Transistor identification?

[DaveC]

Help with identifying a TO-92 transistor?(at least I think it’s a
transistor…)

Pretty vanilla-looking with National Semi logo. Marked:

NS 444
ST
13903

(The “NS” is the logo.)

Not enough 4’s to make up a standard 2N number, and 13903 looks like a date
code?

Ideas?

Thanks,

[DaveC]

> Help with identifying a TO-92 transistor?

A photo:

http://i.imgur.com/z78xG5D.jpg

[Tim Williams]

How's it measure?

(Or if it's dead, is there another one probably working that you can test?)

Tim

--
Seven Transistor Labs
Electrical Engineering Consultation
Website: http://seventransistorlabs.com

"DaveC" <n...@home.cow> wrote in message
news:0001HW.1B9B87FD0...@news.eternal-september.org...
> Help with identifying a TO-92 transistor?(at least I think it’s a
> transistor…)

>
> Pretty vanilla-looking with National Semi logo. Marked:
>
> NS 444
> ST
> 13903
>

> (The “NS†is the logo.)
>
> Not enough 4’s to make up a standard 2N number, and 13903 looks like a
> date
> code?
>
> Ideas?
>
> Thanks,
>

[DaveC]

> How's it measure?
>
> (Or if it's dead, is there another one probably working that you can test?)
>
> Tim

Dead. Other with identical markings measure as NPN. What else to determine
from measurements?

Thanks.

[John Robertson]

Looks like a house number - made by National of course. House numbers
can sometimes cross reference to other parts if you can find a master list.

How made the item in question.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
(604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

[DaveC]

> [Who] made the item in question. [?]
> John :-#)#
An American offset printing equipment company. It’s a paper/envelope feeder
for a small offset presses. Extinct company.

I can say that the bandwidth requirement of the circuit is very small—DC-10
KHz—and low current. Don’t know about gain.

Thanks.

[N_Cook]

Or maker ST and date code for 1994 or 2004 or whatever goes with NS as
plant/batch code for a particular decade

[John Robertson]

There are pretty simple circuit diagrams out there to measure gain if
you have more transistors with the same markings.

One example:

http://electriciantraining.tpub.com/14179/css/14179_111.htm

The only other criteria are breakdown voltage and current. If it only
drives other transistors and resistors to Vcc are over around 2K then it
is likely 100ma or less.

If you know the Vcc you know the minimum breakdown voltage unless it is
being used to control solenoids that have their own B+ source, then that
voltage is your absolute minimum (times at least 2).

[piglet]

On 06/09/2015 07:05, DaveC wrote:
> I can say that the bandwidth requirement of the circuit is very small—DC-10
> KHz—and low current. Don’t know about gain.
>
> Thanks.
>

You could try a 2N3904 or other jelly bean type and see if the thing
then works?

piglet

[Tim Williams]

"DaveC" <n...@home.cow> wrote in message

news:0001HW.1B9BFB310...@news.eternal-september.org...

> Dead. Other with identical markings measure as NPN. What else to determine
> from measurements?

Find a modest to high voltage supply, add a large series resistor
(10k-100k?), and measure the reverse breakdown.

Wire it as a common emitter amplifier, not really to amplify but just to
check bias. Play with base resistors until you get the edge of saturation
for some collector current. Then double the collector current and do it
again. That'll more or less let you plot hFE(Ic). Where it drops by half
or so is near Ic(max).

Now you know Vebo, Vcbo, hFE and Ic(max). Vebo is generally less than 5V
for RF parts, so you'll likely have a high fT in that case. You can set it
up with a square wave input to measure t_r, t_f, t_stg etc. if you like.

[John Robertson]

His next post gave a link to a photo of the transistor. Sure looks like
National Semiconductor to me. My guess is the date is from the 90s based
on the edge of the capacitor shown in the same photo - makes the date
1994 and 44th week.

It will be Silicon then and Ted's recent post in this thread gave a very
good explanation on how to identify its parameters using an identical
transistor.

The OP could tell us if there are other transistors using the 13903, but
a different code beside the NS symbol - that would pretty much show that
three digit number is the date code if it is something like 4xx, 3xx, or
5xx which would be a reasonable spread for date codes for a special run.

[DecadentLinuxUserNumeroUno]

On Mon, 7 Sep 2015 13:29:29 -0700, John Robertson <sp...@flippers.com>
Gave us:

>John :-#)#
>--
>(Please post followups or tech inquiries to the USENET newsgroup)
>John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
>(604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
> www.flippers.com
> "Old pinballers never die, they just flip out."

Ever use "MAME"?

[Ian Field]

"DaveC" <n...@home.cow> wrote in message

news:0001HW.1B9BFB310...@news.eternal-september.org...

You can usually distinguish the B/E junction because it will zener somewhere
around 5 - 8V ish.

You need to keep the test current pretty low - you can get ultra-efficient
LEDs that give a useable indication at only 2mA. Hook one of those up with
an A23 12V keyfob battery and a current limiting resistor.

[krw]

InGaN (true) green LEDs are quite bright at only 1mA. They can be
seen at much less than that.

[Jeroen Belleman]

On 2015-09-17 21:16, Ian Field wrote:
>
> "DaveC" <n...@home.cow> wrote in message
> news:0001HW.1B9BFB310...@news.eternal-september.org...
>>> How's it measure?
>>>
>>> (Or if it's dead, is there another one probably working that you can
>>> test?)
>>>
>>> Tim
>>
>> Dead. Other with identical markings measure as NPN. What else to
>> determine
>> from measurements?
>
> You can usually distinguish the B/E junction because it will zener
> somewhere around 5 - 8V ish.

> [...]

You do that with low noise transistors and they'll be useless.

Jeroen Belleman

[Phil Hobbs]

An ordinary 4-3/4 digit DVM on the lowest volts range (400 mV usually)
makes a poor man's picoammeter. The 10 megohm ones read 1 LSB (0.01 mV)
for 1 pA. Some of the older ones, where the lowest range doesn't have
the 10M resistor in parallel, can be much more sensitive than that.

You aren't going to hurt a transistor with a nanoamp of reverse base
current, and the measurement is pretty simple.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net