volts, and one, probably, to the marriage, and this
due to the reduced fall in the early
characteristics. But the emitter junction
try! But this was to be expected.
Stress on them was about 7 V
at a current around 14 mA. And slope characteristics
characteristics (like zener diodes) at a very
good level (voltage sits less
than polvolta at currents in microamperes).
Not surprisingly, the emitter junction
used as a zener mikromoschnye
(Seen in magazines several circuits). By
the same reason, it is clear how easy it is povre
dit transistor static potential
or touching the accumulated charge.
Not surprisingly, the chip tranzyuki fly at
mistreated ...

   If we calculate the percentage growth of the Pade
of, it becomes clear that at low current
it is clearly superior to the next, and POE
what we can say that the growing part
characteristics becomes flatter - CO
lasno exponentially. But first low-power
Sample the beginning of a table closer to the flat frequency
In particular, than the next two transitions that
Again, due to their large working
current and the shift characteristics of the end of the table
Litsa. But the greater the fall in the early
Table of KD104A due to their Vysokovo
ltnostyu rather than low current. Lida
ramie are transistor transitions - from
their biggest drop in at the beginning of Tabley
particle. However, they have seen (although to a lesser
degree - they are the low-voltage) exponential
exponential dependence of the beginning of the table.
The latter explains the decrease in the amplifying
abilities transistors work in weak
sneeze currents (up to percent of maximum)
revealing the benefits of a "weak" tranzyukov
in mikromoschnyh sites ...

   Somewhat surprising identity of the emitter
rnoy and collector curves. But this explanation
Nimo small currents of experience. At the end of the table
You can see how the high-transition
amplitude and, correspondingly, small operating currents. Y
those transitions, where the resistance is higher -
increase more (added to the drop with
resistance of the junction). It is evident that the slope of
2 nd and third samples is much larger (respectively
respectively, higher operating current), while the transition
moves considerably smaller (this is natural, ve
qb base current is weak, while the main work
current flows from the emitter to the collector).


                  Conclusions


   Leaders for the low-current bias NE
lyayutsya transitions of low-power transistors,
and the emitter can be used as
zener with a small napryazhenie.No
This should not be carried away by currents, it is better not
more than a couple mA and not change the current more
than doubled, otherwise the bias will not dostato
chno stable. Because of noise in the early hara
kteristiki should not use the transition
offset / current less than poluvolta, although
it all depends on the application of circuitry and
specific cases.

   In some cases, to determine a suitable
leads whether the sample is sufficient to measure drop
voltage across it at two different currents (from
difference of no more than twice) about the representation
polagaemogo working value. Thus,
Thus it becomes clear what the limits of deviation
voltage. But we should not judge strictly -
this is not a diode (not stabistorov), and
use at higher multiplicity of current
(More than twice) is unlikely to be determined
ravdano.

March 15, 2006 By KSA-7G